Implementation Guide

Success by Design
Microsoft Dynamics 365
Success by Design Implementation Guide

Published by Microsoft Corporation, 1 Microsoft Way, Redmond, Washington 98052.

Copyright © 2021 Microsoft Corporation.

All rights reserved. No part of this book may be reproduced or used in any manner
without permission of the copyright owner.

Azure Cosmos DB, Bing, BizTalk, Excel, GitHub, Hololens, LinkedIn, Microsoft 365,
Microsoft Access, Microsoft Azure, Microsoft Dynamics 365, Microsoft Exchange Server,
Microsoft Teams, Office 365, Outlook, Power Automate, Power BI, Power Platform,
PowerApps, SharePoint, SQL Server, Visio and Visual Studio are trademarks of Microsoft
Corporation and its affiliated companies. All other trademarks are property of their
respective owners.

The materials contained in this book are provided “as-is” and Microsoft disclaims all
liability arising from you or your organization’s use of the book.

Information contained in this book may be updated from time to time, and you should
refer to the latest version for the most accurate description of Microsoft’s FastTrack for
Dynamics 365 program.

Case studies used in this book are fictional and for illustrative purposes only.

First Edition

Printed in the United States of America

ii
This book is dedicated to Matthew Bogan, our colleague and friend. The way you think
and the way you approach everything you touch embodies a growth mindset—that
passion to learn and bring your best every day to make a bigger difference in the world.
Thank you for being a shining example.

iii
 What’s inside
Strategy................................................1
1 Introduction to Implementation Guide............................................. 2

2 Success by Design overview.................................................................12

3 Implement cloud solutions..................................................................27

4 Drive app value.........................................................................................60

5 Implementation strategy......................................................................85

Initiate............................................. 106
6 Solution architecture design pillars................................................. 107

7 Process-focused solution.................................................................... 122

8 Project governance................................................................................151

9 Environment strategy........................................................................... 188

Implement..................................... 224
10 Data management............................................................................... 225

11 Application lifecyle management................................................... 260

12 Security..................................................................................................... 283

13 Business intelligence, reporting, and analytics............................ 318

14 Testing strategy..................................................................................... 346

15 Extend your solution............................................................................ 380

16 Integrate with other solutions...........................................................413

17 A performing solution, beyond infrastructure........................... 462

Implementation Guide: Success by Design iv

 Prepare........................................... 500
18 Prepare for go live................................................................................. 501

19 Training strategy................................................................................... 539

Operate.......................................... 579
20 Service the solution.............................................................................. 580

21 Transition to support.............................................................................611

Conclusion............................................................................................................ 652

Acknowledgments............................................................................................. 655

Appendix................................................................................................................657

Implementation Guide: Success by Design v

Section
Strategy
1 Introduction to Implementation Guide
2 Success by Design overview
3 Implement cloud solutions
4 Drive app value
5 Implementation strategy

1
1 Guide
Introduction to
Implementation
Guide
Implementation Guide: Success by Design: Introduction to Implementation Guide 2
“Our success is dependent on our customers’
success, and we need to obsess about them—
listening and then innovating to meet their
unmet and unarticulated needs. No customer of
ours cares about our organizational boundaries,
and we need to operate as One Microsoft to
deliver the best solutions for them.”
– Satya Nadella, Chief Executive Officer of Microsoft

Overview
Microsoft believes that every business is in the
business of creating great customer experiences.

As a precursor to the Dynamics

To achieve that, business applications must do more than just run your
365 implementation resources in back office, marketing, supply chain, or even field operations as discrete
this book, this chapter will: entities. Organizations need to tap into data signals from across their
Explore the opportunities presented as business—and use them to drive ongoing improvements in the products,
organizations shift into the predictive era.
services, and experiences they intend to deliver to their customers.
Outline our vision for Dynamics 365.

Introduction
Help readers understand what an
investment in Dynamics 365 invites
in terms of enabling holistic, unified
digital transformation.

In this chapter, we explore how the shift from the reactive past to the
predictive era is currently affecting our industry—and will eventually
affect every industry. We also look at Microsoft’s vision for Dynamics
365 and how our apps allow holistic, unified digital transformation. In
our view, these topics are essential to anyone about to embark on an
implementation of Dynamics 365.

Implementation Guide: Success by Design: Introduction to Implementation Guide 3

 Armed with these perspectives, the rest of this book—which aims to
share the mindset and guidance needed to successfully implement
Dynamics 365—is sure to take a more meaningful shape for readers.
This chapter’s content is adapted
from the combined thinking of James For Microsoft, this is especially important, considering the investment
Phillips, President of Microsoft’s Digital
that you are making in us.
Transformation Platform Group; Rolf
Harms, Corporate Vice President of
Microsoft’s Cloud and AI Strategy; and

Every industry
Muhammad Alam, Microsoft’s Corporate
Vice President of Dynamics 365.

turned upside down

We find ourselves at a turning point: over the coming decades, the role
of technology will be turned upside down with profound implications
Fig. 1-1
for every aspect of business—from customer experiences to core
operational processes to the very nature of products and services.
Reactive past
Picture this: in the reactive past (Figure 1-1), a customer purchases a
Applications
car from a dealer. One day, while going down the road, the car stops
running. The customer calls the service center and reports a problem.
Application platform At this point, a service advisor opens an application, types a few notes
about the problem into a form, and then schedules a repair visit. It
may be the first time in years that any data related to the customer
or the vehicle has flowed—a few kilobytes trickling down, and only in
reaction to the car’s malfunction—into the dealer’s relational database.
Only now the dealer can provide service to the customer, but the car’s
Relational database manufacturer might never see a single byte of data from this event. As
this example illustrates, in the reactive past, data flows downward from
applications at only a drip.

Now consider what is possible today: the same customer purchases a

car from the dealer. With appropriate permissions granted by its new
owner, the car provides a continuous stream of data back to the dealer
and the manufacturer. It may be a self-driving car, with computer
vision models trained in the cloud via data captured by the entire
fleet’s cameras. Every operating parameter of the engine, every press
of the brake, every trip route—all are signals that provide a rich basis
for predictions to ensure a higher-value relationship between the
customer, the dealer, and the car’s manufacturer.

Implementation Guide: Success by Design: Introduction to Implementation Guide 4

 In Microsoft’s view, the possibility of a higher-value relationship is
present for every product, every service, and every business process.
In the predictive era (Figure 1-2), the paradigm is that everything is a
source of data and potential insights. As organizations embrace this
shift, they enable an upward data flow that originates from everywhere
and, at the same time, informs their products, services, and applications.

For companies that “grew up” in the predictive era, that paradigm is
second nature. But for mature enterprises, the path can be more diffi-
cult, as it requires transformation of existing capabilities and processes.

For Microsoft, this isn’t just theory or prediction. We have been pursuing
our own transformation and have faced many of these challenges
ourselves. At the heart of our transformation is the way we build products.
Instead of the multiyear product cycle of traditional, on-premises
products, Dynamics 365 is delivered as a service that’s continuously
refined and always up to date. It also allows us to capture rich telemetry
on bugs and usage to inform ongoing feature development.

Beyond its influence on the way we run our own business, this digital
transformation paradigm has affected how we think about success for
our customers. Microsoft’s vision has always been about democratizing
Fig. 1-2

Predictive era

Dynamics 365 apps

Power Platform

Azure
Data ingestion (Planet Scale
Foundation)

Implementation Guide: Success by Design: Introduction to Implementation Guide 5

 software and technology capabilities to empower everyone. Looking
forward, we have renewed our purpose in the context of the cloud in a
world where everything is a source of data and potential insights.

Microsoft’s mission is “to empower every person and every organization

on the planet to achieve more.” Harnessing data to help you predict
the future and intelligently guide actions represents a core capability
that invites organizations to achieve more by innovating everywhere
(Figure 1-3).

Microsoft’s data-first
cloud strategy
The opportunity contained in the shift from reactive to predictive—know-
ing that the car is on its way to failure before it fails—is hard to overstate.
For Microsoft and for our customers, every product, service, and business
process is ripe for reinvention. But the reinvention requires a wholesale
reimagination—a literal turning upside down—of the systems that power
the interconnected fabric of products, services, and applications.

To help companies reinvent their businesses and thrive in the predictive

era, Microsoft has invested in applications (such as Dynamics 365
Sales), an application platform (Microsoft Power Platform), and an
infrastructure platform (Microsoft Azure) spanning from enterprise and
Fig. 1-3
independent software vendor (ISV) developers to citizen developers.

Innovate everywhere

Suppliers Assets Operations Employees Products and Customers

Create a resilient, Get the most from Optimize your Empower your services Delight your
efficient supply your assets. operations. people and teams. Connect your customers from
chain. products and first touch to
services. delivery to success.

Implementation Guide: Success by Design: Introduction to Implementation Guide 6

 The entire Microsoft cloud (Figure 1-4) comes together as a unified
digital-transformation platform with consistent security, identity, and
compliance boundaries.

With everything as a source of data and potential insight, Microsoft’s

goal is simple: help customers take the right action at the right time
with the right message via the right channel to achieve the right busi-
ness outcome. (Throughout this book, you may find that we refer to
this concept as the digital feedback loop, as shown in Figure 1-5. It
Fig. 1-4 demonstrates how organizations tap into data sig-
Microsoft cloud nals from across their business, and use those signals
to drive ongoing improvements in the products, ser-
vices, and experiences delivered to their customers.
Microsoft 365 Microsoft Dynamics 365 The customers then respond to these improvements,
thus fueling a self-reinforcing cycle.)

Dynamics 365:
Power Apps Power Automate Power BI Power Virtual Agents

Microsoft Power Platform

Cloud-based
business apps
Microsoft Azure

for holistic,
Identity, security, management, and compliance

unified digital
transformation
Fig. 1-5

Engage Empower
customers people Keeping the opportunities of the predictive era top
Intel
l i ge nt of mind, we want to take a step back and focus on
what Microsoft got right with Dynamics 365, further
Data & AI
explain our data-first strategy in terms of Dynamics
365, and provide advice to customers embarking on
s
sy

te
ce

ms n
ie
s

per
and ex
an implementation of Dynamics 365.
Optimize Transform
operations products Looking back, it’s clear that Microsoft’s company-
wide push into the cloud set in motion the trajectory

Implementation Guide: Success by Design: Introduction to Implementation Guide 7

 of what we know today as Dynamics 365. But also notable was Microsoft’s
decision to embark on a business applications journey into the cloud
with its previously named Dynamics CRM and Dynamics AX products.

Although they were separate applications at the time, Microsoft’s deci-

sion to move forward with Dynamics CRM and AX foreshadowed the
importance we place on developing a cloud-based business application
solution that offers customers a holistic, unified digital-transformation
platform covering the front office, the back office, and beyond.

Fast forward to today, and few of Microsoft’s competitors can offer a

complete cloud business application with the breadth, depth, and level
of integration with existing Microsoft products that Dynamics 365 has.

Dynamics 365 is a portfolio of business applications that meets

organizations where they are—and invites them to digitally transform.
Each app is designed to remove the barriers and eliminate silos within
organizations by working together with existing systems—or the entire
portfolio of Dynamics 365 apps—for outcomes you simply can’t get
unless every part of the business is connected seamlessly.

Microsoft’s data-first Dynamics 365 strategy

If you’re reading this book, then your business likely has already invested
in Dynamics 365. However, understanding Microsoft’s future vision for
Dynamics 365 will help you take full advantage of your investment.

Building on Microsoft’s cloud strategy (which includes Dynamics 365),

several other factors contribute to Microsoft’s data-first strategy to put
Dynamics 365 into a category of one:

Dynamics 365 is a ▪ Dynamics 365 provides front-office and back-office cloud applica-
tions that are consumable in a composable manner, which means
portfolio of business that our customers can maximize their investment without having
applications that to do an extensive rip and replace of what were historically behe-
meets organizations moth customer relationship management (CRM) and enterprise
resource planning (ERP) processes.
where they are—
▪ Dynamics 365 is built upon the low-code Power Platform to
and invites them to enable pro and citizen developers to build solutions, automate
digitally transform. processes, and generate insights using Power Automate, robotic

8
 process automation (RPA), and Power BI—which no other vendor
can offer with its core platform.
▪ All of these are natively built on Azure, the cloud with an unmatched
level of security, trust, compliance, and global availability.
▪ Add to this the assets of Office 365 (such as Microsoft Teams,
LinkedIn, and Bing), which can be used to enrich the business
application experience.

With Dynamics 365, organizations have the only portfolio of intelligent

business applications that enables holistic, unified digital transformation
and empowers everyone to adapt and innovate—anywhere.

Advice to customers
embarking on an
implementation of
Dynamics 365
Microsoft’s ideas on what customers should keep doing—whether they
are embarking on an implementation of Dynamics 365 or any other busi-
ness application—are detailed later in this book, but they include many
of the traditional activities, such as business-process definition, quality-
requirements management, end-user readiness, and change management.

However, Microsoft recommends against project teams that structure

The ability to their engagements as massive 9-month to 18-month waterfall efforts.

start quickly in From a cloud services perspective, this old, on-premises implementation
style needs to be jettisoned in favor of agility.
the app allows
organizations to The ability to start quickly in the app allows organizations to bring value
bring value to to their business in an accelerated fashion. The focus of project teams

their business in should be on finding the function or area of the application that the
business can start using as quickly as possible, and then expand and
an accelerated enhance from there. The application and the business are going to change
fashion. anyway, so it is in the best interest of organizations to claim value early.

9
 More pointedly, Microsoft’s view is that being agile does not mean
engaging in sprint cycles that still require a 9-month to 18-month team
effort to deploy part of the application to end users.

Traditional business application models can no longer keep up. Long

setup times, expensive and complicated customizations, and slow
adoption curves make it nearly impossible to meet the needs of a
changing and demanding customer base.

Think beyond the implementation—

and beyond the technology
The key question as you embark on an implementation of Dynamics
365: why are you doing it? Is it just a rip-and-replace effort, or does the
implementation represent an opportunity for digital transformation,
advance your business model, allow you to leapfrog over the competi-
tion, or something else?

In making an investment in business applications, organizations are

called upon to clearly identify the value they intend to drive toward.
Will the technology get you there? Does the software or the imple-
mentation partner have the vision to get you there?

An investment in business applications requires a level of organizational

maturity to think about what you and your business are after. It also
requires you to think beyond just the implementation of technology.
Organizations that
understand the Organizations that understand the power of data—and want to
power of data—and harness it so that they can disrupt the market—are the business appli-

want to harness it cation projects that Microsoft wants to be a part of. This is where we
believe the world is going, and our products reflect that.
so that they can
disrupt the market—
are the business The journey forward
application projects
We understand that every organization comes with a unique set
that Microsoft wants of challenges, opportunities, and priorities. No matter where you
to be a part of. are in your digital transformation, Implementation Guide offers the

10
 guidance, perspective, questions, and resources to help you drive a
successful Dynamics 365 implementation, and harness data for insights
that lead to the right actions and the right business outcomes across
any business process.

The first section of Implementation Guide provides an overview of the

Success by Design framework, which Microsoft created to help our
customers and partners drive successful implementations of Dynamics
365. Additional chapters in the first section focus on the mindset and
thinking required to implement in the cloud.

Subsequent sections of Implementation Guide address the major topics

in the Dynamics 365 implementation lifecycle, including Microsoft’s
point of view on solution design, environment strategy, data, security,
business intelligence, integration, performance, user readiness, support,
and much more.

Whether you’re a first-time implementer or a been-there-done-that

Dynamics 365 architect, we hope you’ll bring us along on your journey.

Implementation Guide: Success by Design: Introduction to Implementation Guide 11

2 Guide
Success
by Design
overview
Implementation Guide: Success by Design: Success by Design overview 12
 Success by
Design overview
This chapter introduces Success by Design—a
framework and practice created by Microsoft to
help project teams implement Dynamics 365.

Based on thousands of real-world customer projects, Success by Design

is the sum of our Dynamics 365 implementation experience. It offers
topic-specific reviews and prescriptive guidance (approaches and rec-
ommended practices), which provide a reliable path to project success.
Success by Design is intended to be used by Dynamics 365 system inte-

One of the primary grators, independent software vendors (ISVs), and customers as a means
to better architect, build, test, and deploy Dynamics 365 solutions.
goals of this book is
to democratize the For our customers, Microsoft recognizes that Success by Design
practice of Success doesn’t guarantee implementation outcomes, but we’re confident that

by Design by making it will help you achieve your project’s goals while enabling the desired
digital transformation for your organization.
it available to the
entire community For our partners, Microsoft is confident that Success by Design, cou-
of Dynamics 365 pled with your implementation methodology and skilled resources, will
increase your team’s effectiveness in delivering successful Dynamics
implementers.
365 projects to your customers.

Implementation Guide: Success by Design: Success by Design overview 13

 Success by Design history
As demand for Dynamics 365 cloud services increased across the
enterprise, Microsoft identified the clear need to change the way we
thought about evolving our services and our responsibility to customers
and partners. We recognized that it wasn’t enough to design a platform
containing a set of features. We needed to also understand what it
takes to deliver a fully functioning, end-to-end solution that runs the
mission-critical processes of our customers’ businesses. From this need,
Success by Design was born.

Microsoft believes that customer success is the precursor to every

Dynamics 365 product development decision. Because of this, questions
fundamental to product development now challenge our engineers at
every step of the process:
▪ In what way must each product feature be delivered to ensure
customer success?
▪ How do we make Dynamics 365 apps valuable, durable, reliable,
and performant?
▪ What aspects of the product should be jettisoned because they
don’t meet fundamental customer success criteria?

For the first time in the product’s history, such questions have led to
the successful transition of 100 percent of Microsoft Dynamics 365
online customers to one version, to a dependable and safe deployment
process, to a steady and coherent feature release cadence (including
many overdue feature deprecations), and to a reliable and performant
platform that delivers customer value. But product is only one half of

Microsoft believes the customer success equation.

that customer Microsoft has put equal emphasis on the need to provide you the
success is the prescriptive guidance and recom­mended practices to ensure a smooth
precursor to every implementation project and a properly designed and built Dynamics 365
solution that successfully transforms business operations. This push has
Dynamics 365 also resulted in the transformation of Microsoft’s FastTrack for Dynamics
product development 365 service, which makes certain that our community of Dynamics 365
decision. implementers—customers and partners—has access to Success by Design.

14
 This chapter focuses on the fundamentals and practice of Success by
Design and its desired result: Dynamics 365 projects whose technical
and project risks are proac­tively addressed, solutions that are roadmap
aligned, and project teams that are successful in maximizing their
organization’s investment in Dynamics 365 cloud services.

Make Success by
Design your own
Enabling Dynamics 365 project teams to practice Success by Design
means an ongoing commitment by Microsoft to provide the commu-
nity of implementers—customers and partners—best-in-class business
applications, along with the latest in Success by Design project resources.

It also means project team willingness to incorporate the practice

of Success by Design into their overall project approach—regard-
FastTrack for Dynamics 365 is a
customer success program run by less of implementation methodology or the Dynamics 365 product
Microsoft’s Dynamics 365 product
being implemented. At its best, Success by Design fortifies a project
engineering team that helps
customers implement Dynamics 365 team’s chosen implementation methodology with a model for prod-
apps and realize business value faster.
The practice of Success by Design is
uct-aligned project governance.
fundamental to FastTrack’s approach,
but the Success by Design framework
is available for use on any Dynamics No matter where you find yourself in your Dynamics 365 implemen-
365 implementation project. tation, fundamental to Success by Design is the willingness of project
teams—despite the common pressures of time, budget, and resources—
to pause in order to understand and address technical and project risks
before it’s too late in the project lifecycle.

To this end, this chapter focuses on the what and the why of Success by
Design, as well as how project teams can use it to accelerate customer
success throughout the implementation of Dynamics 365.

Success by Design
objectives
Success by Design is prescriptive guidance (approaches and recom-
mended practices) for successfully architecting, building, testing, and

15
 deploying Dynamics 365 solutions. Success by Design is informed by
the experience of Microsoft’s FastTrack program, which has helped our
customers and partners deliver thousands of Microsoft’s most complex
In this book, we use the term “Finance
Dynamics 365 cloud deployments.
and Supply Chain Management” to refer
to the collective category of Dynamics
365 apps that includes Finance, Supply
Chain Management, Commerce, Human
Success by Design reviews are exercises in reflection, discovery (obser-
Resources, and, in some cases, Project vation), and alignment (matching to known patterns) that project teams
Operations.
can use to assess whether their implementation project is following
recommended patterns and practices. Reviews also allow project teams
to identify (and address) issues and risks that may derail the project.

Success by Design should be used as an adjunct to the project team’s

For interested customers, Microsoft chosen implementation methodology for the following benefits:
recommends that project leaders team
▪ Reduced risk due to early detection of problems
up with their implementation partner
to enable Success by Design within their ▪ Alignment to recommended practices
project. In addition to the Success by Design
resources available in this book, it’s highly
recommended that project teams ready The result is a roadmap-aligned solution architecture that is performant,
themselves by enrolling in the Success by
Design training on Microsoft Learn. scalable, and future-proof. ¹

Success by Design phases

Success by Design maps the Dynamic 365 implementation lifecycle
into four methodology-agnostic phases: Initiate, Implement, Prepare,
and Operate (Figure 2-1). In this section and the following sections, we
outline the Success by Design phases, their relationship to Success by
Design reviews, and the desired outputs and outcomes.
Fig. 2-1 In the Initiate phase, the project
team is in discovery mode—gath-
ering and validating business
requirements, finalizing the high-
level solution approach, making
inroads to define all in-scope
workstreams, and updating the
Initiate Implement Prepare Operate
project plan to reflect these up-
dates. When the project team has
¹ Acknowledging the possibility that feature produced the high-level solution
deprecations or other changes to product roadmap
in compliance with Microsoft policy may occur. design and the related project

Implementation Guide: Success by Design: Success by Design overview 16

workstreams are more or less defined, Success by Design begins with
the Solution Blueprint Review. (More on the Solution Blueprint Review
and other reviews later in this chapter.)

In the Implement phase, the project team is focused on building the

solution per the agreed-upon solution design and scope. Implementation
Reviews are introduced in this phase, having been informed by the
findings and recommendations of the Solution Blueprint Review. As we
learn later in this chapter, Implementation Reviews are used to more
deeply address questions related to the specific aspects of the solution
design (data model, security, integration) and implementation practices
(ALM, testing strategy). Implementation Reviews are meant to fully ad-
dress the risks identified during or after the Solution Blueprint Review
but before the solution build is too far along.

By the Prepare phase, the solution has been built and tested and the
project team is preparing for the final round of user acceptance testing
(UAT) and training. Additionally, all necessary customer approvals have
been granted, information security reviews completed, the cutover
plan defined (including go/no-go criteria), mock go-lives scheduled,
the support model ready, and the deployment runbook completed
with tasks, owners, durations, and dependencies defined. At this point,
the project team uses the Success by Design Go-live Readiness Review
to identify any remaining gaps or issues.

In the Operate phase, the customer solution is live. The goal of this
phase is stabilization and a shift in focus towards functionality and
enhancements that are earmarked for the next phase of the project.

Success by Design reviews

With a high-level understanding of Success by Design phases, we now
turn to Success by Design reviews.

Each review raises questions that serve as points of reflection that

project teams can use to generate important discussion, assess risk,
and confirm that best practices are being followed.

17
 The Solution Blueprint Review serves as the starting point of Success
by Design. We suggest that the Solution Blueprint Review be a man-
datory review for the project because findings that come from it lead
to Implementation Reviews, which offer project teams the opportunity
Later chapters of this book will cover each
Success by Design review in greater detail.
to drill down into topic-specific areas where deeper dives are deemed
For the most comprehensive coverage of necessary for further understanding project and solution risk. Finally,
Success by Design, refer to the Success by
Design training on Microsoft Learn. the Go-live Readiness Review, which we also suggest as a mandatory
review, is the last stop for assessing any remaining risks before go-live.

Figure 2-2 illustrates that Success by Design reviews are not to be

conducted as abstract exercises separated from the project. Rather, the
scheduling and implementation of each review relies on the availability
of key project artifacts and the readiness of the project team to discuss
them. (For a more in-depth look at Figure 2-2, visit the Appendix,
which provides Microsoft’s view of the typical deliverables, activities,
tasks, and stakeholders involved in each Success by Design phase.)

Success by Design outputs

With a basic understanding of Success by Design’s objectives, phases,
and review flow, it’s important to pause to understand the makeup of
Fig. 2-2 review outputs and their purpose—findings and recommendations.

Initiate Implement Prepare Operate

Get ready to start Design and build Deploy Live

Requirement analysis Code Testing and acceptance Solution health

Project governance Configure Go-live planning Usage

Fit gap analysis ... User readiness Maintenance

Customer kick-off Data modeling Cutover planning ...

... Solution performance ... Post Go-live Readiness Review

Solution Blueprint Review Integration Go-live Readiness Review

Success by Design
review and workshops
...

Implementation Guide: Success by Design: Success by Design overview 18

Fig. 2-3
As described in the previous
Matched to Lead to section, reviews are informed by
Findings known patterns recommendations project artifacts (project plans,
requirements, fit gap, solution
architecture, design documents,
and so on) produced by the project
team, and other information such
as formal or informal project
touchpoints. The availability of
such information points to the readiness of the team to schedule and
conduct Success by Design reviews. Additionally, project teams may rely
on Microsoft for telemetry and other tools to inform review discussions
and generate review outputs.

The primary review outputs fall into two related categories: findings
and recommendations.

Findings come in three types:

▪ Assertions  Findings that capture noteworthy aspects of the
solution or approach. Assertions highlight what the project team is
doing right, typically in line with best practices.
▪ Risks  Findings that could potentially impact the implementation
negatively if not mitigated.
▪ Issues  Findings that are currently impacting implementation
negatively or will do so if not resolved.

Findings should include as much detail as possible and be matched

to known patterns. Findings matched to known patterns often yield
insights that lead to recommendations or actions necessary to resolve
the issues identified (Figure 2-3).

To better illustrate this point, consider this customer example:

A global corporate travel company is implementing Dynamics 365

Customer Service to drive its call center transformation. As the project
team digs into the business’s requirements, they learn that the Dynamics
365 solution must account for integration with multiple legacy systems

19
 (many with high transaction volumes). Additionally, the requirements
point to a business-mandated customization that the project team
agreed could not be achieved using out-of-the-box functionality.

In preparation for the Solution Blueprint Review, the project team

parses these and other details, including confirming that solution
performance testing was purposely left out of the project scope on
the assumption that Microsoft’s Dynamics 365 cloud service should be
performant on its own. (Alt­hough it’s true that Microsoft is responsible
for delivering a reliable and performant cloud service to its customers,
our experience is that solution design and the resulting configurations,
customizations, and ISVs to achieve that design may play a role in
impacting overall Dynamics 365 solution performance.)

Considering that the Dynamics 365 solution in question is projected to

In Success by Design, findings link support 4,000 users at scale (including the multiple integrations and
observations to known patterns that invite
a key customization mandated by the business), the project team’s
actions necessary to address project risks
and issues. findings are clear: keeping solution performance testing out of scope
is a risk that negatively impacts the project. Among other findings
and recommendations not covered in this example, the project team’s
architect (who led the Solution Blueprint Review) recommends that
the project Steering Committee approve adding solution performance
testing into the test cycle.

The architect’s findings are summarized as follows:

▪ The solution requires custom development to meet its
business requirements.
▪ Performance testing is not included in the test cycle.
▪ Following best practices, solution performance testing should be
added to the test cycle (and if necessary, the solution performance
workshop should be scheduled to further explore the risk and report
any additional findings back to the project Steering Committee).

Track success measures

One more step remains in the Success by Design process: tracking success
measures. After completing Success by Design training, you may decide

Implementation Guide: Success by Design: Success by Design overview 20

 to create your own success measure tracking template. Otherwise, project
teams can use Success by Design tooling (as served by FastTrack or your
implementation partner) to track success measures. But what are success
measures, and why are they important?

What are success measures?

By its very nature, Success by Design is a project-specific endeavor. For
anyone practicing Success by Design across one or many Dynamics
365 projects, the question arose: How do we measure the health of just
one of those many projects? Success measures allow us to do just that.

Success by Design empowers project teams to track the health of

projects across seven categories and over 30 success measures. Figure
2-4 highlights some of these catego­ries and related success measures.

Tracking success measures is simple: After a Success by Design review

or other compelling project event, your FastTrack Solution Architect or
partner architect will access Microsoft’s Success by Design tooling and
update the relevant success measures for the project. Success measure
updates are either red, yellow, or green, and include project-related
details relevant to the success measure.

Fig. 2-4
Why are success
Review Findings Recommendations measures important?
Success measures are important
because they provide access to
micro and macro project health
trends. Tracking success mea-
sures for a single project allows
Success measures stakeholders to assess the overall
health of the project at a glance.
Similarly, the benefit of tracking
Risk Assertion Issue success measures over 10, 20, or
100 projects is that Microsoft, the
partner, or the customer project

Implementation Guide: Success by Design: Success by Design overview 21

 team can see patterns that may be impacting them. For example, mac-
ro-level tracking may yield an application lifecycle management (ALM)
problem. Access to this data allows Microsoft or the partner to under-
stand whether it’s a project or product problem, and gain insights on
how to fix it.

Conclusion
Success by Design equips project teams with a model for technical and
project governance that invites questions and reflection, which leads to
critical understanding of risks that might otherwise go unnoticed until
too late in the project.

Considering the pace of cloud solutions and the investment that organi-
zations make in Dynamics 365 software and implementation costs, even
the most experienced customers and partners with the best methodolo-
If you’re using Success by Design within gies will benefit by incorporating Success by Design into their projects.
your Dynamics 365 implementation
project, we want to hear from you! To
share your Success by Design feedback As covered in Chapter 1, Microsoft believes that every business is in
and experiences, reach out to us at
[email protected]. the business of creating great customer experiences. To achieve that,
business applications must do more than just separately run your back
office, marketing, supply chain, or even field operations. They must give
you the agility to remove every barrier in your way. When this happens,
business applications become more than just operational solutions.
They become resilient solutions that enable the digital feedback loop
and adapt to customer demands, delivering stronger, more engaging
experiences around the products you make and services you provide.

22
 Case study
An inside look at the
evolution of Success by Design
When our Solution Architects got their start with Dynamics 365, they
found themselves hooked on the challenge of effectively aligning an
organization’s business processes and requirements with the (software)
product in hopes of delivering solutions that resulted in actual value
for users.
The content of this case study is based on
interviews conducted with Dynamics 365
Because of this, FastTrack Solution Architects also know the challenge of
FastTrack Solution Architects. The goal is
to provide readers with an inside look at working with people, understanding the business, shaping software to
Success by Design. We hope you enjoy
this insider view.
meet the needs of the business, and delivering a solution that is accepted
by users. They know that projects don’t always go well and often fail.

It’s for exactly this reason that the FastTrack for Dynamics 365 team
and Success by Design were created.

A shift to Dynamics 365 in the cloud

Dynamics 365 isn’t software that is purchased and implemented as is.
As such, implementing Dynamics 365 products rely on qualified partners
to get the implementation right.

Complicating matters further, the promise of the cloud has been

that you can focus on the application and forget about everything

Implementation Guide: Success by Design: Success by Design overview 23

 underneath. But the reality is that cloud implementations require a
comprehensive understanding of how design and build decisions may
impact Dynamics 365 cloud performance, scalability, and more.

Accordingly, the shift to implementing Dynamics 365 in the cloud re-

mains challenged by on-premises implementation habits in which the
mindset is to overly customize—“because you can”—without impact.

The shift to implementing Dynamics 365 in the cloud requires not

just technical capability but also a definitive understanding of how to
design and build Dynamics 365 solutions within the constructs of the
online service.

The evolution of Success by Design

FastTrack for Dynamics 365 was formed to help customers maximize
their investment in Dynamics 365 by making sure that they (and
their partners) have access to the right information, recommended
practices, and a connection to the product group throughout the
implementation lifecycle.

When the FastTrack for Dynamics 365 team was formed, these goals
were initially achieved without the use of a framework. In other words,
FastTrack Solution Architects used their experience and connection to
the product team, but this approach contained too many variables to
be consistently reliable.

Over time, Microsoft’s experience revealed that Dynamics 365 customers

were running into very similar problems. For example, FastTrack would
identify that some aspect of the solution was designed and built in
a manner that wasn’t compliant with recommended Dynamics 365
implementation practices. And unfortunately, too often the Solution
Architects would catch the issue too late in the implementation lifecycle.

As a result, Microsoft began to ask itself: How can we change the

FastTrack approach so our Solution Architects can address problems
before they manifest themselves as problems? FastTrack also

Implementation Guide: Success by Design: Success by Design overview 24

 acknowledged that its approach couldn’t be “one Solution Architect
solving the problems of one customer project.” Our Solution Architects
needed to work in a way that would benefit all customers.

To identify and address problems early, our Solution Architects had

to engage with project teams on multiple levels. We needed to un-
derstand project plans, how project teams were thinking about doing
customization, data migration, and more. This need for early visibility
across all dimensions of the project meant aligning to an approach
that afforded our Solution Architects the opportunity to get involved
in these conversations from the start, to identify the problems and risks
early, and to give the right guidance to the project team before more
problems arise.

This required a period of FastTrack experimenting with the right approach.

We started a concept that we referred to as checkpoints, which involved
using surveys, but we found that this approach kept us in a reactive
mode. We kept revising our approach and eventually we got it right.

From this experimentation, Success by Design was born: a proactive

As stated earlier in this chapter, Microsoft practice that invites a purpose-built dialog that touches all levels and
recommends that project leaders team
dimensions of the project (technical and nontechnical) so problems are
up with their implementation partner
to enable Success by Design within their identified before they impact the project.
project. In addition to the Success by
Design resources available in this book,
it’s highly recommended that project Success by Design gets to the heart of how project teams are thinking
teams ready themselves by enrolling in
the Success by Design training on and what they’re doing so that any risks can be identified and mitigated.
Microsoft Learn.

With Success by Design, FastTrack joins the customer, our partners, and
the product team in a manner that ensures alignment between all three.

Success by Design for every

Dynamics 365 customer project
At some point during the evolution toward Success by Design,
Microsoft asked the question: How can we make sure every Dynamics
365 implementation is successful from the start? The key phrase is
“every Dynamics 365 implementation.”

Implementation Guide: Success by Design: Success by Design overview 25

 As this chapter highlights, the FastTrack for Dynamics 365 team doesn’t
have a special lease on the practice of Success by Design. This book’s
goal is to democratize Success by Design and make it available to the
entire community of Dynamics 365 implementers.

The more project teams invest in Success by Design, the more they will
get out of it.

Implementation Guide: Success by Design: Success by Design overview 26

3 Guide
Implement
cloud solutions
Implementation Guide: Success by Design: Implement cloud solutions 27
 Introduction
The cloud—the constellation of connected servers, data
warehouses, and associated software and services—is at
the core of a digital transformation strategy.
Organizations of all sizes recognize that the cloud is the fastest way
to modernize their operations to increase efficiency while better serv-
ing customers. For many organizations, the shift to the cloud simply
means running their applications in a public cloud infrastructure
instead of the traditional on-premises datacenter. For other organi-
zations, however, a software as a service (SaaS) model that uses cloud
applications as well as the underlying cloud platform is a quicker path
Adopt a cloud mindset to digital transformation.

Cloud implementation Whatever model an organization chooses, we find that the key to
success isn’t what implementation looks like, but rather how leaders,
Customize and extend
cloud applications architects, and entire companies approach the digital transformation
journey. Success in the cloud isn’t just about the technology or the fea-
Operate in the cloud tures available—it’s about the organizational mindset. For a successful
digital transformation, your organization must prepare for changes
Evergreen cloud
that span the entire enterprise to include organizational structure,

Upgrade from on-premises processes, people, culture, and leadership. It’s as much a leadership
to the cloud and social exercise as it is a technical exercise.

Implementation Guide: Success by Design: Implement cloud solutions 28

 In this chapter, we provide a perspective on what to expect when imple-
menting your cloud solution using Microsoft Dynamics 365. We introduce
principles relevant to any SaaS application. Concepts introduced here
Supporting research and studies
are elaborated on in subsequent chapters. We hope these principles help
McKinsey  The keys to a successful you set your course and guide you in your leap into the cloud.
digital transformation

Smarter with Gartner  Cloud Shift We start by addressing how to develop a cloud mindset in an organi-
Impacts All IT Markets
zation, which is foundational to your digital transformation. Then we
delve into factors to consider and understand before you shift to a
shared public cloud from an on-premises setup. These include how the
shift will impact controls you have in place and how to think differently
about scalability, performance, shared resources, and more. After
that, we discuss the principles around customization and extending
cloud apps. Finally, we explore the operating model, always up-to date
evergreen cloud models, and the options to migrate to the Dynamics
365 cloud from on-premises.

Adopt a cloud mindset

Adopt a cloud mindset
Cloud implementation Embracing change is a common theme, but infusing it throughout your
organization is vital to your digital transformation journey. Therefore,
Customize and extend
cloud applications we include it as a thread that runs throughout this chapter and book.
Adopting the cloud mindset is also an intentional way to help organiza-
Operate in the cloud tions break the mold and think differently about how applications are
designed, developed, secured, and managed in the cloud.
Evergreen cloud

Upgrade from on-premises

First, let’s understand the “why” behind the pervasive adoption of SaaS
to the cloud business applications by organizations across the globe. Traditional
approaches to business applications mostly involve the IT department
building a custom application or using an off-the-shelf product aug-
mented with customizations to meet the specific business requirements.
The typical approach includes phases to gather requirements, build,
go live, operate, and then follow the change management process to
request changes or add new features. The IT department has full control
of the application, change process, releases, updates, infrastructure, and
everything needed to keep the solution running for years.

On the surface, this sounds ideal. It has been the standard operating

Implementation Guide: Success by Design: Implement cloud solutions 29

 model for deploying and operating business applications. This way of
working, however, is losing relevance in today’s fast-paced world of
shifting supply chains and evolving customer needs and expectations.

Application development and deployment was disrupted not neces-

sarily because of the cloud or new technology—the undercurrent that
made the cloud pervasive was the changing landscape of business.
Unharnessed data from ubiquitous smart and IoT devices and other
sources fueled startups that disrupted whole industries seemingly
overnight. With that, the average lifespan of companies plummeted.
This inflection made the survival of a business dependent on delivering
the best product, the best customer experiences, and constant inno-
vation to meet ever-growing expectations. This new business arena
means businesses must change more often, evolve their processes
more quickly, and become more data driven to understand customer
needs, pain points, and preferences.

The changes we’re talking about aren’t triggered by the cloud or tech-
nology. These are thrust upon us by the changing business landscape.
Adopting a cloud mindset, then, is about transforming your processes,
people, culture, and leadership in such a way that your organization can
embrace change quickly and successfully. We believe the fundamental
organizational characteristics that will determine success in this environ-
ment come down to focusing on delivering business value through a
secure cloud technology platform. This includes the ability to harness the
data that is captured to generate insights and actions. This platform will
also rely on automation to quickly react to changing needs.

In this section, we explore these organizational characteristics that help

drive the cloud-first mindset to achieve digital transformation.

“Every company is a technology company, Focus on business value

no matter what product or service it
provides. Today, no company can make, Technology is a ubiquitous force that influences almost every business.
deliver, or market its product efficiently
without technology.”
Indeed, the IT department within a bank or a manufacturing com-
pany may be larger than entire tech companies. Understanding how
—Forbes, “Why Every Company is a
Technology Company” technology affects customers and business is important to help any
company survive and thrive. It’s also important to not let technology

30
 hinder growth or impede business. The role of technology is to deliver
business value by driving efficiency.

However, we often see enterprises whose technological landscape

has become so large and complex over the years that they fail to
deliver the agility demanded by the business. The IT departments are
consumed with dealing with the complexity of software compatibility,
update cycles, end of life deadlines, aging infrastructure, and antiquated
security policies with little time to focus on delivering business value.

An evergreen SaaS cloud approach to your business application,

however, takes away much of the infrastructure and operational IT
effort to keep the servers running, software patched, and other routine
tasks, and turns the focus of IT on the business. Operating in the cloud
doesn’t mean there are no operational responsibilities, environment
IT has a greater management, updates to SaaS services, feature deprecations, user
opportunity to communications, or coordinating change management for enhance-

focus on delivering ments and new capabilities, monitoring, and support. These functions
are all still required, but with a key difference: IT is now much closer to
positive business the business application layer and has a greater opportunity to focus
outcomes. their energy on delivering positive business outcomes.

Traditionally, business applications depict a set of key business processes

that allows the users to capture, track, and report on the process and
its progress. These business processes usually take the form of having a
user interface (UI) built on top of an underlying data store to capture or
display the information from the database. These “forms over data” soft-
ware applications can help organizations keep track of information and
make it easily available through reports. However, from an end user’s
perspective, they’re a data capture tool that doesn’t really help them
do their job any faster or better. This design of business applications
hasn’t changed in decades. It’s an outdated model that doesn’t bring
any additional value to the business or to the end user, and doesn’t
differentiate a business’s offerings and customer experience from their
competition, which is likely following a similar outdated approach.

Additionally, on an individual level, we want to automate routine,

boring tasks and focus on more creative, challenging work.

Implementation Guide: Success by Design: Implement cloud solutions 31

 For example, take a point-of-sale application that helps you address
your customer by first name and provides them personalized relevant
offers based on their loyalty, or recommends products based on their
purchase history and Artificial Intelligence (AI). This will be better
received by your employees and customers than an application that
can barely determine if a product is in stock or not. Simply offering the
best deal or best quality product isn’t enough in today’s hyper-com-
petitive environment. To win, you need to differentiate between your
customers, respond to their unique behaviors, and react to fluctuating
market demands.

Businesses likely already have the data to do this, but may lack the
technology to turn data into practical insight. For example, Dynamics
Automate the 365 Commerce taps your systems to gather customer intelligence from
routine tasks so a myriad of sources (such as social media activity, online browsing hab-

you can focus on its, and in-store purchases) and presents it so that you can shape your
service accordingly. You can then manage product recommendations,
creative, more unique promotions, and tailored communications, and distribute them
challenging work. across all channels via the platform.

Dynamics 365 applications are focused on delivering value to the busi-

ness and empowering end users, changing the business application
industry paradigm by forcing every business to innovate faster.

The “If it ain’t broke, don’t fix it” mentality common in large organiza-
tions, where business applications can remain unchanged for up to 10
years, puts the enterprise at risk. Unless your business is a monopoly
and can sustain without investing in technology, the hard reality is
most businesses today won’t survive without strategic plans for digital
transformation.

You need to ask yourself some questions:

▪ Has the rest of your industry not transformed and
embraced technology?
▪ Do your customers demand a better and faster process?
▪ Is your competition growing their market share by eating into yours?
▪ Does your competition have a better Net Promoter Score (NPS)?

Implementation Guide: Success by Design: Implement cloud solutions 32

 If you answered yes to most or if you don’t know the answers, your
business might be failing to adapt and react to changes in customer
behavior and expectations. Businesses that don’t invest in technology
will face challenging disruptive competitors, either global behemoths
or innovative startups. You need digital transformation technology
that gives you the command of three essentials: deep, cross-channel
customer insight; synchronized operations from front end to back end;
and scalability to drive rich experiences as needed.

Decision-making around such technology now carries existential

consequences. Lack of strategic visionary thinking, long-term planning,
and matching investments can prove to be catastrophic.

The critical point is your business application shouldn’t be designed

and deployed with an expectation that it will remain unchanged for
long periods of time. Instead, the core design principle for business ap-
Decision-making plications should be ease of change. You should take advantage of the
around digital wonders of automation to deliver changes as often as daily and weekly,

transformation now at the whim of business. This continuous development philosophy is

core to a Dynamics 365 SaaS application with weekly maintenance
carries existential updates and multiple waves of new features delivered every year.
consequences. Realizing the value of these cloud applications, therefore, is about
adopting the latest best-in-class capabilities, continuously delivered.
With this new mindset, SaaS services aren’t just used to meet business
requirements—they’re helping drive business value and adoption.

Increasingly, technology is becoming so powerful and democratized

that non-developers and people with no formal training in software
development are empowered to build applications, create workflows
that automate their work, and use machine learning with ready-made
AI models to analyze data. This concept of citizen application devel-
opment has drastically transformed businesses, where end users are
developing solutions to solve business problems and sharing these
applications with colleagues. This organic approach to app devel-
opment has short-circuited the application development lifecycle,
helping address the perennial issue of software adoption. IT depart-
ments are now responsible for the governance of these applications,
securing the access to data and environment, and focusing on building

Implementation Guide: Success by Design: Implement cloud solutions 33

 the underlying services and connector interfaces that the business can
consume in their applications.

These scenarios all help illustrate how a transition to the cloud can play
an instrumental role in refocusing technology to deliver business value.

Driven by data
This disruptive innovation we’re seeing across several industries is
fueled by data. Almost everything around us generates data: appli-
ances, machines in factories, cars, apps, websites, social media (Figure
3-1). What differentiates one company from another is the ability to
harness this data and successfully interpret the information to generate
meaningful signals that can be used to improve products, processes,
and customer experiences.

Organizations recognize data is an asset, and becoming data driven is

a key motivator for digital transformation. Still, the accounting rigor
that we see in finance is seldom observed when it comes to managing
organizational data. If data is the new currency, then we should
treat it like that. Get an understanding of your organizational data
Fig.
3-1 estate—the databases, data warehouses, and data lakes—that help
your company manage its cor-
porate data. Develop a holistic
Engage Empower view of customer data across
customers people departments, functions, and
applications, while acknowledg-
l i ge nt
Intel ing the regulatory boundaries
on data usage. These steps are
practically a prerequisite to
Data & AI digital transformation.
s
sy

te
ce

ms n
ie
s

an d ex per Depending on the volume of

data, its age, and organizational
Optimize Transform silos, the deduplication, correla-
operations products
tion, and conflation of data from
different systems could become
a complex project on its own.

Implementation Guide: Success by Design: Implement cloud solutions 34

 However, modern AI-powered customer data platform (CDP) technol-
ogies could come to the rescue. A CDP technology creates a persistent,
unified customer database that is accessible to other systems. Data is
pulled from multiple sources, cleaned, and combined to create a single
customer profile.

It’s also important to define your organizational data strategy for

business applications, develop guidelines on how new applications
store their data, what schema they use, and how other applications in
the organization can understand and use this data. You need to use the
most optimal datastore for your use case. For example, storing credit
card transactions in Dataverse (a repository to store and manage data
used by business applications) is suboptimal; a data lake might be a
better choice.

Being data driven is about better understanding your data’s quality

and relevance, and being able to generate valuable insights that are
actionable and can drive meaningful changes to your processes. Going
back to the point of embracing change, your data strategy should
reflect how you design your process, develop your applications, and
operate. If it takes several months to deploy a change request or years
to roll out an improved process, there is little point.

A data-driven business application cloud platform like Dynamics 365

enables you to:
▪ Create a data estate that is intelligent and AI-ready
▪ Connect and consume data from legacy on-premises systems,
other line of business applications, and third-party applications
▪ Have the necessary controls to secure and meet data compliance
requirements
▪ Build smart applications that are continuously improving
▪ Infuse data intelligence directly into the application, empowering
users to make effective and smart decisions
▪ Have agility so applications can quickly adapt to changes in
your process

Overall, the intelligence and insights you generate from your data
will be proportional to the quality and structure of your data. You

Implementation Guide: Success by Design: Implement cloud solutions 35

 can explore this topic more in Chapter 10, “Data management,” and
Chapter 13, “Business intelligence, reporting, and analytics.”

Think platform
An organization can take several approaches towards digital trans-
formation. In many cases, you start with a single application being
deployed to a SaaS cloud. A key responsibility of IT decision-makers

You’re investing and enterprise architects is to deliver a cloud platform for their or-
ganization’s digital transformation. Individual applications and their
in the platform, app feature sets are important, but you should also look at the larger
not just the picture of what the cloud platform offers and what its future roadmap
application. looks like. This will help you understand if the platform can meet the
short-term and long-term objectives of your digital transformation.

Thinking about the platform versus a single app offers clear benefits.
You avoid reinventing the wheel with each additional application,
and you instead deliver a foundation approved by governing bodies,
with clear patterns and practices. This approach limits risk and brings
a built-in structure that enables reuse. Platform thinking doesn’t nec-
essarily mean that other cloud platforms or legacy applications have
to be rebuilt and replaced. Instead, you can incorporate them as part
of an all-encompassing platform for your organization, with well-de-
fined patterns and swim lanes that enable integration and flow of data
between applications. Bringing this “systems thinking” to deliver a
platform for business applications can drastically reduce the amount
of time lost getting security and design approvals for individual apps,
thereby improving your agility and time to value.

Your business applications platform should already provide the following:

▪ Necessary security and design approvals for processing data in spe-
cific categories based on its sensitivity and regulatory requirements
▪ Clear guidelines on extending and customizing the platform to
ensure supportability and agility
▪ Established governance process to manage access and
service operation
▪ Approved integration patterns for communication with other
platforms and on-premises systems, with a clear process for
approving deviations

Implementation Guide: Success by Design: Implement cloud solutions 36

 ▪ Data storage guidelines for apps and patterns for sharing
▪ License entitlements that are available for individual apps
▪ Documented service protection limits and best practices that your
applications should comply with
▪ Resources for additional learning and guidance with access to experts

The due diligence and foundational work you do when choosing a

platform in coordination with IT and business decision-makers can save
weeks, if not months, of effort downstream when developing individual
apps. This thorough evaluation also helps drive predictable success.

Embrace DevOps and automation

Traditionally, the process to manually release a change to applications was
convoluted, involving several teams, change boards, manual regression
Due diligence tests, and approvals. It took weeks. This complexity makes teams averse to
change. Enhancements and updates are then consistently deferred, which
now helps drive affects the adaptability of your applications to shifting business and tech-
predictable success nology priorities. Automating builds and releases and building in testing
in the future. automation is crucial to allow for continuous software updates.

The biggest selling point of the cloud is to stay current and quickly deploy
software. You can’t rely on teams to manually test and deploy each and
every change to achieve that vision.

Traditionally, a time lag existed between code and test and deploy.
A bug in production meant repeating the lengthy cycle. The fear of
breaking code meant that teams tended to delay updates for as long
as possible. With automation, you deploy fast and potentially fail fast.
This is where the culture of fail fast comes in. Automated processes
help companies empower their teams to take risks. You fail but quickly
release a fix. Over time, failures decrease and success rates improve,
instilling confidence in teams to deploy and innovate, and delivering
value to the business faster. At the core of the cloud mindset is un-
derstanding and embracing DevOps. DevOps is the union of people,
process, and technology to continually provide value to customers.
DevOps is a cultural change in which we break down silos between
developers and administrators, and create shared responsibility for the

Implementation Guide: Success by Design: Implement cloud solutions 37

 quality of released software. The DevOps process automates software
delivery and speeds up the technology deployment lifecycle (Figure 3-2).

The idea is to enable development, IT, and quality and security teams
to collaborate to produce more reliable and quality products with the
ability to innovate and respond more quickly to the changing needs of
business. Implementing DevOps inherently depends on automation:
automating the build process to enable continuous deployment,
regression tests to improve speed and reliability, and administrative
actions like user and license management.

At this point, we should also explore continuous integration (CI) and

continuous delivery (CD) in the context of the cloud. The idea is each
code check-in from the developer should go into the main build and
get tested early. This takes CI to the next level by allowing for automat-
ed testing and continuous, automated code deployment.

If a change includes a bug, the build fails, and the cycle repeats. The
Fig. automation allows for changes to
3-2
happen quickly without unneces-
sary overhead—a team can focus
Plan on the business logic and not the
infrastructure. Therefore, with CD
Monitor Code we always have the most up-to-
date working software.

CI and CD bring agility to cloud

DevOps solutions by allowing for the latest
Operate Build
processes changes and bug fixes to move to
production code without delay.
This is a paradigm shift from the
world of on-premises software, in
which each change had to wait un-
Deploy Test
til all user stories were completed,
the full package was deployed, and
Release a testing team tested the change.
With automation, we’re reducing
time and improving quality.

Implementation Guide: Success by Design: Implement cloud solutions 38

 But this isn’t just a technological change. DevOps requires close collabo-
ration between development, testing, and deployment teams. This is a
cultural and mindset change in which the whole team trusts each other,
feels empowered, and is collectively responsible for product quality.
Dynamics 365 offers deep integration
with Azure DevOps with build tools to
Preparing your organization for the fast-paced world of the cloud
help automate your deployment pipeline means that automation isn’t optional anymore. Instead, investments
and effectively manage the application
lifecycle (Figure 3-3). in automation will deliver significant efficiencies and help improve the
overall reliability of your applications.

Cloud implementation
considerations
Implementing solutions from cloud SaaS products can reduce man-
agement overhead and allow you to deliver business value more
quickly by focusing on the application layer. Fundamental application
considerations, however, still apply to SaaS cloud applications. Security,
scalability, performance, data isolation, limits, and capacity are still
critical, but could have a different approach when compared to an
application deployed on-premises. This section focuses on some of
these considerations for SaaS cloud apps.
Fig.
3-3

ALM Powered by Azure DevOps

Initiate Build Release
Getting started, faster Build and walk away Automated, predictive, repeatable

Pack source Run

Provision Deploy Import Run solution Export Run solution Import as Increment Export
code from Run unit test Unpack repo Run unit test Pack solution integration
environment dependencies solution checker solution checker unmanaged version managed
repo test

Initial build pipeline instantiates pristine Building pipeline automates manual steps. No more Automated release pipeline removes manual steps. Weekly, daily, or hourly releases
development environment daily. need to upload to solution checker and manually become the new standard.
export solution, unpack, and push to repo.

Implementation Guide: Success by Design: Implement cloud solutions 39

 Adopt a cloud mindset Security readiness
The responsibility for security in SaaS cloud applications is shared by
Cloud implementation both the service provider and the customer. That means your existing
Customize and extend security policies might not be suitable to meet the security require-
cloud applications ment in the cloud. The SaaS service provider processes customer data
and is responsible for aspects such as securing the backend infrastruc-
Operate in the cloud
ture and data encryption in transit and at rest. As a data controller, the

Evergreen cloud customer is still responsible for securing access to environments and
application data.
Upgrade from on-premises
to the cloud
The IT information security team should clearly understand the
boundaries of these shared responsibilities to ensure the following:
▪ The SaaS service provider meets the organizational security,
privacy, and compliance requirements. This is usually done in the
beginning to approve the platform for use supported by a regular
review or audit process to ensure continued compliance.
▪ The security team is aware of the controls and configurations
required to govern access to data and fulfill the customer se-
curity responsibility. These could include defining the data loss
prevention policies, access control policies, and environment man-
agement policy. They usually have a default setup with a process
to manage deviations for specific apps.
▪ The governance process makes sure the application-level security
requirements are met. This is primarily managed by the appli-
cation teams and driven by business requirements for each app
deployment. You might have additional checks before deployment
to ensure compliance.

A good example of this is General Data Protection Regulation (GDPR)

in the EU, in which the SaaS service itself might be fully GDPR com-
pliant from a data processor standpoint but the customer is still
responsible for implementing data controller processes like a “forget
me” request or managing marketing consent for contacts.

Organizations need to review their information security policies and

ensure that they’re cloud ready. Doing this exercise earlier, before the
app deployment cycle, will help avoid delays. It’s important to closely
work with your cloud SaaS service provider to ensure all your security

Implementation Guide: Success by Design: Implement cloud solutions 40

 requirements are met. Microsoft publishes the details of how this
shared responsibility of securing data is fulfilled from a data processor
perspective on the Microsoft Trust Center, which provides detailed in-
formation on our security and privacy policies as well as the certificate
of compliance for various regulatory norms and internal standards.

Scalability
The scalability of the SaaS platform is a key consideration for business
applications. Being able to scale out and scale up to support seasonal
workloads or spikes in user activity will impact the overall user experi-
ence (both staff and customers) and effectiveness of business processes.

The cloud scalability parameters available in SaaS differ from a

traditional on-premises system. Instead of adding more servers or
increasing the power of machines, these parameters could be trans-
lated as available API capacity. You also shouldn’t assume that cloud
means infinite scale and computing power that can process everything
thrown at it. The good old coding and design best practices are still
relevant, but might need to be adapted. Although managing and
scaling cloud services is complex, increasing or decreasing your data
storage, processing power, and networking requirements can be done
seamlessly. In many cases, you can do so automatically or with simple
configuration changes. This microservices architecture, including
capacity-based routing to resources and storage management, is
transparent to Microsoft customers.

To summarize, the SaaS cloud offers the flexibility to scale horizontally

and vertically to support thousands of concurrent users. Embracing the
capacity-based model for resource consumption in the cloud helps not
only build optimized applications but also a better plan for operating
cost post deployment.

Latency and performance

Performance is a key consideration for business applications—it not only
impacts the end user experience and adoption, but can directly impact
business goals and key performance indicators (KPIs) of success.

Implementation Guide: Success by Design: Implement cloud solutions 41

In on-premises deployments, enterprises had complete ownership and
control of the infrastructure, and could guarantee and monitor appli-
cations’ latency and performance. In the cloud world, it’s not
as straightforward.

Research shows that a few milliseconds of latency lead to a big

percentage drop in page load times. For e-commerce companies,
this could mean a sizable drop in user attention and therefore sales
revenue. Low latency is not a good to have, but a critical deciding
factor in an enterprise’s brand and reputation. The same applies
to back office business applications operating in the cloud—user
experience and productivity can be significantly impacted in a
high-latency environment.

Several contributing factors can impact application performance,

including network latency, user device and browser, application de-
sign, and customizations. SaaS services are inherently scalable, have
vast compute power, and are available from multiple locations around
the world, but that doesn’t necessarily guarantee performance if the
solution is poorly designed, or if users are accessing the service from
environments with high network latency.

Your cloud applications are powered by hundreds of pieces of virtu-

alized hardware running on Azure datacenters sprinkled around the
world. A crucial decision you must make, then, is choosing the data-
center to deploy your application so that all users of the application
get an acceptable level of latency.

Performance also can suffer because of a poorly designed application. If

you’re changing standard code in SaaS or deploying custom applications
in platform as a service (PaaS) or infrastructure as a service (IaaS), your
team must design a high-performing application and carry out proper
performance tests, and only then do you deploy to production. Now, most
cloud providers work on a shared network infrastructure in which your
requests go through the internet and are used and shared by everyone.

Network latency is a crucial factor to consider alongside other architec-

tural and design decisions that impact performance.

42
 One option for businesses looking for more reliability and security is
to use a private connection. Cloud providers offer dedicated channels,
You can test latency using the Azure for example Azure ExpressRoute. These connections can offer more
Latency Test and Azure Speed Test 2.0 for reliability, consistent latencies, and higher security than typical connec-
each datacenter.
tions over the internet.

Isolation in the shared cloud

The cloud operates on a shared common infrastructure across differ-
ent businesses, which leads to economies of scale. But although the
infrastructure and processing may be shared in public clouds, you
have data isolation, which means one customer’s data isn’t shared
with others. Security and data use policies are taken very seriously by
Microsoft; it’s imperative that we maintain the highest data and secu-
Private connections rity standards. These standards and practices are audited and certified
on a regular basis by trusted third-party organizations. Security isn’t
offer reliability, an afterthought, but ingrained into how we develop applications, so
consistent latencies, that the computing infrastructure is shared and other security related
and higher security aspects are isolated.

The other aspect of sharing resources and infrastructure in the cloud is the
possibility of monopolizing resources or becoming a noisy neighbor. While
rare, these cases are more often caused by poorly developed runaway
code than a legitimate business use case. Automated monitoring and
protection throttles are built into the Dynamics 365 platform to prevent
such situations, so it’s important to understand and comply with these
service boundaries and throttles when designing for cloud platforms.

Service protection
and a capacity-based model
Service protection and limits are used in cloud services to ensure
consistent availability and performance for users. The thresholds don’t
impact the normal user operations; they’re designed to protect from
random and unexpected surges in request volumes that threaten the
end user experience, and the availability and performance characteris-
tics of the platform or solution. It’s crucial to understand the protection
limits and design for them with the appropriate patterns, especially
around high-volume workloads like integrations and batch processing.

Implementation Guide: Success by Design: Implement cloud solutions 43

 The cloud provides us the scalability to deal with large workloads, but
a shared public cloud doesn’t mean you have unlimited capacity or
computing power. In many cases, these capacity limits are enforced
Plan for the daily peak and the monthly
maximum order transaction
via licensing. You have to embrace this capacity-based model of cloud
volumes expected to ensure that the infrastructure and plan to operate within the entitlements, taking
service is licensed to satisfactorily
support the peak as well as the total into account usage profile, data volumes, and integration patterns.
maximum expected volumes. Also, plan Understanding these protections and service boundaries that apply
and prioritize for integrations and Open
Data Protocol (ODATA) requests based to a specific service helps bring clarity on available resources. It also
on the volumes, so they’re honored and
drives the right behavior when it comes to design and development so
not throttled due to request limits. These
checks and balances prevent overutilizing the solution is optimized to operate within the allocated capacity, or
resources, preserve the system’s respon-
siveness, and ensure consistent availability
additional capacity is procured to meet the requirements in advance.
and performance for environments
running Dynamics 365 apps.
Check out Chapter 20, “Service the solution,” for more information.

Integration with on-premises systems

The SaaS cloud might be the preferred approach for new applications,
but in most enterprise solutions, you could still have on-premises
elements. Therefore, it’s common to build integrations with these
on-premises systems. Establishing these integration patterns in
compliance with your internal security policies while allowing cloud
applications to authenticate and connect to on-premises services is
critical. This could involve allowlisting URLs and IP ranges on your
firewall, or using technologies like the Azure AD Application Proxy or
an on-premises gateway that allows such integrations without having
to open inbound connections into your customer’s corporate network.

Even from a design standpoint, patterns that worked on-premises

could be challenging in the cloud, due to latency and additional secu-
rity restrictions imposed. An example is the encryption of traffic leaving
a customer network or routing via a peering network.

Cloud and edge

The public cloud is the default and recommended choice of deployment
for the majority of Dynamics 365 customers, but you may want to extend
the cloud to unlock important capabilities for key scenarios where con-
nectivity is an issue. Your industry may need a solution on the ground,
such as brick and mortar commerce, field service, a manufacturing execu-
tion system (MES), warehouse operations, and project operations.

Implementation Guide: Success by Design: Implement cloud solutions 44

 These implementations are called cloud and edge, where cloud is the
system of intelligence and edge is the system of record.

For example, Dynamics 365 Commerce customers use the cloud and
edge model, in which the store operations run on-premises (edge)
and the main instance handles centralized back office functions like
finance, data management (such as products and customers), and
analytics (cloud).

Cloud and edge deployments provide business continuity for

mission-critical functions like point of sales, warehousing, and manufac-
turing. We want you to be able to continue running operations when
disconnected from the cloud. In case of a network outage, Dynamics 365
Point of Sales can sustain itself with basic operations and keep data gen-
Patterns that erated locally. When connectivity resumes, data is synchronized to the
worked on-premises cloud. Also, whenever throughput is high and more heavy processes are

could be challenging run in parallel, user productivity may be impacted. We want to enable
you to scale out and run manufacturing and warehouse processes in
in the cloud. isolation so high user productivity is always supported.

Another reason for choosing an on-premises deployment model is

that these businesses have made significant investments in local IT
infrastructure. We want to ensure you receive a meaningful return on
investment (ROI) before taking the full cloud route.

Another reason is internet connectivity. Regions with poor internet

connectivity see more occurrences of on-premises deployments.
Organizations operating in such regions go with an on-premises model
to mitigate the risk to their business from poor connectivity. They also
want to maintain a stable user experience for their staff.

Customize and extend

cloud applications
A SaaS application with rich features also provides powerful capa-
bilities to customize and extend the app to meet specific business

Implementation Guide: Success by Design: Implement cloud solutions 45

 Adopt a cloud mindset requirements. Customizations to the SaaS application can be at several
levels. Many involve little code (embracing the low-code/no-code
Cloud implementation philosophy) and offer the flexibility to tap into the underlying PaaS
layer. This allows you to take advantage of professional software devel-
Customize and extend
opment techniques that include custom code.
cloud applications
Operate in the cloud Chapter 15, “Extend your solution” covers the details of customization
and app extensions, considerations, and scenarios. In this chapter, we
Evergreen cloud focus on the principles to keep in mind when making decisions about
customizing the application or extending platform capabilities. While
Upgrade from on-premises
to the cloud these principles broadly apply, in some cases a tradeoff is necessary,
in which the approach depends on specific situations, in-house skills,
timelines, budget, user expectations, processes, and other factors.

Know the platform

When using a SaaS cloud platform like Dynamics 365, it’s important
to know that this isn’t a one-time software purchase. It’s an ongoing
partnership between the customer and service provider, with an
operating contract that governs the service-level agreement (SLA),
security, ongoing updates, and more. You need to understand how the
service runs and how updates and changes impact your solution (and
therefore business), and employ the supported way of extending and
customizing the application in harmony with the service contract. With
this knowledge and compliance, you can ensure continued support
and avoid unexpected issues.

Even when using supported extension techniques, you need to adhere

to best practices. A good example is that while the platform might
allow you to run synchronous code for up to two minutes when creat-
ing a record, running to the time limit would block a user’s UI thread
for two minutes when they save a record. Is that acceptable? Similarly,
you could use custom code to limit access to records and implement
a custom security requirement, but certain access mechanisms could
bypass your custom code, leading to exposure. The point is to carefully
design your customizations and assess their impact, particularly when
those customizations affect the end user’s experience or deviate from
the security control constructs provided by the platform.

Implementation Guide: Success by Design: Implement cloud solutions 46

 In general, the low-code/no-code techniques that rely on configu-
ration rather than custom code are optimized by the platform, and
should be a preferred approach. However, custom coding may be re-
quired in some use cases. We have a no-cliffs philosophy with limitless
possibilities to extend the application using the underlying platform.
The ability to address this by using the underlying PaaS is a superpower
in the world of SaaS applications.

Be aware that custom coding and PaaS extensions bring additional

maintenance responsibilities and are best left to professional devel-
opment. Even when using PaaS to extend your solution, serverless
technologies are usually better suited to this approach.

Other factors include solution capacity planning, which should con-

sider the impact of your customizations, and the implications of using
certain PaaS technologies that use a pay-as-you-go model. We recom-
mend your organization establish guidelines for customizations that
Using a SaaS cloud take into consideration the platform, user experience, security, and
platform is an approved patterns for integrations. In addition, enforce best practices
ongoing partnership with automated checks wherever possible in the build pipeline. Power
Platform includes an app checker and app monitor that can help
between customer identify unsupported techniques and anti-patterns that can lead to
and service provider. performance and scalability issues.

Don’t mimic your existing system

A common misstep is trying to mimic a legacy system or the features of a
different SaaS platform when implementing a solution. This mindset can
lead to excessive customization and prevent you from realizing the true
potential of the platform and all its strengths to build your solution. To focus
on business value, you should adopt a business process–oriented approach.
It’s important that business requirements aren’t muddled with implementa-
tion details or dictate a specific approach rather than a business outcome.
A simpler approach is to look at the business processes for transfor-
mation and implement those using the platform capabilities instead of
replicating the existing solution. It’s closer to business, future-proof, less
costly, and easier to maintain than simply lifting and shifting your existing
experiences to the cloud. You also avoid technical debt from excessive

Implementation Guide: Success by Design: Implement cloud solutions 47

 customizations. Aligning to the platform allows you to maximize its capa-
bilities so you get an evergreen, always up-to-date solution.

Future-proofing
One of the key principles we have established in this chapter and through-
out this book is getting comfortable with change. SaaS enables you to
adopt new features to maintain a competitive edge. These features and
capabilities are built on top of an existing baseline of features and tables.
Although repurposing or using a custom table may not be unsupported,
deviations can severely impact your ability to take advantage of future
capabilities or can affect the interoperability of your solution.

In addition to the business process–oriented approach to implemen-

tation, we also recommend that you review your existing business
processes and compare that with what the standard application has to
To focus on business offer. You can benefit from adopting standardized processes backed
value, adopt a by research and feedback from leading customers in their industries,

business process– because you benefit from all that expertise. Using standard processes
also makes your business less dependent on expensive and hard-to-
oriented approach. maintain customizations.

The Common Data Model (CDM) is an open standard schema for

business data created in collaboration with SAP and Adobe, and is
used across Dynamics 365 applications. Aligning to the CDM can help
future-proof your solution. It can also make sure you can take advan-
tage of enhancements as well as other business applications across the
industry that honor the schema and can easily interpret your data.

Another way to future-proof is to understand your roadmap, in-

vestment direction, and trends so you can better align to the future
direction of the platform. This helps reduce technical debt and rework.

Understand the tradeoff

Expect tradeoff decisions to be made when implementing your
solution, particularly when it comes to the user experience (UX). In
Dynamics 365, the platform provides several out-of-the-box controls

Implementation Guide: Success by Design: Implement cloud solutions 48

and UX patterns that can be reused in a model-driven paradigm to
quickly assemble applications. These address common usage patterns
and also meet accessibility standards. It’s common for a business to
make specific requests around application navigation, behavior of a
specific control, or even colors to highlight specific controls. From a
technical perspective, most of these UX needs can be met either by using
custom code to build custom controls or by using a different app para-
digm, in which you start with a blank canvas to develop a pixel-perfect
UI that you could use in combination with model-driven controls.

Several factors could guide these decisions, including user adoption,

but you should also assess the cost-benefit ratio of customization.
In some cases, building and maintaining custom controls can be
time-consuming and require specialized skills. If you can easily address
some scenarios with user training, you can save hundreds of hours of
development and test effort. In addition, specialized UX requests are
often subconsciously driven by the need to replicate an existing system
or just a familiar pattern, even though the requests defy modern UX
best practices and accessibility needs.

Independent software vendors

Independent software vendors (ISVs) are third-party organizations that
exclusively develop software or software solutions that are compatible
with the service to deliver additional capabilities.

An effective way to approach extending your solution is to use an ISV

solution from the app marketplace for your platform. This can save
time and effort in development and testing, and provide you with a
tried and tested solution used by peers in your industry.

Several industry-focused ISVs provide solutions to fill unique product

gaps and address specific business needs in industries such as fashion
and hospitality. Make sure to perform due diligence when looking for
an ISV. For example, an ISV needs to make sure their solution is avail-
able for the version you’re targeting for go live. Otherwise, deployment
timelines can be affected. It’s also important that you’re aware of the
support model for an ISV, their commitment to keep it up to date, and
the protection clause if the solution provider goes bankrupt or is sold.

49
 You will also want to know how the ISV handles deprecation notices for
phased out software and update cycles.

AppSource offers a variety of ISV solutions

to suit your needs.
Chapter 15, “Extend your solution,” provides a deeper dive into making
a solution work for you.

Adopt a cloud mindset

Operate in the cloud
Cloud implementation
Successfully operating in the cloud will drive business value for your orga-
Customize and extend nization. Having the right skills and the change management processes
cloud applications
aligned to your platform will ensure you are set up to operate smoothly
Operate in the cloud and derive continued value from service updates and enhancements.

Evergreen cloud Center of Excellence

You might have a partner and specialized development team assist
Upgrade from on-premises
with the initial deployment, but once in “business as usual” operations
to the cloud
mode, does your team have the skillset to manage the platform, main-
tain the custom components, and assess and implement appropriate
new features? Not having the correct level of expertise and guidance
onboard after you go live can inhibit solution advancements and
undermine your ability to take full advantage of the platform. Creating
a Center of Excellence (CoE) within the organization and investing in
developing the skills to operate on the platform is extremely valuable.
Organizations that embrace organic application development by end
users also need a well-established CoE. This center can administer,
nurture, govern, and guide users to adopt the citizen development
approach to business applications, empowering everyone to create
innovative apps with high standards by using a consistent application
lifecycle management (ALM) process.

Microsoft provides a CoE starter toolkit that includes prebuilt processes:

▪ Administer  Gain insights into your Power Platform adoption
▪ Govern  Establish audit, compliance, and archive processes
▪ Nurture  Accelerate your adoption with a thriving community
▪ Drive innovation  Pick the most impactful and valuable scenarios
for development
▪ Consistent ALM  Apply source control strategies using GitHub

Implementation Guide: Success by Design: Implement cloud solutions 50

 ▪ Standards and theming  Create, manage, and share enter-
prise-branded themes as well as maintain and follow organizational
best practices and standards
You can read more about CoE at
Establishing a Microsoft Power
Platform Center of Excellence. Usage and monitoring
Running your business on the cloud means you want the cloud to be
always available, responsive, and scalable. You also want to be alerted
to unexpected issues in the applications and infrastructure so you can
diagnose and fix them. Additionally, you want to comply with SLAs you
have in place from other businesses. Having a monitoring capability
can prepare your business to meet these objectives.

What to monitor?
Different personas in your company will want to monitor for different
aspects of the system. End users may be more concerned with respon-
siveness, admins may be looking at changes, and the business may be
looking at KPIs like the time taken to place an order.

When running a SaaS application, you want to focus on the business,

while a service vendor assumes the role of maintaining the solution.
As a customer, you still have a level of control over the solution. This is
especially important when you have customizations in place to fine-
tune the business processes for your organization. If you introduce a
new feature in production and it causes unexpected issues, you want to
detect it proactively and resolve it before it becomes a problem. Using
For example, you can configure
Dynamics 365 and Azure features and flags to control new features can serve you well in these
Application Insights to access
situations because you can quickly turn off a faulty feature.
telemetry and diagnostics data
and set up alerts and dashboards
to proactively monitor your
Dynamics 365 environments.
A proactive and reactive monitoring strategy that empowers your staff
to detect and resolve issues should be part of every customer’s arsenal.

Best practices for monitoring

Keep in mind the following best practices:
▪ Decide what exactly to monitor, especially the KPIs, and whether
the service is meeting them.
▪ Monitor your consumption, subscription, and services fees. The
more cloud services you use, the more they will cost.
▪ Use automation to automatically trigger actions and to alert

Implementation Guide: Success by Design: Implement cloud solutions 51

 people when thresholds are reached.
▪ Track user activities for accountability (including the applications, screens,
and features they use), the responsiveness, and the frequency of use.

Adopt and align to a product roadmap

Your cloud solution will continuously evolve and deliver on the prom-
ise of digital transformation by delivering new capabilities that help
you stay competitive in your industry. Cloud vendors do their own
research and listen to customer feedback to improve the product by
adding new features. Being actively engaged in this process by giving
feedback, staying abreast of the latest roadmap, and planning for
upcoming changes is an effective way to operate in cloud. Also, the
key stakeholders from the business need to remain engaged and stay
plugged in to new features development, investments, and product
direction. We recommend developing your own roadmap for business
applications and aligning this to the product direction to get the most
out of your investments.

Dynamics 365 has two release waves per year in which several in-
cremental enhancements and new capabilities are made available
to customers. Adopting these mandatory changes and new fea-
tures—many of which are included in existing license plans—are a
fundamental aspect of operating in cloud.

Stay engaged
Implementing the system is one thing, but getting continuous return
on investment (ROI) from your cloud solution requires engagement.
Companies who stay passionate and curious about the solution and
adopt the latest features are the ones who enjoy most success. We
urge all customers to maintain engagement through conferences and
events throughout the year, and stay connected by using community
You can choose from application-specific groups. Several official and unofficial communities have been formed
forums like the Microsoft Dynamics
Community Help Wiki, Yammer groups,
by customers, partners, and Microsoft engineering. Take every op-
blogs, events, and how-to videos where portunity to influence the product through active engagement with
you can discuss ideas, learn features, learn
roadmaps, and ask questions. product leaders, advisory boards, table talk with product managers,
preview programs, and collaborative development opportunities.

Implementation Guide: Success by Design: Implement cloud solutions 52

 Evergreen cloud
Adopt a cloud mindset

Cloud implementation
One of the benefits of a modern SaaS solution is that every custom-
Customize and extend
cloud applications
er runs on the latest version of the service. It’s the only model that
scales. With this approach, every security patch, bug fix, performance
Operate in the cloud enhancement, and functional improvement accrues to all implemen-
tations across the globe. Microsoft assumes responsibility for keeping
Evergreen cloud the platform current. This means that you no longer have to pull large
teams of people together every few years to perform a traditional
Upgrade from on-premises
to the cloud upgrade over many weeks or months with limited added value for the
business. The evergreen approach and the model of continuous up-
dates gives your business access to the latest capabilities to stay head
of the competition and meet changing customer expectations.

Dynamics 365 has embraced this approach of operating on one

version in a way that delivers benefits without disruption. The improve-
ments and enhancements embrace the CI/CD model, with service
updates delivered on a weekly basis, but these changes aren’t targeted
at the end user experience. You can take advantage of release waves to
preview upcoming changes and test and prepare your users before the
changes are deployed.

Achieving the right balance of delivering improvements and business

capabilities in a safe manner without disrupting the user experience is
paramount to our evergreen cloud approach.

Chapter 20, “Service the solution,” delves further into the updates
wave approach.

Upgrade from
on-premises to the cloud
Organizations over the years might have invested in on-premises
deployment of business applications. These applications could serve
the users and business well in their current state, but keeping them

Implementation Guide: Success by Design: Implement cloud solutions 53

 Adopt a cloud mindset on-premises over time could limit your opportunity for expansion. In
these cases, you should explore upgrading or migrating an existing
Cloud implementation on-premises business application to the cloud and bringing along exist-
ing business processes and data to speed up your digital transformation.
Customize and extend
cloud applications
Dynamics 365 provides the tooling and support to help on-premises
Operate in the cloud deployment upgrades or migrations to the cloud (subject to compat-
ibility and security checks). This option to upgrade to the cloud also
Evergreen cloud helps organizations take advantage of their existing investments in

Upgrade from implementation, as long as the supported implementation techniques

on-premises to are used and compatible with the cloud. Still, everything we discuss
the cloud in this chapter on the cloud mindset, considerations, extension, and
operations is fully applicable.

When upgrading or migrating to the cloud, consider the following:

▪ The most critical factor in deciding on the upgrade-to-cloud ap-
proach is to assess if the current application being upgraded meets
the requirement of the business and is well adopted by users.
▪ The data model, design, and data quality should ensure that the
application being upgraded becomes a stepping stone to acceler-
ate digital transformation. The upgraded application should serve
as a foundation for further growth and adoption of new capabili-
ties in the cloud. It shouldn’t stifle cloud adoption by carrying over
poor design and data.
▪ The cloud platform should be approved by security to store and
process data in accordance with the organization’s security policies
and regulations. The patterns used for authentication and integra-
tion also need to be revisited to make sure they’re cloud ready.
▪ Understanding the data category, flow, transformations, location,
and encryption during and post migration will help you navigate
The Dynamics 365 Migration Program organizational security policies and plan your cutover.
can help you take the first step toward
▪ The upgrade to the cloud can be a multi-staged process that involves
cloud success by migrating your on-prem-
ises solution with expert guidance from stringent compatibility and validation checks, so it’s critical to under-
Microsoft. For more information, refer to
the overview of our program or perform a
stand the process and review the prerequisites early in the process.
standard migration assessment. ▪ Some components in the on-premises application may need to
You can also move from Dynamics AX to be remediated or replaced to ensure cloud readiness. Examples
Dynamics 365 in the cloud.
to look at include deprecation, unsupported customizations, and
non–cloud ready patterns that pose a security or performance risk.

Implementation Guide: Success by Design: Implement cloud solutions 54

 ▪ Existing integration and peripheral applications will require
changes to ensure they work with the cloud applications.
▪ Any impact on process, performance, and usability due to latency,
service protection limits, or license capacity should be considered.

The ability to bring existing Dynamics 365 on-premises applications

into the cloud with proper due diligence can be a game-changer for
your digital transformation.

Conclusion
Embracing SaaS applications to run your business can significantly
accelerate your digital transformation, but it’s also important to rec-
ognize that organizational cloud maturity will play a significant role in
your strategy’s long-term success.

Organizations working on their first major implementation in the SaaS

cloud should expect some level of disruption to the existing ways of
deploying and managing applications. We hope this chapter serves as a
primer for Success by Design thinking by helping set the context for the
changes you can expect and driving you towards greater cloud maturity.

Implementation Guide: Success by Design: Implement cloud solutions 55

 Checklist

Adopt a cloud mindset Ensure that when upgrading existing on-premises solu-
tions to cloud, the data model, design, and data quality
Have a shared understanding at every level, from the make certain that the application becomes a stepping
executive sponsor to the developers, of the business stone and doesn’t stifle cloud adoption by carrying over
impact being delivered. poor design and data.

Ensure the organization has done its due diligence

organizing the data estate and understands the impact Customize and extend
of the new solution on the data estate.
Create guidelines for when to customize and extend
Gain approval for the cloud platform for use with the out-of-the-box apps and only adopt documented
appropriate application data category from information extension techniques.
security and compliance.
Avoid deviations and repurposing the out-of-the-box
Ensure the respective teams understand all administra- tables and data models (Common Data Model) because
tive, operational, support, and monitoring aspects of the this inhibits future adoption of new features and
platform, and the organization policies, processes, and capabilities.
patterns conform to the SaaS cloud.

Implement DevOps and CI/CD pipelines to support Operation

automation for build, testing, and deployment.
Ensure that the necessary expertise is available via the
Design the solution according to service boundaries and partner (system integrator) or internal teams post-de-
available licensing capacity, and ensure the controls to ployment to support and evolve the solution.
further expand and scale the solution are understood.
Engage with the community to keep up with the latest
Design the solution to meet the nonfunctional re- innovation and help influence the product direction
quirements, considering factors like network latency, using channels, events, blogs, and other mediums.
end-user environment, and devices.

Implementation Guide: Success by Design: Implement cloud solutions 56

 Case study
Rev up for speedier service
A family-owned business based in South Australia is one of the largest
retailers in the region, with ventures that span fuel, convenience, quick-
serve restaurants, and real estate.

One of its key businesses is a chain of retail stores that serves busy
guests across Australia.

Their focus on customer service has led the company to reimagine

the way its store systems are designed and connected. As their retail
chain has grown, they have faced common challenges: outdated
technologies, changes in customer behavior, and the pace of change
with existing suppliers. The company was ready for an overhaul and an
infrastructure solution that could serve its goals now and in the future.

Given the high volume of traffic that the retail chain processes and the
diversity of its retail store locations, the company was seeking a hybrid
solution with reliable, flexible connectivity.

Scalability and speed are critical for their retail business, and they
needed an infrastructure design that optimizes for both.

“We serve millions of guests each month,” says the company’s store
systems specialist. “To achieve speed of service, we know that we need
to have something on-premises to retain our workloads in the store.
And, at the same time, we need to have that cloud connectivity to the
back office.”

Implementation Guide: Success by Design: Implement cloud solutions 57

 A cloud engineer at the company adds, “We have stores in rural areas
where they don’t have high internet speed connections. We wanted
to have something that is both in-store and in the cloud that would
synchronize when it comes to failovers and redundancy, so that we can
have the best of both worlds.”

When it came to renewing store hardware and to keep up with the

pace of innovation, the company chose to go with a hybrid model that
integrates Azure, Azure Stack Edge, and Dynamics 365 Commerce.

The new design impacts the stores’ day-to-day connectivity. “What

we are planning to do is a direct connection to Dynamics 365, and the
failover needs to happen in Azure Stack Edge,” one engineer said. “We
expect to have a minimum of 99 percent of uptime.”

One of the team’s other top considerations was how easy the infra-
structure would be to manage, so they factored in what the company
was already using across its retail business.

“We have already heavily invested in Microsoft products. We use

Dynamics 365, Microsoft 365, Power BI, and multiple other products,”
explains one engineer. “We wanted to have a single pane of glass
where we could administer and monitor all of the systems.”

The hybrid deployment model using Azure Stack Edge and Dynamics
365 Modern Point of Sale provides store operations redundancy in case
of network outage.

The new infrastructure design provides the connectivity, consistency,

and efficiency that the company needs. Over time, the company ex-
pects that it will help them avoid store downtime, maintain real-time
stock levels, simplify its infrastructure management processes, and
reduce maintenance and compute costs.

The new infrastructure design will impact the customer experience at

the store level. “We offer a comprehensive omnichannel experience to
guests walking into our stores,” says the store systems specialist. “We
enable our guests to pay at the pump without having to come into the

Implementation Guide: Success by Design: Implement cloud solutions 58

 store by using the [retail chain] app. We also enable our guests to click
and collect their food, coffee, or supermarket needs. To enable that, we
need real-time connectivity to our head office systems—that is key for
us. With Azure Stack Edge and an in-store database, we are hoping to
get that real-time connectivity.”

This orchestration ensures connectivity and consistency, as a cloud

engineer explains. “We deployed three virtual machines. One is used
to monitor the in-store cameras, and the other is used to store a
Dynamics 365 database backup. If the store loses its internet connec-
tion to Dynamics 365, it would be able to operate normally while that
virtual machine continues with the operations. Then, once the connec-
tion is restored, it will be able to synchronize back to Dynamics 365.”

As the company continues to roll out its new infrastructure, it’s piloting
a solution to recognize vehicles’ number plates based on real-time
CCTV feeds. One benefit of this is loss prevention, in situations where a
number plate is tied to a known offender in a published database.

They also plan to use the in-store cameras to do machine learning and
AI on Azure Stack Edge virtual machines to predict stock levels and
alert the staff, all locally, without the cloud.

Implementation Guide: Success by Design: Implement cloud solutions 59

4 Guide
Drive app
value
Implementation Guide: Success by Design: Drive app value 60
 Introduction
In the constantly evolving world of Dynamics 365,
software as a service (SaaS) cloud applications, and
changing business needs, it’s important that your
business applications evolve and are consistently
delivering value.
This chapter outlines the approach to digital transformation and dis-
cusses strategies on how to shorten the time to value, cut through the
noise, and focus on business outcomes.

The digital transformation journey of an organization usually involves

various phases and levels of maturity. A green field implementation could
start with a pilot or even a minimum viable product (MVP) targeted
Key topics in this chapter: at a specific business process, then incrementally expand the solution
with more capabilities for the users and onboard new business pro-
Think like a startup: start lean, build a
successful MVP, evolve quickly cesses. Alternatively, replacing an existing system could mean that a
significant amount of functionality needs to be made available at the start
Phases and incremental strategy
in order to have parity. No matter where you start, realizing the long-term
Manage the change stream: business,
user, and product
goals of a SaaS-based digital transformation requires effective planning.

Achieve hyper-scale with expansion

approaches and enable organic growth In this chapter, we explore how to create a digital transformation road-
map for your organization that can help accelerate the time to value
Better together: satellite applications
and aggregations while maintaining relevance in a constantly changing technology and
business landscape. This guide can help you navigate the challenges

Implementation Guide: Success by Design: Drive app value 61

 of expanding your solution and drive continuous, long-term business
value through digital transformation.

Approach to digital
transformation
Throughout this book, we discuss several concepts related to Success
by Design, including how to initiate, implement, and deploy a project,
but the scope of Success by Design isn’t limited to a successful go live
or operational excellence. The long-term goal of Success by Design is to
create the right foundation for an organization to evolve their business
application landscape and expand their digital transformation footprint.

Success by Design Every implementation related to business applications doesn’t nec-

creates the essarily qualify as a digital transformation program aligned to the
foundation for business; several systems are migrated to the cloud in their current
form as a part of infrastructure modernization or to ensure sup-
an organization portability. In some cases, this is just the first step towards enabling
to evolve their digital transformation; opportunities could arise to take it further and
business application deliver transformational value to the business rather than looking at it

landscape. as just a tactical upgrade to projects run by IT with little involvement

from the business.

The discussion in this chapter targets business-focused digital trans-

formation programs and their expansion. Before we go into the details
of expanding a solution to further drive value, let’s discuss the funda-
mental approach to transformation: depicting your current business
model, the factors that affect your business model, how they can
trigger transformation, and defining your goals and scope of business
processes for a transformation roadmap.

Represent the business model

Before you define a digital transformation roadmap, a good strategic
exercise is to holistically develop an understanding of your business
model. You can do this at an organization, department, business unit,

Implementation Guide: Success by Design: Drive app value 62

 or team level to understand the value that you currently deliver, who
your customer is, and how to deliver that value and the key activities
and resources involved.

You could use several different techniques or frameworks to describe

an organization’s business model and how they create value. One
framework that we discuss in more detail is the Business Model Canvas,
developed by Alexander Osterwalder. It offers a simple, holistic, and
easy-to-understand representation of a business model. This model
isn’t officially endorsed by Microsoft or part of Success by Design, but
you can use the model to not only map your existing business model
but also highlight what aspects of your business model are changing—
which your digital transformation strategy will help you adapt to.

The business model represents the “why” and what it takes to deliver
the value to customer; how we go about doing it is the business
process definition.
Disruptions in the
business model
Triggers
create opportunities
Disruptions and inflections in the business model create big opportunities
for transformation. for transformation. In a world that’s changing at a pace greater than ever,
businesses have to reinvent themselves more often, improve customer
experiences, and attract and retain talent. The opportunities for impact
are plentiful (for a deeper discussion, see Chapter 1, “Introduction to
Implementation Guides”).

Let’s explore some examples of how changes in the business model,

disruption in an industry, new regulations, or even external factors like
global events can be a catalyst that triggers changes to business pro-
cesses and applications.
▪ Changes in customer segment  As businesses evolve their
products and expand their target customer base and demograph-
ics, they often see that their approach to nurturing their potential
customer base also needs to evolve. Targeted messaging that
resonates and is delivered through relevant channels becomes
extremely critical. Does your marketing automation and customer
data platform enable this?

Implementation Guide: Success by Design: Drive app value 63

 ▪ New channels  Customer channels that the business uses to
deliver their product and services are evolving. For example,
customers expect remote delivery of services via digital channels
instead of an in-person visit to a store, which is further necessi-
tated by evolving circumstances such as the COVID-19 pandemic.
What does this change in channels mean for your order manage-
ment systems or other related business processes?
▪ Changes in value proposition  Businesses might differentiate
themselves by coming up with unique customer value proposi-
tions in highly competitive sectors, such as offering a lucrative
subscription model, to drive recurring revenue and shareholder
value. For example, customers aged 25–40 are more open to
subscription-based purchases versus outright ownership. The
automobile industry now offers cars on a subscription basis and
charges based on miles driven, as opposed to customers paying
full purchase price. How are your processes and business applica-
tions affected by this change in business model?
▪ Changes in customer relationships  An organization’s ap-
proach to engaging with customers evolves with the business.
For example, an organization could change from an aggressive
customer acquisition phase to retaining and upselling within their
existing customer base after acquiring a fair market share. How is
your customer relationship management (CRM) strategy chang-
ing, what does this mean for your CRM system?
▪ Changes in revenue streams  As businesses evolve, they create
new revenue streams. For example, an energy provider in a highly
competitive market might provide home services to expand their
revenue stream, which triggers changes to customer processes
and related business applications. Do you know how your revenue
stream changes impact your business applications?.
▪ Resources and sustainability  More organizations are commit-
ting to environmental sustainability. How do your manufacturing
and operations track your carbon footprint, impact on emissions,
and procurement from sustainable sources?
▪ Custom manufacturing  With the advances in 3D printing tech-
nology, a manufacturing company can offer custom, made-to-order
options to their customers. How does this impact your operations
processes and enterprise resource planning (ERP) application?

Implementation Guide: Success by Design: Drive app value 64

 ▪ Partnerships and acquisitions  Inorganic, acquisition-based
growth is common in several industries, and so is the strategic
partnership between organizations to grow market share where
they have complementary services. How do these changes affect
process alignment and your supporting business applications?
▪ Cost and efficiency  New technologies offer efficiency advan-
tages that can reduce operational costs. For example, delivery
companies use electronic signatures instead of paper-based
solutions to get real-time notifications and provide new value
to customers. Are there opportunities to drive process efficiency
using the latest technology?
▪ External factors  Changes in the socioeconomic situation, po-
litical climate, and regulations can affect the way organizations
conduct business. For example, the EU’s decision to bring in the
General Data Protection Regulation (GDPR) affected every busi-
ness with employees and customers in EU, which required changes
to their business applications to ensure compliance. How are new
Changes to the and changing regulations impacting your business?
business model can
kickstart digital These triggers can all disrupt your existing business model. You can use
the Business Model Canvas to highlight what aspects of your business
transformation. model are changing and align your digital transformation strategy to
these changes, which will eventually help your business adapt and grow.

Digital transformation goals

Changes to the business model (caused by a variety of triggers) can
kickstart digital transformation. These transformation goals must be
defined in business terms, with measurable metrics that could poten-
tially involve several milestones. Let’s look at an example in which an
organization’s goal is to create a 360-degree view of their customer.
They want a holistic view of their customer, their activity history, prod-
uct purchases, subscriptions, and more. But this goal doesn’t relate to a
business outcome or a measurable key performance indicator (KPI). A
measurable goal for your digital transformation could be the business’s
ability to increase upsell to existing customers by X percent or reduce
the time to address customer queries by Y percent. This requires a
360-degree view of the customer.

Implementation Guide: Success by Design: Drive app value 65

 Although most programs begin with a set of business goals, it’s equally
important to communicate and reiterate these goals to the IT and
technical teams implementing solutions. In long-running transfor-
You can use objectives and key results
mation programs involving several technologies, teams and partners
(OKR) as a goal-setting framework for
digital transformation. might lose sight of the actual business goals by letting technology
become a distraction.

Business process
The next stage of the process is discovery, which involves key business
stakeholders. With a holistic view of the evolving business model and
transformation goals in place, you can identify the related business
processes and applications that need to change. The discovery exer-
cise should be focused on creating clarity around how the business
processes and applications identified need to evolve to meet the
corresponding transformation goals.

The scope of change could be automating a manual activity to

improve productivity, capture data accurately to improve the effec-
tiveness of strategy, and help your users do their job better and more
efficiently. This will enable the business to generate insights about your
customers by eliminating data siloes.

Prioritizing the changes to maximize and deliver measurable business

impacts continuously (without having to wait for a multi-year trans-
formation program to complete) keeps digital transformation in the
foreground, with engagement from the business, end users, and exec-
utive sponsors maintaining its relevance. Business agility should be the
key focus for long-running, comprehensive transformation, or you risk
missed business opportunities and loss of market share.

For more information, refer to Chapter 7, “Process-focused solution.”

Change streams
So far, this chapter has discussed the approach to digital transforma-
tion and how to develop a transformation plan for your application

Implementation Guide: Success by Design: Drive app value 66

 and the process involved. This has traditionally been a top-down
approach, starting at the business model and then narrowing down
to business processes and applications that deliver the most impact.
However, you should always keep in line with the broader theme of
constant change and embrace the mindset of getting comfortable with
frequent and inevitable change. In this section, we explore the change
streams that impact the roadmap and scope of planned activities.
Planning for how to deal with these changes throughout the program
is a key determiner of success.

Business
As we embark on the journey of transformation and start building the
business application based on the goals set by the business, it’s also
important to appreciate that during design and implementation, you
may need to accommodate for further adjustments and refinements.
Embrace the This support for agility and change needs to be fundamentally built

mindset of getting into the program. This is where some of the iterative methodologies
could be beneficial, enabling us to respond to the change without it
comfortable with turning into a disruption.
frequent and
inevitable change. Those leading or adopting transformational change often find that
it’s not well defined in the early period, so adopting changes quickly
and early is key—but so is flexibility as clarity reveals the resulting
business model. When transformation occurs in an industry, what the
industry will look like following the transformation is often not known.
Companies must be able to adapt quickly in the middle of the project
to incorporate these inevitable changes.

User
A key stakeholder in your transformation plan is the business user, who
interacts with the application daily. If the process being implemented
in the application doesn’t meet the needs of the user, or the applica-
tion doesn’t consider working patterns and end user usage, it could
lead to poor adoption and lack of engagement. Incorporating user
feedback continuously throughout the process using well-defined and
frequent touchpoints is key to achieving your transformation goals.

Implementation Guide: Success by Design: Drive app value 67

Your approach to designing and developing the application needs to
be user-centric; it must clearly articulate how the change being imple-
mented results in value for the user.

You should expect that user experience expectations constantly evolve

as users are exposed to a variety of software products both in the enter-
prise and consumer space. For example, the predictive text fill feature
popularized by web search engines is now the expected search expe-
rience in business applications. This makes it even more important to
adopt newer features and improvements delivered by SaaS applications,
which takes us to the next change stream: product and technology.

Product and technology

In the world of cloud-based SaaS applications, you can expect a con-
stant flow of new capabilities, improvements to existing features, and
transitions to the latest technology paradigms, which can impact your
current application. Traditionally, business applications were built by IT,
with features requested by the business and end users. With the SaaS
cloud, the service providers themselves are invested in understanding
the latest trends in business and delivering innovative capabilities to
help their customers stay ahead of the competition. For example, a
business trying to transform their sales application can assemble an
application using the best-in-class sales capabilities offered by the
SaaS provider instead of spending years in research, design, and de-
velopment to build from scratch, which would make the application
obsolete by the time it launches.

Continued investment in microservices by Dynamics 365, such as

planning as a service, can transform a core business model for companies
allowing them to run material requirements planning (MRP) as needed
to drive a significant and transformative way of managing demand and
the supply chain.

The enhancement delivered via continuous updates has a much

shorter adoption timeframe when compared to traditional major
version upgrades.

68
 SaaS application providers that are competing to build and deliver
business capabilities for customers are helping advance business appli-
cations in way that was unimaginable just a few years ago. Applications
that were just forms over data (mostly passive data capture systems used
to track, view, and report on known data) have evolved into applications
that can automatically capture data, learn from the data, deliver insights,
and guide users to the next best action. This makes it extremely import-
ant for business to watch for new capabilities being made available and
adopt them to accelerate their transformation to be a differentiator in
the industry. The key value proposition of SaaS is also the ability to tap
into the enhancements and features that are based on broad market
research. Activating these features can accelerate your transformation
with minimal effort and without any development cost.

Your SaaS application provider is no longer just a software vendor

providing an off-the-shelf product—they’re a strategic partner to your
business with significant impact on your ability to implement your
strategy. Developing a strong relationship with the SaaS providers
to not just learn about the roadmap but also influence the roadmap
through strategic feedback ensures that you’re getting the most value
out of the product.

External
External economic, social, and political drivers can disrupt your
transformation plan. The COVID-19 pandemic is an example of how
supporting remote work and creating online collaboration channels
became a top priority for most customers in 2020. This required co-
ordinated changes from infrastructure, to network, to the device and
application layer across the IT level of organizations. Although it’s dif-
ficult to foresee and plan for external events, the iterative approach to
delivering continuous value in smaller batches allows you to pause and
pivot as needed. For example, consider an organization on a long-term
transformation program that releases a major capability to users once
a year versus an organization that has adopted the DevOps culture
of delivering value with regular bi-monthly or monthly updates. The
latter company can realize value from investments and is better posi-
tioned to pivot when change demands it.

Implementation Guide: Success by Design: Drive app value 69

 Transformation map
So far, we have discussed a high-level approach to digital transforma-
tion and the various change streams that you need to consider. The
ever-growing product backlog of feature asks, change requests, and
product capabilities can overwhelm and potentially distract from the
drivers for transformation. To determine time to value, you need a shared
understanding between the business, IT, and users on how a product
feature or change aligns with the goals of transformation and impacts
the maturity of process implementation.

One way to visualize and communicate your digital transformation

goals is to plot the application’s capabilities in a transformation map
(Figure 4-1). You can adjust the definition of each quadrant to your
own goals and ambitions for the process, but you can easily get started
with the template we offer here. If all the feature asks, changes, and
SaaS product capabilities are plotted on this quadrant, you can deter-
mine the current state of your business process implementation and
agree on the future state aligned to your transformation goals.

For example, a sales system that only has the basic features of contact
Fig.
4-1 management and opportunity management may be considered
essential, whereas a system that can use AI-driven
insights with advanced forecasting might be a
Enhanced Differentiator

Predictive
differentiator that gives you competitive advantage
forecasting
Forecasting in your industry.
Sequences

Premium autocapture
Creating a dashboard for stakeholders that organiz-
Assistance
cards and studio
es the requirements and features in this fashion can
Essential Innovative bring great clarity—not just during initial implemen-
Contact
Conversion
intelligence tation, but for future increments. It’s also important
management
to prepare for time-based decay, in which a feature
Standard cards
Opportunity
that is considered a differentiator today might be
Advanced

pipeline
management
Quotes
Autocapture
considered essential in a few years. Similarly, some-
thing that is innovative or even considered ahead
Innovative of its time could become a differentiator in the near
future. You should plan to revisit and refresh this

Implementation Guide: Success by Design: Drive app value 70

 transformation map for each phase, product release, pivot, or strategic
inflection in your business.

Phases and increments

Planning and scoping distinct phases of your digital transformation pro-
gram directly impacts the time to value realization. What is delivered and
when it’s delivered also has an impact on the level of engagement from
the stakeholder and the end user’s adoption. These phases and increments
are based on the goals agreed upon with the business and allow you to
select the appropriate capabilities in the transformation map.

It seems reasonable to start with the essentials and basics in Phase 1 of

a solution and then deliver enhancements with most the innovative,
differentiating features in later phases of the program. But if the most
valuable parts of the transformation are planned for much later in the
lifecycle (Figure 4-2), you risk stakeholders losing interest, or user
perception changing (which is hard to correct). Additionally, the fea-
tures that were considered differentiators could now be essential due
to time decay. A transformation program that fails to deliver beyond
the essentials or the initial MVP drives little transformation.
Fig.
4-2

Without the core features, you can’t deliver en-

hanced capabilities. You should plan the phases so
Enhanced Differentiator

Predictive
you can pick up elements from various quadrants
Phase 2 Phase 4 forecasting
and deliver a holistic solution that has the key
Forecasting Sequences
process elements, integrations, and reporting that
Premium autocapture

Assistance cards delivers value (Figure 4-3). Look for quick wins: the
and studio
latest SaaS features that you can easily adopt with
Essential Innovative little effort to bring innovative and differentiator
Phase 1 Phase 3
Conversion
intelligence elements into the earlier phases.
Contact
management
Standard cards
Opportunity Digital transformation is as much about people
Advanced

pipeline
management
Quotes
Autocapture
as the process and technology. The phases and
incremental design should also help product own-
Innovative ers maintain the right level of engagement from
users and stakeholders. This drives excitement and

Implementation Guide: Success by Design: Drive app value 71

 curiosity, but most importantly feedback. You could draw an analogy
from a TV series that shows teasers at the end of each episode or the end
of each series to attract viewers back to the program. Do the same for
your digital transformation story by delivering value beyond the essen-
tials in each phase so customers come back for more after each sprint.

In the next section, we look at how the right approach to MVP strategy
when getting started with your digital transformation journey can help
get early feedback and drive value sooner rather than later.

Minimal viable
product strategy
The concept of MVP, which enables you to quickly build, measure,
learn from feedback, and pivot if necessary, has been well established
in the startup world.

Bringing the invaluable lean startup mindset to digital transformation

can help organizations deal with uncertainty from different change
streams and find the value that drives maximum business impact in
Fig. the least amount of time. Even in the context of digital transformation,
4-3
MVP is often used to test the value of a solution to a business and
gather feedback.
Enhanced Differentiator

Forecasting Phase 2 Phase 4

Predictive
forecasting When translating the lean startup methodology to a
Premium autocapture business application, MVP should not be a solution
Assistance cards
and studio with just the basic and essential features; you should

Phase 1
Sequences create a solution that will help transform your
process and deliver on the most important transfor-
Essential Innovative Phase 3

Conversion
mation goals with the least amount of effort. Most
intelligence
Contact importantly, an MVP is not the end state; programs
management
Standard cards
may be stuck in an MPV state for years.
Opportunity
Advanced

pipeline
management
Autocapture
Going back to the example of a sales application,
Quotes
let’s consider if users are currently using a home-
Innovative grown sales system with a form to enter and view
opportunity data. The system has a backend to

Implementation Guide: Success by Design: Drive app value 72

 store this information, which is also used for management reporting.
The business has invested in a Dynamics 365 SaaS-based sales ap-
plication after assessing its rich features, out-of-the-box AI insights,
and integration capabilities with Power BI for reporting and analytics,
“MVP is not the which allows them to improve their opportunity closure rate and get
smallest product accurate growth forecasts. A common approach is to first implement

imaginable, though the out-of-the-box opportunity management process and reports,

and plan future phases to implement AI-based insights with Power BI.
it is the fastest way Although you can easily follow this route, it doesn’t achieve the actual
to get through the goal of sales implementation. The system delivers an improved user
build measure learn experience but fundamentally does the same thing as the home-grown

feedback loop with sales system, with little direct value to the business. Alternatively, using
embedded intelligence with opportunity scoring and relationship
minimum amount health can help users target the right opportunities and improve their
of effort.” close rate, which directly impacts business.
– Eric Ries, author of The Lean Startup

A good MVP has the following characteristics (Figure 4-4):

▪ Delivers on the most important of the digital transformation goals.
▪ Is production-ready and uses supported customization techniques
(this is different from a proof of concept or prototype designed to
prove feasibility or get specific usability feedback).
▪ Goes beyond the essentials and delivers on top of the existing
capability of the incumbent system.
▪ Focuses effort on delivering maximum business value in the short-
Fig.
est amount of time, (you could compromise on some experiences
4-4
that aren’t directly aligned to business outcomes).

Delivers on Production ready with Beyond the essentials Maximum value to business
goals for supported customization and delivers on in shortest time, minimal
transformation techniques (not proof of existing system compromises
concept)

Characteristics of a good MVP

Creates the foundation, Test hypothesis Test, learn, Built to pivot and
supports building blocks and deliver on activate change after feedback
for success expectations

Implementation Guide: Success by Design: Drive app value 73

 ▪ Is built in a way that allows for quick changes based on feedback,
making it possible to pivot. This translates to a greater reliance on
low-code configuration versus highly customized, professionally
developed, code-heavy experiences.
▪ Allows you to test and learn on the latest SaaS product features
that could be activated (this requires some level of change man-
agement but almost no development effort).
▪ Allows you to test the hypothesis of a specific feature or solution
delivering the expected business outcome and adjust based on
what you learn.
Fig.
4-5
▪ Creates the foundation for long-term success—it should support a
building block mindset in which you keep adding without disrupt-
Build
ing previous steps. (For more information about the architectural
Code
skills to realize this approach, see Chapter 6, “Solution architecture
Ideas Emotional product
Reliable
design pillars.”)
Accessible

Functional
Although the MVP approach naturally works with new greenfield im-
Learn Measure
plementations, you may have existing apps that are being replaced or
Data migrated to another platform. Your scope should consider the existing
functionality to ensure parity, but shouldn’t try to mimic or replicate
the experiences of the outgoing application. Focus your MVP strategy
on delivering the most value to the business sooner without limiting
Fig.
the scope to essentials (Figure 4-5). An MVP with a very broad scope
4-6
that takes years to deliver may cease to be an MVP.

Users
Feature New
adoption workloads
Drive expansion
Incremental Satellite
changes applications All the approaches and techniques we have dis-
cussed so far in this chapter can help you create an
Expansion effective digital transformation roadmap for your
Application Integrations business application. They’re relevant to the Initiate
standardization
phase, but you can apply these methods iteratively
or revisit them for further expansion. This section fo-
Mobile Aggregation cuses on expanding the usage and scope of business
Pivot applications in different areas (Figure 4-6) to drive
even greater business impact, while considering
additional impact on the following:

Implementation Guide: Success by Design: Drive app value 74

 ▪ Security and compliance
▪ Application lifecycle management (ALM)
▪ Administration and governance
▪ User readiness and change management
▪ Environment strategy
▪ Limits and capacity
▪ Data and integrations

Users
The most common expansion scenario for a business solution is add-
ing more users. This could be the result of onboarding new regions
or countries onto the same application. User-only expansion sounds
straightforward, but you could have data migrations on a live system,
security and compliance challenges based on the region, or other
impacts (Figure 4-7).

Mobile
Ensuring support on multiple devices, at times with offline capabilities,
could be critical to adoption. In some cases, this needs to be a part

Fig.
4-7 Users

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

High Low Low Medium-high Low Low Medium

Potentially high If the solution Based on the Full-fledged Creation of a Although each Usually, data
impact based remains the environment user change new production user usually migration can
on regional same, impact on strategy and management environment comes with be complex on
regulations ALM is minimal the creation and readiness impacts the their own a live system
of additional effort is required environment capacity and in case
production strategy integration API, of additional
environments, capacity, and environments,
you could see storage could integrations
medium impact be impacted might have to be
replicated across
environments

Implementation Guide: Success by Design: Drive app value 75

 of the initial release (such as in warehousing systems in which goods
must be scanned when picked up), whereas in others it could be a
quick follow-up. Most out-of-the-box capabilities work as is on mobile
apps, with additional optimizations to support the move experience.
Our example sales mobile application offers the core sales application
features with additional targeted mobile experiences for field sellers
(Figure 4-8).

Incremental changes
Incremental changes are primarily driven through the feedback and
change requests from users and the business. These changes help ensure
that the application meets the expectation of ongoing enhancement
and continues to build on top of the initial MVP to maintain relevance.
It’s still important to look at these improvement requests through the
business value and transformation lens (Figure 4-9).

Feature adoption
As we have seen when discussing change streams, SaaS products make
significant investments in researching and implementing business
capabilities that their customers can easily enable. This is a key value
proposition of the SaaS cloud, and can play a key role in deriving value
from your investments. With continuous updates in Dynamics 365,
Fig.
4-8 Mobile

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

High Low Low Medium-high Low Low Medium

Potentially high If the solution Additional User readiness Should not have It should not Custom
impact based on remains the device policies effort is any significant have any embedded
the regulations same, impact on could have an required to impact significant integrations
and need for ALM is minimal impact ensure adoption impact on user might require
mobile device on mobile and capacity work to make
management or account for any them responsive
mobile applica- variations for mobile
tion manage- consumption
ment features

Implementation Guide: Success by Design: Drive app value 76

 new features are made available to customers. A business analysis of
the new features with technical impact analysis on the existing solution
and customizations (Figure 4-10) is a key business as usual (BAU)
activity that must be conducted in collaboration with key stakeholders.
Most of these feature improvements are made available to customers
with no additional cost and can help advance your application to the
innovative or differentiator quadrant.

Fig.
4-9 Incremental changes

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

Low Low Low Medium Low Low Low

Changes within No major No expected User readiness Usually remains Additional Assuming no
the approved impacts if the impact as long effort is unchanged integration and integration
SaaS service solution doesn’t as the data required app workloads changes, impact
boundaries have new PaaS boundaries could have is minimal
and data loss components don’t change some impact
prevention
policies should
have limited
security and
compliance
implications

Fig.
4-10 Feature adoption

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

High Low Low Medium Low Low Medium

Usually No major No expected User readiness No impact Usually no Features might

no impact impact expected impact as long effort is required impact, generate
except when as the data to use the new sometimes additional data
connecting to boundaries features features might but shouldn’t
external data don’t change require additional impact
sources storage and integrations
capacity

Implementation Guide: Success by Design: Drive app value 77

 New workloads
Dynamics 365 provides several business applications that are built
on the same scalable power platform. As end users adopt your ap-
plication, you will find additional opportunities to onboard related
workloads that could help eliminate data siloes by enabling stronger
collaboration and data sharing. For example, to achieve seamless
collaboration when you’re onboarding a marketing workload in addi-
tion to sales, the sales team can get visibility into the nurture programs
their customers are part of, extend invite to marketing events, and
so on. Similarly, the accounts payable team could expand and move
into a centralized accounts payable processing system. However, new
workloads can have significant impact (Figure 4-11).

Satellite applications
When deployed, your business application covers core use cases and
scenarios, some of which may only be relevant to a specific team, region,
or role. In such scenarios, you can deliver such targeted capabilities via a
satellite application. Satellite apps are a good opportunity for user-
developed apps and automations while using the data model of the core
application. You could also use these applications for incubation before
incorporating it in the core app. Regarding areas of impact (Figure 4-12),
it’s important to have strong governance around the data model and
creation of additional data stores, which can lead to data fragmentation.

Fig.
4-11 New workloads

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

Low-medium Low-medium Low-medium Medium-high Low-medium Low-medium Medium-high

This could be App- Some apps User readiness New apps Additional apps Data and
medium impact specific ALM could require effort is on existing can impact integration
based on the app; requirements additional required, but environments storage and needs for the
core Dynamics can impact your admin tasks existing users won’t have a tenant API new app can
365 apps might release and will benefit from major impact capacity have an impact
have been build pipelines a familiar UI
already approved

Implementation Guide: Success by Design: Drive app value 78

 Integrations
Integrations can play a key role in eliminating data duplication and im-
proving data quality. Most importantly, they reduce the need for users
to navigate multiple applications, which prevents context switching.
The core integrations should be part of the initial phases and not be
left to future expansion phases. Even the out-of-the-box integrations
with existing productivity applications like Microsoft 365 can positively
impact adoption and make the information seamlessly available across
apps while enabling stronger collaboration. However, an abundance of
caution is needed; integration can also lead to performance issues and
increase the overall complexity and cost of maintenance (Figure 4-13),
so using the right patterns and practices is critical.

Aggregation
Aggregation can help consolidate multiple business applications with
significant overlap of data and processes into a single application.
Aggregation is different from integration—it completely replaces the
application and brings the core feature set into an existing application
instead of exchanging or embedding information. For example, a
commercial banking sales system could become an aggregator for all
non-personal banking sales systems.
Fig.
4-12 Satellite applications

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

Low-medium Medium Medium-high Medium Medium-high Low-medium Medium

If the data flows Managing As the satellite User readiness Depending Potential impact You should
in accordance the lifecycle apps grow effort is low on the ALM to capacity; avoid
with data loss of satellite organically, the for citizen- and citizen you may also fragmenting
prevention apps requires appropriate developed development need licenses the data with
policies, impact changes to governance community strategy, the to support the additional apps
should be low existing ALM policies need to apps, but might environment makers and using their own
processes be in place to require readiness strategy is consumers if data stores,
manage them effort if they’re impacted they’re not but it may be
adopted more already covered required in
broadly by existing some cases
licenses

Implementation Guide: Success by Design: Drive app value 79

 The cutover approach for old applications is critical; make sure to avoid
parallel usage leading to fragmentation of data and other aftereffects
(Figure 4-14).

Application standardization
This model of growth is driven by creating a generic application that
isn’t targeted at a specific process but a host of similar processes. The
application could even provide a framework that enables the business
to onboard themselves. This can achieve hyper-scale, in which you can
broadly t-shirt size hundreds of processes or applications that serve
more than one business process.

Fig.
4-13 Integrations

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

Medium-high Medium-high Medium-high Low-medium Low-medium Medium-high High

Depending on ALM process Additional Depending You may need Potential Expect impact
the data flows, could be monitoring on frontend to have stubs impact to API on data
additional impacted and integration versus backend or downstream consumption movement and
security depending on telemetry integration, integration make efforts
approvals may complexity is needed the impact and environments to ensure
be needed for to support readiness effort integrations
integration reconciliation of could vary are built to
pattern failures standard

Fig.
4-14 Aggregation

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

Medium-high Medium-high Medium-high Medium-high Low Medium-high Medium-high

Level of impact Existing ALM Admin User readiness is This will have Additional users High impact
depends processes are processes are required impact on and increased with respect to
on data changed consolidated environment data volumes data alignment
classification strategy, impact capacity and integration
requiring a consolidation
potential merge

Implementation Guide: Success by Design: Drive app value 80

 A general-purpose application allows scale, but can require additional
training and user readiness, so it must be thoughtfully done to be successful.

For example, consider a generic case management system that can

accommodate hundreds of different case processes from different
internal business functions that broadly involve similar steps. The busi-
ness can create a team of resolvers, configure service-level agreements
(SLAs), provide email notification templates, and create Microsoft
Word or Excel templates to capture non-standard data, but they can
still use the generic application for core case management capabilities.
All these actions have important considerations (Figure 4-15).

Pivot
Pivots aren’t necessarily expansion, but can trigger major change to an
application, which could fundamentally change and further broaden
its scope (Figure 4-16). For example, a targeted Power App used by a
small team could go viral within an organization, and you must pivot
to make it an organization-wide application.

Conclusion
As we have seen throughout this chapter, achieving true digital trans-
formation is not a sprint, but a marathon—it’s a continuously evolving

Fig.
4-15 Application standardization

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

Medium-high Medium-high Medium-high Medium-high Low Medium-high Medium-high

Once approved Existing ALM There could be User readiness is Should have Additional users There could
for appropriate processes process specific required for each minimal impact and increased be an increase
data shouldn’t configurations new process data volumes in the data
classification, change for a required for impact capacity generated by
more processes specific template each new each process
shouldn’t have workload on the
an impact template

Implementation Guide: Success by Design: Drive app value 81

 Fig.
4-16 Pivot

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

Medium-high Medium-high Medium-high Medium-high Low Medium-high Medium-high

Application Existing ALM Admin and User readiness Potential Additional users Expect high
will need to processes will governance might be changes to and increased impact to data
undergo further potentially scope will required the overall data volumes and integration
scrutiny change need to be environment will impact based on the
reevaluated strategy capacity pivot

journey with iterations and changes that need to be effectively man-

aged. It starts with understanding the changes to your business model,
using triggers to accelerate transformation, and adopting the lean
startup mindset to build with an intention to test, learn, and change
while staying focused on driving maximum business value in the short-
est amount of time.

Going live with your MVP is the first step to driving value. You will iter-
atively refresh your transformation roadmap, revisit and refresh your
program goals, adopt new product capabilities, and drive meaningful
expansion while staying true to the core program value and transfor-
mation goals. Following this approach can establish your organization
as a leader in your industry.

The holistic value of SaaS is much broader; it goes beyond just the
technology and features of your application. You’re investing not only
in the technology but also in the research and innovation that drives
the product changes. Simplifying experiences and improving interop-
erability further reduces costs and drives business innovation.

As a final takeaway, we can’t emphasize enough the importance of

staying engaged and investing in learning. Stay on top of key product
announcements, industry trends, and research studies on technology
trends. Engage with product leaders via events, meetups, and advisory
boards. This will significantly boost your ability to achieve success.

Implementation Guide: Success by Design: Drive app value 82

 Checklist

Approach to digital Transformation roadmap

transformation Create an effective MVP scope that goes beyond the
bare minimum to accelerate value realization.
Have a clear understanding of the core business model
and how the application being developed impacts the Define your transformation roadmap and structure the
key elements of the business model. phases to maximize time to value and drive maximum
business impact.
Clearly articulate and communicate business transfor-
mation goals with specific measurable KPIs. Identify the expansion models and incorporate that into
your program phases.
Ensure the implementation team understands the
business processes in scope and the business teams are Adopt the latest capabilities to deliver continuous
closely engaged to optimize the process where needed. innovation to business.

Change stream
Account for the various change streams that will impact
the program and ensure budget and timelines are
adjusted to accommodate.

Implementation Guide: Success by Design: Drive app value 83

 Case study
Dynamics 365 helps
adapt business
processes to change
One of our clients, an accounting network, is facing unique and unex-
pected challenges. The COVID-19 pandemic has inhibited face-to-face
work, threatened global supply chains, and shifted regulatory and
political landscapes.

To maintain business continuity and stay connected with customers,

the company is adapting their digital selling techniques with the help
of Dynamics 365. With 360-degree views into each profile, sales staff
can work remotely without interrupting the customer experience.
The solution also works with the firm’s risk assessment tool to identify
vulnerable areas caused by the pandemic, allowing staff to create
proactive plans for their clients. The company is keeping their business
personal, up-to-date, and resilient by optimizing digital solutions in a
time of uncertainty.

This successful pivot clearly highlights the power of Dynamics 365 as

an adaptable SaaS cloud platform that can not only deliver strategic
solutions but also adapt to changing business needs quickly.

Implementation Guide: Success by Design: Drive app value 84

5 Guide
Implementation
strategy
Implementation Guide: Success by Design: Implementation strategy 85
 Introduction
Organizations embarking on a digital
transformation journey consider implementing
business applications as key critical milestones
for their business success.
The journey from a business vision to a successful implementation
of it requires the business and IT owners to define an overarching
strategy that acts as a framework to plan for predictable outcomes.
An implementation strategy defines these strategic choices and
decisions made in planning, implementing, and deploying Microsoft
Dynamics 365 applications.

Along with new technology comes substantial change to a company’s

business processes. It is critical to choose a process-focused, user-
Key concepts that will help in defining
a successful implementation strategy.
centric implementation strategy and methodology that not only
1 Key aspects supports digital transformation but also helps drives change manage-
2 Business drivers ment and user adoption. All of these factors are critical to making an
3 Success metrics implementation successful.
4 Methodology

5 Deployment
In this chapter, we explore common industry methodologies and look
6 Change management

7 Conclusion
at common deployment strategies and high-level views on change
management strategy.

Implementation Guide: Success by Design: Implementation strategy 86

Key aspects of
implementation strategy
The foundation of a successful implementation begins with planning
for key factors that define the overall implementation strategy. Let us
explore these key considerations.

Understand vision and business drivers

Project teams should be aware of the vision and strategic intent of the
customer stakeholders. Having a clear grasp of the key reasons behind
the implementation of a project and a collective understanding helps
the project team stay aligned with the project’s vision of addressing
significant business challenges.

It is important that senior management and project sponsors are clear

about the business requirements and help to establish the scope and
expected outcomes of the project. These goals help drive any decisions
that may be needed during the implementation cycle— for example,
prioritizing delivery of certain functional features over others.

This helps deliver a solution aligned with business needs and allows for
change management and a higher user adoption.

Identify key success metrics

After establishing a common understanding of the business challenges
and priorities, it is equally important to have an agreement on what the
key performance indicators (KPIs) that measure success look like. The
KPIs should be tangible metrics, for example opportunity conversion
rate, or average call handling time.

Understanding and identifying the KPIs is critical to ensuring the scope

being delivered is aligned with the expectations from the business
stakeholders and generates a tangible return on investment on the
Dynamics 365 investments.

87
 Some of the examples of success metrics/KPIs include:
▪ Improve the opportunity conversion rate from X % to Y % over
z period.
▪ Reduce sales cycle from x days to y days over z period.
▪ Increase net new customer adds from x to y over z period.
▪ Reduce average call handling time from x mins to y mins over
z period.
▪ Improve service request turnaround time from x days to y days
over z period.

The KPIs mentioned above are not exhaustive, customers have specific
KPIs that cater to their business scenarios. It is important to acknowl-
edge that each KPI the customer identifies and prioritizes has a direct
impact on the functionality that would be deployed on the Dynamics
365 Business Applications solution. Consequently, having a definitive
list of metrics to refer back to enables prioritization of the right use
cases and allows customer stakeholders to gauge the project success in
a tangible way.

Identify and allocate right resources

Setting up a productive team structure is critical to the success of the
project, which includes identifying the stakeholders, business process
owners, the team that delivers the projects, and the end-user representatives
who can be engaged from the beginning of the project lifecycle.

The key roles required for all projects can often be grouped into two
types of team structures.
▪ Steering committee  The group of senior stakeholders with the
authority to ensure that the project team stays in alignment with
KPIs. As they monitor the progress they can help make decisions
that have an impact on overall strategy, including budget, costs
and expected business functionality. Steering committees usually
consist of senior management members and group heads whose
business groups are part of the identified rollouts.
▪ Core implementation team  This is the team doing the actual
execution of the project. For any Dynamics 365 implementation
project, core implementation teams should include project

Implementation Guide: Success by Design: Implementation strategy 88

 manager(s), business subject matter experts, solution architects,
functional and technical consultants, developers, testers, change
management advisors, and user champions.

Process owners engagement is critical to ensure complete understanding

of business processes. Their understanding is required in case of any
process changes that may result from alignment with out of the box
features. User representation should also be planned properly to
adequately cover all the personas. Lack of availability of any of these
may result in delays and/or incomplete understanding of processes
and also impact user adoption. The involvement of process owners and
users can be staggered depending on the implementation phase. This
should be planned early on to allow business teams to make the required
changes to their schedules and be able to contribute efficiently during
the project lifecycle.

The project may require additional roles depending on the scope and
methodology. Some roles may need to join temporarily during the
project lifecycle to meet specific needs for that period.

For more details on team setup considerations, refer to Chapter 2,

“Success by Design overview,” and Chapter 8, “Project governance.”

Allocate budget and funds

Planning a budget for Dynamics 365 cloud applications typically
includes costs such as subscription, storage, implementation lifecycle
activities of designing, testing and deploying the solution, migration
from legacy applications, training, change management, and
continuous improvements.

Project lifecycles require constant assessments with careful considerations

and actions for the variance between budgets versus actual costs.

A common decision that most teams face is finding the right balance
between delivering a high value solution faster by using the off the
shelf capabilities versus extending the product capabilities to implement
the business requirements and needs. Extending the Dynamics

89
 365 applications not only requires initial development costs but also
maintainability and supportability costs in the long run. This is an area
Forrester’s Total Economic
Impact (TEI) of deploying various
where implementation teams should carefully revisit the cost impact.
Dynamics 365 apps is a good In the cloud world with rapid availability of new features and capabil-
recommendation to assess the
costs and benefits associated with ities, there is a need to rethink the investments required for extending
Dynamics 365 Implementation. the application. Refer to Chapter 15, “Extend your solution,” for more
Some of the exclusive TEI done by
details on assessing this impact.
Forrester are:

The Total Economic Impact of

Microsoft Dynamics 365 for Project management teams should give appropriate weight to areas that
Finance And Supply Chain
Management help them stay aligned with the overall project goals and objectives. It
has often been seen that as project scope and timelines change, the
The Total Economic Impact of
Microsoft Dynamics 365 Field activities that fall towards the end of the project lifecycle are usually
Service
descoped to meet budget constraints. These activities can include
The Total Economic Impact of performance testing, data validation testing, training, and change
Microsoft Dynamics 365
Customer Service management, which, when compromised, can significantly reduce the
quality and success of the implementation. It is important to revisit the
project goals and success factors that are defined during initial stages
to carefully revise the scope and timelines without compromising on
overall quality and user centric approach.

An overall approach is recommended since it helps define the total

impact by identifying the cost, benefit, flexibility, and risk factors that
affect the investment decision, helping organizations improve their
overall business goals of winning, serving, and retaining customers.

Choose a methodology
Before you discuss choosing a methodology for your Microsoft
Dynamics 365 project, we need to understand why a methodology is
important for Dynamics 365 projects.

Why methodology is important

A methodology is a prescriptive definition of activities to be undertaken
for a given project or engagement. It describes the use of a collection of
methods to achieve predictable outcomes.

Implementation Guide: Success by Design: Implementation strategy 90

 Methodology is the core of project execution that allows the team to
take the right path towards:
▪ Improved consistency and predictable outcomes.
▪ Setting clear definitions of project phases, milestones, deliverables,
and entry and exit criteria to each phase.
▪ Setting clear definitions of roles and responsibilities required for
the project.
▪ Reducing risks of missing critical activities required for a
successful outcome.

Following a methodology is akin to taking a planned journey; the

roadmap to success (as illustrated in Figure 5-1) involves the following:
▪ Plan ahead  Ensure you have clearly defined the destination
(project scope), and have all the right tools (resources, governance,
and methodology) and checklists (project monitoring) before
starting this journey.
▪ Know the road  Define the route (solution blueprint) to set clear
expectations on start and end dates at milestones so you can
meaningfully monitor progress and the team understands what is
expected from them.
▪ Avoid pitfalls  Utilize proven best practices, as advocated in the
methodology, that minimize risks, costs, and scope creep.
▪ Get there faster  Take advantage of strategies that help you save
Fig. 5-1

Plan ahead Know the road

Make sure you have all the right tools Set clear expectations around start and
and checklists for the trip. end dates and project milestones.

Start

Go live Go
live

Avoid pitfalls Get there faster No surprises

Minimize risks, costs, and scope creep Take advantage of strategies Delight executives and customers by
with proven best practices from that save time and money. delivering the right results on-time
thousands of employees. and on-budget.

Implementation Guide: Success by Design: Implementation strategy 91

 time and money. Plan, action, and regularly monitor and adjust to
ensure you are taking the most efficient route to your destination.
▪ No surprise  Delight users by delivering a user centric solution
and business stakeholders by delivering results on-time and
within budget.

Choose an implementation methodology

Let us take a look at the specific methodologies that can be used for a
Iteration and sprint is a timeframe successful Microsoft Dynamics 365 implementation.
in which teams deliver incremental
value in the form of working tested
software. Iterative approaches use We often see three types of models used to implement Dynamics 365
the term iterations while Scrum
Business Applications projects.
approaches use the term sprints.
▪ Waterfall model  A delivery execution model based on sequential,
steadily flowing, like a Waterfall, series of phases from conception,
initiation, analysis, design, development, testing, deployment,
operations, to maintenance.
▪ Agile model  Agile development is not a methodology in itself,
but an umbrella term that describes several Agile methodologies.
Some of the industry known methodologies are Scrum, XP, DSDM,
Sure Step for Agile (for Business Applications).
▫ Scrum methodology  A delivery execution methodology
based on Agile principles where requirements are in flux or
unknown prior to beginning a sprint. Sprints involve
requirements analysis, design, build, test, and user review.
Sprints are a fixed duration (usually 30 days), as illustrated
in Figure 5-2.
▪ Hybrid model  A delivery execution model based on Agile prin-
ciples where requirements within the context of business processes
are defined, analyzed, and prioritized and an overview design
solution blueprint is generated prior to beginning the sprints.
Sprints involve refining designs, build, test, and user reviews.
Sprints can be of varying durations.
Fig. 5-2

Product backlog Sprint backlog Sprint Working increment

of the software
24h
30 days

Implementation Guide: Success by Design: Implementation strategy 92

 Now let’s take a deeper look at each of these methodologies.

Agile
Success by Design Framework is
An Agile implementation is an iterative approach that uses a number
methodology agnostic and is aligned
to ensure proactive guidance and of iterations or sprints.
predictable outcomes irrespective
of chosen methodology. For more
information, refer to Chapter 2, In Microsoft Dynamics 365 Business Applications projects, the majority
“Success by Design overview.”
of requirements are delivered by the packaged software. There are
specific challenges and tasks that are out of the box and need to be
managed and mitigated as part of the project methodology with the
user stories contained in the form of a backlog. Using those stories, you
carve out plans and within them a number of iterations or sprints of
development and testing are executed. Within each sprint, you have a
number of user stories outlining the requirements.

Agile promotes a collaborative process between the resources that

own and specify the requirements for the solution and the resources
responsible for the development and rollout of the solution.

The idea is to release software faster, take early and regular feedback,
adapt, and release again. This cycle continues until all the requirements
are met or the backlog is cleared.

An example of Agile practices using the methodology prescribed by

“Sure Step for Agile” is shown in Figure 5-3.

Waterfall
The Waterfall approach to solution delivery is a sequential process
that depicts a linear flow of activities from one phase to another,
culminating with the solution being promoted to production and
then into operation.

This is a traditional methodology of implementing on premises business

applications. It is a linear, noniterative approach of implementing and
delivering projects.

Implementation Guide: Success by Design: Implementation strategy 93

 Fig. 5-3
1.1.1 4.1.1 5.1.1

management
Project management Project Project
Analysis

Program
management management
1.2.1 1.2.2
Conduct Gather user
solutions training
overview requirements
1.3.1 1.4.1 Design
Conduct detailed Gather
Agile preparation

4.2.2
business process business

Training
Conduct end
analysis requirements
user training
1.7.1
Set up DEV Development
and other
non-production 1.4.2
environments

process analysis
Conduct fit gap
analysis

Business
1.8.1
Establish Deployment
integration
strategy

1.9.1
1.4.3
Gather data
Define

Requirements and
migration Operation

configuration
solution backlog
requirements
5.4.2
4.4.2
Transition to
Go live
support

30-day sprint cycle

2.1.1
Define sprint

Custom
coding
backlog

2.1.4 2.1.2
Conduct sprint Conduct sprint
post mortem planning meeting

and testing
4.6.2

Quality
Daily sprint cycle Conduct UAT

3.1.1 3.1.2
Agile execution

Conduct daily Analysis and

sprint cycle design for
meeting sprint cycle

Infrastructure
3.1.3
3.1.6 4.7.1
Create scripts
Conduct Build production
for testing
sprint testing environment

3.1.4
3.1.5
Sprint
Generate
configuration
daily build
and development
and interfaces
Integration

2.1.3
Conduct sprint
technical preview

4.9.3
3.6.1 3.7.1
migration

Final data
Conduct Finalize
Data

migration to
solution production
production
testing specification

Implementation Guide: Success by Design: Implementation strategy 94

The phases, milestones, and deliverables are clearly defined, and you
only move to the next phase when a prior phase is completed.

In the modern cloud world, early stakeholder buy in is critical which

requires delivering quicker results, which makes waterfall technique
not so suitable.

An example of the Waterfall model as prescribed in Sure Step

Methodology is depicted in Figure 5-4.

Hybrid
In the current landscape of rapidly changing business needs and
changing technology trends, projects commonly adopt a hybrid
approach. This is a combination of Waterfall and Agile practices that
implementation teams use to meet their project needs. It allows teams
to tailor an approach that enables a successful implementation of
Dynamics 365 applications. It is also our recommended approach for
Dynamics 365 engagements.

Figure 5-5 shows an example of a hybrid approach, based on

Dynamics Sure Step 365.

With this approach, we can manage initial activities, like initiation and
solution modeling, and final activities like system integration testing,
user acceptance testing, and release to production, in a noniterative
way. Then the build activities, such as requirement detailing, design,
development, testing, are completed with an iterative approach.
This helps provide early visibility into the solution and allows the team
to take corrective actions early on in the overall cycle. This approach is
considered to use the best of both Waterfall and Agile approaches and
is a win-win for any size of implementation.

In the hybrid approach, the focus is on a regular cadence of iterative

and incremental releases of software features.

We stress for customer stakeholders to participate in the process, espe-

cially in the business validation reviews that happen after each sprint

95
 Fig.5-4

Analysis Design Develop Stabilization Deploy

Organization

Training plan

Deployment plan User training

Baseline
project plan

Functional design Go live

Process and
functional UAT scripts UAT
Solution

Transition to
requirements
Technical design support
Solution testing
Fit gap analysis
Solution design

Data migration
Technology

requirements
Integration and Production Production
interface design specifics environment
Environment
specification

Fig. 5-5

Noniterative Noniterative Iterative Noniterative Noniterative

Solution Solution Support

Mobilization Initiation Build Deployment Operation
modeling testing transition

Analysis Process
Design End-to-end
Development Performance
Iteration testing UAT

Once per
Once per project Once per release iteration Multiple cycles per release Once per release
in a release

Implementation Guide: Success by Design: Implementation strategy 96

 cycle. These reviews are placed at regular intervals during the project
schedule, and allow for faster valuation and better stakeholder buy in.

Through these reviews, customer stakeholders get an opportunity to

see working software, have discussions with the team, and provide
early feedback. The readiness of business processes is validated for
production use by an invited business audience by showing the
incrementally improving readiness across all the components in the
system as working software, like design, data configuration, data
migration etc.

The idea is to get early feedback and validation from the business team
on the business requirements scope and understanding. This approach
allows the project to showcase working solutions faster than other
methodologies and achieves a higher rate of implementation success.

Define deployment
strategy
A deployment and rollout strategy describes how the company is
going to deploy the application to their end users. The strategy chosen
is based on business objectives, risk propensity, budget, resources,
and time availability, and its complexity varies depending on each
implementation’s unique scenarios. There are several key factors to
consider prior to choosing any of the approaches detailed below.

▪ What is the MVP (minimum viable product) needed for faster val-
ue to customers and then later plan for continued enhancements?
Refer to Chapter 4, “Drive app value,” to define an MVP strategy.
▪ Is this a multi-country rollout or single-country rollout?
▪ Is there a need to consider pilot rollouts prior to wider
team rollouts?
▪ What are the localization requirements, especially in scenarios of
multi-org or multi-region rollouts?
▪ Do we need to phase out a legacy application by a certain key
timeline or can it be run in parallel?

Implementation Guide: Success by Design: Implementation strategy 97

 ▪ What is the extent of impact on end users’ day-to-day
productivity?
▪ What training needs do end users have?
▪ What is the complexity of integrations during the rollout?

Taking these factors into consideration, you should define a rollout

strategy that helps the organization deploy the application in the most
effective manner.

The following are some typical deployment strategy models.

Big-bang
In the big-bang approach, as the name suggests, the final finished
software solution is deployed to production and all users stop using
the old system and start using the new system on the go-live date.
The term “big-bang” generally describes the approach of taking a
large scope of work live for the entire user base of your
organization simultaneously.

The selling point of the big-bang approach is that it concentrates the

focus for a shorter rollout period compared to a phased rollout. You
trade the cost and disruption of staying in rollout mode for a longer
period of time for a longer wait to get value from the solution.

An example of big-bang rollout is when an organization that was using

a legacy business application for all of its corporate office finance team,
communicates to all end users to stop using the old system and start
using the new Dynamics 365 Finance on the go-live date.

The whole implementation can be very large and can encompass

multiple countries, departments, or business units. During transition,
the new solution is deployed through a number of activities while the
old system is turned off so the new system can go live.

The risks of this approach are that the entire project could be rushed,
minor but important details could be ignored, and business processes
transformed may not be in the wider interest of the organization. In

98
 an aggressive big-bang rollout, the risks are magnified due to the dual
combo of a large scope and shortened rollout period.

Change management can also suffer as people are less inclined to

use the system that may not be solving the business problems it was
meant to solve.

With such large changes happening, there is a potential for issues to

arise post go-live and falling back on the older system is very hard,
if even possible. It is critical to ensure proper contingency planning
for business continuity, and support mechanisms to deal with post
go-live issues.

Fig. 5-6
Large project scopes are at much greater risk by using the big-bang
approach since the delivery of
Pros Cons finished software to the produc-
▪ Shorter, condensed ▪ May not accommodate for rapidly
tion takes up a longer timeline,
deployment times, minimizing changing market scenarios and
there is more to organize during
organization disruption product capabilities, leaving a
lack of solution alignment at transition. If the final rollout runs
▪ Lower costs from a quick rollout
without loss of overall work, but deployment causing a negative into challenges, the wider user
with an added need to account impact for end users community is impacted due to
for resource costs to manage the ▪ Daily task delays from users the resulting delays. If Waterfall
complexity learning to adapt to the new methodology is used as the
application without any prior
▪ Same go-live date for all users implementation, then the end
experience
▪ Reduced investment in users don’t have a feel for the real
temporary interfaces ▪ Transitioning back to legacy
system when it is finally rolled out.
systems can be costly and
difficult, if even possible
However, if hybrid methodology
is used, it is possible to involve end
▪ Any failures during the rollout have
a high impact on maximum users users from the beginning, keeping

▪ Heightened probability of risks

them updated on what is land-
emerging from introducing large ing on go-live and keeping any
changes for large groups of users surprises in check.
▪ Heightened probability of issues
coming up in go-live For large and complex imple-
▪ No opportunity to learn from mentations, big-bang projects
early launches to change things can require many months or even
▪ Costs of ramping up a large rollout years to complete. Given the pace
team for a short period of time of change in most industries it may

Implementation Guide: Success by Design: Implementation strategy 99

 not be feasible to wait that long for implementation, especially since the
business that was analyzed for requirements could have different needs
and priorities a year later, as illustrated in Figure 5-6.

Phased rollout
In a phased rollout approach, we are releasing the software in a
number of phases or releases. The phases can be planned in several
ways and can be based on modules, business priority, business units,
or geographies. One approach for a phased rollout can be to release
an initial set of capabilities in the first phase and build on that by
releasing the rest of the requirements in subsequent phases. At the
release of the first phase for a business unit or module, it is expected
that the specific business unit’s end users stop using the legacy system
Fig. 5-7

Pros Cons
▪ Higher business value features can be prioritized. The ▪ Longer implementation and deployment timeline due
implementation team may also choose to prioritize lower to multiple go live events.
complexity features to adapt to the new functionality. ▪ Complexity of data migration from legacy application
▪ Unlike big-bang, phased rollout releases can bring to the target application increases as it needs to be
new use cases gradually, allowing for more buy in planned in a staggered approach.
from end users as they get used to the new system. ▪ Organization needs to plan for continuous disruption
▪ Project teams can optimize later phases by incorporating over longer periods of time, like parallel efforts of
the learnings from earlier ones. supporting deployments while working on future

▪ Risk of disrupting business is less as errors found in phase implementation.

the initial phase are limited to specific business areas ▪ Employee morale may suffer as they face change fatigue.
and users and rest of the business is not affected. Phased projects need lots of focus and coordination as

▪ When phases are based on functionality, rather than the team might start losing momentum after first few

module, geography, or business unit, it typically phases. These projects don’t have the benefit of focused

results in a faster time to value. intensity as the big-bang approach does. It is important
that the change management team keeps a tab on
▪ For large implementations, it reduces the level of
employee morale and plans initiatives to keep their
investments and resources needed for ramp up for
interest alive.
each deployment relative to a big-bang approach.
▪ Need to implement temporary scaffolding solutions to
manage integrations between new and legacy systems
until the new solution is fully rolled out.

▪ Scope creep can occur since project timelines are

drawn out. There can be a tendency to incorporate
changes and rework, which can affect the work for
subsequent phases and their timelines.

Implementation Guide: Success by Design: Implementation strategy 100

 and move to the new one. There could still be temporary scaffolding
integrations setup to integrate the new system with the legacy systems
until all phases are rolled out.

For example, a team can start the implementation with one business
unit. Once implementation is complete, they travel to the next business
unit. As the team gains experience and learns from past mistakes, the
subsequent rollouts become smoother. This leads to less risk and more
employee adoption, as illustrated in Figure 5-7.

Parallel rollout
In a variant of phased rollout approach, parallel rollout, the legacy
system is not discarded but kept alive along with the new system. More
than a rollout approach, it is a validation technique allowing users an
opportunity to learn, test, and get comfortable with the new system.
Typically, for few months both systems are actively used and users are
required to key in information in both systems.

▪ There is much more effort required from the users as they double
key information.
▪ This rollout may be preferred by projects where the business risk is
very high and cannot be easily mitigated by testing. At the same
time, we need to be mindful of making sure team is not cutting
corners on testing and training efforts.
▪ This rollout is less and less popular due to the extra efforts and
costs involved in keeping two systems.

Rollout strategy goes hand in hand with change management strategy

as several activities of change management play a key input to
successful rollout, as illustrated in Figure 5-8.

Fig. 5-8

Pros Cons
▪ Less risk ▪ High efforts and high costs

▪ Users take their gradually plan to migrate to the ▪ Needs for data duplication can get complicated. We
new system recommend having a clear exit plan that is time or
workload based, for example, all existing cases closed
in old system

▪ May mask poor training and testing efforts

Implementation Guide: Success by Design: Implementation strategy 101

 Define a change
management strategy
Adoption and change management planning is a critical aspect of
implementation strategy. Change is never easy for any organization and
requires thoughtful planning and a structured approach to achieve high
adoption and usage. A change management strategy with a clear action
plan ensures smooth rollouts and alignment with stakeholder and end
user expectations. In the absence of a planned change management
approach, many business applications tend to become purely a reporting
tool rather than a value driven user centric application.

Bridge the gap between business

requirements and results
All too often, projects meet requirements without a true focus on adoption
and business outcomes. The gap that exists between technology solutions
and business outcomes directly relates to the users who experience a
change in their day to day work. Change management focuses on bridging
this gap by effectively supporting and equipping the user community
to successfully adopt the transition. A change management mindset is
required across different functions, including the project development
teams, to keep the end user experience as a primary design consideration.

Increase likelihood of project success

PROSCI’s correlation data from over 2,000 data points and 10 years
shows that initiatives with planned change management are six times
more likely to meet objectives than those without a formal change
management strategy. The data is abundantly clear. The better we
apply change management, the more likely we are to deliver on
project objectives.

Take the change out of change

Change is difficult. To the degree that we can, we want to remove the
chance or variability associated with change. Project management has

Implementation Guide: Success by Design: Implementation strategy 102

Change is never accomplished this by providing direction on sequencing milestones,
deliverables, activities, and resources over the lifecycle of an effort.
easy for any Unless we proactively support and guide people through the changes
organization and our projects bring, we leave the likelihood of them embracing change
requires thoughtful to chance. Change management addresses this by providing

planning and a employees with the preparation, support, and skills they need to
succeed in change.
structured approach
to achieve high PROSCI’s framework that is followed at Microsoft describes the three
adoption and usage. sequential steps that are followed in change planning.

Preparing for change  The first phase in PROSCI’s methodology

helps change management and project teams prepare for designing
their change management plans. It answers these questions:
▪ How much change management does this project need?
▪ Who is impacted by this initiative and in what ways?
▪ Who are the sponsors we need to be involved to make this initia-
tive successful?

Managing change  The second phase focuses on creating plans that

integrates with the project plan. These change management plans
articulate the steps that you can take to support the users being
impacted by the project.

▪ Communication plan  Communications are a critical part of the

change process. This plan articulates key messages that need to go
to various impacted audience. It also accounts for who sends the
messages and when, ensuring employees are hearing messages
about the change from the people who have credibility with them
at the right time. Communication is an understated yet very signif-
icant aspect that keeps the various teams connected throughout
the entire journey. Some of the key communication aspects to
consider are:
▫ Providing a sneak peek into what is in it for end users
▫ Launch and availability communications
▫ Project sponsor vision and direction communication
▫ Communication frequency and approach
▫ End user feedback incorporation

103
 ▫ Status reporting, including steering committee reports
▪ Sponsor roadmap  The sponsor roadmap outlines the actions
needed from the project’s primary sponsor and the coalition of
sponsors across the business. In order to help executives be active
and visible sponsors of the change it identifies specific areas that
require active and visible engagement from the various leadership
teams, what communications they should send, and which peers
across the coalition they need to align with to support the change.
▪ Training plan  Training is a required part of most changes and is
critical to help people build the knowledge and ability they need
to work in a new way. The training plan identifies the scope, the
intended audience, and the timeframe for when training should
be planned for. It is important that the training plan be sequenced
in a way that allows for awareness and desire building before they
are sent to training. A common aspect that tends to be overlooked
is cloud focused training for IT administrators. As organizations
implement Dynamics 365 cloud applications, it also may mean a
significant change for IT teams, so it is important to plan training
that keeps their persona in perspective. They may require training
not only on Dynamics 365 administrative activities, but also for a
general understanding of cloud concepts.
▪ Coaching plan  The coaching plan outlines how you engage with
and equip managers and people leaders to lead the change with
their own teams. Managers can play a significant role in aiding
the change management efforts, but they need to be engaged as
employees themselves first and allowed to work through their own
change process. Once that’s done, you can give them the information
and tools to lead the same change process with their own teams.
▪ Resistance management plan  Resistance to change is expected,
so proactively defining the activities to mitigate the areas of concerns
should be initiated early on in the project lifecycle. Engaging user
champions from the end user community early on to build a user
centric solution contributes significantly towards addressing this risk.

Reinforcing change  Equally critical but often overlooked, this third

phase helps you create specific action plans for ensuring that the
change is sustained. In this phase, project and change management
teams develop measures and mechanisms to measure how well the

Implementation Guide: Success by Design: Implementation strategy 104

 change is taking hold, gauge user adoption, identify alignment with
KPIs, and correct gaps.

A successful change management strategy requires continuous

executive sponsorship, integration with the project management teams,
employee engagement, and frequent and open communication.

Conclusion
When a team defines an implementation strategy, they set clear
expectations and guidelines for the team and wider business on how
the project goals are going to be attained. This helps define a clear
translation of business goals into a structured implementation plan.

To summarize, a well-rounded implementation strategy takes the

following into consideration.
▪ Build a common understanding of vision.
▪ Identify the methodology based on the specific factors that are
driving the implementation.
▪ Follow the tried and tested approach that is defined by the meth-
odology—do not leave activities out just for convenience.
▪ Help to take feedback early in the implementation journey by
doing a show and tell after each sprint/phase depending on the
chosen methodology.
▪ Keep organizational culture under consideration while choosing a
methodology. Organizations that have followed a traditional ap-
proach such as Waterfall may need additional change management
planning to embark on a modern approach such as Agile or hybrid.
▪ Consider which industry practices may also have a bearing on
methodology. Some industries may have their own tailored ap-
proach and that needs to be kept into consideration.
▪ Formulate a focused approach towards user adoption by creating
a change management plan.
▪ Plan an approach of communicating to various stakeholders,
including leadership and end users.
▪ Continue to improve throughout the process—solicit end-user
sentiment and act!

Implementation Guide: Success by Design: Implementation strategy 105

Section
Initiate
6 Solution architecture design pillars
7 Process-focused solution
8 Project governance
9 Environment strategy

106
6 Guide
Solution
architecture
design pillars
You must be sure about the
solution before you start building,
because failure is not an option.

Introduction
You can’t have a solution without first having a vision.
When you know the answer you’re looking for, only
then you can find a solution to get you there.
But it’s not always easy to know and articulate what you want, let
alone identify the elements that are essential for creating a blueprint
of your solution.

This chapter outlines the pillars of a successful solution architecture

design that support your vision, and provides you with the building blocks
you can use to create a coherent and satisfying outcome. By utilizing
the Success by Design framework, we present a structured approach for
designing a solution that’s enabled by Dynamics 365 apps.

Solution architecture
design pillars
Most of us know what “architecture” means in the building industry—
it includes the job of the architect, the scope of the work, and what’s
ultimately delivered. The architect starts with the project requirements

Implementation Guide: Success by Design: Solution architecture design pillars 108

 and the budget, and then works on a blueprint of the design. Once the
physical labor begins, the foundation is laid, the house is built, and the
interior is completed. Many years of maintaining and improving the
building may follow. But why and when do you need an architect to
design a solution for your organization?

Let’s take the simple case of building a new house for a dog. In most
cases, this is a single-person job with a general vision of the final
product. Building a doghouse is cheap, and the risk of failure is low,
so a trial-and-error approach is one option.

But what if you were asked to build something as complex as the Sydney
Opera House in Australia? The approach would be completely different:
the job can’t be completed by a single person, the design needs to
be carefully planned, and architects, construction workers, plumbers,
electricians, and decorators have to coordinate their workflow. With so
many moving parts, there must be well-defined processes, an explicit
design, and powerful tools. When all the right pillars are in place, the
likelihood of failure is dramatically reduced.
Fig.
6-1

Solution architecture follows a systematic approach by identifying the

It all begins desired solution and the necessary building blocks, where each builds
with a vision. on the previous block and provides a foundation for the next block.
Solution architecture breaks up the work into logical sections, and
Processes
coordinates those sections, managing and mitigating the risks along
The vision leads to the
business strategy, where
the way.
the processes, people,
and data are joined.

People Data Building blocks of

The solution strategy solution architecture design
outlines how those
elements can be You start with a vision, which leads to a business strategy where the
addressed by using
processes, people, and data are joined (Figure 6-1). A solution strategy
the technology of
Dynamics 365 apps. outlines how those elements can be addressed by using the technology
of Dynamics 365 apps. Project and change methodologies, supported by
governance and control methods, define the workflows and the delivery.
The project and
change methodology
Solution architecture design incorporates the business vision and its
define the workflow
and delivery. implementation into a blueprint. The first version is created as part of

Implementation Guide: Success by Design: Solution architecture design pillars 109

 the pre-sales process and forms the initial high-level understanding of
what you plan to build.

Solution architecture enables implementation of a vision, and typically

begins after the business strategy has been crafted, but before the
solution is built and implemented (Figure 6-2). Architecture creates a
picture of the finished product and includes the specifications needed
to build it.

A solution architect provides the necessary quality control by ensuring

that key requirements are met, with all relevant interdependencies
considered and managed. A solution architect also serves as a translator
between the business and IT teams to articulate the impact of
technology on the business, and vice versa.

Vision
A vision is the desire to achieve something—to change the present and
improve the future.

When an organization decides to go through a transformation, it’s

usually because one or more people had the ability to predict what was
coming and expressed a desire for change, or a change was forced by
industry disruption or regulations. Such a change could be in the form
of a mission statement or a business case listing its objectives, which
Fig.
might include:
6-2

Implementation Guide: Success by Design: Solution architecture design pillars 110

 ▪ Finding unified and faster ways of working.
▪ Earning higher profits.
▪ Achieving better service or product quality.
▪ Improving the user experience.
▪ Empowering users to drive greater value by building apps.

As the vision comes together, the plan for achieving it can start
taking shape.

Business strategy
Every vision serves a purpose, as does every organization, and any
solution needs to be aligned with this purpose. A business strategy
supports your vision by answering fundamental questions, such as:
▪ Why are you making this change, and what are the anticipated
benefits? What is the business value sought by the solution? Where
do you imagine the organization will be in five, 10, or 20 years?
▪ What business capabilities can your organization offer with
the new solution? What business processes can you run? What
information and data would you like to record and report on, in
line with your organization’s services or products?
▪ Which clients, customers, or people inside the organization will be
served by the new solution, and who will be affected by it?
▪ Would you like to improve your current line of business or are you
open to a new industry?
▪ When do you plan to have the vision materialized? What is the
high-level timeline? And do you intend to deliver the solution at
once or grow in stages?
▪ Where are the regions—geographically and in terms of business
functions—to which the solution will apply? Will you apply it to
all or just some of them?
▪ Who is going to plan, design, and deliver the solution? Do you
have a preferred partner or do you need to select a vendor?
▪ How will you incorporate technology into your solution? (This
is the first step of solution design and the link to your solution
strategy, as well as the business case for IT transformation.)

Implementation Guide: Success by Design: Solution architecture design pillars 111

 Why should you
define your processes?
Processes
It creates consistency by Process architecture is a commonly understood, shared view of all
allowing for consolidation and business processes that an organization uses to deliver a product or
unification of processes across
service. It represents how an organization operates, in a structured
locations, teams, and systems.
order. Complete, accurate, and well-organized process architecture
It reduces complexity by is the first and most important pillar for a sound solution design. It
standardizing and rationalizing confirms end-to-end business understanding, provides a structure
previously complicated and
cumbersome procedures. for understanding the scope, and serves as the basis for testing
and training. It also forms a map for requirements gathering. The
It eliminates guesswork and processes are constructed at different levels to reflect specific areas,
potential ambiguity by
functions, locations, and teams. Process architecture is often referred
streamlining operations and
improving communications. to as a target operating model, a process library, or a catalog.

It promotes productivity by
Any transformation starts with defining your processes, which is also a
eliminating inefficiencies and
establishing one workflow for good time to revisit and improve them.
all users.

It guarantees quality by Dependent activities

maximizing attention to detail
and ensuring work is done in While we’re not going to take a deep dive into the business process
a pre-defined, optimized way management life cycle and capability maturity, following the Success
each time.
by Design framework can help you identify important elements of your
It boosts morale by helping solution design.
team members take pride in
mastering the process, refining
their skills, and avoiding mistakes Process architecture includes multiple dependent activities:
and missed deadlines. ▪ Scope management  Defining the scope of the solution is the first
step of the design. Ideally, you have a baseline process taxonomy with
It encourages continuous
improvement by giving team
at least three levels of the process architecture. Start by mapping
members who follow the requirements against it and marking which processes are in
processes a chance to give input scope. You can add to and take away processes from your initial
on how to improve them.
process library.
It increases user acceptance ▪ Business analysis  Initially, it’s likely that business users don’t
by providing a higher-quality know or understand the application, and the tech partner doesn’t
experience at each launch.
understand the business, so a business analysis helps to join them
together. A business analyst gathers requirements and links them
to processes; conversely, a good processes structure drives the
requirements gathering by asking all necessary questions.

Implementation Guide: Success by Design: Solution architecture design pillars 112

 Key deliverables related to processes
Process architecture map  Linear and cross-functional process flows
These show the input, output, and activities at process and
sub-process levels to help you figure out how a process must
change to meet requirements; you can also add details such as
assumptions, metrics, timelines, and costs.

Linear flow

This is a visual representation Product Product Product Product

of your business processes. ideation design development testing

Product
prototyping

Process catalog/ Product ideation phase

inventory and taxonomy Collect customer

suggestions

Test against
Conduct Put ideas into ideation Select ideas for
strategic
competitive analysis pipeline conceptual design
imperatives

Conduct ongoing
research

This is a sequenced list

of the processes, usually Process flows
Route Load Outbound
uploaded to Microsoft planning building logistics

Dynamics Lifecycle Services

(LCS) and synchronized with
Azure DevOps.

Inbound Wave
Receiving Put away Count Pick Pack
logistics processing

Planning

Implementation Guide: Success by Design: Solution architecture design pillars 113

 ▪ Requirements management  Every functional and nonfunctional
requirement needs to be linked to at least one process. If an
applicable process doesn’t exist, you must slot it into the relevant
section of the process architecture. DevOps is a good tool to use
for this success measure, as shown in Figure 6-3.
Fig.
6-3

Requirements
traceability matrix (RTM) Fit gap assessment

This links requirements throughout This is where you mark which

the validation process, and requirement and respective
is a key deliverable for process can be addressed by
requirements management. the system, and which ones
require customization.

Fig.
▪ Solution design  Once you have a better end-to-end business
6-4
understanding, it’s time to translate it into your system processes.
People With the Dynamics 365 solution design, a Microsoft partner works
with the process leads to develop the solution, and it can be helpful
Solution to create process flows that show how the processes can be run in
design the system.
s

ta ▫ As part of the process, data goes in as input exchanged

Da

se
es

oc
Pr between people and applications, and data goes out as output
in the form of documents, analysis, and reports (Figure 6-4).
▪ Test management  Once the solution is designed and developed,
the process architecture, along with the RTM, establishes a baseline
for testing. When you test every process, you ensure that every
requirement is addressed and the solution is appropriate. For more
information, refer to Chapter 14, “Testing strategy.”
▪ Training  Process architecture also defines and drives your
training content. You can use your process flows and guides as
a first draft for your training materials. Learn more in Chapter
7, “Process-focused solution,” and Chapter 2, “Success by
Design overview.”

Implementation Guide: Success by Design: Solution architecture design pillars 114

 Key deliverables
related to people People
Organizational structure
Even though not every process or activity is performed by a person,
Cross-functional there’s always a step or output that involves people, whether internal
process flows and maps
(employees and contractors) or external (customers and vendors). This
List of business users is the key reason why and how the solution design is influenced by the
or personas
people pillar. People shape the solution design in many ways, including:
Dynamics 365 ▪ Geographical location
security roles list ▪ Time zones
Mapping between personas ▪ Languages
and security roles ▪ Customs
▪ Internal and external hierarchies
▪ Organizational structure

This pillar of solution design is usually represented by the organizational

architecture, which visually intersects with process architecture in process
maps, and includes:
▪ Geographical structure
▪ Organizational groupings
▪ Line of business
▪ Reporting lines
▪ Segregation of duties

Data
The third pillar of solution design is data.

A system with inaccurate, misleading, or partial data will lead to a failed

implementation, so you must understand your data and how it fits into
the overall solution. With the right data, you can identify actionable and
analytical information that improves business intelligence. Think about:
▪ Master data  This is static information, such as details about
customers and products, that you use and exchange in your
Refer to Chapter 10, “Data management,”
to gain a broader understanding of the business activities.
various data functions you’ll need before
starting a project.
▪ Documents  This refers to documentation, such as sales orders and
customer invoices, that you create as part of your business process.

Implementation Guide: Success by Design: Solution architecture design pillars 115

 Key deliverables ▪ Reports  This is organized or filtered information, such as a trial
related to data balance or aged debt, serving as input or output for your business
activities.
Data governance strategy
▪ Transactions  This refers to recorded business information, such
Data architecture as customer payments and expense reports, that you create, store,
and report as part of your business process.
Data migration and
integration strategy
At a minimum, you must start with an understanding of how your data
Data quality strategy is currently utilized and how it supports your future vision. Planning
your data strategy upfront also allows for better scalability.

The world is evolving from silos of data to connected enterprise data

that enables the digital feedback loop. That is why a data strategy that
applies AI and analytics to make the data actionable is so critical to the
overall success of your design.

Solution strategy
Your solution strategy is a consolidated view and approach that defines
your overall solution. A solution blueprint is a living document with sev-
eral review points (Figure 6-5) during a project’s lifespan to help you
identify and take necessary actions, mitigate risks, and resolve issues as
they arise. In the Success by Design framework, the blueprint is consid-
ered essential for a project’s success, and provides a view of the overall
solution architecture and dependent technologies.
Fig.
6-5

Kickoff Go live
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52

Requirements

Design/build iteration

Stabilization

Go live prep

Solution SBR 2 SBR 3 SBR 4

Blueprint
Review ALM workshop Performance workshop Go live review
(SBR) 1 Environment workshop Test workshop Cutover plan workshop
Scheduling workshop Gap workshop
Customer kickoff BI workshop
Partner kickoff Integration workshop
Data migration workshop

Implementation Guide: Success by Design: Solution architecture design pillars 116

 The Solution Blueprint Review workshop is a mandatory part of the
solution design experience, and helps you keep track of your solution
design’s progress to ensure that your vision and objectives are still viable.
For more information and a list of activities
in each Success by Design phase, refer to It also allows solution architects and the implementation team to review
Chapter 2, “Success by Design overview.”
and gain an understanding of your:
▪ Program strategy
▪ Test strategy
▪ Business process strategy
▪ Application strategy
▪ Data strategy
▪ Integration strategy
▪ Intelligence strategy
▪ Security strategy
▪ Application lifecycle management (ALM) strategy
▪ Environment and capacity strategy

Capturing these details helps you understand the project, and validates
your solution design via the complete and well-organized processes, data,
and people enabled by Dynamics 365.

Technology
While technology does not drive all the processes, it provides the
backbone of products and services required to fulfill your businesses
strategy. Your processes dictate which technology to use and when, and
that technology will bring your digital transformation to life. For exam-
ple, the Dynamics 365 Sales app provides value-added details about
your organization’s lead-to-opportunity pipeline. Other examples of
technology that’s under the hood to support users include:
▪ Infrastructure as a service (IaaS)
▪ Software as a service (SaaS)
▪ Integration tools
▪ Business intelligence tools
▪ Azure AI
▪ Azure Machine Learning
▪ Data connectors
▪ Customer portals
▪ Mobility solutions

Implementation Guide: Success by Design: Solution architecture design pillars 117

 The technology and system selection are usually performed during the
sales process as part of the first high-level fit gap assessment.
For more on this topic, refer to Chapter 15,
“Extend your solution.”
Finding the right balance
For more information about the Success
by Design framework, refer to Chapter 2, It is easy to say that technology can be used to solve any business process
“Success by Design overview.” scenario, but technology is simply what turns the gears for your processes,
people, and data. It’s important to include all stakeholders in the design
decisions, because processes don’t always align with the technology. This
is what we call the “gap” in a requirement, and it’s where you might decide
to configure or customize an out-of-the-box solution to meet the needs of
the process. Think through the different options to find the right balance for
your solution, but be sure to challenge the business requirements before
customizing the system. You might be surprised to find out that what
seemed like a “must-have” customization is more of a “nice-to-have” feature.

Methodologies
A methodology is a collection of methods used to achieve predictable
outcomes. Good methodology also demonstrates why things need
to be done in a particular order and fashion. The Success by Design
framework is a higher-level abstraction through which a range of
concepts, models, techniques, and methodologies can be clarified.
It can bend around any methodology, including The Open Group
Architecture Framework (TOGAF), the Microsoft Solutions Framework
(MSF), and the Zachman Framework.

To achieve your solution, the workflows and delivery need to be

planned, organized, communicated by various methods, including:
▪ Project management
▪ Change management
▪ Governance and control

For more detailed information about Project management

project management approaches, review
Chapter 5, “Implementation strategy.” There are multiple project management methodologies and approaches,
and methodology for enterprise resource planning (ERP) is a topic that

Implementation Guide: Success by Design: Solution architecture design pillars 118

 can generate more heat than light. Partners often rely on their own
branded methodology to reassure customers about their governance.
Project management approaches can be grouped into three
core categories:
▪ Waterfall
▪ Agile
▪ Hybrid

As we move increasingly toward a low-code world, there are faster

deployments and fewer customizations, and your approach should
match the changing world and your customers’ expectations.

Change management
Project management (the “how”) and change management (the
“who”) are both tools that support your project’s benefits realization.
Without a change management plan in place, your organization’s
objectives are at risk. To drive adoption with your end users and
accelerate value realization, apply a vigorous change management
approach, which is most effective when launched at the beginning of
a project and integrated into your project activities.

The benefits of change management:

▪ It focuses on the people side of organizational change.
▪ It seeks individual and organizational perspectives.
▪ It requires action and involvement by leaders and managers
throughout the organization.

Governance and control

“When you have
The term “governance” often brings to mind just project management
bad governance, or possibly even stakeholder management, but such a limited view of
resources are governance may cause an organization to ignore core elements
destroyed.” required for success (Figure 6-6).

– Wangarĩ Muta Maathai, Every project component contains an element of uncertainty and
Nobel Peace Prize winner is based upon assumptions made before details are known or fully
understood. So, if we know this, why would we expect a project

119
 manager to be solely responsible for project governance? The
project manager is of course accountable for the project outcome,
Read more in Chapter 5, but 360-degree governance requires everyone—especially the
“Implementation strategy,” and
architects and leads—to play a role.
Chapter 8, “Project governance.”

A 360-degree governance method directly affects how well the

project team performs and the ultimate result, and is executed on
multiple levels:
▪ A governance board or steering committee has final
responsibility for meeting the project goals. It typically is
comprised of representatives or stakeholders from each of the
major participants.
▪ A project manager leads the team and administers project elements
that include alignment, risk identification and mitigation, escalation,
and cross-Microsoft collaboration.
▪ All project leaders, including the project manager, solution
architect, and technical architect, closely communicate and align
toward the common goal of solution deployment.

Fig.
6-6

Governance domains
Project Stakeholder Solution Risk and ​issues Organizational change
management engagement management management Change control and communication

▪ Weekly status ▪ Executive steering ▪ Oversight of ▪ Objective risk ▪ Identification and ▪ Intentional and
reports​ committee​ functional and assessment and prioritization of customized
nonfunctional mitigation planning​ requirements and communication to
▪ Project plan and ▪ Communication
attributes solution attributes​ stakeholders​
schedule​ and promotion​ ▪ Issue identification,
▪ Data, integration, ownership, and ▪ Tracking and ▪ Periodic
▪ Resource ▪ Escalation of
infrastructure, and resolution reporting via ADO effectiveness check​
management​ unresolved issues​
performance
▪ Prioritization and ▪ Organization and ▪ Solution adoption
▪ Financial ▪ Solution decision
▪ Customizations and severity assessment​ process changes​ and operational
management​ making and
process adaptation​ transformation
collaboration​ ▪ Risk response
▪ Risk management​ management​
▪ Security and management​
▪ Business alignment
▪ Change compliance​
and tradeoffs​
management

▪ Project status ▪ Project change ▪ Architecture board​ ▪ Project status ▪ Statement of work​ ▪ Communication plan​
meetings and reports​ management​ meetings and report​
▪ Governance ▪ Project artifacts​ ▪ Stakeholder
▪ ADO (project tools)​ ▪ Solution framework ▪ ADO (project tools) engagement​
▪ Project status report​
management​
▪ Governance ▪ Organizational
▪ Governance
framework ▪ Organizational change management
framework
change management

Implementation Guide: Success by Design: Solution architecture design pillars 120

 ▪ The technical architect leads and orchestrates the technical
architecture across all application components to ensure
optimal performance.
▪ The solution architect guides users toward the new vision
and transformation.
▪ The change manager ensures all internal and external
communications are in place.

Conclusion
As we’ve explained in this chapter, solution architecture follows a
systematic approach that identifies your desired solution and the
building blocks needed to construct it. Solution architecture design
takes your business vision and breaks it into logical sections that
become a blueprint for building your solution. Here is a list of the key
steps for solution architecture design:
▪ Define your vision.
▪ Create a business strategy.
▪ Outline your business processes.
▪ Determine how people are connected in and around
your organization.
▪ Know your data.
▪ Develop a solution strategy using technology and tools.
▪ Apply project management, change management, and
governance and control methodologies.

Implementation Guide: Success by Design: Solution architecture design pillars 121

7 Guide
Process-focused
solution
Implementation Guide: Success by Design: Process-focused solution 122
 Introduction
Fast changing technologies and markets demand a
rapid rhythm of delivering products and services.
Business applications improve inefficient processes by automating,
optimizing, and standardizing. The result: A more scalable
business solution.

An approach based on business process mapping and management in an

implementation project is fundamental if organizations are to achieve their
In this chapter we cover the goals. When implementations use this approach, there is more consistency
following topics related to a
in project phases, communication flows better, and processes run better.
process-focused solution:

Start your implementation project

with business processes This chapter introduces the importance of a process-focused approach
when implementing Microsoft Dynamics 365 Business Applications. The
Opportunity for optimization
goal is to emphasize the benefits of employing business processes as the
Defining the scope of
primary framework for a project implementation cycle.
the implementation

Defining your requirements

Fit to standard and

fit gap analysis Start with
Implementation lifecycle
connected to processes business processes
A solution that helps you to
operate your business
Business processes are the main drivers to start defining the solution
implemented in your project.

Implementation Guide: Success by Design: Process-focused solution 123

 The business language
Every organization has an internal language they use to describe daily
operations. This language is immersed in business processes and is
often largely based on terminology common to the industry. When
implementing business applications like Dynamics 365, this business
language is the best, most familiar way for the business to think about
their needs and the tools and technologies that they use for work. This
business language is not simply terminology. It includes the processes
used to conduct day-to-day transactions and activities. Curating and
utilizing these processes in the language of the business is important to
keep sight of the basic truths of business requirements. It also reduces
the risk of the implementation team getting lost in the jargon and
nomenclature of the technical tasks involved in the implementation.

When the project starts, putting the business process view of the
organization at the core pays dividends.

Start with a business process future vision

Creating a vision of how a business is transformed through technology
makes it is easier to focus on the activities needed to achieve that goal.
This vision needs to include the understanding of how the business
currently works and the state that it wants to reach after.

Successful implementations start with a clear understanding of an

organization’s business model. That includes understanding how the
organization creates value through its products and services, as well
When discussions about the project and as its relationships with customers and suppliers. There are multiple
business processes begin with third parties,
business processes strategically aligned to make that business model
it’s helpful for those who may not be
familiar with the organization’s business to work. Chapter 4, “Drive app value,” describes how to define a business
begin learning this language.
model and the importance of connecting the business model to business
processes, and the business processes provide the baseline to draw the
roadmap of your solution for the digital transformation.

The processes derived from the business model provide insight on

what it takes to reach the desired future. Business processes define the
functional scope of the solution being implemented as illustrated in

Implementation Guide: Success by Design: Process-focused solution 124

 Fig.
Figure 7-1. Any set of business processes being implemented needs
7-1
definitions before you can start working on requirements and other
Requirements
aspects of the solution.
Business
process
When defining your business model, assess the state of your processes.
Some organizations may want to use new technology to optimize existing
processes through automation. Others are moving to e-commerce or
outsourcing parts of the operation, which requires new business processes.
Solution
Still others might be moving to lean manufacturing.

Many organizations also seek to consolidate operations and introduce

standard processes across different parts of the business. Understanding
the state of the processes allows you to understand how to reach your
Business process model goals. The effort required to implement an existing process using new
technology is very different than what is required to define new processes
or gain consensus from stakeholders.

New business processes won’t always be the same across a business.

One division, for example, might decide to ship inventory directly to
customers rather than to a warehouse, while others continue to use a
warehouse. On the other hand, it may be that the model for employee
purchasing is the same, so a shared product catalog and ordering
Start your implementation process is implemented. In any case, business model analysis, process
project with business processes
reengineering, and standardization strategies should be considered
Opportunity for part of the project definition. That firmly anchors the processes to the
optimization business strategy and also helps generate process-based project goals.

Defining the
scope of the
implementation Opportunity for
Defining your
requirements
optimization
It is easier to find good opportunities to optimize your process using
Fit to standard and
fit gap analysis business process mapping.

Implementation
lifecycle connected
to processes Mapping processes
A solution that helps you to Processes are the heart of every business. They describe how the business
operate your business operates. The fact that your products or services reach your customer

Implementation Guide: Success by Design: Process-focused solution 125

safely and on time is due to a series of tasks executed by different roles
and different business units. Business mapping represents these steps.

Baseline business processes are called “as-is” processes. Depending on

the organization, these may be legacy processes or they may be new
processes engineered as part of a new solution or business model.

Regardless of where an as-is process set comes from, it is critical that

they are described at the appropriate level. As-is processes that con-
tain details about how current implementations are in legacy systems
introduce a risk of reimplementing the legacy system. Keeping as-is
processes at the more abstract level helps drive the scenarios in scope
and enables the team to focus on the desired business outcomes.

It’s worth reminding ourselves that the objective of as-is business

process mapping is not to define step-by-step how the solution is
designed in the new solution.

The goal of mapping the as-is business process is to:

▪ Provide structure/skeleton for the definition of the scope of
the project.
▪ Express the project aims in the natural business language of processes.
▪ Enable a natural way to define the main end-to-end business
scenarios (process variants).
▪ Provide the best language to identify and highlight areas for process
innovation and improvements, as well as risks and constraints.
▪ Become the basis for mapping to the equivalent process design
within the new system.
▪ Create a workable definition of the project that the business can
directly interact with without going to the level of individually
written requirements.

The value of having business processes mapped early is that it allows

everyone involved to understand how the processes operate by using
them as a common business language. This helps both third parties
and the organization agree on processes across multiple departments
and organization layers. It is common for senior managers to be
surprised by how processes actually work on the ground.

126
 It is essential that the right team is involved in mapping business
processes to the new technology. This team should be guided by the
business stakeholder who owns the process in the business opera-
tion. This individual is charged with making the right decisions about
structure and investment in the process to achieve their goals. The
stakeholder is supported by one or more subject matter experts who
are familiar with how the business process operates in the real world.
These experts can provide depth around the various scenarios under
which a process is executed. Their knowledge, coupled with a mindset
open to new ways of working, helps drive the right conversation on
process mapping.

Here’s an example. Let’s say a company is looking to move to a B2B

business model via e-commerce and seeks to deploy their telephone
salespeople strictly to consumers. There may not be existing patterns
within the business to replicate, so the new processes should be based
on expected outcomes rather copying legacy steps. This mindset
should be applied to all processes, as the key deliverables should be
the expected “business scenarios” and the desired business outcomes.

When reviewing new process diagrams, the team should consider

the following:
▪ Do the business SMEs recognize their business in the processes?
▪ Do they provide adequate coverage of the business activities in scope?
▪ Do they show the key interconnections between the processes?
▪ Do they show meaningful end-to-end business flows?
▪ Will they represent a good level at which the project team can
communicate with business stakeholders on risks, progress,
decisions, and actions?
▪ Do they illustrate processes sufficiently to identify opportunities
for improvement?
▪ Are they sufficiently well-defined to be able to be used in the next
stage where the processes can be applied to drive the design in the
Dynamics 365 system such that they can reflect the improvements
and the project goals and vision?

These process diagrams help define the baseline for the organization’s
current business and are a good place to start when plotting optimization

Implementation Guide: Success by Design: Process-focused solution 127

 objectives. They should help to identify the processes or parts of processes
that are not efficient or well understood.

The key to getting the processes mapped well and quickly is to ensure
the following.
▪ The right people are in the room to direct the mapping.
▪ Any mapping software or tools that help facilitate the rapid definition
of the process and do not slow things down.
▪ The level to which the process is described is directly proportional
to its importance to the business. For example, once the end-to-
end scope of a widely used and very standard finance process is
defined to a certain level it should not be further broken down to
step-by-step processes.
▪ The process mapping effort is interactive, not a series of documents
that go through long handoffs and approvals.
▪ The tools for the mapping can be anything from sticky notes, to
Microsoft Visio, to specialized process drawing software. The pro-
cess maps are visual and not hidden in wordy Microsoft Excel lists.

This is a vital step that needs to be executed early. Plan so that processes
are defined as fast as possible using the simplest tools in the workshop,
like sticky notes, pens, and whiteboards. The drawings and comments
can then be transferred offline to more sophisticated applications.

Remember that doing this mapping at the start of the project

implementation is important. It could become a fundamental part
of helping the analysis of the opportunities for process improvement
and creating the right baseline for the project.

Modeling your solution for the future

Digital Digital transformation is fast paced. Processes that work today may not
transformation is fast remain competitive tomorrow. Technologies like the cloud, artificial
paced. Processes that intelligence, machine learning, robotic process automation, virtual

work today may not reality, the Internet of Things, big data, virtual conference rooms, and
many others create possibilities that did not exist a few years ago.
remain competitive
tomorrow. When reviewing your business processes, it is important to keep in

Implementation Guide: Success by Design: Process-focused solution 128

 mind how innovative technologies can improve process outcomes and
collaboration. Similarly, there may be multiple areas in your processes
which would benefit from modern options now available. Technology is
For example, the Microsoft
HoloLens is being introduced to a key differentiator. Remember, chances are that your competitors may
some manufacturing processes with
well be considering these options and looking to incorporate them.
Dynamics 365 applications. Using the
HoloLens, the user can understand how
to operate a machine. Or if a machine
is broken, and support is on the other
side of the world, virtual reality enables
Standardize your features with a business
collaboration between the user and
support to repair the machine.
processes approach
Earlier, we discussed how a business model analysis can introduce
Or if you have a workflow with
approvals that require a signature, opportunities for process reengineering and standardization.
you could add eSignature software to
automate the process and reduce the
time for approvals. Assuming there is a good plan for process standardization grounded
in the business strategy, this standardization can be aided with a new
business application system.

Of course, standardization across business units assumes that the

business models are sufficiently similar and there are no local market
constraints or competitive reasons for there to be differences.

Various benefits can come from standardized processes assuming the

business models are sufficiently similar between the business units so
that the other barriers to standardization are low.
▪ Enables reduction in cost by reducing the variations of business
processes that need to be implemented, trained, and maintained.
▪ Facilitates comparative analysis across multiple groups of users
that are executing the same process.
▪ Facilitates mobility of resources within the organization through
standard processes to deal with changes in demand.
▪ Promotes the adoption of best practices across the organizations.
Optimization for anyone improves outcomes for everyone.

Standardization also helps with rolling out the implementation to

other business units as it reduces the time and effort compared
to building a process from scratch. After the first implementation,
there is a proven case of it working within the business. The next
business unit to be implemented can discuss the pros and cons of
the standard process with business peers.

Implementation Guide: Success by Design: Process-focused solution 129

 Similarly, it is important to confirm that the constraints and implications
are identified and well understood prior to starting design work.
▪ Standardization of a process can bring benefits at a group level,
but may be seen as additional work by a local business unit.
▪ Standardization can require significant change management.
▪ Is the standardization being defined at an industry level and then
implemented in the system? This can create a conflict if standard
processes are not a natural fit with Dynamics.
▪ Will the process be designed in Dynamics 365? And will these
processes become the new standard?
▪ Is the entire end-to-end process to be standardized or will there
be areas and functions for local variations? If so, does that align
with the standard capabilities in Dynamics, or could it need
customizations?
▪ Does the business case for standardization remain sufficiently
strong after all the constraints and implications are balanced
against the benefits?

In multiple, phased rollouts, process standardization often takes the

shape of a “core template” with the expectation of rolling out this
“template” to subsequent business units.

Creating process maps is essential to identifying the opportunities and

Start your implementation to realizing the potential benefits outlined previously. The use of the
project with business processes
common business language of processes is the vital mechanism that
Opportunity for optimization brings all of these aspects into focus.

Defining the
scope of the
implementation Defining the scope of the
Defining your
requirements implementation
Fit to standard and
fit gap analysis A business application implementation is essentially the delivery
of a new capability for end-to-end business transactions using the
Implementation
lifecycle connected application. Processes are the foundation for the definition of the
to processes solution scope. Processes embody many of the properties that make
A solution that helps you to for a good scope definition.
operate your business ▪ Business processes are well understood by the customer as they

Implementation Guide: Success by Design: Process-focused solution 130

 are expressed in the languages they know best.
▪ Processes are richer in context than a requirements list.
▪ Processes show the connections between different business units,
roles, functions, task-level steps, and systems.
▪ They are naturally hierarchical and can be expressed in different
levels of detail, depending on their intended purpose.
▪ Processes are the connectors that show the data and people
feedback loop.

Process mappings can be collected in a process catalog. This

catalog helps turn the visual process diagram into data. The business
processes in the catalog can then be numbered, labelled, and uniquely
referenced. Process hierarchy can be managed as part of the data
structures. Typically, these can then be managed in Azure DevOps
(or similar tools) where they can then be assigned to various tasks
such as configure, build, design, or test and to different project roles.

Business processes are usually interwoven with other business

processes. These interconnected processes can span systems, and
the connection can be a simple endpoint or an entire subprocess
performed in the external system. If the latter, it should be consid-
ered for your solution scope. Where the process is in the external
system is directly relevant to the end-to-end process, include the
full process in the process mapping. This not only helps in under-
standing the process, but also with making better decisions on
customizations and selecting ISVs, which are also stated in your
solution scope.

Using process mapping as part of the definition of the scope also

helps when working with implementation partners. The process flows
helps express the ultimate outcomes for the project and having them
reflected the statement of work can help with a more comprehensive
definition of the project that results in fewer change orders.

Keep in mind that business processes are always subject to change.

As you move forward you are enriching the business process flows,
and possibly adjusting the solution scope. That is exactly the flexibility
that a business process flowchart offers.

Implementation Guide: Success by Design: Process-focused solution 131

 Fit to standard and fit
Start your implementation
project with business processes

Opportunity for optimization

gap analysis
Defining the scope of the
In many projects, inertia drives the old ways of working. There are several
implementation
dangers in assuming that what works now also works just as well in a
Fit to standard and new Dynamics system. Recreating processes based on legacy systems
fit gap analysis (often very old systems) can lead to the following:
▪ Unnecessary and excessive customizations and costs related to
Defining your requirements
design, coding, testing, training, documentation, etc.
▪ Stagnation in the business as the natural opportunity for innovation
Implementation
lifecycle connected and improvements is lost
to processes ▪ Creating unique, siloed, inefficient processes when Dynamics 365
could provide more standard and recommended ways of working
A solution that helps you to
operate your business based on the learnings of thousands of customer organizations
▪ Damaging the natural usability engineered into Dynamics 365 by
forcing processes designed for legacy software into Dynamics
▪ Creating processes within Dynamics that were not designed to be
used in that way and therefore reducing the ability to use other
related functions because the processes are no longer a natural flow
or fit for the purpose
▪ Reducing the ability to immediately use the documentation, training,
sample code, and sample data available for the Dynamics system
▪ Reducing the ability to leverage market and industry knowledge,
insights, and experience in Dynamics 365 for the areas of interest for
the business
▪ Reducing the ability to directly apply tools and apps in
the marketplace
▪ Reduced capacity and resources available to apply customizations
that can provide significant business value as the available budget
and resources get used on customizations that do not add the
same value
▪ Creating barriers for end-user developers (citizen developers) to
create rapid, low-cost application using the Power Platform by
creating a more complex, customized data model/process model

When it comes to understanding exactly what is going to be implemented

in your future solution, it is necessary to perform a deep analysis about

Implementation Guide: Success by Design: Process-focused solution 132

 the current processes and what needs improvement. To determine what
functionalities of your business applications to keep, and what others
could be built or purchased from external vendors, we suggest under-
going the following analysis. We suggest this because it helps to have a
standard solution, with minimal customizations, which minimizes costs
and therefore maximizes value to the business.

Adapting to the standards of new system

It is recommended to start with a fit-to-standard approach for every
project. As part of the definition of the scope there should be a business
process catalog created in the early stages of the project. This can now
be used as the set of processes that can be configured and enacted in
Dynamics 365. Of course, initially, not all the configuration is completed
perfectly. However, starting with the core business processes and iterating
until there is a good, high-level understanding of how the processes could
be implemented within the system helps create the solution blueprint.
Putting the fit-to-standard approach at the front of the analysis helps
to set the right culture of “adopt wherever possible, adapt only
where justified.”

In most projects this analysis is not starting from a blank piece of paper;
the implementation partner has conducted some part of this analysis as
part of providing estimates on cost and time and a solution overview.
It is recommended that the customer project team get deeply involved
in this process to confirm that the key business processes were correctly
interpreted and to start gaining knowledge of the standard processes
embedded within Dynamics 365.

The continuation of the fit-to-standard leads to the definition of

requirements, fits, and gaps, but this is from the perspective of leveraging
the processes enabled within the system first rather than trying to recreate
the legacy system in Dynamics 365.
Processes are the
foundation for the The advantages of starting with this approach using the process catalog
are as follows.
definition of the ▪ It promotes the reduction of customizations by supporting the deliv-
solution scope. ery of the underlying business needs through configuration rather

133
 than going directly into fit gap analysis with detailed require-
ments that may be influenced by the existing or legacy system.
▪ It helps reduce the risk of missed requirements as the evaluation
of the fit with the new system is based on the richer and broader
context of business processes. As these business processes are the
natural language of business users, their evaluation is more com-
prehensive, meaningful, and effective compared to working with a
list of requirements.
▪ The process catalog can direct the fit-to-standard assessment
by working iteratively through the processes, starting with the
higher-level core processes, and then working with the more de-
tailed sub processes and variants. This also helps the business users
more clearly see how well their underlying business requirements
are being met within the system.
▪ The project is more likely to adopt modern recommended
standard processes embedded in the system.
▪ It creates higher-quality solutions as the processes are tested
by Microsoft and are more likely to be market-tested by others,
whereas custom developments and variants, especially complex
ones based on the legacy system, will need to be specifically
validated by the customer.
▪ The standard solution allows for more agility in adopting related
technologies and by keeping to standards where possible, make it
easier to add the real value-add custom extensions.
▪ It enables a faster delivery of the new solution; there is no need to
wait longer for a custom solution to be developed.
▪ Standard processes are more easily supported by internal and
external support teams, including Microsoft Dynamics Support.

The benefits of staying within the standard product wherever possible

are clear. The implementations following this approach, often called
vanilla implementations, adopt the Dynamics 365 system with its
standard configuration as shown in Figure 7-2.

Some businesses may have specialized business processes or an

innovative idea to improve their competitiveness in the market. In
such cases, a process-centric description of the system allows these
improvements to be adopted with more transparency.

Implementation Guide: Success by Design: Process-focused solution 134

 Fig.
7-2
Dynamics 365

Process Process Process Process

Start End
step 1 step 2 step 3 step 4

Gap analysis
As discussed in the previous section, adopting a process-centric
solution within Dynamics 365 has clear benefits. However, there may
be specialized functions that are not part of the standard solution as
shown in Figure 7-3. That is identified with the fit gap analysis. After
having the configurations set, you can look for gaps in the processes
and make a decision whether to customize.

Extending a solution can vary in terms of time, cost, and complexity.

There are many things to consider when extending. But it offers the
benefit of a solution tailored to business needs. See Chapter 15, “Extend
your solution,” for details on extending a solution. Here we look at the
implications of taking a process-centric approach.
▪ Some extensions can bring innovation into business processes. They
can add business value and provide good, long-term benefits.
▪ It is important to consider potential customizations in light of the
rapid pace of innovation in Dynamics 365 SaaS applications. New
features are released regularly and keeping an eye on the Dynamics
release plans avoids creating extensions that are quickly redundant.
▪ When evaluating a potential custom extension, using the process
maps can help determine the impact, not just on the process being
directly changed but also the impact on connected processes.
▪ When looking at potential solutions for requirements that cannot be
met directly in the standard system, consider the various different
Fig.
7-3
Dynamics 365 methods available:

Process
Process Process Process End
Start step 3 with
step 1 step 2 step 4
Extension

Implementation Guide: Success by Design: Process-focused solution 135

 ▫ For simpler designs, consider citizen developer-led applications
using the Power Platform, which can generate large value very
rapidly and at a low cost.
▫ For more complex or global solutions, some combination
of citizen and professional development through the Power
Platform or other technologies may be a more effective and
productive route. Using the process catalog can help provide
an excellent backdrop to communicate and collaborate
between the project team, the citizen developer, and the
professional developer.

Third-party solutions
An alternative to extending is to buy an existing solution from a
third-party vendor, also known as an independent software vendor
(ISV). This option is more common when there is a significant gap
and developing a functionality in-house can be complex and costly.
Sometimes you don’t have enough resources, budget, or expertise to
develop such a significant solution.

ISV solutions tend to focus on more specialized process areas,

normally covering some end-to-end industry-specific process. In
addition, ISVs provide specific functionality to cover gaps, as illustrated
in Figure 7-4. Successful ISVs have expertise and experience in the
vertical industry processes they cover and can therefore add value to
your business processes.

Dynamics 365 has been designed to meet standard business processes.

It enables projects to adopt the standard application more easily and
Fig.
7-4

Process 4 in
external app
or ISV
Dynamics 365

Process
Process Process Process End
Start step 3 with
step 1 step 2 step 5
Extension

Implementation Guide: Success by Design: Process-focused solution 136

 minimize the number of perceived gaps. It also has the flexibility to
customize and integrate external applications. Rather than choosing one
or the other, buying some third-party solution along with a standard
implementation can build an optimal solution.

Representing the fits and gaps in a business process map in terms of the
processes drives a better understanding within the implementation team
and the organization. This is also needed to enlist the requirements for
the design of the future solution.

Defining your
Start your implementation
project with business processes

Opportunity for optimization

requirements
Defining the scope of the
implementation Some projects start with lists of requirements derived as a set of
atomic functional requirements, but without reference to mapped
Fit to standard and
and documented business processes this can pose risks like the following:
fit gap analysis
▪ Late discovery of missing process steps
Defining your ▪ Misinterpretation of the requirement by parties not experts in the
requirements related process
▪ Lower business engagement in the project
Implementation
lifecycle connected ▪ Longer review cycles as the requirements and designs are less clear
to processes

A solution that helps you to Business process mapping helps draw the as-is processes to understand
operate your business how the business is running right now, and the to-be processes to show
how they work in the future. This also emphasizes the importance of
business process mapping early in the project.

After determining what you keep or build, create the list of requirements,
aligning them to the business processes for an accurate picture. You can
use tools equipped for the traceability of the requirements.

Typically, in a process-focused solution, once the processes have been

defined and agreed upon, the detailed requirements are derived and
relevant designs mapped. There are tools within Dynamics 365 to help
with managing the requirements. For example, Azure DevOps can
work to create and track requirements, and if you are implementing

Implementation Guide: Success by Design: Process-focused solution 137

 Dynamics 365 for Finance and Operations, you have the Business
Process Modeler (BPM) as part of your lifecycle services (LCS).

The list of requirements is easier to be crafted starting from the business

process mapping and it can be iteratively revised and refined.

Process-centric
Start your implementation
project with business processes

Opportunity for optimization

implementation lifecycle
Defining the
scope of the The value of taking a process-focused solution does not end with a
implementation better definition of requirements. Adapting all the activities in the
Defining your subsequent phases to be more process-based helps deliver better
requirements outcomes, regardless of underlying methodology.

Fit to standard and

fit gap analysis At go-live, a business application such as Dynamics 365 sees system
users performing tasks and activities as part of a business process.
Process-centric The users are not executing on gaps, fits, or isolated requirements.
implementation It is important to keep this in mind when thinking of the full project
lifecycle lifecycle. The ultimate test of the project success is when the system is
A solution that helps you to in operation and the design build testing actions are intermediate steps
operate your business towards that goal. It is recommended to drive these phases with the end-
to-end business process as the framework upon which all the related
activities are planned, performed, and measured.

See the prospect to cash with Dual-Write integration to high-level

end-to-end process map in Figure 7-5.

Design
When creating the solution blueprint, the vision of the overall solution
architecture gives solidity and certainty to the project scope. This is the
transition from scope as business processes to scope as designs within the
system. As it is expressed in business language, it helps to ensure that busi-
ness users can be usefully engaged in the review and approval of the scope
and proposed solution. There are, of course, other views of the solution
architecture in terms of data flow, systems landscape, and integrations.

Implementation Guide: Success by Design: Process-focused solution 138

 Fig.
7-5 Dynamics 365 Sales

Product and
Account Contact Quote Order Invoice
pricelist

Dual-write

Customer Contact person Product Quote Sales order Invoice

Finance and Operations

As part of the design phase in a Waterfall methodology, or during
design work in sprints in a more Agile methodology, breaking down the
end-to-end business processes into meaningful subprocesses provides
manageable, bite sized units of work. It also allows better communication
of the project tasks with business users.

The specific user stories, or configurations, and build work at a task

level can be related to the subprocesses so that they have the correct
business context. This makes it easier for business users to better
understand the design and the business reviews and approvals of
designs are meaningful.

If the business processes are collected in a structured and hierarchical

process catalog, this catalog can be used to do the following.
▪ Help more easily and logically determine the sequence of work
starting with the more foundational process designs before tackling
the lower-level designs.
▪ Help highlight the interrelationships and dependencies between
the different business value streams, departments, roles, and
functions—this helps with designing a more cohesive and
integrated overall solution.
▪ Help better manage the distribution of the actions and activities by
leveraging the sequence and interrelationship of the designs in scope.
▪ A process flow can directly be mapped to the system design from
configuration through data setup and functional run through.

Implementation Guide: Success by Design: Process-focused solution 139

 ▪ Help deliver working processes in Dynamics 365 software, by
implementing processes, including end-to-end process as early as
possible, and by taking a process-centric view.
▪ Help provide a better understanding of the solution by looking at
the current capability of the emerging system to execute end-to-
end processes.
▪ View the system in the context of a process, reducing unnecessary
customizations .Better engage the business subject matter experts
(SMEs) and business stakeholders by using the business language
of processes.
▪ More rapid delivery of working software that better reflects and
makes better use of the Dynamics 365 SaaS cloud world.

Deliver processes in Dynamics 365 as early as possible, taking a

process-centric view. The benefits are practical and real, and often the
only real barrier to taking a process-centric view of design tends to be
the legacy of previous, more technical, and custom development-centric
approaches, better suited to the pre-cloud and SaaS world.

Having the high-level process, you can start breaking down the end-
to-end business processes into meaningful subprocesses, as shown
in Figure 7-6.

Development and configuration

A process-centric view provides an excellent basis for converting de-
signs into software. It can help directly address the question of how to
execute the business process within Dynamics 365. The configuration
and development tasks have the background of the business process
context, and do not need to be treated as an isolated delivery of a task
from a long list of such tasks.

This isolation is still frequently seen in projects where the development

and configuration work becomes highly technical and is divorced
from the business process. This activity can become an end in itself,
which does not serve the project well. Instead, when a data- and
process-centric view is taken, the configuration and development
tasks can be planned and managed in service of delivering a process
embedded in Dynamics 365 software.

Implementation Guide: Success by Design: Process-focused solution 140

 Fig.
7-6

Dynamics 365 Sales

Product and
Account Contact Quote Order Invoice
pricelist

Dual-write

Customer Contact person Product Quote Sales order Invoice

Finance and Operations

Dynamics 365 Sales

Add products
Start Create order
to order
Price order

End Invoice order Pack order Pick order Confirm order

Finance and Operations

A process-first view allows the team to collaborate more productively. The

process designs, which now reflect the functions and configuration and
flow within the Dynamics system, help the various teams communicate
the context and better understand the delivery of any development.
Compared to working from a task in DevOps, or from a paper design,
the process background provides a better context and implications for
a build task. For example, taking a process-centric view can help with
the mapping of the access necessary for performing process tasks. That
helps build a picture of the related security roles.

A process-focused view of the emerging solution also helps avoid

the dangers of a primarily “fit gap” approach which can create point
solutions. Instead, with a process-first background, there is a high
probability that the tasks and activities crystalize into a coherent and
well-functioning solution.

Implementation Guide: Success by Design: Process-focused solution 141

 The project can also use the process-focused solution to conduct reviews
of the solution with the wider business to confirm that the project is
delivering on expectations. These reviews are much harder to do well
without the supporting process definition. That’s because without
the processes in focus, the project cannot speak the language of the
business. Many organizations are expecting some degree of business
process transformation as part of the implementation of Dynamics 365.
The reviews of the solution when the new processes are first explained
in business process terms helps ground the subsequent discussion of the
current state of the solution in clear, well understood business language.
This significantly improves the quality and usefulness of the business
review by minimizing the technical implementation jargon and instead
concentrating on the “business of business.”

Issues with resource constraints are well known. Showing incremental

progress of the build and development using business process language
improves the confidence of the wider business in the project. This can be
key for project success.

Testing
A project that has taken a process-centric view reaps the benefits during
testing. Other than the necessarily narrow focus of unit testing and some
non-functional tests, almost all other test types should be rooted in
some form of process definition. When there is an existing set of process
definitions and the associated designs, and a system with the designs
implemented, there are many advantages to the testing methods.
▪ It helps ensure full testing coverage and enables earlier detection
of any missing or inadequately defined functions or processes.
▪ The testing can more easily follow the process-focused path already
created earlier in the project.
▪ Testing has a natural focus on evaluating business process outcomes
which tends to be a closer match to the intention of the design and
eventual production use.
▪ It enables incremental testing of processes and subprocesses
which in turn helps engineer quality into the solution.
▪ Testing end-to-end processes, which is how the system is used in
production, is enabled.

Implementation Guide: Success by Design: Process-focused solution 142

Overall, testing in a business application like Dynamics 365 is a natural
match to a process-focused view of the solution. Having a process-
centric approach throughout the lifecycle allows testing to become the
natural and obvious progression of the project delivery.

See Chapter 14, “Testing strategy,” for more details on the overall
approach to testing.

Training
Training to use business systems like Dynamics 365 applications is
fundamentally learning how to conduct day-to-day business processes
using the system. Most of the functional roles in the system interact
with multiple functions and processes. Rarely do roles exist in a vacuum
of their own process, but instead interact with other processes. Having
business process flows defined and available in the process catalog,
including flows across the seams of system roles, helps both in the
collation of process-based training materials and to guide the training.

During the design process, if the roles are mapped against the processes
as part of security design, they can be directly used for testing and for
generating role-based training.

Even where some of the system roles may be very specifically restricted to
a specialized function in the system, there is often a need to understand
the upstream and downstream processes to perform a function well, and
to understand any implications of any delays or changes in the flow.

If whole project approach is process-focused, including the configure/

build activities, then the process of generating materials for training
(such as Training Guides for Finance and Operations and Guided Tasks for
Customer Engagement applications) goes smoothly, as there is a direct
correlation between the system processes and the training processes.

The process-based training not only helps prior to go-live—it also can
be used with new hires and when users change roles. It allows those
new to the business to absorb the business process simultaneously
while understanding how that process is performed within the system.

143
 Roles can be easily defined in a business process and use the correct
security roles for testing and training as shown in Figure 7-7.

Support
The process catalog created in the project provides more than the
framework for implementation. It can also help with supporting the
solution. Functional issues raised to support usually need to be repro-
duced by the support team to enable the team to investigate the root
causes. A robust process-based definition of the flow allows support
to recreate the steps defined by the agreed standard process flow. The
visual depiction and description of the process allows the support team
to speak a common language with the business user when discussing
the issues.

This ability to place yourself in the business process helps reduce the
number of back-and-forth cycles of communication and the overall
time taken to understand the issue in business terms. It increases the
confidence of the user that their reporting of the issue has been under-
stood. This reduces anxiety and improves user sentiment of the system.

A solution that helps you

to operate your business
To design and build a solution that has a great positive impact is not an
Fig. easy task. There are many aspects to consider, and everything is so
7-7

Create order - E2E process

Dynamic 365 sales Finance and Operation apps

Add products
Sales manager Start Create order
to order
Price order

Sales clerk Confirm order

WHS worker Pick order Pack order

AR clerk End Invoice order

Implementation Guide: Success by Design: Process-focused solution 144

 connected that even missing a little piece can significantly affect the whole
Start your implementation
project with business processes project. The business processes flows are the common playground where
we gather all the implementation artifacts together. They keep all the
Opportunity for optimization solution components and phases aligned. They are the line connecting
the dots and the main instrument for communication that helps to drive
Defining the
scope of the changes on how the business runs and helps to minimize missing those
implementation little pieces.

Defining your
requirements As you have read through the chapter, business processes are involved
during all the journey of the implementation and they also remain useful
Fit to standard and
fit gap analysis for future improvements. They have a perpetual impact in your business.

Process-centric The creation and documentation of business process flows is to prepare

implementation lifecycle
the fertile ground where the seed will be for all those implementation
A solution that helps activities that allow you to harvest optimized and efficient processes, risk
you to operate depletion, cost minimization, role compliance, integral solution, and, as
your business an ultimate goal, a successful implementation.

Implementation Guide: Success by Design: Process-focused solution 145

 Checklist

Business processes focus Ensure the business process definition is complete and
considers all activities and subprocesses.
Ensure the business process view of the organization is
Take advantage of the latest SaaS technology to drive
at the core of the definition of the project.
efficiency and effectiveness for the process optimization.
Clearly articulate the key business processes that are in
Ensure future readiness when mapping your business
scope and the respective personas so they are under-
process to the solution by incorporating configurability
stood by all involved parties in the implementation.
by design.
Ensure business model analysis, process engineering,
and standardization strategies are considered part of the
Fit gap analysis
project definition and deliver a strong process baseline
before implementation starts. Adopt a fit-to-standard approach and align to the
philosophy of adopting wherever possible and adapting
Collect the business processes in a structured and hierar-
only where justified.
chical process catalog during the requirements phase.

Process-centric solution
Use business processes for each phase of the project to
deliver better outcomes (all phase activities are better
planned, performed, and measured).

Opportunity for optimization

Explore opportunities to evolve, optimize, and consolidate
your processes as part of the implementation to meet
the digital transformation goals and drive user adoption.

Implementation Guide: Success by Design: Process-focused solution 146

 Case study
The journey to a
process-focused solution
A large, multinational organization that provides contract-based
servicing with some manufacturing and distribution services was
planning to replace parts of its 15-year-old tier-one ERP and CRM
systems and other related legacy systems.

The wrong road

The project started with the implementation process that they had
traditionally applied to other IT projects. The traditional approach was
strictly Waterfall with a heavy documentation bias. The initial phase of
requirements gathering was conducted by using the old systems as the
reference point. A very long list of requirements was gathered, with the
assistance of an implementation partner, across the various functions.
The requirements gathering phase was extended a few times as the
requirements review took significantly longer than expected.

The reviews needed multiple parties to read, understand, and validate the
requirements and there was a concern among the business approvers that
if they missed any requirement, or even a nuance, their system would not
function well.

So, there were multiple document-based reviews where comments

Implementation Guide: Success by Design: Process-focused solution 147

 from the approvers were formally submitted and the project team
wrote responses to them, which in turn generated even more
comments from the approvers. The spiral of these reviews finally
concluded, but the business was not entirely convinced that they fully
understood what they had approved and hoped that the design phase
would provide more clarity.

The design phase was similarly based on writing and reviewing complex
design documents and was running late. The focus tended to be on the
“gaps” identified, and as these were not always placed in context, the
discussion with business users was not always productive and was often at
cross purposes. As further delays accrued and the business users were
becoming even less confident about the proposed solution, the
stakeholders decided to review the project direction and the reasons
for the continuous delays and general dissatisfaction.

The change of direction

As part of the review, which included third parties, several
recommendations were made. The primary recommendation was to
adopt a more process-focused approach and apply the processes as the
framework for a more Agile way of working. The first step in the process was
to define the end-to-end process at the highest level in order to establish the
boundaries of the project scope in business process terms. The subsequent
steps then generated two further levels of visualization of the process detail
for the business process streams. The project took deliberate steps to try
to define the processes in terms of the business flow and desired business
outcomes, rather than replicating the existing systems. This process mapping
exercise was done at an accelerated pace, working as combined business,
project, and implementation partner workstreams.

The right road

Once there was a reasonable map of the processes, the requirements
were mapped to the processes so that they could be understood in the
context of a process. As a result, many of the requirements were restated
to be better anchored in the context of a business transaction and thus
less likely to be misinterpreted or removed for being redundant.

Implementation Guide: Success by Design: Process-focused solution 148

 The processes at level two were put into a storyboard by using the
expertise and leadership of the partner solution architect and the
customer’s lead functional expert. The customer’s lead functional
expert used the process flows to create logical end-to-end process
flows across workstreams so the delivery of software would be
meaningful. The partner solution architect provided the Dynamics
365 application view of the embedded standard processes,
dependencies, and constraints to ensure that the sequence of
process delivery would be efficient within the Dynamics 365
applications. This storyboard would drive the sequence of work
from realizing foundational processes to more peripheral ones.
Furthermore, the level two end-to-end processes (such as “Prospect
to Cash”) were prioritized by considering the core/most frequent/
baseline path through the process as a higher priority. The more
specialized and less frequent variations were sequenced to be
designed and delivered after the core processes.

This set of processes in the storyboard were then mapped to an overall

design within Dynamics 365 applications, generating a process-focused
solution blueprint. It was reviewed against the existing technical, system,
and data solution blueprint to create a rounded solution blueprint.

The delivery of the processes was distributed into sprints of four

weeks each, so each sprint delivered a meaningful part of the
storyboard for each of the workstreams, with an emphasis on
delivery of end-to-end processes as rapidly as possible. A high-level
plan was constructed based on the sequence, dependencies, and
estimated effort related to process delivery.

The sprint level planning was performed on the processes in scope

for that sprint, defining a more detailed process flow when required.
Documentation was kept to a minimum and the processes were designed
in the Dynamics 365 system in collaborative workshop environments. All
related activities such as data migration, integrations, testing, training,
change management, etc. were performed based on the processes in
scope. Each sprint culminated in a “conference room pilot” (sometimes
called “working demo” or “playback”), where the business SMEs
presented the new logical business process and how the processes

Implementation Guide: Success by Design: Process-focused solution 149

 were designed and implemented in the system via a demo to an
invited business audience.

At each sprint end, the business attendees reviewed and approved the
incrementally complete and functioning processes in the Dynamics
365 system instead of reviewing and approving complex and technical
design documents. Individual gap designs that had previously been
circulating on paper for weeks and months were getting translated
into end-to-end working software. Business engagement significantly
increased as the project was talking their language and they were
working directly with the emerging Dynamics 365 system rather than
with abstract design documents and lists of requirements.

Arriving at the destination

The project further reinforced the process-centric project approach,
beyond design and build, by using the processes to script and drive
end-to-end testing, reporting status and progress as “ability to execute
a process within the system.” The senior business stakeholders on the
steering group also connected with the project more meaningfully as
they were finally able to understand the project readiness and business
operations implications.

The project successfully went live, and the customer continued to adopt
a process-centric view throughout the remainder of their go-lives in
other countries. The implementation partner decided to adopt this
process-centric approach as their new standard implementation approach
for their other projects because they could clearly see the benefits.

Implementation Guide: Success by Design: Process-focused solution 150

 8
Implementation Guide: Success by Design: Project governance
Guide
Project
governance
151
 Introduction
As the footprint of cloud-based solutions increases,
we need to work with and address different business
and technology challenges.
For example, Dynamics 365 applications are increasingly able to use
low-code/no-code technology. Also, customers have higher expectations
on speed to value. The number of custom development-heavy, multi-year,
big-bang projects is diminishing; customers are looking for early and
incremental value, even on long-term, multi-country rollouts. Cloud-
For a successful project gov- based Dynamics 365 software as a service (SaaS) applications with regular,
ernance strategy, consider non-breaking system updates are changing the way we deliver business
these main areas: applications. Many of the limitations and sluggish project governance
Project goals processes of multi-year, on-premises implementations should not be
Project organization carried into the cloud.

Project approach
We need to ensure that we’re creating a project governance model
Classic structures
that is fit for the world today and tomorrow. This includes considering
Key project areas new governance disciplines and reviewing the classic governance
Project plan model for effectiveness. We also need to consider that many partners,
system integrators, independent software vendors (ISVs), and
We explore each of these areas in more
detail in this chapter. customers may have their own approach and governance standards
that they may have used previously.

Implementation Guide: Success by Design: Project governance 152

 Our objective in this chapter is to provide an overview of the importance
of good project governance. We discuss recommended practices on
how to define the different areas of project governance and analyze
and assess the effectiveness of your governance. We also describe in
more detail some of the key areas of project governance and the specific
considerations that can help stakeholders, project leaders, and
implementation teams increase their chances of success.

The project governance topics discussed in this chapter are relevant to any
implementation methodology; we focus on the underlying principles and
provide guidance in a way that allows customers, partners, and others to
evaluate their own approach and adjust as necessary.

Objectives of project governance

Project governance should be designed to provide the framework and
techniques to do the following (Figure 8-1):
▪ Map business goals into actions and measures that deliver the goals
▪ Structure, plan, and reliably, safely, and efficiently drive the project
▪ Anticipate and avoid common issues
▪ Detect emerging risks and divergence from the planned path and

Fig. 8-1
support corrective actions
▪ Provide the right amount
of flexibility and agility to
respond to unexpected events
and to adjust to the specific
constraints of the customer’s
business or project

When designing or reviewing

your project governance model,
you should examine how well it addresses each criterion. Furthermore,
it’s worth reexamining this in the specific context of the goals and
constraints of your current project.

In this chapter, we explore what project approaches, governance models,

disciplines, and techniques project teams should consider when preparing

Implementation Guide: Success by Design: Project governance 153

 for and implementing a Dynamics 365 project. Similarly, we use
Microsoft’s Success by Design framework to highlight critical questions
to consider throughout the project lifecycle to help drive success.

Why is project
governance important?
All system implementation projects, including Dynamics 365 applications,
need good project governance to succeed. However, business application
Fig.
8-2
projects have specific needs and challenges, and aren’t theeasiest of
projects to implement. Business applications directly impact and are
Typical directly impacted by the processes and the people in the business. For a
business application implementation to be successful, it’s not sufficient
governance to have good software and good technical skills; the project also needs to

issues have good governance processes that include the business.

Unclear or ever-changing A Dynamics 365 project needs to understand the business requirements
project scope deeply and at a domain and industry level. It also needs significant and
sustained participation from business users representing many different
Late discovery of roles, disciplines, and skills. Many, if not most business participants,
project slippage
may not have previous experience implementing business application
or business transformation projects, let alone Dynamics 365. This puts
Disputed areas of
accountability an additional burden on ensuring that the methodology, approach,
and governance models are sufficiently robust to support and drive the
Low or sporadic user and Dynamics 365 business application project. It also requires business users
business engagement
to gain sufficient knowledge of the standard capabilities of the Dynamics
365 application to better map their underlying business requirements to
Delays to go live at
a late stage the out-of-the-box solution.

Technical issues hiding It’s perhaps worth reminding ourselves of the most common problems
underlying governance issues
related to implementing business applications (Figure 8-2):
▪ Unclear or ever-changing project scope
Mismatched expectations
between customer and partner ▪ Late discovery of project slippage
▪ Disputed areas of accountability or project responsibility
Stakeholders blaming ▪ Low or sporadic user and business engagement
each other
▪ Delays to go live (often at a late stage)

Implementation Guide: Success by Design: Project governance 154

 ▪ Technical issues hiding underlying governance issues
▪ Mismatched expectations between customer and partner
▪ Stakeholders blaming each other

The list is long, and you may have seen other issues, but most of these
issues result in project delays. However, the root cause of the issues
tends to lie in gaps in the definition of the governance model, or in the
effectiveness of operating the project governance processes. Even after
the project goes live and meets most of the business requirements, if the
project delivery isn’t smooth, it can create stakeholder dissatisfaction
and a lack of confidence in the project.

In the context of the implementation lifecycle, you should define the

project governance model as early as possible, and certainly as part of
the Initiate phase of the project, because it directly influences the project
approach, oversight, resources, and project plan. During the Initiate
phase, if the customer or the partner has an underlying governance
model, you should review it and adapt it for this specific project, with
consideration for any specific business requirements and constraints.

Next, we explore the various areas that you should think about as part
of establishing your project governance model.

Project governance areas

Project Project governance is a wide topic and can encompass multiple different
governance practices and disciplines. Various methodologies exist in the market,
including those that are customized by partners and customers.
Project goals
Irrespective of the methodology, you should consider some fundamental
Project organization principles when designing your governance, or when reviewing and
revising it.
Project approach
The vast majority of projects have some form of governance; we’re
Classic structures
not describing setting up the common governance disciplines here.

Key project areas Instead, based on our experience across thousands of projects, we look
at the areas where we often see the need for reinforcement and expan-
Project plan sion to drive a more reliable and successful project delivery experience:

Implementation Guide: Success by Design: Project governance 155

 ▪ Project goals  We look at whether the goals deliver solid foundations
for the implementation, are well defined, and provide the necessary
direction to the project
▪ Project organization  We explore how the right project organization
and related roles and responsibilities help or hinder the efficiency of
the project
▪ Project approach  We examine the impact of the right methodology
and the wider approach on project success
▪ Classic structures  We review how we can examine the most
common governance processes to ensure they don’t hide the real
issues and provide false reassurance
▪ Key project areas  We look at how we can improve effectiveness in
areas of the project that get a lot of technical attention, but frequently
suffer from insufficient direction, planning, and oversight
▪ Project plan  We examine the critical importance of good project
planning and extract some of the planning-related lessons from
the previous sections

Project goals
Project goals Well-defined project goals are essential for steering a project and for
defining some of the key conditions of satisfaction for the stakeholders.
Project organization Often, the project goals are described in the project charter or as part
of the project kickoff. In any case, it’s worth shining a spotlight on them
Project approach
as part of the Initiate phase of the project and regularly throughout
Classic structures the project. We recommend taking deliberate actions to reexamine the
goals in the context of your latest understanding of the project.
Key project areas
When reviewing or crafting project goals, consider the following:
Project plan

Are the goals clear and realistic?

If your goals are overly ambitious or vague, the project may expend
unreasonable effort and time in the attempt to meet them, only to fall
short of the expectations set. Of course, no project intends to create
unclear or unrealistic project goals, yet there is a long history of such
examples, so we recommend conducting an honest review to confirm
these against the latest project scope.

Implementation Guide: Success by Design: Project governance 156

 Where possible, create objective goals, which are easier to measure, work
with, and track. Some qualitative goals such as “the new system should
work at least as well as the old system” are open to interpretation and can
drive the wrong project behavior. They can be misinterpreted and force
the team to design customizations in the new system to try and recreate
the old system. Instead, consider the goals in terms of the overall business
outcomes you desire; this framing allows the project to deliver them in a
way that is most efficient in the new system architecture.

Are the goals aligned with the business priorities?

If the goals aren’t well aligned with the priorities of the business stake-
holders and the business units in scope, the project won’t receive the
necessary business ownership, participation, or attention. Consider
whether the IT priorities and the business priorities expressed (or
implied) by the project goals are complementary. We also recommend
explicitly verifying that the priorities given by the organization are well
understood and aligned with those of the operating business units.
Sometimes group-led policies and requirements implied by the project
goals, which are expected to deliver additional controls and value at the
group level, may generate additional work at the operating unit level.

It’s essential to have all the stakeholders pull the project in the same
direction—conflicts at the level of misaligned goals are extremely
hard for the project team to solve. For example, if Finance leadership is
looking to improve compliance by adding more checks and approvals
in a purchase process, and Procurement leadership is looking for a
faster, less bureaucratic and more streamlined purchasing process,
It’s essential to have unless the goals are balanced, the project delivery will falter. The end
all the stakeholders result will probably disappoint both stakeholders. Another common

pull the project in example is when IT leadership has a goal of a single platform or
instance for multiple business units, but the business unit leadership
the same direction— has no goals to create common business processes. This mismatch
conflicts at the level can remain hidden and undermine the feasibility and efficiency of a
of misaligned goals single platform.

are extremely hard

Again, we recommend reviewing the goals with the stakeholders to
for the project team confirm that they will promote the right project delivery and business
to solve. outcomes. Successful projects have not only ownership of the project

157
 goals from the business, but also ownership of the successful delivery
of the project goals.

Are the project goals well understood by all the project members?
Some projects rely on a single kick-off meeting to communicate the
goals. However, many goals would benefit from more in-depth discussion
(especially with project members from outside the business) to better
explain the underlying business reasons. Consider how you can reinforce
this communication not just during the initial induction of a new project
member, but also throughout the project lifecycle.

Have the project goals been correctly translated into

project deliverables?
Once a project starts the Implementation phase, the necessary

Once a project starts attention needed for the day-to-day management and delivery can
sometimes overshadow the importance of the strategic project goals.
the Implementation This is one of the key findings from project post go-live reviews—the
phase, the necessary original aims of the project faded into the background as the project
attention needed battled with the day-to-day challenges.

for the day-to-

In many cases, projects are driven by the fit gap list, which takes a very
day management narrow view of the business expectations. Consider specifically reviewing
and delivery the initial scope of the project and initial solution blueprint (and project
can sometimes plan) with the business stakeholders to assess how well the goals are

overshadow the mapped to the project deliverables and take any corrective actions.

importance of the Is the governance process structured to formally monitor

strategic project goals. progress against the goals?
This is an often-neglected area; sometimes projects only seriously review
the goals or success criteria as part of the final go-live assessment, which
is too late. Try to avoid this by defining specific structured processes to
ensure the assessment is ongoing. For example, look at how the solution
design blueprint helps meet the business process objectives and how the
data migration and reporting strategy meets the information access ob-
jectives. Monitor to confirm that the project priorities are aligned with
the project goals and aren’t being diverted by other considerations or
day-to-day challenges.

Implementation Guide: Success by Design: Project governance 158

 Project organization
Projects implementing business applications tend to have common
Project goals
structures and roles (such as project manager, solution architect, subject
matter expert, functional consultant, and technical consultant) that are
Project organization recognizable across different projects. The difference in the effectiveness
Project approach of project teams comes from the way in which the project organization
functions in practice compared to the theoretical constructs (Figure 8-3).
Classic structures

When assessing your project organization, consider the following:

Key project areas

Project plan How well is the project team aligned to the business?
Teams that have good alignment between the business streams and
project functional workstreams tend to have more effective and high-
er-velocity projects. A common model is for each key business stream to
be matched with a corresponding project workstream. An experienced
leader from the business stream is usually seconded to the project and
takes on the role of the lead subject matter expert (SME). For larger
projects, multiple SMEs may be appointed for a single workstream.
Fig.
8-3

The direct involvement of a busi-

ness stream with the project is
usually the most successful model
because it engenders trust both
ways—from the business to the
project and vice versa. If you have
multiple organizational steps
between the business and the
project team SMEs, the level of
business ownership of the solution
design diminishes, as well as the
quality of the input from the SMEs.

Does the project organization

include active and appropriate
senior business stakeholders?
Projects are far more successful

Implementation Guide: Success by Design: Project governance 159

 when senior business stakeholders have clear roles and are active and
deeply engaged in driving the project. A good sign for the project is
when the senior stakeholders are keenly interested in the status of
critical processes and are making a direct impact on the prioritization of
tasks and making the corresponding resources available.

Projects in which the senior stakeholders are more passive and just
occasionally asking “How is it going?” or “Let me know if you need
something” tend to have poor outcomes.

Does the structure promote cross-team collaboration?

Examine if the project organization structure, reporting lines, team
leadership and composition, communication channels, team objectives,
and delivery approach naturally generate sufficient cross-team
collaboration, or if they naturally encourage silos.

Danger signs are project organization structures with the following:

▪ Team members performing the delivery work on the ground have
to navigate multiple layers of management to overcome issues
▪ Members of workstreams rarely work with other workstreams
▪ Responses to ad hoc questions on a team member’s task status
and issues are routed via the project manager—this also introduces
mistrust within the team
▪ Workstream communication channels are all hierarchical, with few
organic, cross-workstream objectives

A good way to determine the level of cross-stream collaboration

is, for example, to scrutinize how closely, and how often, the data
migration workstream works with the various functional and technical
workstreams’ design. How closely entwined are the ongoing delivery
objectives of the data migration team and those of the functional
team? Do they primarily communicate hierarchically or across to their
teammates in other areas?

Is accountability and authority well defined at the project

leadership level?
Projects where the wrong role is accountable for delivery or where
the accountability is diffuse tend to have low velocity, with project

Implementation Guide: Success by Design: Project governance 160

decisions stagnating. Projects where the accountability is given to
someone too junior in the organization to have the corresponding
authority may also languish without sufficient direction and resources.

How well are the team roles and responsibilities defined?

Every role on the project needs to be well defined, including roles that
may be part-time on the project. The critical role played by business
users that aren’t seconded to the project is often ill-defined. They
may typically be necessary for defining a detailed process or data
requirements, for approving business processes, or for preparing and
performing testing. These are critical activities, and the corresponding
roles should be well defined so the resources can be planned and their
tasks and status are visible to the project leadership.

How well are the team roles aligned to the solution complexity
and system design requirements?
You should conduct an honest analysis of the experience and ability
of the resources in key roles when compared to the complexity of the
design and its constraints. Be wary of job titles that don’t match the
experience and authority that normally accompany such titles.

This is particularly important for the key roles of lead solution architect,
lead technical architect, and project manager.

Are the levels of resources proportional to the level of

effort and complexity?
By the end of the Initiate phase of the project, you should have a reasonably
credible high-level solution blueprint, a reasonable estimate of the backlog
and effort, and a high-level project plan. This should provide a good
grounding for reviewing the level of resources planned for the key roles.

This honest review is especially recommended for the roles that are
likely to be the most constrained, so that mitigation plans can be
initiated in a timely manner.

During the implementation, you should regularly assess how the day-
to-day working of the control, communication, and feedback functions
are being helped or hindered by the project team organization

161
 (Figure 8-4). How well does the project organization structure facilitate
or constrict the undiluted and timely flow of direction, guidance, and
decisions from the business to the project workstreams? In the other
direction, does the structure enable or hinder the flow of accurate,
actionable, and timely feedback from the workstreams to the project
leadership and business stakeholders?

A typical setback on projects is the leadership team not receiving timely

and accurate feedback on project issues, therefore adding unnecessary
cost and delays. This is often due to structural issues created by a project
organization that doesn’t help the flow of meaningful information.
Projects with a lot of hierarchical communication and reporting-line
hoops to clear (especially on large projects) tend to have difficulties in
surfacing accurate data promptly. This can result in project challenges
stagnating for too long, which leads to delays and additional cost.

Many projects start with the ambition of some model of partnership

and shared leadership with the various parties involved. The projects
that better achieve this ambition are the ones that meaningfully
map the commercial and partnership arrangements (or expectations
if informally agreed) and their respective expertise to the project
organization. Projects tend to have better success when the custom-
er has a clear leadership (and ownership) role in the organization
Fig. structure driving the overall proj-
8-4
ect. Similarly, unless the customer
Accurate, honest, has experience and capability in
timely status: implementing business applica-
Progress, risks, tions, having the implementation
opportunities, blockers
partner in a leadership (and
accountability) role for provid-
Workstreams ing strategic guidance on the
implementation methodologies,
Project
system design, and technical
leadership
procedures allows for the best
Based on a close mapping of leadership roles to
understanding of
those with the best experience. Of
the project:
course, depending on the specific
Goals, direction,
decisions, resources expertise between the customer

Implementation Guide: Success by Design: Project governance 162

 and partners, the actual balance of responsibility lies somewhere in
the spectrum of these arrangements, and the organization structure
should be mapped to that agreement.

In business application projects, some critical project roles have a

disproportionate impact on the success of a business application
project. For most projects, the role of the solution architects and the
project manager is particularly critical to success. The project manager
and solution architect roles are important both from the implementa-
tion partner and from the customer; the customer solution architect
is usually described as the lead business SME or the business lead.
Depending on which areas of your project are particularly significant or
risky, you may want to add other roles to the list of the critical project
roles. You should then pay extra attention to ensure that the individ-
uals chosen for these critical roles have the right level of experience
and capability. For example, an ineffective solution architect may not
be able to direct the solution design blueprint such that the different
elements of the solution design come together into a coherent and
efficient solution that matches the business needs.

You should also confirm that the project organization allows all roles to
be successful by enabling them to properly assert their expertise and
strongly influence the design and implementation, thereby delivering
the full positive impact of their roles.

Project approach
Project goals When talking about project approach, one of the dangers is that it
can be assumed to be synonymous with the project implementation
Project organization methodology. This can then leave a vacuum in the processes that need
to be defined outside of the implementation methodology. This can
Project approach be especially true if the implementation partner is providing a limited,
technical implementation methodology. If you’re the customer, you
Classic structures
should consider the wider set of processes required to define your
Key project areas project scope, manage your resources, and manage the changes,
tasks, and processes. Prioritize areas that aren’t directly covered by the
Project plan partner but are necessary for every business to perform in support of

Implementation Guide: Success by Design: Project governance 163

 a business application implementation. You should explicitly identify
areas that would be the responsibility of the customer to manage.
Then confirm that you have adequate definition of the approach and
the right level of governance planned for each of these areas.

The areas excluded from the implementation partner’s responsibility

will vary, but the following areas are typically involved:
▪ Defining the scope of the project (the to-be business process and
system requirements)
▪ Managing internal project resources and liaising with wider
business teams
▪ Managing the process for the proper participation of all the
relevant business units and roles
▪ Managing the source data quality for data migration
▪ Training the internal project team and the end users
▪ Data, system, and process validation by the non-project business
users, such as user acceptance testing (UAT)
▪ Security and access definition and validation
▪ Implications on the wider IT enterprise architecture and wider
business processes
▪ Interpreting, applying, and validating the requirements related to
relevant regulatory bodies
▪ Cutover management
▪ Managing the non-Dynamics 365 systems involved in system
integrations
▪ Managing communication within the business and with customers
and suppliers
▪ Budget management
▪ Business and organizational change management
▪ Non-functional requirements such as performance management,
system security, and business continuity planning
▪ Transitioning and operating the support model and other post-
operational duties

This is not intended as an exhaustive list, more of an indicator that

the overall project approach needs to consider a much wider set of
functions than what may be covered by a typical implementation
methodology or a commercial contract. Some projects define the

Implementation Guide: Success by Design: Project governance 164

 overall approach in a project charter or similar document. You should
examine if the project approach adequately covers the missing areas
from the implementation methodology.

One of the discussions that regularly occurs about project approach is

the debate on waterfall versus agile. Methodologies based on either
of these principles have their advantages and disadvantages. For more
detailed discussion on implementation methodologies, refer to “Chapter
5, “Implementation strategy.” From a project approach perspective,
whichever methodology (or hybrid) is adopted, it’s important to identify
and record the specific risks that logically emerge from that methodolo-
gy and have a mitigation plan for each one.

For example, if your chosen methodology doesn’t include early anal-

ysis and defining a solution blueprint that reflects the whole design
scope, the project may stumble from designing specific solutions from
sprint to sprint. You may find that the designs in later sprints can’t be
built upon the previous sprints due to the constraints of not having
analyzed and considered the high-level design for the whole end-to-
end process. We often see this in pure agile projects.

Similarly, if the chosen methodology is strictly linear and mostly

document-based, you risk spending an enormous amount of time and
effort in analysis, design, and coding phases with little exposure to the
working solution and limited feedback on how the solution performs
in the system. This is a risk we see in many pure waterfall projects. You
need to identify such risks in your chosen methodology so you can
discuss, accept, understand, and specifically address them in the
project approach.

Sometimes when a methodology is being selected, there is an

assumption that a methodology that works well for a bespoke
software development project will apply with minimum changes
to a highly configurable Dynamics 365 cloud-based business ap-
plication implementation. Similarly, methodologies that suited the
old world of on-premises business application implementation
with long implementation durations aren’t best adapted to today’s
more rapidly evolving cloud world. Consider the very different

Implementation Guide: Success by Design: Project governance 165

circumstances of a modern, cloud-based Dynamics 365 business appli-
cation implementation:
▪ The implementation is for a cloud-based, packaged SaaS
software application
▪ A significant (if not a majority) of the activities aren’t coding, but
business process design, setup, configuration, data migration,
validation, and more
▪ Some critical activities for a Dynamics 365 application implementation
(such as extracting a fit gap report from a fit-to-standard analysis)
aren’t covered by more generic software development implemen-
tation methodologies
▪ There is ever-increasing functionality in the Dynamics 365 busi-
ness application of powerful configuration options or low-code/
no-code options and the ability to use the Dynamics 365 Power
Platform in citizen-developer and professional developer modes
▪ The methodology must address the fact that a very significant part
of the implementation process revolves around the understanding
and configuration of business processes and working closely with
business users throughout the project lifecycle
▪ Although Dynamics 365 applications provide a platform for custom
development, the chosen methodology must address the fact that
the development process is building on, or extending, the standard
business process flows designed in the system
▪ Dynamics 365 applications are a cloud-based, SaaS business appli-
cation with a regular update rhythm that needs to be recognized
as part of the new way of working

All of these factors (and more) mean that the implementation

methodology needs to be directly relevant to the nature and needs of
a Dynamics 365 business application project.

When looking to determine what the overall project approach should

address, consider the specifics of your whole project solution build life-
cycle and make any additions, adjustments, or highlights to the standard
methodology. Confirm that the key processes and controls you expect
are adequately addressed as part of the governance of the project:
▪ Does the methodology and approach enable a good definition of
the scope and a solution blueprint?

166
 ▪ Is the testing strategy fit for purpose?
▪ Do controls such as regular and appropriate reviews by the project
team and the business exist?
▪ Is a good process for project status analysis and reporting in place?

You may want to define a set of critical processes, deliverables, and

controls to ensure you have the right coverage from the project ap-
proach ( just as you would for functional and technical requirements).

Once you are in project implementation mode, and the focus is on

day-to-day tasks, it’s easy to lose sight of the ideals agreed in the gov-
ernance model at the start of the project. You should establish regular
checkpoints to go through the formal discipline of honestly comparing
the actual ways of working on the project with the governance model
and taking corrective actions as necessary.

For example, analyze a random sample of how a given feature, require-

ment, or user story is actually progressed through the lifecycle, or take
a scope change item and trace how it was handled to resolution. Did
they progress through their lifecycle as you expected, and did your
governance model help deliver a fast and quality result? Taking a (pref-
erably independent) regular pulse of the real, practical implementation
of the processes and controls helps keep the project approach relevant
and reduces risk.

Classic structures
Project goals Most, if not all, business application projects have the common, classic
governance structures in place. This section doesn’t give a general
Project organization overview of historically well-known disciplines around these common
governance areas; instead we look at how to assess the true effective-
Project approach
ness of these processes and controls in practice. The mere presence of
these classic governance structures such as steering groups, program
Classic structures
boards, or risk registers can sometimes lull projects into thinking that
Key project areas they have adequate active governance. Let’s dig deeper into these
areas to explore how we can get better insight into their function and
Project plan evaluate their effectiveness.

Implementation Guide: Success by Design: Project governance 167

 Steering groups
Most projects have some form of steering group in which the project
sponsor, senior business stakeholders, and project leaders meet regularly
to discuss and review the project. A lesson learned from participating
in multiple such steering groups across multiple projects is that the
effectiveness of these meetings varies hugely. Common factors that can
impact the effectiveness of a steering group are when the purpose of
steering group meetings is diluted or unclear, and when project status
reporting isn’t easily understood, accurate, or actionable (Figure 8-5).

Steering group meetings

Steering group meetings can negatively influence project success if the
project manager presents the status and the steering group is simply
checking in with very little understanding of the details of the project,
no direct engagement, and a lack of knowledge to challenge and ex-
plore issues more deeply. Projects that are considered able to manage
themselves despite some issues are often surprised by a sudden crisis,
resulting in delays and dissatisfaction.

An efficient and effective steering group has leadership team partici-

pants that understand and perform their primary function of steering
the project. This requires them to invest the time to understand the con-
Fig. tent of the project at a sufficient depth so that they can read between
8-5
the lines of the project status report, ask the right questions to
understand how and where they
can help, and proactively direct
the project.

Apply similar, regular assessments

of the effectiveness of other such
boards, like the change control
board or the design authority
board.

Project status reporting

Accurate, meaningful, and action-
able project status identification

Implementation Guide: Success by Design: Project governance 168

 and reporting is the lifeblood of a project. Business application imple-
mentations often involve complex, multiple strands of work, all of which
need to come together seamlessly and on time. Multiple parties are often
involved in the project delivery—business SMEs, business IT, business
project managers, and their equivalent roles from the implementation
partner, as well as other third parties. In addition to the main func-
tional and technical delivery, there are multiple other disciplines, like
data migration and security. All these roles have their own specialist tasks
to perform, and managing and monitoring all these different types of
tasks to provide a coherent project status isn’t easy.

Additionally, specialist teams may provide an overly technical view of

the status in terms that only that team understands. Often, project-lev-
el status reporting is driven by fit gap progress, which isn’t necessarily
the critical part of the project. Consider the following when looking for
effective project status reporting to the steering group:

Is the project status meaningful and actionable for

the intended audience?
For example, when presenting internally to your own specialist team, a
detailed and technical view is very appropriate; however, when presenting
the status to a steering group or to non-project team members, consider
how to express the status in terms that are understood and actionable.

Is there evidence to believe that the project status is accurate?

Consider how the project status data has been gathered, analyzed, sum-
marized, and presented. Understand the level of uncertainty associated
with the data. Is there a history of the project status data and related es-
timates versus actuals that can provide some idea of certainty? Steering
groups have to consider the degree of uncertainty associated with the
status when assessing risks and making critical decisions on moving or
adding resources, approving budgets, and setting timelines. This means
testing the accuracy of the status by looking at it from multiple angles to
ensure that it provides robust and actionable data.

Is there a comparison between the planned and actual status?

Project status reports only have meaning when they show an easily
understood comparison between the planned status and the actual
progress—the report must be able to answer questions like “Are we

Implementation Guide: Success by Design: Project governance 169

 on track?” “What is the size of the remaining work to completion?”
“What do we, as the steering group, need to do to get the project on
track and keep it on track?” The absence of comparing planned work
to actual and remaining work doesn’t allow the steering group to take
necessary actions to help recover.

Risk register
Most projects have some form of a risk register. When used well, this
register is a good way to communicate risks and help teams focus their
attention on removing barriers to success. The following are examples
of ineffective use of risk registers:
▪ Risks that are of little practical value to the project are being used
to shrug responsibility or provide cover against blame.
▪ Risks remain on the register for a long time with just new comments
and updates being added weekly at each risk review meeting, but
with no resolution. This implies that either the risk isn’t getting the
attention it deserves or it’s difficult to resolve and consuming more
and more resources without results. In any case, you should either
treat risks stuck in this loop urgently with focus or accept them with
a mitigation so they don’t drain the project over time.
▪ Risk priorities don’t reflect the project priority at that stage of
the project. You should take care to ensure that the risk register
doesn’t create a parallel project effort with its own priority and
path. Risks and issue resolution should be incorporated as part of
the main project delivery.
▪ The risk register has a very large number of risks, many of which
are stagnant. Consider how many risks the project can realistically
work on at any given time and trim the register according to real
project priority.

Stage gates
Stage gates or milestone-driven planning and reviews are a common
feature of the majority of business application projects, including more
agile projects. These milestones are regarded as important checkpoints
spread throughout the project timeline, which the project can only
pass through if they have met certain criteria. The reality of many

Implementation Guide: Success by Design: Project governance 170

 projects is that the checkpoints don’t always perform their intended
function. There may be several reasons for this, and projects should
examine if they are suffering from the following limitations:
▪ Criteria for the milestone are unclear  You should strive to
create very explicit criteria for entering and exiting a milestone. If
the criteria are difficult to measure, it’s difficult to take unequivocal
decisions based on data.
▪ Exit and entry criteria aren’t respected  Projects are typically
under great pressure to get past one milestone and move into the
next phase, either because of resource usability or commercial
payment triggers, or because the project optimistically believes
that it will somehow catch up with the stragglers that didn’t meet
the milestone criteria. Sometimes a project is also pressured to
move to the next phase because it implies progress. This self-de-
lusion can create additional risks in the project; uncontrolled
disturbances in project deliverables can delay the ultimate goal of
a successful go live. The debt incurred by skipping the criteria has
to be paid back at some point—the later it’s resolved, the higher
the cost. When considering allowing a milestone to pass when it
hasn’t met all its criteria, project leadership should strongly consid-
er the nature and size of the incomplete tasks, precisely how these
tasks will be addressed, what impact the lagging tasks will have on
A milestone dependent tasks. This allows for a managed transition between
should have the milestones with the risks fully understood and a clear and realistic

primary purpose action plan for the stragglers.

▪ Milestones don’t reflect true dependency  From a project
of confirming lifecycle perspective, a milestone should have the primary purpose
that the project of confirming that the project has satisfactorily completed the
has satisfactorily precursor tasks so that it’s safe, efficient, and meaningful to start

completed the the tasks of the next stage. There may be other perspectives, such
as commercial ones to trigger stage payments or just periodic
precursor tasks points for the project to review progress or cost burndown, but
so that it’s safe, we’re focusing on a project lifecycle point of view. Some common
efficient, and examples of stage gate criteria are as follows:
▪ To-be business processes defined  An approval step from the
meaningful to start business to confirm the business process scope helps ensure that
the tasks of the prior to spending a lot of effort on detailed requirements and de-
next stage. sign, the process scope has been well established and agreed upon.

171
 ▪ Solution blueprint defined  This helps ensure that the require-
ments have been analyzed and understood sufficiently well to be able
to define an overall, high-level design that is confirmed as feasible.
Additionally, the key design interrelationships and dependencies are
established before working on the individual detailed designs.
▪ Formal user acceptance testing start agreed  Starting formal
UAT with business users tends to be the final, full formal test phase
before go live, with the expectation that go live is imminent and achiev-
able. Prior to starting this phase, it makes sense to validate the readiness
of all the elements to ensure that this test phase can complete within
the allocated time period and meet the necessary quality bar.

Consider the various milestones that help with a true understanding of

readiness, especially those that help you better understand when to start
dependent tasks, such as feature complete, code complete, code frozen,
business data validated in Dynamics 365, or solution ready for user ac-
ceptance testing, and determine what entry and exit criteria would help.

Design and change boards

A project with multiple parties involved in the design and delivery tend
to have some form of design review board that ensures compliance
with the enterprise IT policies and standards. This board also ensures
that the proposed designs from the project (and especially design
changes) are fit for purpose.

Effective design review boards (also called architecture review

boards or design change boards) tend to have a good grasp of the
overall solution blueprint. In addition to reviewing any individual
design, they can ensure new designs and changes will operate
within the boundaries of the solution blueprint and won’t adversely
affect other designs.

The design review board should also have representatives, not just
from the project, but also from the wider business and IT, to help
ensure the business case for the design is sound and the impact on
related systems (integrations) and overall enterprise architecture are
considered. Another characteristic of a successful design review board

Implementation Guide: Success by Design: Project governance 172

 is that they have minimum bureaucracy and their focus is on helping
the project to move as securely and as fast as possible.

The process by which a design review board actually functions is as

important as its constituents and roles and responsibilities. Design
review boards should communicate the solution blueprint boundaries
and design standards expected so projects can proactively make sure
that they’re working within the expectations.

For the best project velocity, you should set an expectation that the
reviews will be interactive and expected to be resolved within a single
iteration. This requires the design review board to be pragmatic and well
prepared. It also requires the project to have the discipline to provide
sufficiently detailed proposed design information. This process needs to
be organized and communicated as part of the project governance in
the Initiate phase so you’re not trying to establish these practices during
the intensive pressures of a running project implementation.

Key project areas

Project governance is often thought to be confined to the classic areas
Project goals we discussed in the previous section, but in Dynamics 365 business
applications, we should consider the key activities that also need to
Project organization have governance processes embedded. We see many projects where the
strategies for the more specialist project delivery areas have their own
Project approach
processes and disciplines that don’t always work in concert with the
project approach or to the benefit of the overall project. These areas are
Classic structures
often managed and driven by specialists with specialist processes and
Key project areas internal goals.

Project plan It’s worth considering what governance is appropriate for these areas to
ensure the project (and hence the business) derives the best value. In this
section, we discuss the key areas of application lifecycle management
(ALM), test strategies, data migration, integration, cutover, and training
strategies (Figure 8-6).

Implementation Guide: Success by Design: Project governance 173

 Fig.
8-6 Application lifecycle
management
Does the defined application
lifecycle match the project
methodology? For example,
how will the configuration data
be analyzed, defined, approved,
secured, managed, and distribut-
ed to the different environments?
Is this process also governed by
the requirements of the project
approach, and will it be designed
to meet the expectations of the
project timetable? The risks here
are that ALM processes are some-
times considered as a technical
topic and not rolled into the
overall project governance, which creates mismatched delivery and
project planning.

The build process (configuration, data, code) in a business application

can be presented as a complex, technical, and jargon-filled process.
Business stakeholders may retreat and let specialists take complete
oversight. However, project leadership can get meaningful insight into
the (necessarily technical) build process by devising a layer of gover-
nance that provides the necessary understanding and helps keep the
project on track. For example, ask the following questions:
▪ Are the build requirements directly mapped onto the end-to-end
business processes, so that any discussions related to the require-
ment can be well understood in that context by non-technical
stakeholders?
▪ Are the right controls in place so that progress can be tracked and
reported in a way that is meaningful outside of the technical teams?
▪ Do controls exist to ensure the right quality, standards, and
compliance needs will be applied?
▪ Does the project have a practical and efficient method to raise

Implementation Guide: Success by Design: Project governance 174

 business process questions with the business area owners, and
a way for the business to understand the system delivery of a
business process?

A more in-depth exploration of the ALM and build process is available

in Chapter 11, “Application lifecycle management.”

Test strategy
A well-defined test strategy is a key enabler for project success. A high
level of governance is required to ensure that the right test strategy is
created and implemented, because multiple cross-workstream parties,
business departments, IT departments, and disciplines are involved.
When evaluating the suitability of the test strategy, in addition to the
technical angles, consider how well some governance areas are covered:
▪ Does the right level of governance exist to ensure that the testing
strategy is mapped to, and proportionate with, the project scope
and the business risks?
▪ Is the business sufficiently involved in the test coverage, test case
development, and approval of the testing outcomes?
▪ Is the project approach in line with the test strategy and test planning?
▪ Does the right governance exist to make sure the business pro-
cesses are tested end to end?
▪ Is the data to be used in the various testing phases and test types
sufficiently representative to fulfill the test objectives?
▪ Is the test coverage across the functional and non-functional areas
adequate to help assure the business of safe and efficient opera-
tion in production use?

More details on defining a successful test strategy are in Chapter 14,

“Testing strategy.”

Data migration
For data migration, examine the related governance coverage to
ensure that this process is well understood at a business level:
▪ Do you have the right level of business ownership and oversight
on the types and quality of data being selected for migration?

Implementation Guide: Success by Design: Project governance 175

 ▪ Is a data stewardship process in place to make sure that the
cleansed and migrated data is kept clean during the project and in
operational use?
▪ Will the data from the existing source systems be faithfully
transformed so that it’s still meaningful and fit for purpose in
Dynamics 365?

Review the data migration strategy from a governance view to ensure

the workstream is delivering to the needs of the project and the busi-
ness and isn’t becoming a silo with only technical oversight.

Integration
A common area missing governance is the management and own-
ership of non-Dynamics 365 systems involved in the integrations,
because they tend to be managed by those outside of the project.

The cutover from the Make sure that the business impact on either side of the integration is
understood and managed. Confirm that the data exchange contracts
previous system to are defined with the business needs in mind. Examine if the security
the new Dynamics and technology implications for the non-Dynamics 365 systems are
365 system is a time- properly accounted for.

critical and multi-

You also need project-level governance on the definition and manage-
faceted process. ment of non-functional requirements, such as expected volumes and
performance, and clear processes to resolve issues that may stem from
the other systems in the integration.

Cutover
The cutover from the previous system to the new Dynamics 365 system
is a time-critical and multi-faceted process. It requires coordination
from multiple teams for the related tasks to all come together for a go
live. You almost certainly need to include business teams that aren’t
directly part of the project. Therefore, cutover needs to be shepherd-
ed with a deep understanding of the impact on the wider business.
Preparation for the cutover needs to start early in the project, and the
governance layer ensures that the cutover is a business-driven process
and not a purely technical data migration process. For example, early

Implementation Guide: Success by Design: Project governance 176

definition and agreement on the related business calendar sets the
right milestones for the data migration to work with.

The legacy systems shutdown window for the final cutover is typically
short, perhaps over a weekend. For some cutover migrations, that window
may be too short to complete all the cutover activities, including data
migration. In such cases, the project team may perform the data migration
as a staggered, incremental migration, starting with slow-moving primary
data and frozen transactional data. This leaves a smaller, more man-
ageable remainder to address during the shutdown. This needs careful
governance, and the strategy needs to be decided early because the data
migration engine needs to be able to deliver incremental data loads. You
should also carefully consider what business activities you may need to
throttle or perform differently to reduce the number and complexity of
changes between the first migration and the final cutover. The changes
need to be meticulously recorded and registered (automatically or
manually) so that they can be reflected in the final cutover.

Training strategy
Most projects have plans to train at least the project team members
and business users. All too often though, training is seen as a lower pri-
ority. If any project delays put pressure on timelines or budget, training
can be one of the first areas to be compromised.

This trade-off between respecting the go live date and completing the
full training plan can be easier for the implementation team to rational-
ize because the team is aiming for a go live and the risk of poor training
can be seen (without sufficient evidence) as manageable. The worst
consequences of poor user training are felt in the operational phase.

You can mitigate this (mostly unintentionally) unbalanced trade-off

with the following actions:
▪ Clearly defining and communicating the implications on project
success in operational use of poor training delivery.
▪ Making the review and approval from the business reflect people readi-
ness. This check should be part of the entry criteria into user acceptance
testing and be a meaningful part of the go/no go for production use.

177
 The test for people readiness needs to be meaningful; it should be an
evaluation of the effectiveness of the training, and not just that it was
made available. The test should be equivalent to assessing whether
enough people in key roles can conduct their critical day-to-day busi-
ness process safely and efficiently using the Dynamics 365 application
and related systems.

More details on how to put together a good training strategy are in

Chapter 19, “Training strategy.”

Project plan
Project goals Project plan analysis is where the outcomes of a lot of governance top-
ics become visible. The effects of good governance are felt on a project
Project organization by noting that the project plan is resilient to the smaller changes and
unknowns that are a reality for all business application projects. Poor
Project approach
governance, on the other hand, manifests itself as continuously missed
targets, unreliable delivery, and repeated re-baselining (pushing back) of
Classic structures
the project plan milestones. The key is for the project planning process
Key project areas to have its own governance mechanisms to avoid poor practices, detect
them early, and provide the agility and insights to fix and adjust quickly.
Project plan
When determining how the project plan should be constructed, the
project plan should be able to demonstrate the following:

Is the project plan designed to cover all the critical areas?

Does the plan cover (at least at a milestone level) all the key areas of
activity? If not, key dependencies and effort will be missed.

Is the project plan actionable?

A project plan should be able to fulfil one if its primary purposes of di-
recting a project. If that isn’t the reality on the project, you should look
at how to restructure it so that it’s the main driver of project activity at
least at the milestone level.

If a plan is thousands of lines long, it’s unlikely to be directly actionable,

and will probably be in a state of continuous update and adjustment.

Implementation Guide: Success by Design: Project governance 178

 Similarly, if the project workstreams aren’t paying any attention to
the plan, it’s a clear sign that the project plan may not be actionable.
Actionable project plans help direct the project teams to deliver the
right activity at the right time and reduce uncontrolled activities.

Projects using agile practices may be using some form of prioritized

backlog management for much of the build and test activities, but there is
normally still a high-level milestone plan to allow for the communication
and management of related tasks that aren’t covered by the backlog.

Whatever methodology is in place, the plan should be regularly evalu-

ated to confirm that the teams are taking their direction from the plan.

Is the project plan up to date, accurate, and realistic?

Compared with the traditional on-premises business application
projects of a few years ago, today’s cloud-based Dynamics 365 imple-
mentations tend to move faster with shorter durations. A project plan
that isn’t up to date is ineffective.

Projects should institute thresholds for how out of date a project plan
is allowed to become. A project plan that is inaccurate in terms of the
estimated effort (and remaining effort), duration, and dependencies
will promote the wrong decisions and allow risks to remain hidden, and
ultimately give an inaccurate picture of the project status.

For a project plan to remain accurate, it needs a robust task estimation

process and a reliable and easy means of tracking tasks in progress. Con­
sider how to make this possible structurally and with efficient processes.

Does the project plan give a clear indication of what tasks

are critical?
Business application projects have many moving parts and many spe-
cialist areas. The project must be able to accurately derive the critical
path so that project leadership can focus their attention on those tasks.
For agile projects, agile crews prioritizing the backlog and managing
the critical path within short-duration sprints and iterations can pro-
vide the necessary insight.

Implementation Guide: Success by Design: Project governance 179

 Accountability for outcomes
Project planning should promote the clear identification of account-
ability for the outcomes (both at the workstream level and overall).
Often, projects create a RACI (responsible, accountable, consulted,
informed) matrix at the start of the project, but this may not provide
the right level of accountability for all areas. Additionally, this matrix
isn’t always referenced during the project implementation. Asking the
question via the project plan if the accountability for the outcome (not
responsibility) is clear helps keep the accountability fresh in everyone’s
mind. It keeps the project more agile with faster decision-making.

Status reporting
A well-constructed project plan facilitates accurate project status
reporting. However, it needs to be deliberately designed into the
project plan with the right dependencies, estimated effort, and mile-
stones so it can be easily extracted from the plan. This means that the
project should be deliberate in the detail to which tasks are scheduled
so dependencies and milestones are created with this in mind. Because
status reporting often has milestone-level status as a key indicator, the
meaning of a milestone’s completion must be explicitly defined so the
implications are clear to the intended audience.

Project plans also need a regular and easy-to-use process to keep

task progress updated with an accurate status. You should also define
a uniform and reliable method for units of progress so you get the
same quality of updates from all teams. For example, tasks defined as
80 percent complete often under-estimate the effort to complete the
remaining 20 percent; instead, consider reporting the remaining effort
in hours for a given task.

Some projects, especially ones using more agile practices, may use
alternative or additional analysis and presentation methods such as a
backlog burndown, remaining cost to complete, or earned value, but
these principles apply nevertheless and they all rely on having accurate
underlying data on actual versus expected progress to date and the
remaining effort expected.

Implementation Guide: Success by Design: Project governance 180

 A useful way to report the status
of a business application project
is to present a functional heatmap
view (Figure 8-7), which shows

Fig.
the readiness of the key processes
8-7
for the next major milestone (such
as sprint review, SIT, UAT, or go
Business process heat map example live). The functional heatmap is a
business process perspective of
Legend
Process not complete Process complete, the project plan status.
and not on track in repair, or testing

Status reports can also include

Process design blocked Process solution in Process complete and
or not started progress and on track ready for next test cycle status from other project controls
such as risks and issues, sentiment
Finance surveys, or resource availability in
prior months. These aren’t directly

HR derived from a project plan and

can provide a useful supplemen-
tary view. However, it’s important
Operations
to ensure that actionable con-
clusions are extracted from these
Logistics
alternative views; these conclu-
sions should generate tasks that
can be measured, planned, allocat-
ed, and subsequently reflected at
Legend
Process not complete Process complete, some level in the project plan.
and not on track in repair, or testing

Process design blocked

or not started
Process solution in
progress and on track
Process complete and
ready for next test cycle
Project
feedback loop
Finance Project status should provide a
feedback loop to the project stake-
HR holders, project management,
and the project team to adjust the

Operations project based on the findings. This

should be a defined process in
the overall project governance to
Logistics
create the discipline to regularly

Implementation Guide: Success by Design: Project governance 181

 review the status and generate explicit actions for the project that
directly address the status findings.

The project status should not be the only source for this feedback loop—
consider all the other controls and procedures in the project that can
generate actionable information, if only there were a systematic mech-
anism defined to diligently listen and extract the actions. For example,
updates from daily stand-ups, design reviews, sprint playbacks, or even
informal risk discussions can help provide useful feedback.

In summary, create a culture from your governance processes that ac-

tively recognizes the feedback loop and translates feedback into actions.

Conclusion
The primary intention at the start of this chapter was to provide an
overview of the importance of good project governance and provide
guidance on how to assess the effectiveness of your existing or pro-
posed governance model.

We then moved on to describe how many of the critical issues on a

Dynamics 365 project that initially may manifest themselves as tech-
nical issues actually have their roots in failures of project governance.
We also saw that the governance requirements for a Dynamics 365
business application needs to address the fact that the project is es-
sentially delivering the capability to perform business processes using
the application. So the governance model must include strategies and
techniques that can help realize the project goals through the right
engagement with business users and business processes.

We discussed how defining a project organization model that has an

engaged leadership and facilitates rapid and accurate communication is
more aware of the reality of the project. Similarly, project organizations
that enable the right accountability, naturally encourage cross-
collaboration, and reflect the complexity and effort of the project are
much more likely to have good velocity and be successful.

Implementation Guide: Success by Design: Project governance 182

We also discussed how the approach is much more than a technical
methodology, and that the approach must be practical and tailored to
the specific needs of a Dynamics 365 business application and to the
customer’s project and business needs and constraints.

We showed that the mere presence of the usual governance structures

doesn’t assure project success. We can evaluate the effectiveness of
the governance model in very practical ways and with a focus on the
specific needs of Dynamics 365 projects.

We also brought awareness to how a governance perspective on the

specific technical implementation strategies (such as test and build
strategies) can help a Dynamics 365 project achieve better, faster
results. We proposed that governance should be an organic part of
the strategy. These strategies shouldn’t be considered as being purely
subject to technical oversight and control.

Lastly, we discussed how a good project plan and planning process

will expose the outcomes of a lot of these governance topics. Project
planning shouldn’t just be an overall status reporting tool, but be
deliberately engineered to gain early and deep insights into the
effectiveness of the different elements of the project work and allow
rapid responses to the deficiencies.

You can use the key objectives of good project governance to judge
the overall suitability of a given governance model. Ideally, you should
test a proposed governance model against these objectives during the
Initiate phase of the project and throughout, but it’s never too late.

In summary, project governance for Dynamics 365 projects should be

a framework that permeates all aspects of our projects, not just the
typical areas that were historically or conventionally regarded as the
domain of governance. Good project governance isn’t a set of
generalized bureaucratic or administrative procedures to follow—it’s
about creating a Dynamics 365-specific governance culture of driving
the project to reliably and efficiently deliver the project goals.

183
 Checklist

Project goals Identify relevant project roles and areas of ownership and
assign them to team members with the right expertise
Ensure goals are clear, realistic, mapped to actions and and experience.
measures, correctly translated into project deliverables,
and regularly monitored throughout the project lifecycle.
Project approach
Align goals and ensure they are sponsored by relevant
Analyze, review, and confirm that your chosen imple-
stakeholders across the organization, and properly com-
mentation methodology works well with the business
municated and understood by the implementation team.
and project constraints and circumstances.

Project organization Classic governance

structure
Establish an effective steering group that understands
Align business streams with functional workstreams
enough of the project details to actively steer the project
for better efficiency and structure, and make sure the
based on meaningful, actionable, and evidence-based
business stream is directly involved in the project.
information.
Secure a strong executive sponsorship and active
Create a risk register with meaningful and actionable risks
engagement from senior business stakeholders within
that align with project priorities and are actively addressed.
each business stream.
Implement stage gate or milestone-driven planning and
Ensure cross-team collaboration in which members of
reviews as checkpoints to better track and communicate
each workstream are involved in other workstreams to
implications of the project status.
avoid working in silos.
Implement design review boards to ensure new designs
Plan and budget the project resources in proportion to
and changes operate within the boundaries of the
the effort and complexity of the project.
solution blueprint and don’t adversely affect
Define accountability and responsibility at the project other designs.
leadership level. Each stream identifies an owner with
autonomy who is empowered to make decisions.

Implementation Guide: Success by Design: Project governance 184

 Case study
Project governance as a
critical success factor
A large enterprise that was using legacy home-grown applications for
managing their sales and customer service business processes decided
to embark on a new cloud application, Dynamics 365. This was a mis-
sion-critical and complex implementation for this company; it would
affect their key business of providing customer service to their large
enterprise customers, and was the first cloud deployment within a
traditional IT managed landscape. The plan was to release a minimum
viable product (MVP) within a timeline of six months and follow a
phased approach for their users to adopt the new solution smoothly
and without too many challenges.

Although the implementation team ensured that the scope was

aligned with business priorities and clear project goals were agreed
upon, no clear processes were defined for the following key aspects of
the project:
▪ Realistic timelines for the activities planned within the engagement
▪ Understanding of technical and business complexity
▪ Well-defined project organization structure
▪ Assignment of adequate and qualified resources
▪ Effective project updates for the steering group
▪ Full understanding of the product’s out-of-the-box capabilities,
thereby avoiding unachievable product requirements

Implementation Guide: Success by Design: Project governance 185

 As the project began, complexities surfaced early in the cycle. However,
the team didn’t revisit and realign their activities. This resulted in the
following impacts:
▪ Poor quality of deliverables
▪ Mismatch in customer versus partner understanding of requirements
▪ High number of defects identified during initial testing cycles
▪ Incomplete implementation of requirements due to lack of skilled
resources on the project
▪ A continuously shifting end date for final go live, resulting in a
delay of 6–8 months
▪ A complete lack of trust between the customer and partner

After several discussions between customer and partner stakeholders

and a high impact on cost and morale, the team agreed to clearly de-
fine a project organization structure and an approach to continuously
monitor and manage the overall project. The following activities were
incorporated as project governance processes were defined:
▪ Considering the new technology, training needs for both the custom-
er’s business and IT teams were identified so they could understand
both the SaaS world and Dynamics 365 product capabilities.
▪ An architecture board was established that was responsible for
diligently reviewing architectural decisions coupled with the gap
analysis. For any gaps identified, this board provided input to
the change control board to ensure that a change management
approach was followed without any further impact on timelines.
▪ Both industry and product architects were assigned to the project
to avoid any further lack of domain understanding and perform a
clear fit gap analysis.
▪ The steering committee was empowered by the customer’s senior
management to take the relevant and necessary decisions through
the project.
▪ A key challenge that impacted the deliverable quality was that
several activities were planned almost parallel to each other and
were to be done by the same project team. As part of the project
reorganization, activities such as performance testing and vulner-
abilities testing were outsourced. This allowed the existing team to
stay focused on their core functional scope.

Implementation Guide: Success by Design: Project governance 186

 As the project team implemented these various controls and processes,
they were able to redefine their project timelines and have now gone
live within that timeline. This has led them to ensure that they can
apply and tailor these learnings to other projects to avoid similar high-
cost impacts.
As illustrated by this and other case
studies, project governance is the most In conclusion, project governance must not be thought of as an af-
critical factor that entails all the key
elements to make a project successful. terthought. It must be set up from the beginning to work as a flexible
backbone for the entire project.

Implementation Guide: Success by Design: Project governance 187

9 Guide
Environment
strategy
Implementation Guide: Success by Design: Environment strategy 188
It all starts with the foundation.

Introduction
Defining your environment strategy is one of the
most important steps in the implementation of your
business application.
Environment-related decisions affect every aspect of the application, from
application lifecycle management (ALM) to deployment and compliance.

At a fundamental level, a good environment strategy is about obtaining

the right balance with multiple layers of compliance and security,
productivity, collaboration, isolation, performance, and maintainability.
A strategy to manage environment provisioning and access—and
controlling resources within them—is key to:
▪ Securing data and resource access
▪ Organizing and structuring solution components
▪ Governing and managing capacity

In this chapter, we explore several deployment and transition scenarios.

By the end of this chapter, you should understand the factors to consider
when defining your environment strategy, and the available patterns
that will enable you to confidently deliver an environment plan for
your applications.

Implementation Guide: Success by Design: Environment strategy 189

 Environment strategy
Environments are containers that store, manage, and share your organiza-
tion’s data. They also store the data model, application metadata, process
definitions, and the security constructs to control access to data and apps.

A successful environment strategy defines the core principles and

policies for creating, granting access to, and decommissioning
environments of different types, along with the necessary governance
process. These policies should provide a consistent approach for
centrally deployed, IT-driven implementations and for citizen-
developed applications for individuals or small business teams.

A data strategy is likely to inform or impose constraints on an environ-

ment strategy. For example, an organization could build a unified data
platform that different apps and services consume, or there could be a
fragmented legacy data estate that must be accounted for in the cloud
environment strategy. Chapter 10, “Data management,” delves into the
patterns and scenarios that are relevant to environment planning.

It should be apparent from the sidebar questions that environment

To get started with creating strategy is not the responsibility of a single person or team. It can
an environment strategy,
involve IT admins, the architects responsible for the design and imple-
ask the following questions:
mentation of the solution, the information security and compliance
• Where are my users physically located?
teams, and the business stakeholders.
• What are my organization’s data com-
pliance and residency requirements?

• What types of environments are re- Environment-related decisions are hard and expensive to change. As
quired, and when?
we explain throughout this chapter, environment strategy can affect
• What are the different roles (such as
developers, testers, admins, business how solutions are implemented. This exercise needs to be done at the
analysts and trainers), and which envi-
ronments do they need access to?
start of the project—even before beginning implementation—and is
• What’s the access procedure for not something that should be left to the end as a part of a go-live plan.
an environment?

Tenant strategy
• Which apps will be installed in
the environment?

• Will the solution require more than one

production environment?

• Which services and third-party applica- Understanding tenants is fundamental to defining an environment
tions do I need to integrate with? strategy for your deployment. A tenant is a logical structure that rep-
resents your organization. Every cloud service to which an organization

Implementation Guide: Success by Design: Environment strategy 190

Fig. subscribes will be associated with the customer’s
9-1
tenant. A tenant is not dedicated to a specific ser-
Services vice; it has subscriptions for Microsoft services, such
Dynamics 365 as Exchange, Dynamics 365, Power BI, and Azure.
Every tenant is given a Microsoft domain—such as
Microsoft 365
Tenants yourorganization.onmicrosoft.com—but admins can
Azure microsoft.onmicrosoft.com
also associate their own custom domains with the
tenant. For example, as shown in Figure 9-1, contoso.
yourorganization.onmicrosoft.com onmicrosoft.com can be associated with contoso.
com and other domains owned by the organization.
contoso.onmicrosoft.com
You can also federate access to your tenant via your
organization’s on-premises Azure Active Directory (Azure AD) to enable
single sign-on (SSO) across on-premises and cloud services.

A tenant provides the highest level of isolation for a cloud service,

as it represents an organization and differentiates one customer
from another. Tenants never share resources or licenses because
they’re meant to represent an organization and not a department
or business unit. (In some cases, subsidiaries could share a tenant
with the parent organization, depending on the organization
structure, and the tenant model might mirror that structure.)

Isolation approaches also may differ between services. For example,

as shown in Figure 9-2, you could have different subscriptions in
Azure within the same tenant—and each subscription could host
different applications and have a different set of users—because
Dynamics 365 and the Microsoft Power Platform have environments
that isolate data and manage user access.
Fig.
9-2

Test
Operations
environment
Dynamics 365
and Power Platform
Production
Tenant Sales app
environment

[email protected]

SharePoint
Microsoft 365 SharePoint
site

Azure Subscription Web roles

191
 Your environment strategy should provide a balanced level of isolation
to meet the security and compliance needs of your organization. It also
should take into consideration the administration, ALM, and collabo-
ration needs of the project. To define the right tenant and environment
strategy for your organization, it’s necessary to understand the con-
trols that are available for each service to isolate code, configuration,
data, and users. It’s also important to note that the isolation provided
Fig. by services doesn’t necessarily reflect the underlying infrastructure or
9-3
design of a cloud service. A separate tenant doesn’t mean it’s a sepa-
Global single rate server farm—and having a separate environment doesn’t give you
a different front end.
tenant vs. global
multitenant setup Global single-tenant setup
Global single tenant Using a single Microsoft tenant to represent the organization in
the Microsoft cloud (Figure 9-3) is a common setup. It provides
unified administration for user access and licenses, enables seam-
Product
less integration between services, and lets the organization share
tenant resources.
Account Contact Addresses
All Dynamics 365 and Power Platform environments will be a part of
the same tenant. There could be different apps deployed in different
environments, or they could have different users for each app and
Global multitenant
environment, but all would belong to the same Azure AD that is
associated at the tenant level. Using the Multi-Geo setup, you could
Product #1
create environments in different countries or regions to meet your
compliance and application needs, but they would remain a part of the
Account Contact Addresses same organization-wide tenant. Sovereign cloud deployment requires
a separate Azure AD and might have additional regulatory restrictions
Product #2 that limit your ability to create environments in different countries
or regions.

Account Contact Addresses

Let’s examine some of the pros and cons for a global single-tenant setup.

Product #3 Pros of a global single-tenant setup:

▪ It aligns to the concept of tenant and the way cloud products
are licensed.
Account Contact Addresses
▪ It allows sharing of licenses and quotas across the tenant per
service. (Microsoft Dataverse storage is one such example.)

Implementation Guide: Success by Design: Environment strategy 192

 ▪ It provides a central place for IT admins to manage the services,
users, and licenses.
▪ It is easier to deploy organization-wide IT policies such as data loss
prevention (DLP) and conditional access.
▪ It is simpler to manage service account configurations and connections
when moving from one environment to another in the same tenant.
▪ It facilitates easier and safer integrations between tenant services,
without the need to elevate privileges to cross service boundaries.
(For example, Power Platform and Exchange Server-side sync is
only possible within the tenant.)

Cons of a global single-tenant setup:

▪ A tenant can only be associated to a single org Active Directory forest.
If there are independent subsidiaries using a separate Active Directory
forest, they can’t be added to tenant, but can use alternatives such as
Azure business-to-business (B2B) for guest-user access.
▪ In a single-tenant setup, you can create tenant-wide policies. All
environments, including the production environment, are governed
by the same policies. Creating exceptions for trying out new services
or testing might require the team to go through multiple information
security hoops. (For example, a proof of concept or pilot requiring a
different Active Directory conditional access policy will attract scrutiny
from security.)

Global multitenant setup

Usually, the tenant structure for the Microsoft cloud software as a
service (SaaS) mirrors the structure of the organization. But some
organizations operate as separate business entities in a country or
region—even though they might be part of a single umbrella brand. In
such cases, an organization might adopt a multitenant setup (Figure
9-3), as each business entity might have its own contractual needs,
regulatory constraints, and licensing agreements.

Alternatively, organizations can have separate tenants for development,

testing, and live systems. This primarily supports the information security
requirements to isolate data and keep users who aren’t involved in pro-
duction from accessing production systems. Still, using separate tenants
is generally not recommended because there could be data and access
isolation at the service level to meet information security requirements.

Implementation Guide: Success by Design: Environment strategy 193

(For example, you might use security groups on environments to
limit access.)

User accounts, identities, security groups, subscriptions, licenses, and

storage can’t be shared among different tenants. So, using separate
tenants might require replication of all other services or creation of
stubs for external systems—per tenant—for all the integrations. Also,
duplication of licenses and quotas across tenants may be needed
to support ALM and performance testing. Admins would have the
additional responsibility of ensuring that all the policies and settings
between tenants are in sync to avoid deployment failure between
tenants. Some organizations do go down this path because of their
historical setup or a lack of understanding of the controls available to
restrict access at the service level.

Let’s examine some of the pros and cons for a global multitenant setup.

Pros of a global multitenant setup:

▪ It works when different subsidiaries are fully independent, with
different systems to manage user identities in separate domains,
and have separate licensing needs.
▪ It provides the highest level of isolation that will simplify the
security approvals but is not generally required.
▪ It is easier to manage and monitor the cost per tenant without the need
for internal cross-charging and tracking usage of shared resources on a
single tenant. (For example, storage is shared on a single tenant.)

Cons of a global multitenant setup:

▪ Licenses can’t be shared across different tenants, which means
that storage, API add-ons, and other tenant-level quotas must be
purchased for each tenant, and you may have to purchase
separate licenses for the same user to access separate tenants.
▪ Overhead to maintain tenant-level Active Directory and DLP
policies consistently across tenants can lead to surprises during
production deployment.
▪ Service-level admin actions, such as ability to copy an environment,
may not be available across tenants, which can make testing or
troubleshooting difficult.
▪ The build automation and continuous integration and continuous

194
 delivery (CI/CD) pipelines that span multiple tenants can be more
complicated and might require manual intervention, especially
when managing connections to the service.
▪ You may have to purchase a significant number of licenses for con-
ducting testing. With load testing, for example, you can’t reliably
simulate a load from 1,000 users using five test-user accounts.
▪ If you’re using capacity add-ons, you will have to make duplicate
purchases for each tenant to develop and test your solutions.
▪ Integrations with other services, such as Exchange, can’t be done
across tenants, which means potentially purchasing licenses for
other Microsoft services for each tenant.

Overall, managing your ALM process across several tenants brings

unnecessary complexity to licensing and technology, and delivers little
value in terms of security, because you might already have the necessary
controls at the service level to limit access to business data.

Key factors affected by

an environment strategy
In this section, we explore the key factors directly affected by an
environment strategy, including security and compliance, ALM,
operations, and functional and nonfunctional aspects of the solution.
Considering these key factors will help you define an environment
strategy that meets the needs of your organization, and set the
direction for the future growth of these applications to meet evolving
business requirements.

Compliance
Security and compliance are critical considerations for an environment
strategy. Each organization needs to ensure that data is stored and pro-
cessed in accordance with local or regional laws, such as data-residency
requirements for a specific region. For example, the European Union (EU)
enforces the General Data Protection Regulation (GDPR) for EU residents,
as well as its citizens outside of the EU.

Implementation Guide: Success by Design: Environment strategy 195

At the onset of the project, your organization must determine your
environment’s compliance requirements. These can vary widely de-
pending on the industry, regulations, business type, and user base, and
will need to be approved by your internal security and compliance teams.

The most common elements affecting compliance are:

▪ The physical location of the environment(s), which usually
determines data residency
▪ The specific features and apps in the scope and their data-
processing requirements
▪ The encryption (at rest and in transit for confidential information)
▪ The disaster recovery procedures to restore the service and data
▪ The data-retention policies, and the ability to back up and restore data
▪ The personally identifiable information (PII) standards for data
protection, and ensuring the service is compliant with the regula-
tory requirements for the industry and region

Application design
The environment strategy can affect the application design. Conversely,
the needs of an application can drive the environment requirements,
and it’s not uncommon for IT systems within an organization to reflect
the actual structure of the organization. Depending on your organiza-
tion’s strategy for data isolation, collaboration, and security between
different departments, you could choose to have a single shared envi-
ronment or create isolated environments. For example, a bank might
allow data sharing and collaboration between commercial and business
banking divisions while isolating the personal banking division, which
reflects the bank’s application design and environment strategy.

Environment policies should not be created in isolation. Creating a generic

organization-wide policy, such as “Every application should use a separate
environment” or “All applications should share a single environment”—
without considering the business requirements or understanding the
underlying data and integration dependencies—could lead to unneces-
sary complexity and fragmentation.

The data store for an application and the supporting business process

196
 plays a key role in the environment decision. If multiple apps for different
departments can benefit from leveraging each other’s data, a single envi-
ronment with integrations can improve consistency and collaboration.
The user experience can be tailored via dedicated apps for different
personas and secure data access using the security model.

For sophisticated systems that need to use multiple environments and

potentially sync data between them, it could be technically possible,
but such patterns require careful consideration of performance and
API capacity—and are best avoided.

App integrations also play a key role, as multiplexing with different

environments may not work. For example, a user may not be able to
simultaneously sync emails to several environments.

Performance
Microsoft cloud services provide a high degree of scalability and
performance. Based on considerations such as network latency,
firewalls, network traffic monitoring, organizational proxies, and
routing by internet service provider (ISP), globally distributed users
can experience higher latencies when accessing the cloud. This is why
we recommend creating a latency analysis matrix (Figure 9-4) for
solutions that have a globally distributed user base.

This exercise gives a fair idea of how the network latency will
affect the user experience based on the location of your environment.
This information can be used to make a balanced choice so most
users have an acceptable level of latency, and application design can
be optimized for users in high-latency locations. For example, using
Fig.
9-4

Environment
User location Device Network Latency Bandwidth
location

North America Corporate

Canada Browser 80 ms 6.4 Mbps
(NAM) network

North America Corporate

Amsterdam Tablet 120 ms 8 Mbps
(NAM) Wi-Fi

Implementation Guide: Success by Design: Environment strategy 197

 lightweight forms or reducing the number of records shown on a page
can enable faster loads.

Also, environment types should be chosen carefully, based on the

workload and the capacity needed. Using trial environments, for exam-
ple, will not deliver the same level of performance as live production
environments, so it may not be suitable to develop or test the solution
on a trial.

Scalability
Scalability of the SaaS platform is a critical consideration for business
applications. Traditionally, scaling up in on-premises deployments was
about adding more servers or more CPU, memory, or storage capacity
to existing servers. In a cloud world with elastic scale and microservice
architecture, the server could be replaced by an environment and the
compute and data transfer units by the API capacity. (This is just used
as analogy—it’s not a one-to-one mapping where one environment
corresponds to a server in the SaaS infrastructure.)

The scalability parameters for a SaaS could be:

▪ How many parallel connections or concurrent users can you sup-
port with one application or in a single environment?
▪ How many API calls is a user entitled to make?
▪ What are the limits on workflow or code executions?

The following is a potential SaaS cloud interpretation of vertical and

horizontal scalability.

Vertical scalability
Organizations commonly operate single environments supporting
thousands of users, and each user has a defined API entitlement based
on the license type assigned. The environment’s storage grows as more
users and applications store their data. Dynamics 365 SaaS services
are built on a scalable cloud infrastructure that can store terabytes of
data to meet the requirements of large enterprises. When it comes
to workflows automation, each environment can have any number
of Microsoft Power Automate flows, each with thousands of steps

Implementation Guide: Success by Design: Environment strategy 198

 performed per day, which can be further scaled using appropriate
license types and add-ons. There are no hard limits on the number
of apps you can have per environment, but you can’t have multiple
first-party apps of the same type in the same environment.

Horizontal scalability
With horizontal scalability, organizations can have several separate en-
vironments, with any number of Power Automate flows on the tenant.
There are no native sync capabilities between environments, and you
still need to take license entitlements into consideration, especially
when it comes to tenant-wide storage and API entitlement.
The effort to
maintain the
Maintainability
solution is directly
Maintainability measures the ease and speed of maintaining a solution,
proportional to including service updates, bug fixes, and rolling out change requests
the number of and new functionality.
environments
involved. If you have several applications sharing common components in
the same environment, you should consider maintainability when
upgrading or deploying a new release. It’s also important to invest in
automation testing so you can run regression tests and quickly identify
any dependencies that could cause issues.

The effort to maintain the solution is directly proportional to the number

of environments involved. For example, testing a release wave or analyzing
the impact of deprecations is easier when there is just one production
environment with the Dynamics 365 Sales app and the Dynamics 365
Customer Service app, compared to a solution that uses different
production environments for the Sales and Customer Service apps.

ALM
ALM includes the tools and processes that manage the solution’s
lifecycle and can affect the long-term success of a solution. When
following the ALM of a solution, consider the entire lifespan of the
solution, along with maintainability and future-proofing. Changes to
your environment strategy will directly affect the ALM (and vice versa),

Implementation Guide: Success by Design: Environment strategy 199

 but it’s important to be clear that environments are not the repository
of your code and customizations.

Environments can be used to separate business functions or for different

purposes, such as development, testing, and production. Typically, at
least three environments are necessary to support a mature
release-management process.

Types of environments
Production environments are meant to support the business. By default,
production environments are more protected for operations that can
cause disruption, such as copy and restore operations. Sandbox
environments can be used to develop, test, and delete as required.

Purposes of environments
▪ Development  One or more development environments are
usually required, depending on the customization requirements and
time available. Development environments should be set up with
proper DevOps to allow for smooth CI/CD. This topic is covered in
more detail in Chapter 11, “Application lifecycle management.”
▪ Quality assurance (QA)  Allows for solution testing from both
a functionality and deployment perspective before the solutions
are given to the business teams in a user acceptance testing (UAT)
environment. Only managed solutions should be deployed here.
Depending on project requirements, there can be multiple testing
environments, including regression testing, performance testing,
and data-migration testing.
▪ UAT  Generally the first environment that business users will have
access to. It will allow them to perform UAT before deployment.
▪ Training  Utilized to deliver training. It allows business users to
practice in a simulated environment without the risk of interfering
with live data.
▪ Production  The live system for business users.

Citizen development
One of the key value propositions of the Power Platform—the underly-
ing no-code/low-code platform that powers Dynamics 365 Customer
Engagement apps—is that it enables people who aren’t professional
developers to build apps and create solutions to solve their own problems.

Implementation Guide: Success by Design: Environment strategy 200

 Central IT still has governance responsibility and creates guidelines
to secure data with DLP policies. But business users should be able to
build apps and automate their work while accessing data in a secure
way. This organic growth of applications by business users facilitates
digital transformation at a massive scale. Thousands of organizations
have adopted this “citizen developer” philosophy to quickly roll out
hundreds of apps across the organization, creating a community of
engaged employees who are helping realize the vision of an agile busi-
ness that can evolve to meet customer needs in days or weeks instead
of months or years.

The environment strategy that enables this kind of organic growth

will be different from a traditional IT-developed solution. It’s crucial
to deliver an agile and automated process where business users can
request environments to enable the maker experience and connect
When planning to data in a secure way while conforming to the standards for data
your environment modeling and application design set by the organization.

strategy, consider
any future phases Future-proofing
and rollouts of Future-proofing is the process of developing methods to minimize
any potential negative effects in the future. It can also be referred to as
the solution, as resilience of the solution to future events.
well as changing
requirements. Your environment strategy needs to take into consideration any future
phases and rollouts of the solution, and allow you to deal with changing
requirements and build a solution that will not limit the business or take
away necessary control. As an example, consider country-specific needs,
as well as variations to and deviations from the standard process.

Environment app
strategy
If you’re deploying multiple business apps on the Dynamics 365 platform,
you will need to decide on an environment app strategy. Should all apps
be deployed in the same environment? Or should each app have an
environment of its own?

Implementation Guide: Success by Design: Environment strategy 201

 There isn’t a standard answer or blanket approach that will work for all
apps within your organization. The best approach for you is the one that
facilitates collaboration and cross-pollination of information between
apps—while also reducing data silos and meeting your organization’s
security, compliance, and data protection requirements.

Multiple-app environment
In a multiple-app environment scenario (Figure 9-5), a single production
environment is used for multiple apps. For example, the production
environment might have the Dynamics 365 Marketing and Sales apps
to enable collaboration between the marketing and sales teams, and
facilitate a quick transfer of qualified leads to sales representatives.
Similarly, having the Sales and Customer Service apps in the same
environment gives the sales team insights into customer support
experiences, which could affect ongoing deals.

With multiple-app deployment, the app data, security models, and

data models are in a common repository, allowing the apps to reuse
any integrations. The security model design will be used to limit access
to data for each app, with the user experience defined by the individu-
al app. For example, the Sales and Marketing apps might use the same
lead table, but security roles and the app definition will control access
to records, fields, and processes.

Let’s examine some of the pros and cons for the multiple-app
Fig.
9-5 deployment model.
Pros of a multiple-app
Multiple-app environment deployment model:
▪ It enables stronger collabora-

Production environment tion between business teams

by decreasing data silos.
▪ It reduces the number of
Dynamics 365 apps
environments to manage, as
well as the amount of effort
needed for platform updates.
Sales Customer Marketing Field Project
▪ It allows reuse of integrations,
Service Service Operations
and its simpler design low-
ers API consumption while

Implementation Guide: Success by Design: Environment strategy 202

 avoiding the need for data sync across environments.
▪ It simplifies reporting and the business intelligence architecture.

Cons of a multiple-app deployment model:

▪ Its security model could become complex, depending on access
restrictions.
▪ Its ALM and release process needs to be coordinated for each
app and will require full regression testing when changing shared
components, making automation testing even more critical.
▪ Its capacity utilization for individual apps is more difficult to track.
▪ Its security, data models, or integrations, if poorly designed on one
app, can affect other apps.
▪ It might have limitations on deploying several apps of same type in
a single environment. (For example, you can’t have several Power
Apps portals with the same template in a single environment.)
▪ It can’t be used if you need to segregate the environments for a
globally distributed user base due to performance, compliance
regulations, or process differences.

Per-app environment
In a per-app deployment model (Figure 9-6), every application gets
its own production environment, with a separate set of environments
to support the underlying ALM and release process. There is complete
isolation of the data, schema, and security model. A per-app environ-

Fig.
ment might seem simpler from a deployment perspective, but it can
9-6
create data silos such that one business process cannot easily benefit
from sharing information with an-
Per-app environment other, leading to additional effort
in building complex integrations.
Production Production Production
environment environment environment Also, the security model is defined
per environment, so you can’t use
the platform security constructs
App App App
to define which data from one
environment will be accessible to
a user in a different environment.
Sales Customer Field
Service Service Consider an example where
the Sales and Customer Service

Implementation Guide: Success by Design: Environment strategy 203

 apps are deployed in separate environments, requiring core customer
records to be synchronized across the two. In this scenario, providing
sales representatives with visibility into support experiences will require
some level of integration, either by copying over the data periodically
or integrating the visual user interface (UI).

This approach might be more relevant for isolated functions within

an organization where there is little need for data sharing and col-
laboration. For example, an internal HR ticketing system for employees
has little to do with the sales process. Network latency, compliance,
and region-specific requirements could mandate the use of per-app
environments, but will introduce data-replication challenges, integra-
tion redundancy, and increased costs for storage, integration, and
maintainability.

Let’s examine some of the pros and cons for the per-app
deployment model.

If you’re deploying Pros of a per-app deployment model:

▪ Its security model is separately managed for each app, which
more than one app, simplifies configuration when there is no need for data sharing
you will need to between environments or development of custom
carefully consider security components.

these factors in ▪ Its ALM and release process for each app can be independent.
▪ Its capacity utilization for individual apps can be tracked easily for
coordination with cross-charging within the business.
the multiple- ▪ Its design issues in one app will not directly affect other apps.
app and per-app ▪ It is the preferred approach if there is a need to segregate the

environment environments for a globally distributed user base due to network

latency, compliance regulations, or process differences.
strategies covered in
Cons of a per-app deployment model:
the previous section.
▪ It can create data silos between business units and departments,
potentially reducing collaboration and value to the business.
▪ It means a linear increase in the number of environments to support
ALM for each app and the administration effort for platform updates.
▪ It largely prevents reuse of integrations and is a potentially complex
solution requiring data sync and complex custom security logic. In
some cases, integrations are not possible across several environments.

Implementation Guide: Success by Design: Environment strategy 204

 (For example, you can’t sync users’ Exchange mailboxes to several
environments at the same time.)
▪ It potentially brings higher API and storage costs due to data and
integration redundancy.
▪ Its reporting and intelligence architecture will be more complex and
require data unification in an external data store, such as a data lake.

Global deployment
scenarios
This section will focus on three common app-deployment models you
could use—along with the tradeoffs, pros, and cons when choosing
one approach over the other—to help you find the best option for
your deployment.

We’re discussing global deployment scenarios in the context of a

single-app deployment. If you’re deploying more than one app, you
will need to carefully consider these factors in coordination with the
multiple-app and per-app environment strategies covered in the
previous section.

Global single environment

A global single environment is just one environment deploying an
app that’s accessed by users across the globe. All the data, processes,
code, customizations reside in a single environment. Based on the
app’s security requirements and regional data-sharing policies, your
organization will have to manage access via security roles. This is the
most common approach, as it enables strong global collaboration and
a unified view of data and processes across different regions, business
units, and legal entities.

While a global single environment simplifies deployment and

maintenance of the solution, centralized process governance is needed
to ensure that different regions conform to a unified process, with
deviations kept to a minimum. If processes are not unified globally,

Implementation Guide: Success by Design: Environment strategy 205

 handling region-specific requirements will require a complex
solution—likely with poor performance—which ultimately will affect
usage and adoption.

Your organization Your organization might have to adjust the security model for each
region to meet local regulations and accommodate cultural differences
might have to around sharing and collaboration.
adjust the security
model for each Let’s examine some of the pros and cons of a global single environment.

region to meet Pros of a global single environment:

local regulations ▪ Data unification across regions and countries enables greater

and accommodate collaboration and helps standardize processes

▪ Settings, updates, configurations, and ALM are easier to manage
cultural differences ▪ Costs of storage and integrations are reduced by avoiding duplication
around sharing and ▪ Intelligence and reporting architecture, as well as master data
collaboration. management, are simplified

Cons of a global single environment:

▪ Change management and governance could be harder when trying to
make a process work for everyone while avoiding political challenges
▪ Poor adoption is possible in regions where users aren’t engaged or
if their requirements are deprioritized
▪ Security model configuration could become complex, as each
persona might have a different regional flavor
▪ Some regulatory and data-residency requirements would be hard
to meet in a global single environment, and could slow down
change and business agility if the effort isn’t well coordinated
▪ High network latency for users in different countries or regions can
affect performance and adoption
▪ Maintenance schedules and downtime are difficult to manage

Global multiple environment

Multiple-environment scenarios support different business units,
departments, teams, countries, or other structures inside an organization.

Each environment has its own database, and data isn’t shared across
environments. This builds a physical security aspect on top of logical,

Implementation Guide: Success by Design: Environment strategy 206

 out-of-the-box security based on security roles. In separating environ-
ments by region or country, solutions included in each environment can
support local business and compliance requirements.

If users are distributed over large distances, particularly for global

deployments, multiple environments may be more suitable because
of the implications (such as latency) associated with the connection
infrastructure, which can significantly affect the user experience.
Distributing environments to provide users with more local access can
reduce or overcome network-related issues, as the access occurs over
shorter network connections.

By creating
In terms of ALM, the solution-release process can be unique for each
multiple production environment, which will support specific business units, departments,
environments, and countries. However, it will increase the number of environments
it’s possible to needed to support development, testing, and training activities, as

easily implement each production environment needs its own environment to support
the solution-release process.
specific business
requirements and Because the data is not unified in a global multiple-environment setup,
have different the reporting and analytics strategy is more complex, and typically

teams delivering there needs to be a data-consolidation exercise for reporting on key

performance indicators (KPIs) and insights about the overall business.
the solution within
different time This approach is common when companies have different require-
frames. ments across subsidiaries, business units, or countries. By creating
multiple production environments, it’s possible to easily implement
specific business requirements and have different teams delivering the
solution within different time frames. Each subsidiary, business unit, or
country operation has control over its own environment.

This model makes it easier for local operations to run and maintain
their solutions, but it also increases the costs, and may lead to siloed
work and deployment.

Let’s examine some of the pros and cons of a global multiple environment.

Pros of a global multiple environment:

▪ Individual environments are only responsible for fulfilling their

Implementation Guide: Success by Design: Environment strategy 207

 own requirements, making them simpler and smaller
▪ Local administration minimizes language and time-zone barriers
▪ Local environments may perform better when closer to the end
user, reducing latency
▪ Compliance with local requirements, such as GDPR, is easier
▪ Data can be physically separated, and there’s flexibility in
managing data residency
▪ Maintenance time windows are easier to adjust

Cons of a global multiple environment:

▪ Data must be unified in a central repository for reporting and analytics
▪ Data isn’t updated across environments or may become specific to
each environment
▪ There’s no single 360-degree view of a customer
▪ It’s more costly
Global
▪ Usersmultiple
may need environment
to access different environments for different
purposes, affecting the user experience
n Hu
tio b
▪ Multipleeninstances
tlu Spo must
ke
-a
be managed
so

nd

nm
A

-sp
ke

ok
o

B
po

vir

eB
en
d-s

Spoke A en

solut
viro
Hub-an

Hub environment
Hub-and-spoke model
ion
nment

Hub solution

A hub-and-spoke model (Figure 9-7) depicts the company’s core

unified business
Sp
o
solutionntas the hub. The core is then distributed and
ke e
v iro C nm
extended by theensubsidiaries, business units, or country operations,
Hu n
b-a l utio
which are depicted as spokes.
nd-spoke C so

Fig.
9-7
The hub-and-spoke model could
Global multiple environment Global single environment be a variation of the multiple-
environment model where each of
tio
n Hu
b-
Sp
the regions have an independent
lu Spo a
n
ent ke tio ok
so

nd

nm environment, but share the core

A

-sp

lu
ke

eB
ok
o

so
po

vir

eB
en

data model.
d-s

sol
Spoke A
Spoke A en

solut
viro
Hub-an

ution

Hub environment
ion
nment

Hub solution
Hub solution
Alternatively, this could also be
achieved with multiple apps in a
Sp t
ok en Sp
eCe
n v iro n
m o ke C
s o l u ti
on single environment, where each
Hu
b-an
d-spoke C solut
io n region is able to independently
manage the variations.

Implementation Guide: Success by Design: Environment strategy 208

Global single environment
There are multiple Environment
factors that could
trigger environment lifecycle scenarios
transitions, ranging Environments transition through various states before they are decom-
from the overall missioned. Some of these transitions are natural to support ALM, but
ALM process to there could be business, compliance, and technical reasons that trigger

changes in business these transitions. As described earlier, environment-related changes are

complex, and they require significant planning and (usually) downtime.
structure.
There are multiple factors that could trigger environment transitions,
ranging from the overall ALM process to changes in business structure.
Let’s explore the various types of transitions that you might see
throughout the lifecycle of an environment. These transitions could
require admins with appropriate roles to self-serve and, in some cases,
seek assistance from product support teams.

Creation
An organization might create a new environment for several reasons.
The purpose of the environment—and considerations such as the
environment type, country or region, apps required, access security
groups, URL and agreed-upon naming conventions, and tiers and service
levels—should be well defined and understood before its creation.

Environment transition
An organization might have multiple types of environments, each
targeting specific use cases, including trial environments, default
environments, sandboxes, and production environments. Be aware
of possible transitions between types, as there could be limitations
and implications to the service. For instance, changing a production
environment to a sandbox might change the data-retention policy and
applicable service-level agreements (SLAs).

Copy
Most often used in troubleshooting scenarios, environment copy lets
you create a copy of an existing environment with all its data or only

Implementation Guide: Success by Design: Environment strategy 209

 the customizations and configuration. Be aware of the effect of storage
capacity and compliance requirements, which could restrict copying of
customer data. Sometimes copy is also used to support ALM instead of
using the source control. This pattern should be avoided, and environ-
ments should never be used as a repository for code and customizations.

Restore
Sometimes, you might need to restore an environment. Depending
on the type of environment and service, there could be several restore
points or a feature allowing you to restore an environment to a precise
time. Restore could lead to data loss, so it’s not really an option for
production environments unless performed in a limited window.

Geo-migration
Geo-migrations could be triggered by changes in regulations, or simply
because of a need to move an environment to a lower-latency region
for a better user experience. This would normally involve creating a copy
of an environment and then physically moving it to a different region. It
almost certainly will involve downtime for users, and requires approval
from the compliance and security teams. It also could lead to change in
service URLs and IP ranges, affecting integrations and network security.

Tenant-to-tenant move
Any change in tenant strategy might trigger a request to move an
environment from one customer tenant to another. This is not a com-
mon request and will require several steps before the actual move. Most
importantly, users must be correctly mapped between tenants and the
record ownerships must be restored. It might require a regional migration
first and could involve several hours of downtime.

Merged environments
Merging multiple environments into a single environment takes
substantial planning and could be extraordinarily complex. Merging
involves the data model and the processes in the app, followed by the
actual data with necessary transformations and deduplication, then the
security model and, finally, the integrations.

Implementation Guide: Success by Design: Environment strategy 210

Split environments
Splitting an environment could be required if an organization
transitions to a multiple-environment strategy. This could involve pro-
visioning new environments and migrating all previous customizations,
along with the relevant data subset. Alternatively, an organization
could create the split by copying the environment and moving it to
another region.

Migration from on-premises to cloud

Many organizations have on-premises infrastructure, but the evolving
needs of businesses and the potential and benefits of the cloud
lead them to migrate to a SaaS cloud solution. Traditionally, this has
involved creating new cloud environments, running a cloud-compat-
ibility assessment, moving the solution components and data, and
redesigning parts of the solution to take advantage of the latest cloud
features. To help speed up this process, Dynamics 365 services offer
a “lift and shift” solution to move data into the cloud. It’s fast and
cost-effective, but by moving a legacy solution into the cloud, organi-
zations may miss an opportunity to optimize their processes.

Administration mode
Administration mode can be used for maintenance when only selected
users can access the environment. Before you do this, assess the sys-
tems that will be affected by putting an environment in administration
or maintenance mode, such as a public-facing portal that connects to
the environment.

Deletion
Deleting an environment removes all the data and customizations, as well
as the option to restore and recover. The process of decommissioning or
deleting an environment needs gated approvals.

Governance and control

Establishing a center of excellence (CoE) for business applications that is
responsible for defining guidelines, creating best practices, and nurturing
organic growth plays a fundamental role in driving business innovation.

211
 Good governance is critical during deployment and for long-term
system operation. But the governance and control processes for a
centrally deployed and managed IT solution might be vastly different

A CoE helps enforce from processes used to deliver a secure ecosystem of business-user-
developed applications.
security and policies,
but also fosters A CoE helps enforce security and policies, but also fosters creativity and
creativity and innovation across the organization. It empowers users to digitize their

innovation across business processes, while maintaining the necessary level of oversight.

the organization. With the changing technology landscape and evolution of security
It empowers users mechanisms, organizations should review and update policies on a
to digitize their continuing basis. The better organizations understand the service

business processes, architecture, the greater the opportunity to fine-tune the controls,
rather than creating overly restrictive policies.
while maintaining
the necessary level To define a robust environment governance policy:
of oversight. ▪ Examine your organization’s usage, application types, user personas
and location, data confidentially, and access control
▪ Evaluate your capacity needs in terms of storage, API call, and
other factors
▪ Understand the release and rollout strategy, and how it affects the
environment lifecycle
▪ Enforce audit requirements at the environment and application levels.
▪ Ensure DLP policies apply tenant-wide
▪ Define admin roles and processes to obtain admin privileges
▪ Monitor deployment and usage

Product-specific
guidance: Operations
Now we’d like to offer some guidance specific to the Dynamics 365
Finance, Dynamics 365 Supply Chain Management, and Dynamics
365 Commerce apps. A team implementing a project with these apps
requires environments to develop, test, train, and configure before
production. These nonproduction environments come in a range of

Implementation Guide: Success by Design: Environment strategy 212

 tiers with pre-defined sizing, topologies, and costs. Understanding
these basics is key to charting out a well-designed environment plan.

What is environment planning?

Environment planning is the process of describing the details of all the
environments required during and after implementation. It includes
procuring and preparing all environments that the team will need to
adequately develop and test the Dynamics 365 solution.

For Finance, Supply Chain Management, and Commerce projects, the

FastTrack team provides a sample environment plan in a matrix format
To learn more, check out the Environment
Planning for Finance and Operations
that lists the type, topology, deployment option, and due date for each
Implementations Tech Talk. environment. This plan is a living document that should be revisited
after each milestone of the project.

The importance of environment planning

Implementing an enterprise resource planning (ERP) and a customer
A note on LCS projects relationship management (CRM) project typically requires collaboration
In the case of Dynamics 365 Finance,
Supply Chain Management, and
among multiple teams that will need to access the apps for different
Commerce, all environments are contained reasons—to develop, build, test, and train. So, part of your plan should
in one project on the Microsoft Dynamics
Lifecycle Services (LCS) portal. This LCS cover making environments available for different teams to ensure
project is hosted on one Azure AD tenant, they can deliver their respective tasks on time. Also, if you plan for all
and includes all the environments used in
the implementation, testing, and for going needed environments early, you will know what the costs are going to
live, such as the production environment.
be throughout the project and at which stage(s) you need the funds,
Typically, an implementation project and you will avoid project delays from last-minute discoveries of
has only one LCS project per Azure AD
tenant. In rare cases, an organization
missing environments.
may decide to have multiple LCS projects.
Reasons to have more than one LCS
project or instance include times when Environments and project complexity
a global implementation’s requirements
When you buy the Finance, Supply Chain Management, and
for data residency, latency, or data
volume can’t be met by one instance. Commerce apps from Microsoft, you get two environments. One is the
Or perhaps different business units are
implementing the product separately as
production environment, which is where your company business will
independent applications. be conducted. It’s a robust environment, with high availability (HA) and
Each LCS project must meet the minimum disaster recovery (DR) capabilities.
licensing requirement. Currently, you
need to open a support ticket to request
additional project, but soon this will be a In addition to the production environment, you will receive one sandbox
self-service activity.
environment that’s a standard acceptance testing (tier 2) instance for
the life of the tenant. It’s a nonproduction, multibox instance that can be
used for testing, UAT, integration, and training needs.

Implementation Guide: Success by Design: Environment strategy 213

 (Because it’s a basic tier 2 instance, it may not be enough for every
type of testing. For example, performance testing needs a bigger
instance to handle larger data volumes and more users.)

Depending on the project complexity, these environments may not be

sufficient. Figure 9-8 depicts a sample chart comparing the additional
requirements of two projects.

For less complex projects, you can use the out-of-the-box tier 2 instance
for testing. You still need to consider development environments for
development activities, with one development environment per devel-
oper. These environments have Microsoft Visual Studio installed.

You should also think about the build environment. Source code
developed in the development environment is synched to Azure
DevOps. The build process will use the build environment, which is
a tier 1, to produce code packages that can be applied to sandbox
and production.

Alternatively, you can use Microsoft-hosted build agents to do the

builds instead of having a dedicated build environment. The limitation
Fig. of this approach is that you can’t run unit tests, and there is an overall
9-8

Environment plan sample

Finance and
Environment purpose Supply Chain Dataverse Standard subscription Less complex project More complex project
Management

Testing or training 2 Sandbox

Production Sized Production

Development 1 Development

Build 1 n/a

Golden configuration 1 or 2 n/a

Data migration 2+ n/a

Performance testing 4 or 5 Sandbox

Training 2 or 3 Sandbox

System integration testing 2 or 3 Sandbox

Other/ad hoc n/a Trial or sandbox

Finance and Supply Chain Management environment Dataverse environment

Implementation Guide: Success by Design: Environment strategy 214

 time limit per month. Assess the limitation and decide whether you
need a dedicated build environment.

Next is the “golden configuration” environment, for configurations

that are fully tested and those you want to retain and transfer to other
sandboxes and production environments.

You can consider a data-migration environment to move data from

legacy systems, and a performance-testing environment—which is a
higher tier (4 or 5) environment—to conduct performance tests.

When it comes to more complex projects, you may need more

environments. For example, you may wish to conduct testing in
parallel if multiple teams are developing and testing parallel workstreams.

Production environment and

go-live assessment
When you’re preparing to go live, the first thing you need is access to
your production environment. Project-wise, a production environment
is deployed after all customizations are code-complete, UAT is finished,
the customer has signed off, and there are no blocking issues.

For a Finance, Supply Chain Management, and Commerce project, you

won’t have access to this environment on day one, as Microsoft needs
to complete few prerequisites—including a go-live assessment—
before we can release the production environment.

In a go-live assessment, a solution architect is assigned to review the

project and provide an assessment of potential risks, best practices,
and recommendations for a successful go live. In some cases, the solu-
tion architect might highlight risk factors and ask for a mitigation plan.
When the assessment is completed, the solution architect will indicate
that you’re ready to request the production environment in LCS.

Self-service deployments
For Finance, Supply Chain Management, and Commerce environments,

Implementation Guide: Success by Design: Environment strategy 215

 there is a new type of self-service deployment that’s rolling out
incrementally across the globe. The benefits are easier and faster
deployment times, and there are a few differences from the earlier
Please note that the production environment
infrastructure as a service (IaaS) architecture. For example, there is no
is used exclusively for running your business
operations and shouldn’t be used for testing. remote desktop access; to access the SQL server database, you will
You will be able to perform the
cutover and, if planned, “mock” the
have just-in-time access.
cutover in production.

General recommendations
▪ Plan for environments early in the project and revisit the plan at
regular intervals.
▪ Define a consistent naming standard for your environments. For
example, a gold environment should have “gold” in the name.
▪ Have a regular schedule to deploy updates and import fresh
data (if needed).
▪ Keep all environments in the same region if your business is in
Assign the one region. For example, avoid having a test environment in one
environments to geographical location and production in another.
an owner who will ▪ Deploy environments by using an unnamed account, such as
dynadmin@your_organization_name.com. Assign the environments
be responsible for to an owner who will be responsible for their status and mainte-
their status and nance. We strongly recommend using the same dedicated admin
maintenance. account on all environments.

Test environments
▪ Consider the number of testing environments you will need
throughout the project. A lack of environments may prevent
concurrent testing activities and delay the project.

Training environments
▪ Make sure all users have access with appropriate roles and
permissions, which should be the same roles and permissions
they will have in production.
▪ Plan for downtime and have a communication plan to alert users.
(Zero downtime is the eventual goal.)
▪ Ensure all integrations are set up and working, so users can
experience the end-to-end cycle of a business process.

Implementation Guide: Success by Design: Environment strategy 216

 Gold environment
▪ Make sure no transactions happen in a gold environment. There should
be a process to bring tested configurations into the gold environment.

Data-migration environments
▪ Assess whether you need a dedicated environment for data
migration, which is a disruptive task that can’t generally
coexist with other types of test activities. Multiple data-migration
environments may be needed to avoid conflicts if multiple parties
are migrating data concurrently.
▪ Account for data-migration performance in environment planning.
Depending on the size of the data-migration task, it may be necessary
to use a tier 3 or higher environment to perform data-migration
testing. (You can also use an elevated cloud-hosted environment.)

Pre-production environments
▪ Assess whether there is a need for a separate environment to
test code or configuration changes before they’re applied to
production.
▪ If there will be continuing development of the solution after
you go live, you may need a separate pre-production environment
to support concurrent hotfix and hotfix test activities. (This
environment should have the same code base and data as
production, so a like-for-like comparison can be performed for
any new changes before they’re applied to production.)

Performance-testing environments
▪ Plan a specific environment for performance testing, or you won’t
be able to do performance testing activities in parallel with other
test activities
▪ Avoid doing performance testing in a cloud-hosted environment, the
production environment, and tier 1, tier 2, or tier 3 environments
▪ Use of tier 3 or smaller environments typically won’t provide the
resources required to perform a performance test

Implementation Guide: Success by Design: Environment strategy 217

 ▪ Limitations on the number of threads and database transaction
units (DTUs) available in these environments won’t allow for
modeling of most concurrent workload profiles
▪ Turn off these environments when not in use to save costs

Developer environments
▪ Ensure each developer has an independent development environment

Production environment
▪ Raise production environment requests through support, as you
don’t have direct access to this environment

Product-specific guidance:
Customer Engagement
Throughout this chapter, we have discussed concepts related to envi-
ronment strategy that are product-agnostic and apply to most cloud
deployments. Now we’re going to focus on product-specific resources
that apply to Power Platform and customer-engagement apps.

Environments and
Dataverse in Power Platform
For each environment created under an Azure AD tenant, its resources
can only be accessed by users within that tenant. An environment is
also bound to a geographic location, such as the United States. When
you create an app in an environment, that app is routed only to
datacenters in that geographic location. Any items that you create in
that environment—including connections, gateways, and flows using
Power Automate—are also bound to their environment’s location.
Every environment can have up to one Microsoft Dataverse database
(Figure 9-9), which provides storage for your apps.

Dynamics 365 products and Power Apps

Microsoft business applications are a set of intelligent solutions that

Implementation Guide: Success by Design: Environment strategy 218

 supply a comprehensive view of an organization’s business. These
solutions include the Dynamics 365 products that are connected by
data and intelligence and supported by the Power Platform.

Microsoft unified CRM and ERP within Dynamics 365 and the Power
Platform to make it easier to create apps and share data across all
Dynamics 365 applications. The combination also creates a set of pur-
pose-built apps with threaded intelligence to connect front-office and
back-office functions through shared data. Rich analytical capabilities
provide organizations with deep insights into each functional area of
their business.

Power Apps is a suite of apps, services, connectors, and a data plat-

form that provides an app-development environment where you can
quickly build custom business apps that connect to your business data
stored in either the underlying data platform (Dataverse) or online and
on-premises data sources.

Customer-engagement apps such as Sales and Customer Service are

first-party Power Apps developed by Microsoft.

Fig.
9-9 Tenant: Business
Development Development Development

Testing Testing Testing

Production Production Production

Environment

Business US (default) Business Canada Business Australia

Apps Flows Dataverse Apps Flows Dataverse Apps Flows Dataverse

Implementation Guide: Success by Design: Environment strategy 219

Administration and governance
The Power Platform admin center provides a unified portal for
administrators to create and manage environments, including sandbox
environments, backing up and restoring environments, and copying
an environment. Admins can also manage settings for Power Apps and
Power Automate.

The Microsoft Power Platform Center of Excellence (CoE) Starter Kit is a

collection of components and tools that are designed to help you
get started with developing a strategy for adopting, administrating,
and supporting the Power Platform, with a focus on Power Apps and
Power Automate.

On-premises to online migration services

The Dynamics 365 migration program enables on-premises customers to
simplify and accelerate their move to the cloud. The migration program
also features access to a dedicated team of migration advisors, no-charge
migration assessments, pricing offers, tools (including a CRM platform
assessment tool), and migration support for qualified customers.

Conclusion
As we have seen throughout this chapter, defining an environment
strategy for your organization is a critical step when planning for
deployment. Decisions made about environment strategies are hard
to change and could be very costly later in the program lifecycle.
The wrong environment strategy can create unnecessary data
fragmentation and increase the complexity of your solutions. Early
environment planning aligned to your organization’s digital roadmap
is fundamental to success.

220
 Checklist

Organization environment Assess the impact on the overall data estate. Avoid
excessive fragmentation and promote reuse of existing
and tenant strategy integrations and data flows.
Define environment and tenant strategies and obtain
agreement for them at the program level with all key
Global deployment
stakeholders, including business, IT, security,
and compliance. Account for global deployment scenarios; additional
coordination and agreement may be required to meet
Create a strategy that considers the future growth of the
regional requirements and compliance needs.
solution.
Assess the latency to choose the optimal location—
Create an environment strategy that can support the
global deployments are prone to network latency
ALM processes and necessary automations.
related performance issues.
Create an environment strategy that considers the
short- and long-term impact on licensing, compliance,
Governance and control
application design, performance, scalability,
maintainability, and ALM of the solution. Establish governance processes for provisioning, mon-
itoring, managing the lifecycle, and decommissioning
Create a strategy that considers the potential need for
the environments early on.
citizen development scenarios or collaborative
development with both IT and business users. Ensure the different security personas involved in the
environment management are understood and appro-
Create an environment planning matrix with key
priately assigned.
considerations of the pros and cons to help plan and
visualize the impact. Use the CoE Starter Kit to make necessary adjustments
to citizen development use cases as needed because

Environment app strategy they require further governance consideration.

Create an organization environment strategy that

defines the guidelines for onboarding additional apps.

Implementation Guide: Success by Design: Environment strategy 221

 Case study
Timely environment planning
is a critical factor to success
A multinational retailer with stores in Southeast Asia, Australia, and
New Zealand implemented the Dynamics 365 Finance, Supply Chain
Management, and Commerce apps in a series of phased rollouts
spread over eight months.

The management team was not clear that the environments, services,
and default capacities included in the SaaS subscription license—mainly
the production and sandbox environments and hosted build automation
services—were the minimum needed to run the solution on an ongoing
basis, but were insufficient on their own to complete all the implementa-
tion activities required for the moderately complex rollout.

When the FastTrack team was engaged early in the implementation

process, this issue was addressed as part of the Solution Blueprint
Review (SBR). In discussions with the solution architect, the retailer
and partner members of the implementation team gained a better
understanding of the deployment plans and risks posed by their cur-
rent environment strategy (or lack thereof). After the review, the team
agreed to conduct a deep-dive workshop on environment strategy
with FastTrack, which was one of the recommendations based on the
findings from the SBR.

Implementation Guide: Success by Design: Environment strategy 222

 In the workshop, the solution architect explained that environment
planning is not a one-time, set-it-and-forget-it event. Rather, it’s a
continuing activity that requires close collaboration between key
stakeholders. Careful planning aligned with the project schedule
ensures that the team can be productive, conduct important activities
concurrently without environment contention, and achieve milestones
on time, within budget, and with the desired level of quality.

Next, the project team charted out an environment plan to document

the different environments, their purpose, the exact project phase for
which they were needed, and how long they would be needed. For
example, they acquired a high-grade testing environment for three
months just to conduct their performance testing. They also identified
key activities that might affect the environment plan and incorporated
recurring reviews of the environment plan into the project plan. This
simple but important exercise helped the team feel confident that
they understood and accounted for all important implementation
requirements, and that they were prepared to procure the needed
environments on time to reach all milestones.

The development team also worked on configuring DevOps for multi-

ple release branches, with a separate pipeline of sandboxes to facilitate
go-live phases and ongoing support for each one.

Ultimately, this retailer successfully performed the implementation and

rolled out the solution across multiple countries—and realized that
timely environment planning was a critical factor in that success.

Implementation Guide: Success by Design: Environment strategy 223

Section
Implement
10 Data management
11 Application lifecycle management
12 Security
13 Business intelligence, reporting, and analytics
14 Testing strategy
15 Extend your solution
16 Integrate with other solutions
17 A performing solution, beyond infrastructure

224
10
Implementation Guide: Success by Design: Data management
Guide
Data
management
225
 Deploy faster and
more efficiently.

Introduction
Data surrounds us every day, like a blanket. Whether
you are posting on your social media, scheduling
a doctor’s appointment, or shopping online, the
information collected is one of the most valuable
assets of any business.
With the right data, organizations can make informed decisions,
improve customer engagement, and gather real-time information
about products in the field.

This chapter aims to break down the various functions within data man-
In this chapter, we discuss the many
ways data plays a part in defining a
agement that collectively ensure information is accessible, accurate, and
solution. Data plays a vital role in the relevant for the application’s end users. We focus on the most common
success of every deployment.
discussion points surrounding the lifecycle of data within a project.
You learn about:
• Data governance
• Data architecture Regardless of your role, take the time to consider what is important
• Data modelling
to each person interacting with the data. For example, users of a
• Data migration
• Data integration system focus on their data quality, ability to search, data relevance
• Data storage
• Data quality
and performance, while architects and administrators are focused on
security, licensing, storage costs, archival, and scalability.

Implementation Guide: Success by Design: Data management 226

 Make sure to always ask these questions:
▪ How is the data used?
▪ What use cases require the data?
▪ Where is the data coming from?
▪ Where is the data stored?
▪ Is there a cost (internal/external) to access the data?
▪ Do users need their data on a mobile phone?
▪ Do customers need access to their data?

We use the Success by Design framework to highlight critical questions

like these that must be asked throughout the project lifecycle to ensure
the desired results for everyone.

Governance
Data governance
Architecture
Let us start the discussion with data governance before we start
Modeling unpacking different principles of data management.

Storage
While data management is usually seen as dealing with operational
issues of data, data governance is about taking a high-level strategic
Migration
view of policies and procedures that define enterprise data availability,
Integration usability, quality, and security.

Quality
With artificial intelligence (AI) and machine learning (ML) taking
center stage in most digital transformation projects, and the fact
that the success of these initiatives is highly dependent on data
quality, it is prudent that executives start considering data
governance seriously.

You may have heard of instances where different departments in the

same company do not agree on the definitions of business terms and
the values reported. For example, a sales report generated last month
may experience two departments reporting different values. This could
be happening because the two departments are using two different
systems, which is caused by there not being a common definition of
business terms and values that everyone in the company agrees upon.

Implementation Guide: Success by Design: Data management 227

 With advanced data governance in place, the walls that create data
silos effectively crumble as data is shared in a trustworthy way with a
consistent understanding of business concepts and definitions.

Sharing data within the company is crucial for two economic reasons—
growth and distribution. Data is a non rival resource. It is not a material
resource that if one person uses it, others cannot use it. If data is shared
with everyone in your company, people can start becoming self-suffi-
cient and build on top of it increasing the overall value as it becomes
available to all.

Though data governance is a broad topic, for our discussion we want

to focus on key aspects that can help drive a successful implementation.

Consider the following key pillars when analyzing data governance

within your implementation.

Data stewardship
A data steward is a role within an organization responsible for the
management and oversight of an organization’s data assets, with the
goal of providing that data to end users in a usable, safe, and trusted
way. By using established data governance processes, a data steward
ensures the fitness of data elements, both the content and metadata,
and they have a specialist role that incorporates processes, policies,
guidelines, and responsibilities for administering an organization’s
entire data in compliance with policy and/or regulatory obligations.

Data quality
The quality of the data cannot be thought of as a second or third
process. Data quality is at the front of data governance policies.
It should be thought of as high quality and fit for the purpose
of whatever the intended use is. Driving data’s accuracy and
completeness across the organization is typically managed by
dedicated teams who may use a variety of tools to scrub the con-
tent for its accuracy. Although these tools aid the process more
and more, this is still typically a human responsibility.

Implementation Guide: Success by Design: Data management 228

 The users that are most familiar with their data should typically be
the gatekeepers for cleansing, including standardization and adher-
ence to the policies outlined by the data steward.

Master data management

Most enterprises keep a primary set of data used across the organization
to supplement systems and reporting. The data elements are typically
stored within a master data management (MDM) solution. This includes
customers, accounts, products, and other areas according to business
needs. There are typically rules of engagement that internal teams
requesting access to the master data must follow, which is documented
within the data governance plan and managed by the data steward.

Proper use cases

A data governance plan cannot be set up without understanding the
proper use cases. Every line of business (LOB) has a level of ownership
of their data along with the use cases that drive the solution design. The
same holds true for data governance. The use cases aid in naming the
primary data elements the organization is taking proactive control of.

For example, one retailer wants to better target its customers through
email campaigns, which in the past failed to deliver expected results
due to incorrect contact information being captured. While defining
this use case, the company also set up the right data governance
principles that ensure what data quality is required for email cam-
paigns. They were challenged to define “good” data to satisfy the use
case. This simple use case uncovered other issues. The company found
that online customers, who buy as guests, could enter any value in the
Without data
email field and there was no validation. This led to data stewards and
governance in LOB process owners setting up new validation processes.
place, organizations
struggle to control Without data governance in place, organizations struggle to control
corporate assets needed to ensure data quality. During the require-
corporate assets ments gathering stage of your implementation, start paying particular
needed to ensure attention to the availability of data required for your solution. Early
data quality. discussion and identification goes a long way in defining your use cases.

229
 For an example of a proper use case, see “A Show-Don’t-Tell Approach
to Data Governance.”

Governance
Data architecture
Architecture
After data governance is understood within the organization, the
Modeling next step is to look at data architecture and the different types of
enterprise data.
Storage

According to Data Management Book of Knowledge 2 (DMBoK 2):

Migration
“Data architecture defines the blueprint for managing data assets by
Integration aligning with organizational strategy to establish strategic data re-
quirements and designs to meet these requirements.”
Quality

Data architecture is a high-level concept, traditionally part of enter-

prise architecture. In simpler words, data architecture describes how
data can be provided to meet a business’s strategic objectives.

The right data architecture is central to the success of your business’s

data strategy. It focuses on where and how you use your data across
various sources and what technologies are available to harness. It is
important to know where the data is coming from, if it is accurate, and
if it provides the insights needed by the business within the solution.

If your organization has an enterprise-level data architecture in place,

make sure to research policies, procedures, and restrictions that are in
place for accessing the information you need.

For example, an organization may have a number of disparate applications

to support different business units, and one of the legacy applications (not
Dynamics 365) is traditionally used as a central repository for customer
master data. The enterprise data strategy may be to support a master
version of the customer data in order to supply better analytics and
visibility of customer interactions across all lines of business. However,
interactions with a customer, like sales opportunities, sales transactions,
support cases, or general activities, may be managed within your Dynamics
365 solution.

Implementation Guide: Success by Design: Data management 230

 When there is no enterprise data architecture, if you simply draw out
all the existing business applications, their interfaces, and connections,
you come up with a spaghetti diagram. That is a good start to getting
an overview of your integration and data architecture. In fact, the
resulting data flow diagram is an important input into defining your
integration strategy.

Read more about integrations in Chapter 16, “Integrate with

other solutions.”

The key point to data architecture is to create a holistic view of the data
repositories, their relationships with each other, and ownership. Failure to
implement data architecture best practices often leads to misalignment
issues, such as a lack of cohesion between business and technical teams.

During your design and analysis phases, it is important to capture

the data owners, systems, and conceptual flow between systems. This
information helps to meet the integration requirements, as well as
providing a reference for the broader project team.

When you prepare your data architecture, use the questions below to
guide you.
▪ How do you use your data?
▪ Where do you store your data?
▪ How you manage your data and integrate it across your lines of
business?
▪ What are the privacy and security aspects of the data?

Before you can start working with data, it is important to know what a
typical enterprise data is made up of, as illustrated in Figure 10-1.

Types of enterprise data

There are several distinct types of physical data to be aware of. This
type of data typically falls into one of four categories: master data,
configuration data, transactional data, and inferred data.

Implementation Guide: Success by Design: Data management 231

 Master data
Master data is your source of common business data and represents a
non transactional, consistent, and uniform set of tables. The details are
maintained and controlled for accuracy. For example:

▪ Customers ▪ Company data

▪ Vendors ▪ Contact data
▪ Products ▪ Account holdings

Fig.
10-1

Azure APIs

Telephony Interactions

Corporate Line of
application business
End
users Company’s
Contacts Activities Master relationships
A accounts
Cases Other

Microsoft
Microsoft Master data integration
USD Dynamics
365 Master Employees
B

Email Exchange
Power server-sync
Applications Master Products
C
External
knowledge Knowledge
Customers base base
self-service Master Product
articles
D alerts

SharePoint SharePoint
document
library

Financials
Corporate
accounting
application

Implementation Guide: Success by Design: Data management 232

 Configuration data
Configuration data is data about different setups needed to prepare
your production environment for operational use. Besides importing
master data like customers and products, you need to import the
configurations data. For example:
▪ Currencies
▪ Tax codes
▪ Modes of payment
▪ Address (countries, states, postal codes, etc.)

Transactional data
This type of data is generally high in volume due to the nature of its use.
Transactional data typically refers to events or activity related to the
master data tables. The information is either created automatically or
recorded by a user. The actual information could be a statement in fact
(like in banking), or it could be a user interpretation, like the sentiment of
a customer during a recent sales call. Here are a few other examples:
▪ Communication history
▪ Banking transactions
▪ IoT transactions
▪ ERP transactions (purchase orders, receipts, production orders, etc.)

Inferred data
Inferred data is information not collected by the business or users.
Typically, this information is automatically generated based on other
external factors, which adds a level of uncertainty. For example:
▪ Social media posts
Governance
▪ Credit score

Architecture
▪ Segmentation

Modeling
Storage
Data modeling
With a base understanding of the data governance and data architecture,
Migration we can now focus our attention on how we plan to store the information
we collect. Data modeling should always be completed before any config-
Integration
uration begins. Let us start with a basic understanding of data modeling,
Quality recommended practices, and how it is related to a project.

Implementation Guide: Success by Design: Data management 233

According to the DMBoK 2, “The process of discovering, analyzing,
representing and communicating data requirements in a precise form
is called the data model.” While data architecture is at an elevated level
and involves a data architect looking at the business requirements
broadly, data modeling is at a lower level and deals with defining and
designing the data flows in precise detail.

Data architecture and data modeling need to work together when de-
signing a solution for a specific business problem. For example, if your
organization wants to implement an e-commerce solution, you cannot
do that unless somebody knows and has defined the existing data ar-
chitecture and data models, different integrations in play, existing data
imports and exports, how customer and sales data is currently flowing,
what kind of design patterns can be supported, and which platform is a
better fit into the existing architecture.

Likewise, if your organization wants to implement a data lake, some-

body needs to answer questions like what the existing data flows are
both in and out of the business applications, what is their format, and
what are the existing data warehouses we need to migrate to data lake?

It is also recommended that the data model conforms to the

Common Data Model standard, which does not deviate from or
repurpose existing out-of-the-box tables, to ensure cross application
compatibility and future readiness. Common Data Model provides a
shared data language for business and analytical applications to use
and makes it possible for data and its meaning to be shared across
applications and business processes such as Microsoft Power Apps,
Power BI, Dynamics 365, and Azure. Refer to For more information,
refer to Common Data Model.

What is data modeling?

A data model is a visual model showing how data flows through your
system and how different tables relate to each other. Data models
define the relationship types between tables and abstract a database
to a visual representation that is easy to understand.

234
 If your project requires creation of new tables to support storing data
for some specific business processes, it is recommended that you
create a data model to define and understand the schema and data
flow. But if yours is largely a vanilla implementation or has minimal
Fig.
10-2 customizations, you may not need to do this. It is recommended to
discuss with your system
integrator and solution architect to
1 Technical contact 0
Customer see what works best for your case.
Customer
contact
1 Account manager 0
There are multiple types and
1
standards for data modeling,
To technical
contact
including Unified Modeling
Language (UML), IDEF1X, logical
0
data model (Figure 10-2), and
Monthly 1 Attached 1 Market physical data model (Figure 10-3).
report email report PDF A common type of physical data
model is an entity relationship
diagram (ERD).

Fig.

Data storage
10-3

Supplier Delivery Order detail

delivery Next, we need to talk about data
Delivery ID Delivery ID
Delivery ID storage and what you can do to
Delivery date Delivery date
Delivery date manage the growth and costs
Supplier ID Supplier ID
Supplier ID
involved. The reality is no matter
where your data is stored, there is
Product Order detail Order always a cost involved. Whether
you are managing data on-prem-
Product ID Order detaill ID Order ID
Supplier ID Product ID Order date ise or in Microsoft’s cloud, you
Order ID Headquarters ID should be aware that storing and
Product quantity
moving your data can impact the
overall cost of the solution.

Branch Headquarters There are several factors to

Branch ID Headquarters ID consider when defining a data
Branch ID storage strategy.

Implementation Guide: Success by Design: Data management 235

 Forecasting capacity
The capacity your environment consumes should be part of the
overall environment strategy. Every environment should be assessed
for storage requirements based on usage. For example, the
development environment does not require as much storage as a
production environment to operate.

First, you need to calculate the approximate volume and size of the
data to properly come up with an estimate. You can gather this detail
from your existing data sources. You should come up with a size either
in megabytes (MB) or gigabytes (GB) needed in production.

The reality is no Based on the estimated size for production, you then need to allocate
a percentage for each environment. Figure 10-4 provides an example.
matter where your
data is stored, there The best practice is to build a data storage forecast for a minimum
is always a cost of three years, including an average increased annual volume. We

involved. recommend that when designing and configuring your different

environments you discuss with your system integrator, and Microsoft,
the amount of storage you anticipate you need and when you need
it by. The business needs to track the data growth to make sure the
Fig. overall size stays within your allotted storage entitlements.
10-4

Environments should be
% of
Environment Use managed with a well-planned
Product
environment strategy. For
Production Contains all data required for release 100 example, a developer should not
be able to simply go ahead and
Contains a sampling of data for create new environments on the
Training training
15
fly. Doing so could create
opportunities to go over
QA Contains a sampling of data for testing 15
entitlements, which leads to
enforcement, which in turn could
SIT Contains a sampling of data for testing 15 limit your ability to deliver on
the application lifecycle
Contains limited data only for management (ALM) strategies.
DEV development needs
5

Implementation Guide: Success by Design: Data management 236

 Read more about environment and ALM strategies in Chapter 9,
Governance
“Environment strategy,” and Chapter 11, “Application lifecycle

Architecture
management,” respectively.

Modeling

Storage
Configuration data
Migration and data migration
Integration Once you have a good understanding of your business requirements,
project scope, and have set up proper data governance, data architecture,
Quality and data models, you need to start the preparations for importing data to
prepare for your solution to be ready for go live(s).

There are two types of data we are talking about: Configuration data
and migrated data.

Managing these two data imports is often the most challenging

part of Dynamics 365 implementations. We often see customers
underestimating the efforts required and projects suffering as issues
surface later. In this section, we talk about plans for configurations,
plans for data migration, different environments, identifying data
sources, ETL process, and finally, staffing the right resources who
can manage these efforts.

Configuration data
Configuration data is data about different setups needed to prepare your
production environment for operational use, for example, currencies and
tax codes. Besides importing master data like customers and products, you
need to import configuration data. Managing configurations is an activity
that needs to happen throughout the lifecycle of the implementation. As
the team understands the business requirements and designs required,
they understand different setups and parameters needed to enable the
business processes required by the business.

Since the Dynamics 365 offerings are highly configurable, you need
tight control over the different configurations required to be set. With

Implementation Guide: Success by Design: Data management 237

 so many development and testing environments and different team
members, there is a high chance of incorrect configurations being set,
leading to errors in processes. In the case of Dynamics 365 Finance,
Supply Chain Management, and Commerce, we recommend setting
up a dedicated golden configuration environment to capture and
maintain the correct configuration data. A gold environment can
help ease the pain in tracking, storing, updating, and controlling the
configuration data. Gold environment settings should be kept pristine,
have tight access control, and no one should be allowed to create any
transactions. You can use the golden configuration database to restore
other environments for testing, data migration, and even initial go live.

Use a configuration plan

Next, it is important to discuss the idea of using a configuration plan. This
is especially important in a phased rollout scenario when you need to plan
to setup a number of environments, legal entities, or business units.

You can use a golden configuration environment to restore into

production for the first phase of go live, but for every subsequent
phase of go lives you do not have that luxury. The double whammy of
a large number of setups and slight but crucial variations of each go
live can be highly daunting, so doing it manually is fraught with danger.
We highly recommend that organizations use a configuration plan.

Configuration plans are a structured list of different master and

configuration data that you want to import for go live and any
subsequent go lives. It is organized in the sequence required by
the system and manages the dependencies between entities. As an
example, if you need to enable purchase orders you need vendors
first, if you need projects, you need accounts first, etc.

This plan also represents the functional scope for the solution. As
described in Chapter 7, “Process-focused solution,” you start with your
process definition and scope for the implementation. This process
requires enabling functionality to be enabled in the application, and
this functionality requires configuration and master data and even-
tually some transactional data in connection to the first use of the

Implementation Guide: Success by Design: Data management 238

 application, like some opening balances. The configuration plan rep-
resents this scope in terms of system enablement and the correlation
between all of them.

Another benefit of the configuration plan is to provide early visibility

of how those configurations and master data play a role between the
different business units (legal entities) or apps that share the same
database foundation. You can determine if that data is natively shared,
or if it is unique for an entity. Having this visibility prevents you from
having redundant data replication and improves the speed to enable
the solution.

For example, flagging the setups that only need to be imported once
as they are shared across business units or legal entities versus set-
ups that need to be imported for each phase as they are not shared.
A configuration plan can help you be organized and consider the
requirements for all the subsequent business units or legal entities.
Another example, in the case of Dynamics 365 Finance, Supply Chain
Management, and Commerce, can be whether you want to use data
sharing features to copy the configurations across companies instead
of manually importing those configurations into each company.

The configuration plan is your configuration playbook that shows you

scope, dependencies, and data redundancy and on top of that, it can
help you to track the progress of enabling the solution in terms of
configurations and master data required to be loaded.

Note that managing a configuration plan is an iterative process. As

you make progress in the project and discover new requirements, and
thus new setups, you should come back to the plan and update it. The
list is your one-stop shop to tell you what you are going to do for each
go live for each configuration, what data is already available, and what
data needs to be imported. Just like your golden configuration
environment, it’s crucial to keep your configuration plan up to date.

A proper well-maintained configuration plan makes sure the team

considers all the important configurations for all the planned rollouts and
does not miss them causing unwanted errors in production after go live.

Implementation Guide: Success by Design: Data management 239

 This plan also helps the team be prepared for cutover and calculates
the required downtime window to import all necessary data. The
team should test the configuration plan a number of times, note the
time it takes for all the configurations, and confirm that the total time
required is within the cutover window.

We strongly recommend that teams put some thought in and create

a configuration plan to manage all their imports as the benefits of
reduced errors and faster imports far outweigh the efforts needed to
create and manage the plan.

Since the configuration plan is essentially a list, many customers prefer

to use Excel sheets as most organizations are familiar with the program.
There are also customers who prefer to maintain the list in DevOps. For
example, you can have one work item for currencies and another work
item for tax codes with exported data entities attached to the work items.
It does not matter which option you choose, as long as your team mem-
bers can access and update the list when necessary.

A sample plan for managing configuration imports for Dynamics 365

Finance, Supply Chain Management, and Commerce can be downloaded
from Sample Configuration Plan using Data Entities.

Data migration
The second type of data we are dealing with is migrated data. Data migra-
tion, in simple terms, is the process of moving data from one data model
to a different data model for future use. In the case of Dynamics 365, data
migration refers to moving data from a legacy system to a Dynamics 365
application. Whether your legacy system was another application or dis-
parate Excel spreadsheets, you may have been capturing data to run your
business and you need that data when you move to your new Dynamics
365 application.

Migrated data is either master data such as customers, products, and

vendors, or open transactions such as open sales orders, open purchase
orders, on hand stock, and open balances.

Implementation Guide: Success by Design: Data management 240

 Migration planning
When deploying a solution, data is particularly important. Different
departments within the organization can find it challenging to support
certain activities when there is no data in the system.

Typically, when replacing another business application, relevant data is

migrated into the new system during the deployment so the users can
see relevant business information when they start using the application.
It is especially important to manage the timeline of data migration
activities in the project plan so you can allocate time and people to the
activities. For example, you need to track activities like migration plan-
ning, identifying data sources, data mapping, ETL (extract, transform,
and load), importing data for UAT, and importing data for go live.

Explore the parts of the data migration lifecycle in Figure 10-5. Take time
Fig.
to discover the mandatory data required by your solution and analyze
10-5
the data types and systems that
source the data to be imported.

Migrating data is a time-consuming

Discover and expensive process that often
lacks full visibility and under-
standing of the use cases. The
use cases help define how the
data is used within your
Verify Analyze
Data solutions processes.

migration
When building a plan, keep in
lifecycle mind that data migration activities
can be a disruptive task and should
not co-exist with other testing
activities, so it is advised to procure
Implement Plan a dedicated high tier data
migration environment.

It is recommended to implement
at least one round of system

Implementation Guide: Success by Design: Data management 241

integration testing (SIT) and user acceptance testing (UAT) with migrated
data. During testing, developers and users are able to verify the
migrated data and ensure it meets the requirements.

Defining a migration strategy

Data migration is a major undertaking and is often run as a parallel
activity to the development and testing activities and spans multiple
phases. Several decisions need to be made regarding the scope of
the migration, environments, source databases, and the tools used to
extract and transform.

There needs to be clear business requirements defining the data migration

scope, timelines, and expected cutover plans before a project kickoff. You
should raise concerns immediately if these details are not clarified.

▪ Identify the source for where the data needs to be pulled

and migrated.
▪ Define the environments that are needed (staging environment).
▪ Is there a mapping document for the tables/fields between old
and new systems?
▪ Do you have the right analysts and developers resourced?

Multiple sources data

The source data can come in all shapes and sizes. From a salesperson’s
Excel spreadsheet to a legacy systems SQL database. Knowing all the
sources allows time to analyze the options for extracting the information.

▪ SQL database
▪ Third party database
▪ External webservices
▪ Access
▪ Flat files/Excel

In advance and during the project, it is recommended that the LOB

begin the process of identifying, consolidating, deduping, scrubbing,
and cleansing the data. The goal is to limit the amount of transformation

242
 needed during migration. The more logic required to transform, the
slower the migration takes to load.

Fig.
10-6
Environments
Another missed factor is sizing of the import and staging databases
Extract required for running migration tooling and cleansing of data. You need
to make sure environments used are sized appropriately to handle the

Source volumes of data in scope. It is recommended practice to have all data-

bases and environments running under the same location and region.
database During environment planning, the appropriate environments should
be sourced for data migration.

Data mapping
The process of data mapping can start once the solutions for data
modeling have been defined. You should be organized to keep the
nsform
Tra
process systematic and simple.

▪ Use an Excel spreadsheet with each tab representing a system table.

Staging ▪ Capture all the fields being mapped together along with any

database transformation requirements.

▪ Pay close attention to the target system you are loading data into.
For example, data types should match between systems or trans-
formed accordingly.

Extract, transform, load (ETL) tooling

Once you have identified the sources, set up the environments, and
Load
documented the data mappings, then the logical next step is to extract
the data from the source, transform the data for target, and load the

Target data into the target, as illustrated in Figure 10-6. You can use many
types of tools and approaches when migrating data to your Dynamics
database 365 solution. Some of the very standard options include:
▪ Data import/export wizards
▪ Azure Data Factory
▪ SQL Server Integration Services (SSIS)
▪ Third-party integration products

Implementation Guide: Success by Design: Data management 243

 Refer to the product-specific section later in this chapter for a list of
some of the most common options.

Governance Roles and responsibilities

Customers and partners should staff the project team with the right
Architecture
resources who understand data and the tools in Dynamics 365.
Modeling
Examples of some important roles and responsibilities in Figure 10-7.
Storage

Migration
Data integration
Integration Integration is the connecting of one or more parts or components

Quality
of a system to create a more unified experience or to ensure a more
consistent outcome of a process. Integration allows leveraging of
existing services both internal and external without having to rebuild
or migrate existing functionality.

Data integration is done to bring data into the system or out to other
Fig.
10-7 systems. Typically, this happens through an event or in a batch on a
schedule. For example, when a
record is created or updated it
Role Responsibility
would be event driven and the
Assist in designing, planning, and managing the data nightly scheduled transmission of
migration process. data would be a batch.
Data migration
analyst Work with subject matter experts and project team to
identify, define, collate, document, and communicate the
data migration requirements.
Refer to Chapter 16, “Integrate with
other solutions” for more details.
Maintain data and manage it according to data properties
as required by administration.

Data quality
Data steward
Coordinate with stakeholders and provide all definitions
for data.

Once data is migrated and

Design and develop the environments required for
migration. integrated to your solution, it is
Data migration critical that information remains
architect/ Develop data packages as required to move the data
developer between systems. accurate, complete, reliable, and,
Provide initial testing and validation. most importantly, up to date.

Implementation Guide: Success by Design: Data management 244

 You can use various techniques to manage the quality of the data,
including data profiling, data cleansing, and data validation. The
most important aspect of data quality to understand is that data
quality is the responsibility of everybody in an organization. We
see many customers who overestimate the quality of their data and
underestimate the effort it takes to get it into shape. Keeping data
quality top notch requires following all the principles we highlight-
ed in this chapter and equally strong leadership to drive the habit of
managing data on an ongoing basis.

Make your data quality a habit!

Conclusion
Data has taken center stage in all enterprise transformation projects.
Businesses want to be data driven. Having the right data flowing in
your business means you have set up a well-functioning business in
which you have the most up-to-date and accurate information about
your customers and products and can provide a superior experience to
your staff and customers.

The benefits of data do not just stop there. Having the right data
means you can start leveraging ML and AI today to predict what your
customers need tomorrow.

Having proper data governance, data integration, data security, and

data quality are not just best practices; they are necessary for your
business’s survival because every business wants to be data-driven
and compete for customers’ attention. These concepts can help lay the
foundation on which you can start tapping into data to transform your
processes, gain insights, and predict customer behavior.

Data management covers so many different areas, it is easy to get

overwhelmed. By starting with the right people, process, and data, you
cannot go wrong.

Implementation Guide: Success by Design: Data management 245

Product-specific guidance
Up to this point in the chapter, our data related guidance has applied
to Dynamics 365 Finance, Supply Chain Management, Commerce, as
well as Customer Engagement application projects. While both appli-
cations live in the Dynamics 365 ecosystem and customers frequently
adopt both systems, often simultaneously, there are differences
between the two. This can mean differences in how each application
should manage its data.

In this section we highlight some of the different approaches for

both applications.

Customer Engagement
This section includes a number of recommendations and resources
provided in Customer Engagement to help manage modeling, storage,
migration, and archival.

Data modeling
Data modeling is a science, and there are data modeling professionals and
established standards for data modeling. To be effective with Dynamics
365 data modeling, you do not have to be a professional data modeler
or use any special tools. Popular tools like Microsoft Visio can be used to
quickly create a basic ERD diagram that visualizes the relationships and
flow of data between tables. In this section, we discuss some general best
practices for data modeling for Dynamics 365 deployments.

▪ Data models should be updated continuously during a deployment.

It is common for a data model to be designed at the beginning of
a project, but it is very important that it does not stop there. As you
go through the deployment, new fields and tables are added—it
is important to capture these in the data model to make it a living
data model. Recommend to customers that they continue to update
it in order to enhance the system.

246
 ▪ Do not include every table. Some core tables, such as activities,
notes, and users (record owners), are related to nearly every entity
in Dynamics 365. If you include every relationship with these tables
in your data model, the result is unreadable. The best practice is to
include only the primary tables used in your configuration in your
data model diagram, and include only custom relationships with the
user and activity tables to maximize readibility.
▪ Data models should include tables outside of Dataverse. If you
are integrating with other systems via Dataverse data connectors
or virtual tables, or if data flows outside of the Dataverse via an
integration, this data should also be represented in your data
model diagram.
▪ Start simple with the standard tables, then add custom entity
relationships to your data model.
▪ User experience should influence your data model. Sometimes it is
easy to overnormalize your data, but the process could make the
application more cumbersome to use.

Start with what you need now but design the data model in a way that
supports what you are going to be doing in the future. For example,
if you know that down the road you need to store additional details
about sales territories, using a text field for territory now makes it more
difficult to implement than if you use the territory entity relationship.
Plan for what is coming.

Data storage
This section provides product-specific guidance you can use while
implementing or maintaining your solution.

Storage capacity
The storage capacity is a standard calculation within the Power
Platform that is easily managed by the system administrator. The Power
Platform admin center is the tool you should use to maintain visibility
of storage and consumption. Within the Power Platform admin center,
go to Resources > Capacity > Dataverse for more details about your
capacity entitlements, as shown in Figure 10-8. You can access this by
going to the Power Platform admin center.

Implementation Guide: Success by Design: Data management 247

 Storage entitlements
Dataverse capacity (database, file, log, and add-ons) is pooled across the
tenant and shared among all environments and workloads. The first
subscription provides a one-time default capacity entitlement for the
tenant. Learn more about the Dataverse storage capacity.

Storage segmentation
To better understand how capacity is calculated within Customer
Engagement, the following provides the breakout based on storage
type and database tables.

Dataverse database: All database tables are counted for your

database with the exception of logs and files below.

Dataverse files: The following tables store data in file and database
storage:
▪ Attachment
▪ AnnotationBase
▪ Any custom or out-of-the-box entity that has fields of datatype file
or image (full size)
▪ Any entity that is used by one or more installed Insights applications
and ends in “-analytics”

Dataverse logs: The following entities are used:

▪ AuditBase
Fig. ▪ PlugInTraceLogBase
10-8

Implementation Guide: Success by Design: Data management 248

Data migration
This section provides product-specific guidance you can use while
implementing or maintaining your solution.

Approaches and tools

You can use many types of tools when migrating data to your
Dynamics 365 solution. The following is not an exhaustive list, but it
includes some of the most common options.

Do not migrate data  It is often assumed that large amounts of data

must be migrated into Dynamics 365 to provide a complete view of the
customer. However, just because you need to see data in Dynamics 365
does not mean that the data must live in the Dynamics 365 Dataverse
database. Considerations should be taken for the value of the data
weighed against the quality of data, the volume of the data, the
amount of time required to load the data, and potential storage costs
before migrating data. Not physically migrating data can frequently be
replaced with embedded Power BI reports, virtual entities, embedded
canvas apps with connectors to the source data, and other options. For
example, the email auto-capture option in Dynamics 365 Sales Insights
displays activities to salespeople from their Exchange inbox rather
than having to load the data to Dynamics 365. This can be a viable
option to show activity data rather than loading large amounts of data.
Virtually displaying data can also significantly reduce the time and cost
of deployment, as you do not have to take time to load data and test
your data migration.

Power Apps Excel add-in  This add-in can be used to open entities
directly in Excel and create and update records. Records are updated or
created directly in the Dataverse. Not all entities support updates from
Excel, and lookup fields must be manually edited to correctly match.

Get data from Excel At make.powerapps.com, you can choose

“Get data” when viewing an entity and import data from an Excel or
.csv file. This option gives additional control over the mapping of data
than the Excel add-in, and lets you import data from one file into
multiple entities. To help ensure that the format of fields and mapping

249
 is correct, you can export a template from a CDS entity, populate your
data, then import it. Alternatively, you can import a custom file and
provide mapping. Lookup fields must include the primary key or entity
alternate key valued to match correctly.

Power Platform dataflows At make.powerapps.com, you can select

“Dataflows” from under the Data menu and configure an import from
several data sources. Data sources includes common external services
file sources as well as generic web APIs. Data from these data sources
can be transformed prior to import using Power Query.

Legacy Dynamics 365 data import utility  You can import data to
Dynamics 365 entities from CSV, .xls, .xml, and .zip While the Dataverse
API Get Data option is recommended for most flat file imports, the
legacy data import option has several unique capabilities that might
be useful in some cases.

▪ Legacy data import can be used to create new entities and fields
and option set values in Dynamics 365. While this is convenient,
the best practice is to add these items from a solution rather than
from data import.
▪ Legacy data import can import multiple files at the same time
when multiple files are added a zip file.
▪ Legacy data import can resolve lookup fields using values not
contained in the primary or alternate keys.

Dynamics 365 Configuration Data Migration Utility  This is a

tool provided by Microsoft that is designed to move data between
Dynamics 365 environments. If you are migrating data from sand-
box environments to production, this tool is very useful as it allows
you to keep the record IDs consistent between environments.
Consider this when you have entities like country and specific
records are referenced in workflows or flows, if the ID changes
between environments, the process does not work. Configuration
Data Migration Utility was updated in early 2020, so you can now
provide filters to tailor the records included in the Configuration
Data Migration Utility data packages.

Implementation Guide: Success by Design: Data management 250

Extract, transform, and load (ETL) software  For more complex
migrations, such as migrating an entire legacy CRM environment’s data,
manual import from flat files is not efficient and can be error prone.
For more complex migrations, commercially available ETL tools like SSIS,
Azure Data Factory, or a number of third-party ISPs offer services and tools
that can be used to create a migration data transformation specification
(DTS) that can migrate legacy data to Dynamics 365/CDS. This approach
can have the following benefits:
▪ Reusability of migration, allowing the data migration developer to
test and refine the migration multiple times before go live.
▪ Delta sync loads when moving the data from an in-production system
to a new system. Users still use the legacy system until the switcho-
ver to Dynamics 365 happens. If the data is loaded into production
several days before go live, there is additional data created in the
legacy system after the initial load is finished. ETL tools allow the data
loaded to be filtered to only include data changed since the last load,
ensuring that the migrated data is up to date when the users start
using the new system.
▪ Consistency with data integration. Sometimes data is migrated in
an initial load, and then updated via an ongoing integration. In
these cases, it is optimal to use the same tooling that you use for
the ongoing integration to also migrate the initial data load, and
in those cases the same DTS and field mappings may be used for
the data migration as well.
▪ More complex data transformation. When moving data via flat files,
you can choose what data is mapped to the new system, but if you
wish to change or transform the data, it must be done manually in
the source flat files. With an ETL based migration, you have flexibility
to transform the data during the migration. For example, say there
are five different types of accounts in the source data, but you wish
to consolidate them to three, ETL tools allow for this transformation
in the data translation specification.
▪ Updates and upserts. The flat file and Excel imports support
updating records that match record keys, but sometimes your
matching logic is much more complex. Say you want to both
insert and update records (upsert). This is complex and tedious
to do with the out-of-the-box data import options, as you don’t
always have the record IDs. ETL tools allow the data migration

251
 developer to define record matching logic and update, insert,
or upsert records based on any matching criteria. This is also
helpful for making sure that duplicate records are not being
created in the target.
▪ More flexibility in mapping lookup fields. Lookups to other
entities can be challenging for data imports, especially fields like
“customer” that are polymorphic, or fields like “owner” that have
special properties. If your legacy data has owning users that have
a different format, the ETL tool can transform the values into a
format that can be imported.
▪ Many-to-many (N:N) relationships. ETL based imports can easily
import data to N:N relationships and entities like marketing list
member, which is not available from some of the flat file
import options.
▪ Faster import of large data sets. Good ETL tools can use
multi-threading to import data quickly into the common data service.

Power Automate and Azure Logic apps  These can be used to

import data and provide connectors to over 300 services. These
options provide many of the same benefits as an ETL tool. In addition,
since Power Automate includes more than 300 connectors to many
leading services and application platforms, Power Automate can be a
great option to migrate data from these services. Although we have
these options, they are not really meant to perform large data migration
jobs, they could be considered a great option for low-volume or low-rate
delta migrations that are not too demanding on throughput.

Custom API migrations  These can be used to migrate data to

Dynamics 365. This option gives you more control over the migration
and does not require any extra tools, but it requires that you have a
developer to develop and support the data migration.

Data archival and retention

The challenge for most organizations is not necessarily defining a
strategy but rather implementing one. Depending on your solution,
there is most often a need to be a technical solution required to
satisfy the strategy of moving the data.

Implementation Guide: Success by Design: Data management 252

 As mentioned earlier, over time the volume of your data grows as well
as the cost. Implementing a data archival and retention strategy allows
regular shifting of data from one storage location to another. The fact
is, over time your data grows and the SQL server comes under strain.
This impacts SQL server’s performance and thus degrades user experi-
ence. Even though Dynamics 365 is a SAAS application, and you do not
bother with managing SQL server, it is a clever idea to have periodic
data maintenance routines in place to archive and delete unwanted
data to keep database nimble.

For example, you can set up data export schedules to replicate data
to Azure Data Lake, which is comparatively cheaper than Dataverse.
From a best practice’s perspective, you do not want to keep old
historical data in Dynamics 365, data that the business does not need
for day-to-day operations. Once data is moved to the data lake, you
can setup Azure Data Factory to create dataflows, transform your
data, and run analysis and you can use Power BI to create business
reports and produce analytics.

Figure 10-9 shows a modern data warehouse architecture by using

Azure Data Lake Storage.

For more information, read Ingest Microsoft Dataverse data with

Azure Data Factory and Analyze Microsoft Dataverse exported to
Azure Data Lake Storage Gen2 data with Power BI.

More information about reporting strategy can be found in Chapter 13,

Fig.
10-9 “Business intelligence, reporting, and analytics.”

Other services or Ingest Model and serve Visualize

customer business
Azure Synapse Analytics
apps (structured) (data warehouse)
Up to14x faster and costs 94% less
than other cloud providers
Media
(unstructured) Prep and train
Azure Data Factory Power BI
Code-free data transformation and integration Leader in the magic quadrant for business
Logs and files from 90+ data integration connectors Azure Databricks intelligence and analytics platforms
(unstructured)

Store

Azure Data Lake Storage

Microsoft High-performance low-cost data lake
Dynamics 365 available in 54 Azure regions

Implementation Guide: Success by Design: Data management 253

 Operations
Dynamics 365 Finance, Supply Chain Management, and Commerce have
a number of data management tool sets and data maintenance cleanup
schedules, which we describe in more detail here.

Data maintenance
When it comes to data storage, an important factor to consider is
that not all data needs storing. You have lots of logs and staging data
generated that can be safely truncated after a certain time.

This is where you need to set up a plan for data maintenance.

In Dynamics 365 Finance, Supply Chain Management, and Commerce,

cleanup routines are available across various modules, which can keep
the tables tidy. It is important to note that these cleanup routines should
be completed only after detailed analysis and confirmation from the
business that this data is no longer needed. Always test each routine in a
test environment prior to running it in production. This article provides
an overview on what is available today: Cleanup routines in Dynamics
365 for Finance and Operations.

Best practices with handling PII data

▪ Avoid storing unnecessary PII data in Dynamics 365 Finance and
Supply Chain Management if possible.
▪ Identify, tag, and classify PII data that you need to store.
▪ Dynamics 365 Finance, Supply Chain Management, and
Commerce uses Azure SQL database that allows data encryption at
rest and transport.
▪ X++ APIs/patterns to encrypt and decrypt data at columns level
for added security.
▪ Build Edge applications/integration to help store and mitigate
data residency and PII requirements.

Data management toolsets

Dynamics 365 Finance, Supply Chain Management, and Commerce has a
rich toolset and processes available to support customers’ data movement

Implementation Guide: Success by Design: Data management 254

 Fig.
and migration requirements. Customers can use the features in app and
10-10
LCS to combine different approaches. Figure 10-10 highlights the options.
LCS
lifecycle
services Data management workspace
In the case of Dynamics 365 Finance, Supply Chain Management, and
Commerce, first and foremost is the data management framework. This
Data
is the primary tool, and you access it through the data management
management
import/ workspace.
export
Here you have an administrator managing all data-related activities
• Copy legal entity
through data projects using concepts like data entities, data templates,
• Templates
and data packages.

Database
operations You can use this workspace to run a number of data-related scenarios
Backup / restore like copying configurations between environments or loading data as
/ refresh / point in
time restore part of data migration.

• Product to sandbox For more information, refer to:

• Sandbox to product
▪ Data entities overview
• Tier 1 to sandbox (new)
▪ Configuration data projects
▪ Configuration data templates
▪ Configuration data packages
Data sharing
framework
Database operations
Though you can use data entities and data packages to move small
configurations, this may not be practical always.

You may often find it handy to move entire databases.

Database movement operations are a suite of self-service actions

that can be used as part of data application lifecycle management
(DataALM). These actions provide structured processes for common
implementation scenarios such as golden configuration promotion,
debugging/diagnostics, destructive testing, and general refresh for
training purposes.

You can use database movement operations to perform refresh, export,

import, and point-in-time restore.

Implementation Guide: Success by Design: Data management 255

 For example, once your golden configuration environment is ready

References
and all testing cycles have been completed and signed off in SIT and
UAT, you can choose “Sandbox to Production” type database request to
restore this database to production for go live.
Data Management/Data
Warehousing information, news
For more info go to:
and tips - SearchDataManagement
(techtarget.com) ▪ Database movement operations home page
▪ Submit service requests to the Dynamics 365 Service
Insights-Driven Businesses Set
The Pace For Global Growth
Engineering team
(forrester.com)
Data sharing framework
DMBoK - Data Management Body
of Knowledge (dama.org) Cross-company sharing is a mechanism for sharing reference and
group data among companies in a Dynamics 365 Finance, Supply
Chain Management, and Commerce deployment.

This framework is introduced to allow sharing setups and master data across
multiple legal entities. This facilitates master data management when you are
dealing with multiple legal entities and want to designate one legal entity as
master for some setups and parameters data. For example, tax codes may
not change from company to company so you can set up in one legal
entity and use cross company data sharing framework and its policies to
replicate the data across rest of the legal entities.

For more information, go to Cross-company data sharing.

Copy company configuration within an existing environment

If the deployment is a phased rollout involving multiple companies within one
Dynamics 365 instance, a very good strategy can be to use the “Copy into legal
entity” feature. This allows you to designate one base company as the source
template from which you can seed multiple other companies or legal entities.
All the setups are automatically copied into those destination companies. This
way you can templatize your rollouts, allowing you to quickly stand up
companies when needed. This strategy needs some upfront planning and
efforts in doing the setups, but depending on the scale of a company’s rollouts
can be quite an effective strategy savings time and reducing errors.

For more information, see Copy configuration data between companies or

legal entities overview.

256
 Configuration data and
Checklist data migration
Create, maintain, update, and test a configuration plan
throughout the project lifetime. It accounts for all the
Data governance required configuration data you import to support go live.

and architecture Ensure the data migration analyst, data migration archi-
Establish data governance principles to ensure data tect, and data steward create a plan for data migration
quality throughout the business processes lifecycle, that includes identifying data sources, data mapping,
focusing on the data’s availability, usability, integrity, environments, ETL, testing, and cutover planning.
security, and compliance.
Focus on maximizing the data throughput during
Appoint a data steward to ensure data governance migration by following Customer Engagement apps
principles are applied. best practices.

Define proper use cases and make data available to Optimize for network latency by staging in the cloud
support the business processes. and batching requests.

Define proper rules and select applications for master

data management. Data integration
Define a proper data architecture that depicts a holistic Only store necessary data in the app for key processes
view of the data repositories, their relationships with that actively interact with it and choose the right type of
each other, and ownership. Capture the data owners, data store based on the usage.
systems, and conceptual flow between systems during
your design and analysis phases.
Data quality
Data modeling Create a realistic view of your data quality and estimate
the efforts required to perform the necessary cleanup.
Define, clearly document, and keep a data model up to
Ensure that the apps have the necessary validations
date to serve as a blueprint.
and controls to enforce data quality and that you have
Conform the data model to the Common Data Model processes to measure and report on it.
standard without deviations to ensure cross-application
Maintain high-quality data by following the principles
compatibility and future readiness.
in this chapter, and have leadership drive the habit of
managing data on an ongoing basis.
Data storage
Estimate and forecast data storage needs across different
environments and types of data stores used in the solution.

Implementation Guide: Success by Design: Data management 257

 Case study
Building a rock-solid data
foundation, step by step
A global travel company with thousands of staff in the UK, Dubai, and
ANZ had been a user of a legacy homegrown IT system that had ex-
panded over the years.

The system was hard to maintain and fraught with issues as customer
data was divided in a number of applications and databases.

In order to target productivity, achieve compliance, and reduce costs,

the business started a project to move to Dynamics 365 Finance,
Supply Chain Management, and Customer Engagement.

The first phase of the project targeted the UK business and involved
moving over 150 of their business processes to Dynamics 365.

Since the travel business handles people’s personal data, including

sensitive data like passport information and medical certificates, they
had to be GDPR compliant. The team worked with an experienced
system integrator and took a proactive approach to data governance
and data security starting with charting out the architectural blueprint
of their “to be” design. They wanted to be on top of all data outflows
and inflows to make sure there is consistency, integrity, and security
of data. There is one source of truth, so they defined apps that held
the master data and a companywide policy formulation to let every
department know what to find where and the exact process to request,
update, and delete a customer’s record.

Implementation Guide: Success by Design: Data management 258

 They also built in workflows to manage data retention responsibly, like
disposing of information when requested by a customer.

The previous systems were disjointed with data existing in silos.

Different departments had sometimes widely different views of cus-
tomers, which caused confusion and poor service for customers. Call
centers were constantly busy as a customer’s call was routed from
department to department as no one person knew the whole picture.

One of the senior managers remarked that taking a proactive approach

to data governance and coming up with a design that put data at the
heart of every decision, helped the company to become truly data driven.

With the first phase of the UK deployment successful, the company is

now actively working on going live for Dubai and ANZ businesses.

The system integrator also recommended an approach of managing

configuration data for deployments. With businesses in three countries,
there is lot of commonality in setups in the three legal entities that could
be reused. But at the same time, there are peculiarities that need to be
maintained separately. Instead of managing all this complexity manually,
system integrators had good experience using data templates and pack-
ages in Finance and Supply Chain Management. They worked with the
company and came up with a configuration plan checklist that identified
each of the three businesses’ specific setup requirements. This planning
exercise paid back multi-fold when the business discovered how smooth,
quick, and error free their subsequent deployments became.

As part of the project, they also moved their historical databases of

different applications to Azure Data Lake and connected with Power BI
to provide analytical reporting to the end users.

The company now has a unified view of their customers, which helps
them provide better customer service and allows marketing efforts to
be more targeted.

Implementation Guide: Success by Design: Data management 259

11
Guide
Application
lifecyle
management
Implementation Guide: Success by Design: Application lifecycle management 260
Success starts with
self-discipline.

Introduction
Generally, everything has a life and its own lifecycle.
Your solution goes through the same process, starting
with conception, moving through implementation,
then continuous operation, and finally to transition.
A solid application lifecycle management (ALM) strategy brings a
successful solution to customers with complete visibility, less manual
interaction with automation, improved delivery, and future planning.

In this chapter, we talk about ALM in the context of the overall solution
lifecycle. Your solution may use one or more of the Dynamics 365
Business Applications such as Finance, Supply Chain Management,
Sales, Field Service, or Commerce.

Your entire solution lifecycle may go through several iterations. Your

With ALM, you have defined processes and
practices, a structured team, and tools at application lifecycle may also be part of a larger solution lifecycle. ALM
your disposal.
provides defined guidance for people, processes, and tools. It sets the
ALM is the management of your lifecycle stage for a successful solution outcome. Additionally, innovation and
(from conception to operation) of your
solution.
automation are essential for improving the lifecycle of your solution.

ALM includes disciplines that span throughout

What is ALM?
the entire lifecycle of a solution, such as
governance (decision-making), project
management, requirement management,
architecture, development, test management,
maintenance, support, change management,
ALM is managing the end-to-end lifecycle of your solution, starting
and release management.
from procuring the Dynamics 365 license, to mapping your business

Implementation Guide: Success by Design: Application lifecycle management 261

 requirements in your application, designing your solution, extending
your custom requirements, validating and testing the solution consid-
ering business requirements, deploying the solution to business, and
maintaining it through the lifetime of your solution (Figure 11-1).

The high-level lifecycle of your solution could be one or many imple-

mentations of your solution.

ALM focuses on a few key concepts:

▪ Defining best practices, processes, and templates for
implementation
▪ Collaboration between customer, partner, and Microsoft
team members
▪ Multiple phase rollouts of your entire solution
▪ Innovation and automation using tools

Having well-documented processes connected with teams using

various tools will result in team confidence, a complete picture of the
implementation, and ultimately the success of your solution.

It’s very important that the implementation team has the appropri-
ate ALM tooling. ALM tools (such as Microsoft Azure DevOps) are
required to manage all aspects of the solution, including application
governance, requirement management, configuration, application
development, testing, deployment, and support. The ALM tool should
be well connected with all team members as well as all processes. For
Fig.
11-1
Planning example, when a developer checks in the code, they
should mark the changes with a particular work
Maintenance Requirement item, which connects the development work item
with the business requirement work item to the
requirement validation work item.

Deployment Design
Application vs. development
Testing Configuration lifecycle management
Some people may define ALM as improving de-
velopment processes such as version control, build
Development
automation, release deployments. Although this is

Implementation Guide: Success by Design: Application lifecycle management 262

 part of ALM, we refer to this part of the lifecycle as the software devel-
opment lifecycle (SDLC), or Dev ALM.

ALM isn’t just about the development lifecycle, in which we look at

the solution’s development activities including design, develop, test,
and deploy. Beyond the development lifecycle, ALM covers the entire
lifecycle of your implementation, including decision-making, project
management, requirement management, the application development
lifecycle, test management, and solution maintenance.

Why have an
ALM strategy?
From the time you conceptualize your Dynamics 365 solution, you start
the application lifecycle: from the project Initiate phase, to the Implement
phase, Prepare phase, and finally the Operate phase (Figure 11-2).

During the lifecycle of the Dynamics 365 solution, you identify partner
teams, require project management, gather and map business process-
es, develop new processes, perform testing, deploy code, and finally
maintain the solution in production.

Think about taking a trip. You set dates, book flights and hotels, and
plan places to visit. All that planning will likely result in a great vacation.

In the same way, a well-planned ALM will lead to a solution that grows
Fig. your business. With ALM recommended practices, you’re set for success
11-2

Initiate Implement Prepare Operate

Kickoff Configuration UAT Go live

Requirements Development Mock go live Support

Testing Cutover Bug tracking

Application Lifecycle Management

Implementation Guide: Success by Design: Application lifecycle management 263

in your solution implementation. You gain visibility into several areas:
▪ The current work items (requirement, development, testing)
▪ The work items completed, in progress, or planned
▪ The teams that worked, are working, or will work on specific tasks
▪ Issues and risks associated with solution implementation
▪ Development best practices
▪ Code history or code version control
▪ Build and release automation
▪ A testing plan, test cases, and test results against requirements

Your ALM may not be perfect right from the start. But it’s a foundation—
you can refine your ALM practices over time.

Implementation without defined ALM

ALM is the lifeline of your implementation. Let’s return to the planned
trip. You planned your dates and booked your flights, but you didn’t
book lodging or decide on any sightseeing excursions or activities.
Your trip could be a disappointment, and you might end up spending
more time planning when you should be out having fun.

Like a poorly planned trip, if you don’t have effective ALM, you can
expect a negative impact on your implementation, solution quality,
and business satisfaction.

Without effective ALM, you may have poor decision-making, recom-

mended practices may not be followed, and teams are disjointed in the
implementation, which can cause delays.

Is ALM only for complex implementations?

You might think that implementing well-defined processes and track-
ing work items, individuals, and using automation tools is too much
work for a small or straightforward implementation.

However, every implementation no matter its size should follow

application lifecycle components, such as having a project plan,
defined scope, business requirements, and documentation. Your

264
 implementation might not require development or integration, but it
will still have configurations and application version updates.

With effective ALM and well-defined practices, you can keep your
solution healthy and up to date with the latest application releases.

Steps to successful ALM

You should use ALM from solution conception through operation. In
this section, we dive into some of the areas of the ALM process, includ-
ing before and after implementation, project management, process
management, other configurations, development, testing, and finally
maintenance and support.

During implementation
While you’re implementing the solution, you go through multiple
phases: Initiate, Implement, and Prepare. ALM applies to all these
aspects in your implementation:
▪ Project management
▪ Business process management
▪ Application configuration
▪ Development
▪ Testing
▪ Bug tracking
▪ Ideas, issues, risks, and documents
▪ Release management

After implementation
When you’re live in production, you’re in the Operate phase. ALM
continues with the following aspects:
▪ Continuous updates
▪ Independent software vendor (ISV) updates
▪ Maintenance and support
▪ New features
▪ Next phase

Implementation Guide: Success by Design: Application lifecycle management 265

 Project management
Efficient ALM requires having documented processes, templates, and
tools for all project management activities, such as project planning,
cost management, team management, ideas for improvements or
solutions, issues, and risk management. A project manager performs
their responsibilities efficiently when these areas are defined and
documented. For example, you should have a template for the project
plan that the project manager can apply in the Implement phase.

Chapter 8, “Project governance,” discusses project management in

more detail.

Business process management

ALM processes and tools should include managing business requirements
and processes efficiently and effectively. After you define your templates
and tools, provide a framework for functional team members to work
together to define requirements and business processes efficiently. Teams
can track business requirements for business processes and connect
further with configuration and development.

For example, a functional team member can gather and define business
requirements in a given template and track them in Azure DevOps. They
can have it reviewed by assigning a work item to business users and then
store the business requirement document in a repository.

Chapter 7, “Process-focused solution,” covers requirement

management in detail.

Efficient ALM Application configuration

requires having ALM processes and tools should also include managing application
configurations. Before configuring an application, the business should
documented define and review business processes and requirements. Functional
processes, team members should have defined processes with tools to perform fit
templates, gap analysis, and a defined configuration required in the application.
It’s important to track your configuration setup and changes using
and tools.
266
 tools such as Azure DevOps. Typically, configuration is a one-time
activity without documentation and tracking. This leads to confusion,
no visibility, and questions about why particular decisions were made.
You should always maintain your documentation for application con-
figuration during testing, user acceptance testing (UAT), cutover, and
even after go live during maintenance and support.

Development
Having an efficient development lifecycle is one of the mandatory
aspects of ALM. In general, the development lifecycle consists of the fol-
lowing: Design, Develop, Build, Test, and Deploy. Continuous integration
and continuous deployment (CI/CD) is one of the best and latest practices
to enable delivering code changes more frequently and reliably.

After the business requirements are defined and identified, the devel-
opment team should get involved to work on any gaps in the solution.
The development team analyzes the requirements, reviews the functional
design, prepares the technical design, and gets the technical design
reviewed. The development team should use version control, create
development tasks, link check-ins with work items, and prepare a
unit-testing document.

Finally, use build definitions and pipelines for automation and to

identify and fix any build issues. Build automation is vital for continu-
ous integration or gated check-ins. Manual builds are error-prone and
time-consuming. The build process, either manual or automated, is
mandatory for development.

The following example of a CI/CD approach (Figure 11-3) has a central-

ized version control in place, a build agent using build definitions, and
release pipelines to push code to a sandbox or other environments. In
this scenario, Dev A develops Solution A and Dev B develops Solution
B. Dev A gets Solution B (Dev B code) using version control and after a
successful build automation. When the build is successful, the code is
released to the sandbox environment using the release pipelines. After
successful testing, the code is released to the production environment.

Implementation Guide: Success by Design: Application lifecycle management 267

 Fig.
11-3
Developer A Sandbox
Developer A Code Solution A

Developer B Code Solution B

Solution A Version Control Build Agent

Build
Solution B

Successful Product
Developer B testing in Yes Solution A
Developer B Code
Developer B Code sandbox Solution B

Testing
Testing is an integral part of ALM. Under ALM, test management
processes and tools should be defined with templates to help manage
test preparation, implementation, and reporting. It should also include
what tools to use for these steps. Chapter 14, “Testing strategy,”
provides more information about test management.

Maintenance and support

In addition to the steps taken for planning, developing, and deploying
your solution, as part of effective ALM, you should have a plan to make
sure your solution continues to be optimal and healthy.

Chapter 20, “Service the solution,” and Chapter 21, “Transition to support,”
cover maintaining the solution and the support process in detail.

ALM strategy for

Dynamics 365
ALM can improve every aspect of your Dynamics 365 solution.
Dynamics 365 offers various tools to achieve a successful ALM in
your implementation.

Implementation Guide: Success by Design: Application lifecycle management 268

 Team responsibility
Various team members are involved in an implementation, with sep-
arate roles and responsibilities. Multiple teams such as the customer,
partner, and ISV work are involved and must work together.

Every team member, either directly or indirectly involved, should own

the ALM maintenance of the processes and tools they’re using. For
example, the project manager owns and maintains the project plan,
and the tester owns and maintains the test plan. You should also have
cross-team collaboration for maintaining ALM practices.

Azure DevOps
Microsoft recommends using Azure DevOps as the tool for managing
or maintaining your ALM practices and processes. For some areas of
Dynamics 365 (such as Finance and Supply Chain Management), Azure
DevOps is the only version control tool.

In addition to the standard templates in Azure DevOps, you can build

your own custom templates according to your business needs. Azure
DevOps lets you store all your work items, including features, user
stories, requirements, tasks, bugs, and more.

A standard lifecycle is available for each work item, and you can build
your own custom states and rules for each type of work item.

Dynamics 365 also recommends Azure DevOps as the version control

management tool for your code. You can achieve build automation
and release pipelines using build definitions and release definitions.

You can also use Azure DevOps for project management, test manage-
Take some time to learn more about ment, bug tracking, release management, and many more aspects of
DevOps tools on Azure. your implementation.

Operations
In this section, we provide some Finance and Supply Chain
Management recommendations to achieve an efficient ALM in your

Implementation Guide: Success by Design: Application lifecycle management 269

solution. We’ll cover things such as Microsoft Dynamics Lifecycle
Services, business process modeling, development, and version control.

Microsoft Dynamics Lifecycle Services

Microsoft Dynamics Lifecycle Services (LCS) is a cloud-based solution in
which all participants (customer, partner, and Microsoft) work collabo-
ratively, and you manage your implementation from pre-sales, through
implementation, and finally the Operate phase. LCS provides various
checklists and tools to help you manage your project efficiently:
▪ A single location for all your projects
▪ Project onboarding
▪ Project methodologies
▪ Azure DevOps configuration
▪ Business process modeler (BPM)
▪ Cloud-hosted environments (CHE)
▪ Data application lifecycle management (DataALM)
▪ Management of Microsoft support
▪ Customization analysis
▪ Subscription estimator
▪ Issue search
▪ Continuous updates
▪ Service requests to Dynamics Service Engineering (DSE)
▪ Environment monitoring and diagnostics
▪ Asset library

The goal of LCS is to deliver the right information, at the right time, to
the right people, and to help ensure repeatable, predictable success
with each rollout of an implementation, update, or upgrade.

Business process modeler

Business process modeler (BPM) provides abstract explanations of your
solution’s business processes. Business processes are the underlying
foundation for application configuration and application development.

Effective ALM provides defined, documented, and templated processes

to follow in your implementation. You can model your business pro-
cesses in Dynamics 365 Finance and Supply Chain Management using
BPM under LCS.

270
 You can use the BPM tool to define, view, and edit the Finance and
Supply Chain Management out-of-box business processes (in the form
of libraries), which you can use in future implementations. The tool
helps you review your processes, track the progress of your project,
and sync your business processes and requirements to Azure DevOps.

Every business is different in some ways; this tool helps you align your
business processes with your industry-specific business processes and
best practices. BPM libraries provide a foundation for your business
process, and you can add, remove, and edit your processes according
to your solution requirements.

Development
In Finance and Supply Chain Management, Microsoft Visual Studio
is used as the development environment. The development lifecycle
includes the following steps (Figure 11-4):
▪ Each developer uses their own development environment
▪ Developers write source code and check in their code to
Azure DevOps
▪ Developers also sync code from Azure DevOps to get the source
code from other developers
▪ The build takes the source code from Azure DevOps, uses the build
Fig.
11-4 definition, and creates a deployable package

Environment
Developer Developer
Build User acceptance Product
virtual machine virtual machine
testing
1 2

Source code Source code Source code Application Application Release Release
deployable deployable candidate candidate
package package

Azure DevOps Lifecycle Services

shared assets

Tool
Implementation Guide: Success by Design: Application lifecycle management 271
 ▪ The build pipeline also pushes the deployable package to the LCS
asset library
▪ Azure release pipelines work with Visual Studio to simplify deploying
packages to UAT
▪ When UAT is complete, the deployable package is marked as a
release candidate to deploy to production
▪ The Dynamics Service Engineering (DSE) team deploys it to
production using a service request from LCS

Version control
The primary purpose of version control is storing and maintaining
source code for customizations, as well as ISV solutions. You develop
against local, XML-based files (not the online database), which are
For more information, take a look at our
document on how to develop and customize
stored in version control tools such as Azure DevOps. The following are
your home page. recommendations for version control branching:
▪ Consider a using minimum branching option
▪ Consider the following recommended branching strategy:
▫ Development  Developer check-in and testing with develop-
ment data (Trunk/Dev)
▫ Test  Deploying to the tier 2+ and testing with current pro-
duction data (Trunk/Main)
▫ Release or Release XX  Retesting in the tier 2+ and deploy-
ing to production (Trunk/Release) or v-next (Trunk/Release XX)
Fig. ▪ Use the shelve command or suspend changes to keep work safe
11-5
▪ Request a code review to ensure code quality
▪ Check in code when a feature is complete, and include changes
Release Release XX
branch branch from one feature in each changeset
▪ Merge each feature in a separate changeset
▪ Don’t check in code directly to the test or release branches
▪ Don’t check in changes for more than one feature in a single
Test (main) branch changeset
▪ Don’t mark deployable packages from the development and test
branches as release candidates
▪ Don’t merge untested features into the release branch
Code
Development upgrade
branch branch
Figure 11-5 illustrates a recommended branching strategy.

Arrows show the direction

when creating the branch

Implementation Guide: Success by Design: Application lifecycle management 272

 Build automation
Speed is essential for rapid implementation, and build automation is
the key to achieving this.
To learn more about build automation,
check out our guides on how to deploy
and use a continuous build and test The build process is mandatory for any code to run. This process
automation environment, create a build
automation that uses Microsoft-hosted involves compiling the source code and producing binary files (assem-
agents and Azure Pipelines, and update
blies). A database sync also requires a build first because the schema is
model versions in an automated build.
retrieved from the assemblies (and not the XML files).

The Azure DevOps build system provides the following triggers for builds:
▪ Scheduled builds, such as nightly at 6 PM
▪ Continuous integration, such as:
▫ Starting a build as soon as code is checked in
▫ Gated check-in
▪ Manual builds (on demand)

Here’s a checklist to ensure your build goes smoothly:

▪ Make sure the code compiles without errors
▪ Don’t use build environments for development activities
▪ Create a build definition for each branch and for each build trigger
▪ Consider using continuous integration or gated check-in build triggers
▪ Perform automated testing after successful builds
▪ Create a single deployable package that contains all packages
▪ Don’t use Visual Studio to create a deployable package, for the
following reasons:
▫ Deployable packages contain all the code on the development
machine, including experimental code and incomplete features
▫ Deployable packages may not include all the modified pack-
ages, which means they’re not independent and testing results
could vary
▪ Create a deployable package with a build machine so it’s self-con-
tained and contains all applicable modules
▪ Make sure to keep service updates on the build machine less than
or equal to production
▪ Follow a consistent naming convention for your deployable package

Automated testing
As part of development ALM, testing automation should be in place

Implementation Guide: Success by Design: Application lifecycle management 273

 Fig.
to achieve fast-moving code from development to deployment. In
11-6
Finance and Supply Chain Management, you can integrate testing and
validation two different ways:
Author tests
▪ Unit and component level testing using SysTest framework
Unit tests ▪ Automated testing using Task recorder and the Regression suite
Task recorder tests automation tool (RSAT)
Component tests
To keep up with innovation and constant changes in your solution,
it’s critical to invest and build in continuous validation. Achieving this
Integrate tests
requires different components (Figure 11-6).
Test module setup
Integrate with The RSAT significantly reduces the time and cost of UAT for Finance
build system and Supply Chain Management applications. RSAT lets functional su-
per users record business tasks by using Task recorder and converting
Run tests the recordings into a suite of automated tests, without having to write

Continuous source code.

validation
Deployment
Let’s review the key concepts for Finance and Supply Chain
Management deployment:
▪ Deployable package  A unit of deployment that can be applied
in an environment
For more information about automated
testing, we offer tutorials on testing and ▪ Deployment runbook  A series of steps that are generated to
validations, using the regression suite
apply the deployable package to the target environment
automation tool, and acceptance test
library resources. ▪ AX Installer  Creates a runbook that enables installing a package

You can deploy a package manually or through automated deployment

(which isn’t applicable to production). We recommend the following
when deploying Finance and Supply Chain Management applications:
▪ Consider using automated deployment to achieve
continuous deployment.
▪ Make sure that the deployable package that should be applied
is valid.
▪ Make sure that the package is applied in a sandbox environment
before it’s applied in the production environment.
▪ If you want to apply multiple packages, create a merged package
that can be applied first in a sandbox environment and then in the
production environment.

Implementation Guide: Success by Design: Application lifecycle management 274

 ▪ Production deployment is manual. After successful UAT, mark the
deployable package as a release candidate and schedule a service
request with the DSE team.
For more information about deployment,
refer to our guides on how to create ▪ Clean up the old deployable package periodically from the LCS
deployable packages in Azure Pipelines, asset library.
apply updates to cloud environments,
troubleshoot package application issues,
and uninstall a package.

Customer engagement
When organizations implement their software as a service (SaaS) cloud
solution, some level of customization and extensibility typically exists,
with the goal to provide additional value and adjust the functionalities
to specific business or industry needs.

The following concepts are important to understanding ALM when

using the Power Platform.

Solutions
Solutions are the mechanism for implementing Dev ALM in Power
Platform apps. They’re the vehicle that distributes components across
the different environments.

A component represents what can be customized within the solution,

such as site maps, applications, the components of the data model,
forms, charts, or plug-ins.

Solutions can either be managed or unmanaged. You should use

unmanaged solutions in development environments while changes are
still underway. Managed solutions are used to deploy to any environ-
ment that isn’t a development environment for that solution.

You should have a good understanding of the solution concepts in

the Power Platform—the solution is the baseline for the components
across their lifecycle. For more information, refer to our overview of
solution concepts.

Tools
Several tools are available to help automate the process of managing
and shipping solutions, which can help increase efficiency, reduce

Implementation Guide: Success by Design: Application lifecycle management 275

 manual labor, and reduce human error when working with solutions
throughout their lifecycle.

Applications like Azure DevOps provide developer services that support

the activities required to plan, store, collaborate, build, and deploy.
This includes synchronization of solution metadata between develop-
Check out our additional resources for
ment environments and your version control system, generating build
Microsoft Power Platform Build Tools for
Azure DevOps, Power Apps build tools for artifacts, deploying to downstream environments, provisioning or
Azure DevOps, and key concepts for new
Azure Pipelines users.
de-provisioning environments, and performing static analysis checks
against your solution by using the Power Apps checker service.

Consider using the following tools:

▪ Version control system  This category of software tools helps
record changes to files by keeping track of changes committed to
software code. A version control system is a database of changes,
which contains all the edits and historical versions of a software
project. Version control systems allow you to maintain a single
source of truth and recall specific versions when needed. Git is a
popular example of a version control system.
▪ The Configuration Migration tool  This tool enables you to
move configuration and reference data across environments.
Configuration and reference data is different from user and trans-
actional data, and is used to define custom functionality in apps
based on Dataverse.
▪ Package deployer  The package deployer lets administrators or
developers deploy comprehensive packages of relevant assets to
Dataverse environments. Packages can consist of not only solution
files, but also flat files, custom code, and HTML files.
▪ Solution packager  This tool can unpack a compressed solution
file into multiple XML files and other files, so they can be easily
managed by a source control system.
▪ The Power Apps CLI  The Power Apps CLI is a simple, sin-
gle-stop developer command-line interface that empowers
developers and app makers to create code components.
▪ PowerShell cmdlets  The PowerShell cmdlets for administrators,
app makers, and developers allow automation of monitoring,
management, and quality assurance tasks that are possible
through the Power Apps admin center user interface.

Implementation Guide: Success by Design: Application lifecycle management 276

 ALM workshop strategy
The implementation team should plan for an ALM workshop before
project initiation or before project kickoff. The customer and partner
should gather the required processes, tools, and teams before project
kickoff, such as:
▪ Project plan template
▪ Business requirement document template
▪ Functional and technical spec templates
▪ Test plan and test suite template
▪ Defined tools such as DevOps template
▪ Defined roles required for partner and customer teams

In addition to this workshop, the implementation team should plan

for a Dev ALM workshop before starting development activities. Our
FastTrack ALM workshop focuses on your development lifecycle and
provides guidance and best practices such as development environ-
ments, version control, build automation, and release pipelines. It
underlines the importance of having fast builds, quick testing, and fast
deployments using automation, which helps ensure a rapid develop-
ment lifecycle, especially for agile development methodology.

Workshop scope
FastTrack ALM workshops are designed for implementers who want to
make sure that their development approach meets the requirements
of the implementation and is aligned with typical best practices. The
workshop could cover several topics:
▪ Development work management  Focuses on high-level,
day-to-day developer activities. It reviews that the team has devel-
opment guidelines, development best practices, and work items
such as requirement, tasks, and bugs management.
▪ Code management  Reviews your version control, branching,

An ALM workshop code merging, and code reviews strategy. Branching can be simple
or complex depending on the project and phases. It also looks at
provides guidance how customization is checked in, such as gated check-in.
and best practices. ▪ Build management  Looks at your build strategies, such as

277
 manual build or automated build. It mainly reviews your build

References plan for build definitions, different variables, and build triggers for
your implementation. It also reviews whether you’re using a build
environment or Azure-hosted builds.
DevOps and ▪ Release management  Assesses your deployment plan for
Azure Pipelines customizations, ISVs, and continuous (one version) updates. It also
reviews how hotfixes and your code release, or one version updates,
DevOps solutions on Azure
are applied to production and nonproduction environments.
Release pipelines

Sign up for Azure Pipelines

Timing
It’s important to ensure ALM workshops are conducted at the right time.
Operations Ideally, the ALM workshop should be before project kickoff, and the Dev
ALM workshop should be before starting development or sooner.
Implement application lifecycle management
in Finance and Supply Chain Management apps

Business process modeler (BPM) in Lifecycle You should complete both workshops either in the Initiate phase or
Services (LCS) when starting the Implement phase. If you plan ALM workshops too far
Dynamics 365 Finance and Supply Chain into implementation, any findings and recommendations could cause
Management Tools
significant rework.

Customer Conclusion
Engagement ALM is the management of your solution lifecycle from conception to
Application lifecycle management (ALM) with operation. It includes governance (decision-making), project manage-
Microsoft Power Platform
ment, managing requirements, solution architecture, development,
Create your first pipeline
CI/CD, test management, maintenance and support, change manage-
Microsoft Dynamics CRM SDK Templates
ment, release management, and many more areas.
Solution Lifecycle Management: Dynamics 365
for Customer Engagement apps
Your solution lifecycle may go through several evolutions. Each evo-
lution may use the same SDLC methodology. Basically, SDLC is part of
ALM, ALM is far bigger than Dev ALM, and Dev ALM is part of SDLC.

278
 Checklist

Establish the ALM at the onset and align it to the

environments, source control, implementation
methodology, and automation strategy.

Ensure the ALM provides an efficient, reliable, robust,

and automated mechanism to smoothly deploy the
configurations and customizations across different
environments, including peripheral solution components.

Ensure ALM strategy enforces a source control-centric

approach that enables creation of new environments
with the latest configurations without the need to
replicate existing environments.

Use a build automation and CI/CD pipeline for the project.

Consider the structure of the implementation team to

enable effective collaboration while providing exclusive
control to developers where needed.

Follow documented best practices in general and

specifically to each app.

For Customer Engagement projects, download the

whitepaper Solution Lifecycle Management: Dynamics
365 for Customer Engagement and complete the
self-assessment.

Implementation Guide: Success by Design: Application lifecycle management 279

 Case study
Global transport systems
company finds ALM strategy
to be a cornerstone of
implementation
A global transport systems company embarked on a multiple-phase,
multiple-region implementation of Dynamics 365 Sales, Customer
Service, and Field Service with an aggressive schedule for achieving a
minimum viable product (MVP) go live.

To give business stakeholders a taste of the solution before the official

project kickoff, the customer asked their Dynamics 365 implementation
partner to lead a rapid prototyping phase, which the customer referred
to as “Phase 0.” The Phase 0 team included technical salespeople, as well
as consultants who would leave when the project kicked off.

Considering Phase 0 to be a rapid prototype, the implementation team

didn’t see a need to focus on governance and quality. To meet the
deadlines, the team made several quick decisions without assessing the
impact on the overall application management:
▪ Staffing Phase 0 with resources who wouldn’t remain part of the
project team.
▪ Developing the prototype using separate, unmanaged solutions
with different publishers for Sales, Customer Service, and Field Service.
▪ Pivoting to managed solutions for test and production environments
too late in the design phase, after unmanaged solutions were
already built and deployed to these environments during Phase 0.

Implementation Guides: Success by Design: Application lifecycle management 280

 ▪ Merging the Sales, Customer Service, and Field Service solutions
into a single “workstream” solution during the build phase of the
implementation lifecycle.
▪ Resolving errors too quickly during solution deployments because
of pressure from project leadership, instead of taking the time
to understand each problem’s root cause. (Such pressure led to
developers customizing directly within the test environment.)

As the project team prepared for the go live, the seemingly independent
decisions made during the initial phases resulted in deployment issues
that eventually stalled the go live. Investigations by Microsoft Support
confirmed that there was no ALM strategy in place, and identified the
key issues:
▪ Unmanaged solutions caused solution-layering issues for testing and
production, affecting the sanctity and integrity of the environments
▪ A prototype solution employed for production purposes
introduced quality issues
▪ Failure to use ALM practices such as DevOps caused traceability
issues and prompted developers to build functionality that wasn’t
aligned with customer requirements
▪ Suboptimal code was implemented because tools such as solution
checker weren’t used to enforce code quality
▪ Testing without traceability was insufficient, and buggy code was
deployed to other environments

As the old saying goes, “By failing to prepare, you are preparing to fail.”
The MVP go-live date was delayed by 12 weeks and Microsoft worked
alongside the project team to determine the root cause of each issue.
The team eventually acknowledged that a series of seemingly unre-
lated decisions affected the MVP go live, and they sent every team
member working in a technical role to an ALM refresher training. The
project team also confirmed plans that should have been solidified at
the beginning of the project, including an environment strategy and a
solution management and ALM approach.

After the refresher training, the project team started with a “crawl to
walk” approach. During the “crawl” stage, they implemented mandatory
ALM practices with these governance elements:

Implementation Guides: Success by Design: Application lifecycle management 281

 ▪ Cleaning up their production and test environments, and moving
from unmanaged to managed solutions
▪ Implementing a build process to deploy builds from development
to test to production
▪ Establishing a governance process to restrict developer access to
production and test environments
▪ Adding a bug-triaging process that allowed the development team
to troubleshoot and fix issues in development environments and
use the build process to deploy fixes in higher-tier environments
▪ Mandating the generation of solution checker reports

Once these practices were in place, the implementation team moved

toward a partially automated ALM process that included setting up
DevOps pipelines to automate build deployments and auto-generate
solution checker reports. As this case study illustrates, an ALM strategy is
a cornerstone of the entire implementation—whether it’s a fresh imple-
mentation or a rapid prototype elevated to a production-use solution.

Implementation Guides: Success by Design: Application lifecycle management 282

12
Implementation Guide: Success by Design: Security
Guide
Security
283
“Businesses and users are going
to embrace technology only if
they can trust it.”
– Satya Nadella, Chief Executive Officer of Microsoft

Introduction
In this chapter, we look at the fundamental
security principles applicable to Microsoft
Dynamics 365 implementations.
Next, we discuss in more detail how some of these principles apply
differently to Dynamics 365 Customer Engagement, Dynamics 365
Finance, and Dynamics 365 Supply Chain Management applications.
We then address the importance of making security a priority from
day one, with specific examples from each product that build upon the
concepts we’ve discussed. Finally, we look at how to avoid common
mistakes by examining some key anti-patterns.

Security overview
Security is the protection of IT systems and networks from theft or damage
to their hardware, software, or data and disruption of the service.

Dynamics 365 is a software as a service (SaaS) offering from Microsoft. In a

SaaS service, data and applications are hosted with a provider (in this case,
Microsoft) and accessed over the internet. In this deployment model, the
customer maintains ownership of the data, but shares application con-
trol with the provider. Therefore, security, compliance, privacy, and data
protection are shared responsibilities between provider and customer.

Implementation Guide: Success by Design: Security 284

 Figure 12-1 illustrates the areas of responsibility between customer
Fig.
12-1 and Microsoft based on deployment type.

Shared responsibility model

Physical datacenter SaaS PaaS IaaS On-
Prem
Information and data
Devices (mobile and PCs) Responsibility always retained by customer
Accounts and identities
Identity and directory infrastructure
Applications
Network controls
Responsibility varies by service type
Operating system
Physical hosts
Physical network Responsibility transfers to cloud provider
Physical datacenter

Microsoft Customer

Fig.
12-2 Microsoft takes its commitment seriously to safeguard customers’
data, to protect their right to make decisions about that data, and to
Security Implement
be transparent about what happens to that data. On our mission to
strong security
measures to
empower everyone to achieve more, we partner with organizations,
safeguard your empowering them to achieve their vision on a trusted platform. The
data Microsoft Trusted Cloud was built on the foundational principles of
security, privacy, compliance, and transparency, and these four key
Privacy Provide you with principles guide the way we do business in the cloud (Figure 12-2). We
control over your apply these principals to your data as follows:
data to help keep ▪ Security  Implement strong security measures to safeguard
it private
your data
▪ Privacy  Protect your right to control and make decisions about
Compliance your data to help keep it private
Help you meet
▪ Compliance  Manage your data in compliance with the law and
your specific
help you meet your compliance needs
compliance needs
▪ Transparency  Be transparent about our enterprise cloud services
and explain what we do with your data in clear, plain language

Transparency Explain what we

do with your
Microsoft has been defending against threats and providing security
data in clear, protections for our online services since 1994, and we invest over $1
plain language billion dollars per year to continue our commitment to protecting our
customers. For these reasons, we say that our products and services

Implementation Guide: Success by Design: Security 285

 run on trust. You can place your trust in our expertise. This confidence
allows you to focus on running your business. Ninety-five percent of
Fortune 500 businesses run on the trusted Microsoft Cloud.

Compliance
Every organization must comply with the legal and regulatory stan-
dards of the industry and region they operate in, and many are also
subject to additional contractual requirements and corporate policies.
Figure 12-3 lists some standard compliance goals and their implemen-
tation in Dynamics 365.

Microsoft is responsible for the platform, including the services it

Fig.
12-3
offers, and provides a cloud service that can meet or exceed the se-
curity, privacy, and compliance needs of your organization. Microsoft
Compliance goals complies with data protection and privacy laws applicable to cloud
services, and our compliance with world-class industry standards is
Define and document standard
operating procedures that meet
verified. Detailed information about compliance for our cloud services
multiple certification requirements and solutions that help organizations meet regulatory requirements
for data security is available at our online Microsoft Trust Center.
Run the service in a compliant
fashion and collect evidence
As noted earlier, compliance is a shared responsibility. To comply with
Control and audit access to
laws and regulations, cloud service providers and their customers
environments as well as actions
to data enter a shared responsibility to ensure that each does their part. Tools
available at the Trust Center include compliance offerings that help
you comply with national, regional, and industry-specific requirements
governing the collection and use of data, and audit reports that help
Implementation details
you verify technical compliance and control requirements. Specific
Adhere to strict privacy and security tools you can access at the Trust Center include:
practices when building features
▪ Compliance Manager, a cross-Microsoft cloud services solution
and when operating the service
designed to help organizations meet complex compliance ob-
Pass internal and external audits ligations like the General Data Protection Regulation (GDPR). It
performs real-time risk assessment that reflects compliance pos-
Prioritize cross-industry
certifications ture against data protection regulations when you use Microsoft
cloud services. It also provides recommended actions and step-by-
Authentication and authorization
step guidance.
▪ Service Trust Portal, which contains details and documents such as

Implementation Guide: Success by Design: Security 286

 whitepapers, ISO reports, and other resources detailing Microsoft’s
implementation of controls and processes that protect cloud
services and customer data.

Customer responsibility
As a customer, you’re responsible for the environment after the service
has been provisioned. You must identify which controls apply to your
Refer to Microsoft compliance business and understand how to implement and configure them to
offerings for more information
about regulatory compliance manage security and compliance within the applicable regulatory
standards and Microsoft products. requirements of your nation, region, and industry.

Privacy
You are the owner of your data; we don’t mine your data for advertising.
If you ever choose to end the service, you can take your data with you.

Fig. Figure 12-4 lists some standard privacy goals and their implementation
12-4
in Dynamics 365.

Privacy goals
How we use your data
You own your data Your data is your business, and you can access, modify, or delete it at
any time. Microsoft will not use your data without your agreement, and
You know where your data is located
when we have your agreement, we use your data to provide only the
You control your customer data services you have chosen. We only process your data based on your
agreement and in accordance with the strict policies and procedures
that we have contractually agreed to. We don’t share your data with

Implementation details advertiser-supported services, nor do we mine it for any purposes like
marketing research or advertising. Learn more about how Microsoft
All data is classified
categorizes data in the delivery of online services.
Role-based security puts the
customer in charge We believe you should have control over your data. The Trust Center can

Your own logically isolated data

tell you more about how we handle data requests from government and
repository helps maximize the law enforcement agencies.
security and integrity of your data

Documented data-handling Customer responsibility

practices in the Microsoft Online As a customer, you’re responsible for data classification, identity man-
Services in the Trust Center
agement, and assigning appropriate security roles to secure the data.

Implementation Guide: Success by Design: Security 287

 Security
Security is a shared responsibility in a SaaS deployment. This means
that some aspects of security are shared by both the customer and
the provider, other aspects are the responsibility of the customer,
and others are the responsibility of the provider. For Dynamics 365
deployments, Microsoft as the cloud provider is responsible for
security aspects including physical datacenter security, the oper-
ating system, network controls, and providing a secure application
framework (Figure 12-5).

Figure 12-6 lists some standard compliance goals and their implemen-
tation in Dynamics 365.

Fig.
12-5

Microsoft
protecting you
Intelligent
Security
Graph
Industry
Partners
Antivirus
Network
Certifications ...
Cyber Defense
Opperations Center

Malware protection center Cyber hunting teams Security response center Digital crimes unit ...

Conditional
access
Cloud app
security
Event
management
Rights
management
Key
vault
Security
center
... Active
protection
service
Windows
update
Microsoft 365
advanced threat
protection
Smart
screen

PaaS IaaS

Advanced
threat Terminal
analytics app

Azure Active Microsoft SaaS Windows Outlook

Directory Dynamics
365

Identity Applications and data Infrastructure Device

Implementation Guide: Success by Design: Security 288

 Fig.
The following is a list of core security controls available in Dynamics 365:
12-6
▪ Security Development Lifecycle
▪ Datacenter security
Security goals ▪ Data segregation
Safeguard data using state-of- ▪ Encryption
the-art security technology and ▪ Secure Identity
processes of the industry
▪ Authorization
Use the same identity platform as ▪ Auditing and monitoring
Microsoft 365, so users have the
same username and password for all

Control who can access what data

Security Development Lifecycle
The Security Development Lifecycle (SDL) is a process that helps de-
Customer data is isolated from
other tenants velopers build more secure software and address security compliance
requirements by introducing security and privacy considerations
throughout all phases of the development process. It consists of the
following phases: training, requirements, design, implementation,
Implementation details verification, release, and finally response (Figure 12-7).
State-of-the-art physical security
measures
Datacenter security
24/7 incident response teams to
mitigate threats and attacks Microsoft designs, builds, and operates datacenters in a way that
strictly controls physical access to the areas where your data is stored.
Encryption of data transferred
between our datacenters and at rest Microsoft understands the importance of protecting your data and is
committed to helping secure the datacenters that contain your data.
Single sign-on federated through
We have an entire division at Microsoft devoted to designing, building,
Azure Active Directory
and operating the physical facilities supporting Azure, which delivers
Role-based security Dynamics 365 (Figure 12-8). This team is invested in maintaining state-
of-the-art physical security.
Fig.
12-7

Training Requirements Design Implementation Verification Release Response

Core security Establish security Establish design Use approved tools Perform dynamic Create an incident Implement incident
training requirements requirements analysis response plan response plan

Create quality gates Perform attack Desperate unsafe Perform fuzz Conduct final
and bug bars surface analysis functions testing security review
and reduction

Perform security Use threat Perform Conduct attack Certify release

and privacy rich modeling static analysis surface review and archive
assessments

Implementation Guide: Success by Design: Security 289

 DDoS defense system
A distributed denial of service (DDoS) is an attack in which multiple
compromised computer systems attack a target, such as a server,
website, or other network resource, and cause a denial of service for
users of the targeted resource.

Azure has a defense system against DDoS attacks on its platform services.
It uses standard detection and mitigation techniques, and is designed
to withstand attacks generated from outside and inside the platform.

Data segregation
Dynamics 365 runs on Azure, so it’s inherently a multi-tenant service,
meaning that multiple customers’ deployments and virtual machines
are stored on the same physical hardware. Azure uses logical isolation
to segregate each customer’s data from others. This provides the scale
and economic benefits of multi-tenant services while rigorously pre-
venting customers from accessing one another’s data.

Encryption
Data is an organization’s most valuable and irreplaceable assets,
and encryption serves as the last and strongest line of defense in a
Fig. multi-layered data security strategy. Microsoft business cloud services
12-8

Barriers Fencing
Perimeter

Seismic Security 24/7 Days of backup

bracing operations center security staff power

Building

Two-factor access control:

Cameras Alarms Biometric readers and card readers
Computer
room

Implementation Guide: Success by Design: Security 290

 and products use encryption in transit and at rest to safeguard custom-
er data and help maintain control over it.

At-rest data protection

At minimum, Dynamics 365 environment databases use SQL TDE
(Transparent Data Encryption, compliant with FIPS 140-2) to provide
real-time I/O encryption and decryption of the data and log files for
data encryption at rest (Figure 12-9).

By default, Microsoft stores and manages the database encryption

Fig.
keys for your Dynamics 365 deployments. Finance and Supply Chain
12-9
Management apps use server-side encryption using service-managed

Full database encryption using SQL Transparent Database Encryption

Data protection
Windows OS level Capabilities
API (DPAPI)
DPAPI encrypts SMK
SQL TDE performs real-time I/O encryption and ecryption of the
data and log files to provide data encryption at rest

All data tranmitted between servers within the datacenter for

SQL instance level Service master key mirroring is encypted as well

SMK encrypts the DMK Microsoft manages the keys and handles management
for master DB of encryption

Encrypts the entrie SQL database

Master DB level Database master key No CMROL code changes required

Master keys protect database-level keys

DMK creates cert in
master DB Can change the database master key without having to change the
actual database encryption

Master DB level Certificate

Certificate encrypts DEK

in content DB

Database encryp-
Content DB level Availability
tions key
Certificate encrypts DEK
in content DB Available now for all Dynamics 365 online environments

Organization DB

Implementation Guide: Success by Design: Security 291

 keys. All key management aspects such as key issuance, rotation, and
backup are handled by Microsoft. For Customer Engagement apps,
the optional customer managed encryption key capability provides
administrators the ability to self-manage the database encryption key
for all the instances in their tenant.

In-transit data protection

Your data in transit is protected in several ways (Figure 12-10).
Dynamics 365 uses HTTPS encryption so that you can enable encryp-
tion for traffic between Dynamics 365 and end users. Azure protects
data in transit to or from outside components, as well as data in transit
internally, such as between two virtual networks. Azure uses industry
standard transport protocols such as TLS between user devices and
Microsoft datacenters, and within datacenters themselves.

Secure identity
Identity and access management is critical to every organization. Azure
Active Directory (Azure AD) is a complete identity and access man-
Fig. agement solution with integrated security that connects 425 million
12-10
people to their apps, devices,
and data each month. Dynamics
365 applications are safeguarded
using Azure AD as a seamless
identity solution.

Authentication: Users
Authentication is the process of
proving an identity. The Microsoft
identity platform uses the OpenID
Connect protocol for handling
End-to-end
Data in transit Data in transit
encryption of authentication. By default, only
between a user between
communications
and the service datacenters authenticated users can access
between users
Dynamics 365.
Protects user from Protects from bulk Protects from
interception of their interception of data interception or loss of
communicatin and helps data in transit Azure AD is used as a centralized
ensure transaction between users
integrity
identity provider in the cloud.
To access the system, users must

Implementation Guide: Success by Design: Security 292

 be provisioned into a Dynamics 365 environment and should have a
valid Azure AD account in an authorized tenant. Azure AD is built to
work for apps in the cloud, on mobile, or on-premises, and delegating
authentication and authorization to enable scenarios such as the
following (Figure 12-11):
▪ Conditional access policies that require a user to be in a
specific location
▪ Multifactor authentication (MFA)
▪ Single sign-on (SSO), in which a user can sign in once and then
be automatically signed into all the web apps that share the same
centralized directory

Azure AD provides fully secured identity federation with Active

Directory on-premises. Federation with Azure AD or Microsoft 365
enables users to authenticate using on-premises credentials and access
all resources in the cloud.

Authentication: Conditional access

Users can access cloud applications from anywhere and from any de-
vice such as mobile phones, tablets, or laptops from the office or home.
As a result, there are many scenarios in which access control decisions
shouldn’t be made based solely on who can access a resource—you
also need to consider how a resource is accessed. With Azure AD
conditional access, you can address this requirement.

Fig.
12-11 With conditional access, you can implement automated access control
decisions for accessing your cloud
Access control: Authentication for online services apps that are based on conditions
PC and device (Figure 12-12).

Express route

Authorization
Authorization is the control
Multifactor
authentication
of access to the Dynamics 365

Azure Active Directory

applications. It allows you to
control the functionality and data
Azure Active
Other directory services
Azure Active Directory Microsoft 365 Dynamics 365 Directory accessible to the users. Dynamics
365 uses security roles for

Implementation Guide: Success by Design: Security 293

 authorization. Once a user is authenticated, the security roles assigned
to a user (or groups of users) authorizes the users for access to data,
services, menus, or other Dynamics 365 features and capabilities. Note
that the concept of security roles is different for Customer Engagement
apps (based on Microsoft Dataverse and Power Apps) and Finance and
Operations apps. We examine this difference later in this chapter.

Auditing and monitoring

All Dynamic 365 applications provide audit functionality. Auditing
provides a log of events related to activity for secure datastores, logins,
or other important events. As with authorization, there are differences
related to auditing in Customer Engagement apps and Finance and
Supply Chain Management apps, which we examine later in the chapter.

Customer responsibility
As a customer, you’re responsible for:
▪ Account and identity management
▪ Creating and configuring conditional access policies
▪ Creating and assigning security roles
▪ Enabling and configuring auditing and monitoring
▪ Authentication and security of components of the solutions other
than Dynamics 365
Fig.
12-12

Conditions Actions
Cloud
User / group Allow access applications

Cloud application
Device state On-premises
Enforce MFA
Location (IP range)
User per user/per
Client application application
Sign-in risk

Block access

Implementation Guide: Success by Design: Security 294

 Security for Dynamics 365 integrations
Most Dynamic 365s implementations have integrations with a range of
Azure services, on-premises systems, and cloud systems (Figure 12-13).

Consider the out-of-the-box integrations before building custom solu-

tions. The native integration with the other Microsoft services is a tried,
tested, and recommended solution for these scenarios. The following
are examples of native integration options available in Dynamics 365:
▪ Dynamics 365 and Microsoft SharePoint integration
▪ Dynamics 365 and Microsoft Exchange Online integration
▪ Dual-write for integration between Customer Engagement apps
and Finance and Supply Chain Management apps.

Use server-to-server (S2S) authentication to securely and seamlessly

integrate Dynamics 365 with your custom websites and services. Some
of the native integrations we mentioned such as Dynamics 365 and
SharePoint also use S2S authentication.

Regarding your on-premises systems, Microsoft recommends you

deploy an on-premises data gateway connected to an Azure hybrid
connection where possible.

Azure Service Bus is a fully managed enterprise message broker with

message queues and publish-subscribe topics. Service Bus is used to
Fig. decouple applications and services from each other. In many cases, we
12-13
recommend integrating line-of
Dynamics 365 (SaaS) Azure (PaaS) Office 365 (SaaS) business applications with cloud
(Customer’s subscription)
services by using Service Bus,
Dynamics 365 Sales SharePoint Online
Dynamics 365 Customer Service Azure SQL for use with the Data Export Service Exchange Online which provides a secure channel
Azure Service Bus
Azure BLOB Storage for communicating between dif-
ferent on-premises or cloud-based
Azure Services Azure (IaaS)
Used by Dynamics 365 internally (Customer’s subscription) line of business applications.

Full text search Virtual machines used for OP2OL migration

Azure SQL Many customers implement an
allowlist for IP ranges that are

On-premises systems Other cloud systems

permitted for outbound connec-
(Customer’s system) (Customer’s subscription)
tions. Modern cloud services such

Implementation Guide: Success by Design: Security 295

 as Dynamics 365 don’t have static and specific IP ranges for Dynamics
365 solutions. We recommend you include the whole Azure region on
the IP allowlist for the system to work properly. Note that the list of IPs to
allow is considerably shorter for Azure Logic Apps. In addition, the use
of the on-premises data gateway previously mentioned can reduce or
eliminate this need.

Transparency
Microsoft is transparent about where your data is located. You know
where your data is stored, who can access it, and under what conditions.
Dynamics 365 customers can specify the Azure datacenter region where
their customer data will be stored. Microsoft may replicate customer
data to other regions available within the same geography for data
durability, except in specific scenarios, such as the following:
▪ Azure AD, which may store AD data globally
▪ Azure multifactor authentication, which may store MFA data globally
▪ Customer data collected during the onboarding process by the
Microsoft 365 admin center

Fig.
12-14 Microsoft imposes carefully defined requirements on government
and law enforcement requests for customer data. As described at
Transparency goals the Microsoft Privacy Principles, if Microsoft receives a demand for
a customer’s data, we direct the requesting party to seek the data
Choose where our data is stored
directly from the customer. If compelled to disclose or give access to
Transparent about how we respond any customer’s data, Microsoft promptly notifies the customer and
to government requests for your data provides a copy of the demand unless legally prohibited from doing so.

Figure 12-14 lists some standard transparency goals and their imple-

Implementation details mentation in Dynamics 365.

You select the region where your

Dynamics 365 data is stored
Customer Engagement
apps security
No third-party (including law
enforcement, other government
entity, or civil litigant) is given direct
or unfettered access to customer
data except as you direct
The Customer Engagement apps (Dynamics 365 Sales, Dynamics
365 Customer Service, Dynamics 365 Field Service, Dynamics 365

Implementation Guide: Success by Design: Security 296

 Marketing, and Dynamics 365 Project Service Automation) use the
Microsoft Dataverse to provide a rich security model that can adapt to
many business scenarios.

The goals of Dataverse security model are as follows:

▪ Provide users with access only to the appropriate levels of infor-
mation that is required to do their jobs
▪ Categorize users by role and restrict access based on those roles
▪ Support data sharing so that users and teams can be granted access
to records that they don’t own for a specified collaborative effort
▪ Prevent a user’s access to records the user doesn’t own or share

Security features
We use three main categories of security features to provide appro-
priate end-user access (Figure 12-15): fundamental security controls,
additional security controls, and manual sharing. Most of the security
requirements should be addressed using fundamental security con-
trols; other options should be used to manage the exceptions and
edge scenarios.
Fig.
12-15 Record ownership
Dataverse supports two types of
record ownership:
▪ Organization owned 
Manually sharing records Manually handles exceptions to the model When a record is assigned to
Organization, everyone in
Additional controls the environment can access
Hierarchy security Handles exceptions to the the record
Field-level security fundamental security
Access teams controls more easily ▪ User or Team owned 
Table relationships behavior
If not assigned to the organi-
zation, a record is assigned to
Fundamental security controls Business Unit, Child Business
Business units hierarchy
Security roles
Generally covers Unit, Team, or User
most requirements
Users and teams
Record ownership
Some out of the box tables are
exceptions to the above 2 types
Environment-level security group configuration such as system user record is
owned by a Business Unit.

Implementation Guide: Success by Design: Security 297

 Business units
Business units are a security modeling building block that helps in
managing users and the data they can access. The name “business
unit” can be misleading because the term doesn’t necessarily have any
direct relationship with an organization’s operating business units. In
Dataverse, business units provide a framework to define the organiza-
tional structure of users, teams, and records. Business units group users
and teams by organizational hierarchy and can work in conjunction
with security roles to grant or restrict access to data.

The real power of business units comes from their hierarchical nature.
Users can be given access to records just in their business unit, or their
As a best practice, minimize the number
business unit and the business units under their unit. For example, the
of business units in your implementation
based on access control requirements hierarchical nature of business units can allow you to limit access to re-
instead of mapping the organizational
chart into business units.
cords at the site, district, region, and corporate levels. Business units are
useful to segment data into ownership containers for access control.

Security roles
A privilege is permission to perform an action in Dynamics 365. A
security role is a set of privileges that defines a set of actions that can
be performed by a user. Some privileges apply in general (such as the
ability to use the export to a Microsoft Excel feature) and some to a
specific table (such as the ability to read all accounts).

To access Dynamics 365, users must be assigned one or more pre-

defined or custom security role, either assigned directly to their user or

Customer Engagement apps come with

inherited from a team they’re a member of.
several out-of-the-box security roles. If
you need additional or different security
roles, start by copying existing ones. For example, the salesperson role is assigned a set of privileges that are
relevant to the performance of the tasks defined for that role.

Teams
Teams provide a straightforward way to share records, collaborate
with other people across business units, and assign security roles.
While a team belongs to one business unit, it can include users from
other business units. You can associate a user with more than one
team, which is a convenient way to grant rights to a user that crosses
business units.

Implementation Guide: Success by Design: Security 298

 Multiple types of teams available are in Dataverse:
▪ Owner teams  You can make an owner team the owner of a
record, which is a useful way to link a record to a specific business
unit. Owner teams can also share records and be assigned to
security roles.
▪ Azure AD security group teams and Azure AD office group
teams  These are special types of owner teams. Their member-
ship can’t be managed within Dynamics 365, instead they’re linked
to an Azure AD security group or Microsoft 365 groups. Provided
that the users are appropriately licensed and part of the environ-
ment security group (optional), users who are added to the Azure
AD group are automatically enabled in the system and added to
the team when they connect to the environment. Using Azure
AD is particularly useful because the groups and permissions also
extend to Azure AD-enabled apps outside of Dynamics 365.
Use owner teams when:
▪ Team-level record ownership is ▪ Access teams  These teams are unique because they can’t own
required, rather than user records and can’t have a security role association. However, just
▪ You need to assign security roles to
users through teams like owner teams, access teams can have records shared with them.
▪ Reporting on team progress is required
When enabled at the table level, access teams can grant specific
Use access teams when:
▪ The team members need different record-level permissions to the members of a record’s access team.
permissions on individual records than
the same permission on the record type
This is an alternative to manually sharing the record with a user or
a team.

Field-level security
You can use field-level security to restrict access to high business im-
pact fields to specific users or teams. For example, you can enable only
certain users to read or update the credit score of a business customer.

Field-level security is available for the default fields on most out-of-the-

box tables, custom fields, and custom fields on custom tables. Field-level
security is managed by the security profiles. The scope of field-level
security is organization-wide and applies to all data access requests.

Sharing
Record sharing lets a user give access to a table record to other users or
team. The user must have a share privilege on the record to be able to
share it. Sharing should be seen as a way for users to manually manage
exceptions to the default security model.

Implementation Guide: Success by Design: Security 299

 Keep in mind the following recommended practices:
▪ Share only the necessary records with the smallest set of users or
teams possible
▪ Grant the minimum access required for users to do their jobs
▪ Disable share rights in security roles where possible
▪ Extensive sharing can cause performance issues at scale
▪ Microsoft recommends that you don’t automate sharing

Hierarchy security
You can use a hierarchy security model for accessing data from a user
or position hierarchy perspective. With this additional security, you
gain more granular access to records, for example by allowing man-
agers to access the records owned by their reports for approval or to
perform tasks on reports’ behalf.

Summary of different access control mechanisms

Figure 12-16 summarizes the different access control mechanisms
discussed in this section. It highlights the ideal scenarios, benefits, and
limitations of each option.

Environment security group

You can associate an Azure AD security group with a Dataverse en-
vironment to control user access. Unless users have specific, highly
privileged Microsoft 365 admin roles, a user in Azure AD can’t access
any information in Dynamics 365 even with a valid authentication and
security role assigned unless they are also a member of the correct
environment security group in Dynamics 365.

It’s a best practice to associate a security group with an environment.

This prevents all the Azure AD-eligible (appropriately licensed) users
from appearing as active users in the Dynamics 365 environment.

Audit
Auditing helps you comply with internal policies, government regula-
tions, and consumer demands for better control over confidential data.
Organizations audit various aspects of their business systems to verify

Implementation Guide: Success by Design: Security 300

 that system and data access controls operate effectively and identify
suspicious or non-compliant activity.

Dataverse audit
Audit logs are provided to ensure the data integrity of the system and
to meet certain security and compliance requirements. The auditing
feature logs the changes that are made to records and user access so
the activity can be reviewed later.

Auditing can be enabled on:

▪ Table level (you can select specific fields in a table for auditing)
▪ User access

Fig.
12-16

Bypasses need for

For open access
Organization information to all
security checking, so
optimal performance

Can be used for

Great for large volumes Reduces management and
Allows for optimal maintenance
Business unit of data accessed by large
groups of users
access checking overhead
oversight access with
more granular access
at lower levels

Ideal for large Also drives business

Optimized access Limited to one
volumes of overall unit position so
checks combine type of access i.e.
User ownership data but individual
memory cache and owned by single enables combination
granular access to business unit
record data directly user or team
smaller subset access

Ideal for large Isolate users from

Optimized access Reduces cascading
volumes of overall record business unit
checks combine impact as user
Team ownership data but group memory cache and involvement
structure, enables
granular access to user movement
record data directly changes
smaller subset without data impact

But comes with cost,

Ideal for modeling Provide very checking lots of rules
exceptions over specific unique expensive in perfor- Sharing with teams
Sharing standard model using permissions per mance, storage, and mitigates cost
other features record maintenance
overhead

Use to support Not intended to be

Builds on ownership immediate replacement for
Inheritance and sharing models managers in chains buisness unit access
above granular at more senior or
indivual access broad scope levels

Implementation Guide: Success by Design: Security 301

 Data auditing can be viewed:
▪ Contextually, at the record level, on the form
▪ Globally, from the audit summary view
▪ From the APIs

Don’t enable auditing for all tables and columns. Do your due dili-
gence to determine which tables and fields are required for auditing.
Excessive auditing can affect performance and consume large volumes
of log storage.

Microsoft 365 audit log

Activity logging records user and admin activities across Microsoft 365
and Dynamics 365 apps. Data is stored in a Microsoft 365 unified audit
log. It has a smaller footprint on system resources compared to the
previous Dataverse audit functionality. Microsoft 365 audit logs have
several additional benefits:
▪ Management in a central location on the Microsoft 365 security
and compliance portals  Microsoft 365 admins can manage
settings and access activity reporting for all environments within
the Microsoft 365 security and compliance portals. Dataverse
auditing is set up and stored separately within each Dynamics
365 environment.
▪ All data in the system is logged  This functionality logs all data
transactions, including read operations, plug-in operations, table
operations, bulk operations, user sign in and sign out sessions, and
even Microsoft Support operations.
▪ Configurable alert policies  You can set up the system to notify
administrators or compliance officers of certain events according
to configurable settings.
▪ Audit log search capability  Administrators can easily query
audit logs via predefined or custom filters.
▪ Analyze suspicious behavior with Security Information and
Event Management (SIEM)  Use SIEM functions in near-real time
to analyze and alert administrators of possible suspicious behavior
within the system and provide actions to address these events.
▪ SIEM vendor integration  Dynamics 365 now provides out-of-
the-box integration with multiple SIEM vendors such as ArcLight,
Microsoft OMS, Okta, SumoLogic, BetterCloud, and standard CEF

Implementation Guide: Success by Design: Security 302

 format integration for many others. Microsoft has also released a
connector to integrate activity logs into Azure Sentinel.

By default, activity logging data is retained for 90 days or 1 year based

on the Microsoft 365 license type. We recommend changing the reten-
tion policy or moving these logs to external stores if you need to retain
the data for longer periods.

Finance and Supply

Chain Management
apps security
Finance and Supply Chain Management applications use Azure AD as a
primary identity provider and utilize the security capabilities of Azure. To
access these applications, users must be provisioned into a Finance and
Supply Chain Management instance and should have a valid Azure AD
account in an authorized tenant. Once authenticated, these applications
use role-based security to authorize user access to individual elements
of the applications. Additionally, you can use extensible data security
policies to restrict access to a subset of data.

The security architecture of Finance and Supply Chain Management

consists of the following components (Figure 12-17):
▪ Authentication  Only authenticated users with user rights in
Finance and Supply Chain Management apps can connect
▪ Authorization  Security roles are comprised of duties, duties are
comprised of privileges, and privileges are comprised of permissions
▪ Security roles  Roles grant access to Finance and Supply Chain
Management apps
▪ Duties  Duties allow access to parts of business processes
▪ Privileges  Privileges allow access to individual tasks
▪ Permissions  Permissions grant access to individual
securable objects
▪ Data security  You use data security to deny access to tables,
fields, and rows in the database (extensible data security framework)

303
 ▪ Auditing  The system logs when a user signs in or out of
the application

Role-based security
In Finance and Supply Chain Management apps, role-based security
is aligned with the structure of the business. Users are assigned to
security roles based on their responsibilities in the organization and
their participation in business processes. Because you can set up rules
for automatic role assignment, the administrator doesn’t have to be
involved every time a user’s responsibilities change. After business
managers set up security roles and rules, they can control day-to-day
user access based on business data. A user who is assigned to a security
role has access to the set of duties that are associated with that role,
Fig.
12-17 which is comprised of various granular privileges. A user who isn’t

Authentication (Azure Active Directory)

Users

Authorization
Security roles

Duties Privileges

Permissions for application elements

User interface elements Tables and fields SSRS reports Service operations

Data security
Data security policies Table permissions
framework

Database
Database

Implementation Guide: Success by Design: Security 304

 assigned to any role has no privileges. Privileges are composed of
permissions and represent access to tasks, such as cancelling payments
and processing deposits. Duties are composed of privileges and repre-
sent parts of a business process, such as maintaining bank transactions.
Both duties and privileges can be assigned to roles to grant access to
Finance and Supply Chain Management.

By default, approximately 100 standard security roles are provided

within the solution. All functionality is associated with at least one of
We recommend duplicating the sample
security roles and using these for role the standard security roles. The administrator can duplicate and mod-
assignments for all users before you start
ify the sample security roles to fit the needs of the business or create
creating custom roles.
entirely new security roles.

Security diagnostics for task recordings

The Task recorder for Finance and Supply Chain Management apps is a
utility that lets users record business processes for several different use
cases. This is an invaluable tool for many purposes, including process
definition, training, testing, and security. The security diagnostics for
the Task recorder allows implementers to analyze and manage security
permission requirements in any task recording. When you use the Task
recorder to define and document business processes for testing or
training purposes, the security requirements for implementing those
tasks are also captured. Implementers can then use this tool to tailor
duplicated copies of the sample security roles to better meet their needs.

Segregation of duties
You can set up rules to separate tasks that must be performed by
different users. For example, you might not want the same person to
acknowledge the receipt of goods and to process payment to the ven-
dor. This concept is named segregation of duties. Segregation of duties
helps you reduce the risk of fraud, enforce internal control policies, and
The standard security roles included in
Finance and Supply Chain Management detect errors or irregularities.
apps incorporate segregation of duties. If
a user needs full access, they often need
a combination of roles (such as an HR When an organization defines rules for segregation of duties, you should
assistant or HR manager for full access to
HR features). evaluate existing role definitions and role assignments for compliance,
thereby preventing role definitions or role assignments that don’t

Implementation Guide: Success by Design: Security 305

 comply. Future role assignments that result in conflicts with existing
rules are logged and must be allowed or denied by an administrator.

Security reports
Finance and Supply Chain Management applications provide a set of
rich security reports to help you understand the set of security roles
running in your environment and the set of users assigned to each role.
In addition to the security reports, developers can generate a work-
book containing all user security privileges for all roles.

▪ User role assignments report  Generates a view of the current

user role assignments in your system. By default, the report in-
cludes all users with roles assigned. You can optionally limit the
report to a specific set of users by entering a list of users when
generating the report. In the User role assignments parameters
pane, under Records, choose Filter. From here you can add or
remove filters to the list of users the report will be generated for.
▪ Role to user assignment report  Provides an aggregation of
role assignments. Expanding a role in the report shows the list of
users assigned to the role, and expanding the username shows any
restrictions the role has applied. You can apply the same method
for filtering the set of users to this report as described for the User
role assignments report.
▪ Security role access report  Provides a view of the effective
permissions for each security role. This report provides a flattened
list of permissions grouped by type across all sub-roles, duties, and
privileges contained in the role.
▪ Security duty assignments report  Provides a view of all the
duties contained within a role. You can configure this report to run
on any collection of roles to ensure that segregation of duties is
maintained between roles. By default, the report includes all the
roles. To limit the roles included, use the filtering provided in the
Records to include section.
Familiarize yourself with these reports
during implementation and engage your
compliance team early. There is no benefit To aid compliance with GDPR requirements, you can use the per-
to waiting for an audit to use these reports
son search report when a data subject (the individual EU citizens or
for the first time.
residents that share personal data) approaches the controller (the

Implementation Guide: Success by Design: Security 306

 organization collecting and managing personal data) to request a copy
of their personal data.

System log
System administrators can use the User log page to keep an audit
log of users who have logged on to the system. Knowing who has
logged in can help protect your organization’s data. The user logging
capability allows administrators to identify roles that provide access to
sensitive data. The sensitive data identifier enhances the user logging
experience by letting your organization produce audit logs that show
who in your system has access to sensitive data. This capability is help-
ful for organizations that might have multiple roles that have varying
degrees of access to certain data. It can also be helpful for organiza-
tions that want a detailed level of auditing to track users who have had
access to data that has been identified as sensitive data.

Extensible data security policies

Extensible Data Security (XDS) policies in Finance and Supply Chain
Management apps allow developers to supplement role-based security
by restricting access to table records based on security policies. Policies
can be created that contain a query to select information and a table
or tables that will be constrained by the policy. The query in the policy
applies a filter and only records that satisfy the conditions of the filter
are accessible from the restricted tables. XDS provides a powerful data
security tool for advanced requirements, but you should use it judi-
ciously to avoid unanticipated impacts on performance or usability.

Database logging
Database logging provides a way to track specific types of changes to
the tables and fields in Finance and Supply Chain Management apps,
including insert, update, delete, and rename key operations. When you
configure logging for a table or field, a record of every change to that
table or field is stored in the database log table (sysdatabaselog) in the
environment database.

You can use database logging to create an auditable record of changes

to specific tables that contain sensitive information, or to monitor the
use of electronic signatures. Note that this feature is in addition to

Implementation Guide: Success by Design: Security 307

 audit trail capabilities already available in many modules. Database
logging is intended to track individual changes to specific, sensitive
information or configurations. It’s not intended to track automated
transactions that are run in batch jobs, and is not recommended for
transactions. Although database logging can be valuable from a busi-
ness perspective, it can be expensive with regard to resource use and
management, so it too should be used judiciously.

Microsoft recommends the following when implementing

database logging:
▪ Create a plan for reviewing and using your database logs
▪ Limit log entries and help improve performance by selecting
specific fields to log instead of whole tables
▪ Create a plan for how long you will retain logged data and use the
included cleanup job

Make security
a day one priority
Security should be at the forefront when you start a Dynamics 365
project, not an afterthought. Negligence of security requirements can
lead to significant legal, financial, and business risks. It can also impact
the overall scalability and performance of the solution. In this section,
we discuss some of the security impacts on Dynamics 365.

Impact on scalability
Security design can have a significant impact on scalability. Security
design should consider scalability especially for multi-phased and
global deployments.
Security should
be at the forefront Customer Engagement examples
when you start In multi-phased global deployments, short-sighted security design can

a Dynamics 365 affect the rollout to other regions. The different regions can have their
own compliance, business processes, and security requirements. The
project, not an same model may not work for all the regions. Therefore, you should
afterthought. think about the security model holistically at the beginning of the project.

308
 One such example can be a global deployment where in the first
region (Phase 1) you create all users in the root business unit with
business unit-level access in security roles. Because Phase 1 users were
placed in the root business unit, when you roll out the solution to the
next region (Phase 2), Phase 2 users have access to the users in the first
region. This may cause some regulatory or compliance issues.

Several businesses use Organization owned entities for reference

data tables, which means security roles don’t apply to these entities.
In multi-phased deployments, some departments may not want their
reference data to be seen by other departments, or simply only want
to see their reference data appearing when creating records. This
segmentation is hard to achieve using Organization owned entities
and much simpler to achieve when User or Team owned entities have
Business Unit access. Entity ownership is a choice that happens at the
time the entity is created and can’t be changed. This assignment con-
trols who can access the entity. If you want to change who has access
to these entities, the only option you have is to delete these entities
and recreate them in an owner (Organization, Team, Business Unit, and
so on) that that is properly scoped.

Finance and Supply Chain Management examples

Similarly, for Finance and Supply Chain Management implementations,
different legal entities may have distinct functions and purposes, be
subject to different security constraints, and have completely different
security requirements. Even if legal entities, subsidiaries, or regions
have a high degree of standardization and centralization and are good
candidates for streamlined security implementation, failing to source
requirements from the right stakeholders, design security as modular
building blocks, or otherwise fail to take a long-term view of security
needs can result in subsequent rollouts struggling with ill-fitting global
templates or a need to start over from scratch.

Design of role-based access control should follow best practices, or

subsequent rollouts will be unable to use and further build upon
established security roles. Security roles should be designed based on
business processes and incorporate segregation of duties, not individ-
ual personas. We recommend that you avoid creating a single role for

Implementation Guide: Success by Design: Security 309

each user, and avoid creating nested security roles. The resulting mod-
ular security roles can be assigned to users as they cumulatively give
the access needed for a users’ duties, and no more. Native tools such
as the Task recorder and Regression suite automation tool support the
testing of individual user permissions, which facilitate this approach.
We also recommend following recommended practices for the initial
rollout (such as copying standard roles rather than directly assigning
them). This gives subsequent rollouts the flexibility of building upon
existing role structures, while retaining access to standard roles for
comparison. This prevents the subsequent rollouts from being forced
to choose between modifying roles already in use elsewhere or starting
from scratch because existing roles are too specialized.

Impact on performance
Security design can impact system performance. Dynamics 365 has
multiple access control mechanisms, all with their own pros and cons.
The ideal security design is a combination of one or more of these
controls based on your requirements and the granularity, simplicity, and
scalability they provide. The wrong choice can lead to poor performance.

Customer Engagement examples

Sharing provides particular and unique permissions on a specific record,
but sharing for large volumes of records comes at a cost. To check access
for a user, the system checks user security roles, team memberships,
and records shared with a user and their teams. Checking lots of rules is
expensive in terms of performance, storage, and maintenance overhead.

Sharing can also trigger cascading rules that can be expensive if not set
appropriately. Sharing should be for edge cases and exception scenarios.

You can get more information on this topic from the Modelling
Scalable Security Modelling with Microsoft Dynamics CRM whitepaper.
It’s an older document, but still relevant for Dynamics 365 security.

The volume of data will impact performance in the long run. You
should have a well-defined archival strategy. The archived data can be
made available through some other channels.

310
 Finance and Supply Chain Management examples
Database logging in Finance and Operations can be valuable from a
business perspective, but it can be expensive regarding resource use
and management. Here are some of the performance implications of
database logging:
▪ The database log table can grow quickly and increase the size of
the database. The amount of growth depends on the amount of
logged data that you retain.
▪ When logging is turned on for a transaction type, each instance of
that transaction type causes multiple records to be written for the
Microsoft SQL Server transaction log file. Specifically, one record
is written for the initial transaction, and one record logs the trans-
action in the database log table. Therefore, the transaction log file
grows more quickly and might require additional maintenance.
▪ Database logging can adversely affect long-running automated
processes, such as inventory close, calculations for bills of materi-
als, master planning, and long-running data imports.
▪ When logging is turned on for a table, all set-based database op-
erations are downgraded to row-based operations. For example, if
you’re logging inserts for a table, each insert a row-based insert.

When you’re implementing database logging, be sure to have a plan

for how long you’ll retain logged data, and how you’ll limit log entries
and help improve performance by selecting specific fields to log in-
stead of whole tables.

Extensible Data Security policies allow developers to supplement role-

based security by restricting access to table records based on security
policies. The query in the policy (the policy query) applies a filter so
only records that satisfy the conditions of the filter are accessible from
the restricted tables. The policy query is added to the WHERE clause, or
ON clause, on SELECT, UPDATE, DELETE and INSERT operations involving
the specified constrained tables.

Unless carefully designed and tested, policy queries can have a significant
performance impact. Therefore, make sure to follow the simple but im-
portant guidelines when developing an XDS policy. Performance can be
significantly impacted by poorly performing queries. As a best practice,

Implementation Guide: Success by Design: Security 311

 avoid numerous joins, consider adding lookup tables to improve
performance, carefully tune queries for optimal performance, and
carefully and thoroughly test any XDS policies when using this feature.

Impact on regulation and compliance

Several different security controls are required to meet regulations and
compliance based on industry and region. These controls are a combi-
nation of duty separation, data categorization, identity management,
access control, log management, and auditing. Missing any of these
controls can make your solution noncompliant.

Governance takes time to understand, roll out, and adopt. The earlier
you begin the process, the easier compliance becomes.

Customer Engagement examples

The activity logging data is retained between 90 days to 1 year based
on the Microsoft 365 license type. For compliance reasons, if you need
to keep these records for longer, you should move this data to an
external store.

Impact on rollout
In many projects, security design is created late in the implementation
process. It can lead to a lot of issues, ranging from system design to in-
adequate testing. Keep the following recommended practices in mind:
▪ Test the system using the right security roles from the very beginning.
▪ Use the correct security profile during test cycles. Don’t provide an
Governance administrator role to everyone.

takes time to ▪ When training people, make sure they have access to make all the
required actions. They shouldn’t have access to more information
understand, roll than required to complete those actions.
out, and adopt. The ▪ Validate security before and after data migration occurs.
earlier you begin
Finance and Supply Chain Management examples
the process, the
Design your security roles with both compliance and scalability in
easier compliance mind. By engaging your security and compliance team early in your
becomes. implementation, you can use the segregation of duties framework to

312
 identify conflicts and create rules that enforce separation of duties that
should be separated.

For example, a global company has made efforts to standardize

business processes and is planning to roll out a new functionality in
multiple phases to globally. Rules that segregate duties in your large
legal entities prevents assignments that may result in conflicts.

However, in smaller legal entities, users may need to accomplish the same
tasks with a smaller staff, which makes it challenging to construct declara-
tive rules that enforce segregation of duties. In this case, modular security
roles allow you to enforce the same segregation of duties rules but when
assigning multiple roles to some users, and apply documented overrides
to the detected conflicts. In this manner, your global rollout can reuse
modular security components, and you have a robust segregation of du-
ties framework in place for entities of many sizes, allowing them to grow
without introducing the need to redesign the security implementation.

Impact on reporting
Security design can have a significant impact on reporting. For example,
your security design identifies sensitive data that requires restricted
access to be enforced and monitored. You need to create auditing re-
ports related to access and changes on sensitive data, and designated
personnel responsible for oversight need to review them periodically.

In addition to built-in reporting capabilities, you have multiple options

for exporting Dynamics 365 data into external stores for reporting
and analytics. The two native options are exporting to a data lake or
using the Data Export Service (not available in Finance and Supply
Chain Management). When data is in external stores, it doesn’t honor
Dynamics 365 security permissions (nor would it be useful to do so
in many cases). You need to apply your own security mechanism for
reporting. We recommend having a data management strategy for all
data that is stored outside of a core application. This strategy should
include security requirements (encryption, access controls, row level,
personally identifiable information) as well as data residency and
retention requirements. The strategy should consider the future state

Implementation Guide: Success by Design: Security 313

of Dynamics 365 data in the analytical platform as well as its legacy and
other applications.

You can use Tabular Data Stream (TDS) endpoints with Customer
Engagement for direct query scenarios, which honor the Dynamics 365
security context.

Operational aspect of security

Operation aspects can have a big impact on the security design. Operation
aspects can decide which access mechanisms are the best fit for the
solution, such as in cases of organizational changes or maintenance.

Organizational changes
What happens if a user moves to another team or business unit? Do
you need to need to reassign the records owned by the user? What
should be the process?

Do you reassign all records or just active record? What is the effect of
changing the ownership on the child records, especially closed activities,
closed cases, and won or lost opportunities? Based on these questions,
you can come up with a process and cascading rules in Dynamics 365.

Reassigning a large volume of records can also take a long time and affect
the system’s performance. Learn more about assigning or sharing rows.

Maintenance overhead
Plan out how to assign security roles to new users. Are you going to
automate a process? For example, you can use Azure AD group teams
in Dynamics 365 to assign roles to new team members. This can make
it very easy to assign license and security roles to new team members.

Security anti-patterns
An anti-pattern is a frequently implemented, ineffective response to a
problem. Several security anti-patterns should be avoided for scaling,
performance, and security reasons. We shared a few in the previous

314
 section, such as using Organization owned entities for reference data
tables in Customer Engagement, or indiscriminately logging transac-
tional tables in Finance and Operations apps. In this section, we discuss
a few more anti-patterns to avoid.

Security through obscurity

Project teams may try to implement security through obscurity. One
example is hiding the fields from forms and views through custom
code. This is not security, because anyone with a proper security role
can access those fields in several ways such as with the use of Advance
Find, Power Automate, and APIs.

Always consider all the ramifications of avoiding default security controls.

Credentials stored in Dynamics 365

This is also one of the more common anti-patterns. Projects often use
calls to external systems through plugins and JavaScript. Calling exter-
nal services through plugins is recommended, but the credentials to
Azure Key Vault allows you to securely store call these services are stored in the code of some configuration entities
and manage application credentials such
as secrets, keys, and certificates in a central in the system. This practice creates a significant security risk because
and secure cloud repository. Key Vault
the credentials could be discovered by numerous roles not directly
eliminates the need to store credentials in
your applications. Your applications can related to the use of the service. In addition, if the credentials change,
authenticate to Key Vault at run time to
retrieve credentials.
the call to the external service fails. As a best practice, don’t store any
credentials in the system.

Conclusion
In this chapter, we introduced how the Trusted Cloud is built on the
foundational principles of security, privacy, compliance, and trans-
parency, and outlined basic concepts of information security as they
apply to Dynamics 365. We then took a closer, product-specific look
at how these security concepts are applied to Customer Engagement,
Finance, and Supply Chain Management applications. With that as a
foundation, we then examined why it’s crucial to the success of your
implementation to make security a priority from day one, citing some

Implementation Guide: Success by Design: Security 315

Resources specific examples and considering the impact of security on perfor-
mance. Finally, we wrapped up with a handful of anti-patterns to avoid.

General Equipped with this information, you can be confident in the security
Microsoft Dynamics CRM Online considerations and capabilities of Dynamics 365 products and better
security and compliance planning
guide ensure the security of your implementation.

Trusting the Cloud

Compliance
An Introduction to Cloud
Computing for Legal and
Compliance Professionals

Achieving Trust and Compliance

in the Cloud

Privacy
Privacy at Microsoft

Protecting Data and Privacy in the

Cloud

Security
Dynamics 365 Security Assessment

Security Incident Management in

Microsoft Dynamics 365

Encryption in the Microsoft Cloud

Security in Microsoft Dataverse

Security architecture

Plan and implement security in

Finance and Operations apps

Review the security model for your

Dynamics 365 solutions

316
 Checklist

Privacy and compliance Application security

Understand the responsibilities of the service provider as Understand the app-specific security constructs and
a data processer versus the customer responsibilities as model your security based on the native access control
the owner and data controller to ensure compliance on mechanisms rather than customizing the build.
both sides.
Understand that hiding information from the view
Refer to the Dynamics 365 cloud service agreements doesn’t necessarily remove access—there are alternate
and compliance documentation to understand the mechanisms to access and extract information.
policies, procedures for handling data, disaster recovery
Understand the impact of losing the security context
strategy, data residency, encryption, and more.
when the data is exported.

Identity and access Ensure the security model is optimized to perform and
provides the foundation for further expansion and scale
Establish an identity management strategy covering by following the security model best practices.
user access, service accounts, application users, along
Have a process to map changes in the organization
with federation requirements for SSO and conditional
structure to the security model in Dynamics 365. This
access policies.
needs to be done cautiously in a serial way due to the
Establish the administrative access policies targeting downstream cascading effect.
different admin roles on the platform, such as service
admin and global admin.

Enforce relevant DLP policies and procedures to make

changes or request exceptions.

Have necessary controls to manage access to specific

environments.

Implementation Guide: Success by Design: Security 317

13
Guide
Business
intelligence,
reporting,
and analytics
 Introduction
In a world where so much comes down to reporting
and analytics, business intelligence has become the
differentiator—it’s all about actions and outcomes,
measuring what matters the most, and empowering
users with the right insights at the right time.
Business intelligence solutions have evolved to keep up with an
ever-increasing number of data sources, as organizations seek ways
to expand their global reach to attract more customers, and give their
current customer base access to more content, purchasing channels,
and brand options. In this context, turning collected data into
competitive business knowledge is more important than ever.
In this chapter, we discuss how data helps
businesses make informed decisions.
We also cover the evolution of business While AI and machine learning keep evolving—and are now used natively
intelligence solutions, and how
organizations can build a data estate
by apps—your organization’s intelligence strategy should be based on a
based on modern components, on top modern data estate, with a focus on how to strengthen the organization.
of solutions that support augmented
analytics and embedded intelligence.

The importance of data

Business intelligence solutions strengthen
organizations, prepare them for
challenges and market changes, and guide
them toward the best insights and actions.
As James Phillips, president of Microsoft’s Digital Transformation Platform

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 319
 Group, wrote in the Microsoft Dynamics 365 blog, “There is a
fundamental change occurring across industries and organizations:
Data now comes from everywhere and everything.” As an essential
part of a business solution, data represents a sizable portion of each
user’s engagement. The solution processes and analyzes the data to
produce information, which allows an organization to make informed
decisions, and determines actions that can come from it.

As a part of the day-to-day usage of any application, users generate

significant amounts of process-related and customer-related data—
structured and unstructured. For example, customers generate data by
visiting a website or using a specific product. In turn, businesses generate
data about their products or services, as well as how customers interact
with those offerings.

In addition, the pace of data growth is increasing dramatically. Not

only does user activity create data, but new data sources such as the
Internet of Things (IoT) need to be captured, rationalized, and secured.

The key role of data is to generate information that helps businesses

and customers make informed decisions. For businesses seeking to
stay competitive, the data also must trigger actions demonstrating that
they understand and value their customers.

Resilience—the ability to adapt quickly to changing conditions—has

become a hallmark for business success today. Competitive, resilient orga-
nizations invest in turning the data they collect into business knowledge,
build the right data estate to generate a 360-degree view of their users,
and use business intelligence to determine the best actions and outcomes.

Organizations that have already embarked on their transformation

journey to improve the ways they use data are better prepared to
respond and adapt to change.

Break the silos and use the digital

feedback loop
Organizations are challenged by how to deal with ever-increasing
amounts of data—and how to generate user value with it—especially
when that data becomes siloed by the systems that generate or collect

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 320
 it. Data silos make the goal of having a 360-degree view of each user
even more challenging. Successful organizations are able to digitally
connect every facet of their business. Data from one system can
be used to optimize the outcomes or processes within another. By
establishing a digital feedback loop (Figure 13-1) that puts data, AI,
and intelligent systems and experiences at the core, organizations can
transform, become resilient, and unlock new values for users.

Evolution of business
intelligence, reporting,
and analytics
Practices for gathering, analyzing, and acting on data have evolved
significantly over time. Traditional methods of standardizing and
generating static reports no longer give businesses the agility to adapt
to change. New technology and secure, highly reliable cloud services—
which are poised to meet the needs of organizations that must quickly
manage increasing amounts of data—have given rise to a new era of
digital intelligence reporting.

Traditional reporting
The early generations of business intelligence solutions were typically
Fig.
13-1 centralized in IT departments. Most organizations
had multiple data repositories in different formats
Engage Empower and locations that were later combined into a single
customers people
repository using extract, transform, and load (ETL)
l ig
Intel ent tools, or they would generate reports within siloed
sources and merge them to provide a holistic view of
Data & AI the business.
s
sy

te
ce

ms n
ie
s

an d ex per
Once the data was unified, the next steps were
Optimize Transform deduplication and standardization, so the data could
operations products
be structured and prepared for reporting. Business
users who lacked the expertise to perform these tasks

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 321
 would have to rely on the IT team or specialized vendors. Eventually, the
business would receive static intelligence reports, but the entire process
could take days or even weeks, depending on the complexity of the data
and maturity of the processes in place. Data would then undergo further
manipulation, when required, and be shared across different channels,
which could result in the creation of multiple versions that would be
difficult to track.

Such an approach, which also heavily depended on the business

owner’s analysis to generate business insights and actions, was often a
time-consuming and error-prone process. And if an error was found, it
had to be fixed in the central repository. The process to produce the re-
port would need to be executed again—and repeated multiple times if
the same fix wasn’t applied at the source. This method of reporting was
referred to as “closing activities,” because it typically occurred at the
end of a month, quarter, or year, which meant that organizations were
slow to respond to opportunities because they waited until the end of
a given period to analyze their data and make decisions.

Self-service reporting
The evolution to a more agile approach favored self-service capabilities to
empower users. More user-friendly solutions reduced the IT dependency
and focused on providing business users with access to data and visualization
tools so they could do their own reporting and analytics. This accelerated the
speed of data analysis and helped companies make data-driven decisions
in competitive environments. However, in this model, reporting was
unmanaged—increasing the number of versions of reports, as well as
different representations of the data—which sometimes prevented
organizations from using a standardized method to analyze data and
inhibited effective decision-making.

This updated approach didn’t entirely replace the IT-centric model,

as many organizations started using a combination of traditional and
self-service reporting, but it did provide quicker access to information, so
organizations could react faster to business challenges and opportunities.

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 322
Reporting from anywhere
The growth of IT infrastructures, networks, and business usage of
devices such as mobile phones and tablets launched a digital trans-
formation from legacy systems into more modern solutions for most
organizations. Business intelligence apps allowed reporting from
anywhere at any time, giving users access to data while they were out
of the office or on business trips. These apps improved how organiza-
tions could respond to customers, and gave a 360-degree view of each
customer interaction. They also provided more succinct visualizations
of data, with better features to interpret it.

The new era of business intelligence

solutions
With data now coming from everywhere and everything,
organizations must be prepared to convert that data into business
knowledge so users can make informed decisions and trigger
actions. Many organizations employ highly skilled workers, such
as data scientists, who are responsible for analyzing the data and
producing the necessary insights that can affect the business. This
approach is expensive and adds dependency on specialized roles to
perform tasks that usually can’t be completed by typical
business users.

An approach to reduce this dependency while simultaneously

increasing ready access to data is to use augmented analytics, a
feature of modern business intelligence solutions. According to
Gartner, Inc., a research and advisory firm, “Augmented analytics is
the use of enabling technologies such as machine learning and AI to
assist with data preparation, insight generation, and
insight explanation.”

Embedded intelligence solutions that make use of augmented analyt-

ics allow processes, services, or products to deliver their own insights
to increase quality, efficiency, and customer satisfaction. These are new
types of reporting and analytic experiences where the intelligence is
by-design, and the solution itself analyzes the data to provide insights.

323
 While the data may not yet be unified with such an approach, customers
and organizations can use insights provided by products and services
from the earliest phases of usage or production, which increases their
return on investment (ROI).

Reporting and analytics

strategy
To be a market leader, an organization’s products and services must
evolve continuously and exceed customer expectations. The information
used to improve a product or service is based on data that comes to the
surface via reporting. Reporting requirements can be as simple as deter-
mining the status of an order or as complex as a cash-flow analysis for a
global organization.

What is a reporting solution?

An effective reporting solution is one of the essential elements for
business success. A properly implemented reporting solution provides
a competitive advantage, as users have access to the exact information
they need exactly when they need it, which accelerates and improves the
decision-making process.

Define your analytics strategy

An analytics strategy determines how an organization will provide the ide-
al level of accurate information to users for timely decision-making. The
strategy includes a list of tools and types of reporting (such as analytics,
static reports, and ad-hoc reporting), as well as how the information will
be secured and shared with users. Implementation of the strategy forms a
foundation for defining standards, processes, and best practices.

Your analytics strategy can help transform data collected from different
processes and systems into knowledge to help you stay competitive in
the market.

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 324
 A successful strategy caters to user requirements and how those require-
ments must be fulfilled. For example, some business users may require
printed reports, which can be sent to customers and vendors. Others
may need data to be available for summarized, detailed, or ad-hoc
reporting. Business management may require financial or operational
reports to understand the health of the organization and determine an
overall strategy. Serving such varied user needs often requires different
tools to manage data coming from multiple sources. For example,
financial and ad-hoc reporting requirements may rely on data extracted
and stored in a staging location utilizing Microsoft Azure data services
such as Azure SQL Data Warehouse, as well as manufacturing or sales
information in Dynamics 365 apps.

With a centralized data and business intelligence strategy, the IT de-

partment is usually responsible for data, ETL processes, and reporting
solutions. In a self-service approach, the IT team implements a data
analytics platform that enables users without statistical, analytical, or
data-handling expertise to access and use data. Teams should under-
stand their organization’s goals and strategy before implementing an
enterprise data platform.

Any reporting solution requires ongoing maintenance and updates

as underlying system requirements and business priorities change.
The project team should clearly define the delivery, maintenance, and
operational responsibilities. There should also be a plan for making
historical data from legacy systems available for reporting, business
Every organization needs people who intelligence, or audit purposes.
understand how to analyze and use data.
Training and communication are key to
ensuring that business users know how
to use data and insights to optimize their
Gain deeper insights from data
work, drive efficiency, and make For most organizations, business intelligence and analytics are critical
informed decisions.
to their growth and digital transformation. With technology such as
Data and analytics processes change Azure Machine Learning and Adaptive Insights, organizations can gain
periodically to improve how information
is provided. Make sure users are aware of deeper insights from their data. Having an early understanding of
any upcoming change and understand these technologies also can help organizations design their systems for
how it will affect them. Constant updates
without appropriate change management future-state data consumption.
can confuse users, but effective change
management can improve the adoption
rate of analytics. Public, cloud-based business intelligence solutions are a good fit
for many organizations, as they are on-demand, self-service, and

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 325
 scalable solutions. To choose the best technology to meet requirements
today and in the future, the implementation team must understand their
organization’s current state and the future vision for enterprise business
intelligence and analytics.

Understand reporting requirements

Understanding reporting requirements is key to delivering successful
reports and helping improve the entire business. Reporting require-
ments must be examined in the context of the business process for
the purpose of the report—such as operational reporting, financial or
regulatory reporting, ad-hoc inquiries, and dashboard and analytical
reporting—and should not be an afterthought.

As part of the business requirements gathering process, organizations

must address reporting requirements, and collect information about
how reports will be used, such as:
▪ Printed on paper and sent with a shipment on a truck.
▪ Used to create a pivot table in Microsoft Excel and build another report.
▪ Submitted to government agencies at the end of the year.
▪ Provided to a bank on a periodic basis for audit purposes.

Organizations must also identify any critical reports that require data
mash-up with other sources, such as front-office apps or transportation
systems, to develop an appropriate data integration and solution strat-
egy. Data-volume reporting requirements help determine how reports
will be designed and shared with users.

If reports will be validated by a third party, organizations need to

determine how reports will be generated and shared, and what data
must be included in the report before it can be approved.

Requirements gathering also includes security policies—encryption, access

controls, and row and column level—and requirements related to data
retention, data residency, and personally identifiable information (PII).

Document printing and electronic reporting

Document printing requirements are another factor to consider. To

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 326
 comply with local laws, businesses typically must submit regulatory
and compliance documents in a pre-defined printed or electronic
format provided by government agencies. The data required for
these documents often resides in enterprise resource planning
(ERP) systems.

The Dynamics 365 Finance and Dynamics 365 Supply Chain

Management apps have an Electronic reporting (ER) tool to help
organizations adopt new regulatory requirements and generate doc-
uments in the required format to electronically exchange information
with government bodies, banks, and other parties. Targeted at business
users instead of developers, the ER tool lets users configure different
document formats instead of writing code, making the processes for
creating and adjusting formats for electronic documents faster and
easier. It works with the TEXT, XML, Microsoft Word document, and
OPENXML worksheet formats, and an extension interface provides
support for additional formats.

Customer-engagement apps allow users to use Word templates

to create standardized documents. The Dynamics 365 Customer
Engagement app includes system reports for sales insights, while
Dynamics 365 Sales lets you individually print quotes, invoices, or
other records—and create PDF files from sales records.

Business document management

Business document management is built on top of the ER framework,
and enables business users to edit document templates by using
Microsoft 365 or a Microsoft Office desktop application. Edits might
include changing business document designs and adding placeholders
for additional data without source code changes and new deploy-
ments. No knowledge of the ER framework is required to update
templates of business documents.

Financial reporting
Finance and business professionals use financial reporting to create,
maintain, deploy, and view financial statements. The Finance app’s
financial reporting capabilities move beyond traditional reporting
constraints to help you efficiently design distinct types of reports.

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 327
It includes complex currency-reporting requirements and financial
dimension support, and makes account segments or dimensions
immediately available—no additional tools or configuration steps
are required.

Financial reporting in the Finance app provides default financial re-

ports that organizations can use as is or as a starting point for their
financial reporting needs. In addition to traditional financial reports
such as income statements and balance sheets, the app’s default
reports include examples of the many types of financial reports you
can create and customize.

Dashboards, charts, and analytical workspaces

The Sales app uses dashboards to provide an overview of actionable
business data across an organization, and give insights on sales data
and team performance. Sales representatives and managers can use
an out-of-the-box sales pipeline chart in the Sales app to visualize the
revenue for an opportunity based on each pipeline phase.

Dynamics 365 apps deliver rich, interactive reports that are seamlessly
integrated into application workspaces. By using infographics and visuals
supported by Microsoft Power BI, analytical workspaces let users explore
the data by selecting or touching elements on the page. They also can
identify cause and effect, perform simple what-if operations, and discover
hidden trends—all without leaving the workspace. Power BI workspaces
complement operational views with analytical insights based on near-real-
time information. Users also can customize embedded reports.

Reporting categories
Reporting needs for an organization can be classified into two
categories: operational reporting and business reporting.

Operational reporting
Examples of operational reporting include orders received in a day,
delayed orders, and inventory adjustments in a warehouse. This kind of
reporting supports the detailed, day-to-day activities of the organiza-
tion. It is typically limited to a short duration, uses real-time, granular

328
 data, and supports quick decision-making to improve efficiency. It also
helps organizations identify their issues and achievements, as well as
future strategies and actions that may affect the business. Operational
reports can empower businesses to determine their next steps for
improving organizational performance. Organizations can fulfill oper-
ational reporting requirements using elements such as native controls,
SSRS reports, dashboards, and business documents.

Business reporting
Business reporting refers to reports detailing operating expenses and
financial key performance indicators (KPIs) to business stakeholders so
they can understand the organization’s overall health and make more
informed decisions. This kind of reporting delivers a view of current
performance to enable the stakeholders to identify growth oppor-
tunities and areas for improvement, and track business performance
against the planned goals for the organization.

The data estate and data

modernization
A data estate refers to the infrastructure that helps organizations man-
age all their data, regardless of where or how it is stored. The massive
growth in data and the increasing need for immediate insights demand
a fast, agile, accessible, secure, and highly scalable data platform.

A common risk for businesses is the quality

of their data. If data is weak, unstructured,
Organizations typically used legacy applications to both store data
unsecured, or difficult to use or access, and serve as the interface for data collection. Data modernization
additional work will be required to move,
scrub, protect, and improve the quality of restructures how data is collected, stored, and managed to take advan-
the data. Poor data quality also can lead to tage of innovative technology, and is backed by AI and data analytics
lost opportunities, failure to consistently
attract customers, increased time and solutions that optimize the data-review process. Organizations seeking
expense for cleaning data or processing
improved business performance must design and manage their data
corrections, and inaccurate or inconsistent
KPI measurements. estate based on a modern platform that breaks the data silos and
unleashes the data’s full potential.

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 329
 Build a data strategy
For business solutions to deliver the intended results, an organization’s
data strategy must articulate a clear vision that effectively aligns busi-
ness intelligence investments with business goals to maximize impact,
drive growth, and improve profitability.

To define your data strategy, start with the business questions you
need to answer to meet your organization’s goal of using the data to
make more effective decisions that improve the business outcome.
Some organizations emphasize
collecting data more than analyzing it to
drive insights. It’s also important to identify
gaps and blockers for business objectives
With customers now having access to more information and channels,
and outcomes, and not focus solely on the an organization’s data strategy should reflect the customer journey.
data, structure, analytics, tools,
or technologies. From a data management perspective, all channels and checkpoints
across the customer journey should be represented.

Modernize your data estate

Your data estate modernization strategy should map the current data
estate, goals, business processes, and regulatory requirements to aid
gap analysis of the existing and desired future state. It should identify
key analytics and metrics to improve the business and effectively align
business intelligence investments with business goals.

Cloud-based solutions offer modern capabilities based on machine

learning and AI to analyze portions of data and automate other pro-
cesses, such as identifying and solving data quality issues. In terms of
analysis, prebuilt models can serve the most common use cases, elimi-
nating the need to build custom models.

The modern data estate and the digital

Implementing or investing in sustaining
legacy solutions brings the risk of a future feedback loop
gap. AI and machine learning solutions
take a modern approach to data and an- Gartner, Inc., predicts that investments in data and analytics strategies
alytics that grows as technology evolves, will critically affect businesses in the years ahead.
making it possible for organizations to
add new features and capabilities in the
future—and reduce the risk of being out-
How organizations modernize their data estate is influenced by the
performed by competitors.
four data signals found in the digital feedback loop (Figure 13-1).

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 330
 Empower people
To do their jobs more efficiently, employees need tools and resources,
as well as timely access to information. Utilizing AI to further automate
business processes contributes to better and faster results, which then
empowers your people (Figure 13-2) to make the best decisions and
deliver value to customers.

Engage customers
Modern applications are already capable of delivering insights by
using AI and data analytics to optimize business processes and shed
light on customer activity. For example, Dynamics 365 apps can pro-
vide a customer service agent with insights into a customer’s interests
and purchasing behavior information in real time, allowing the agent
to make suggestions tailored to the customer. These types of insights
help businesses intelligently engage customers (Figure 13-2) to
Fig.
provide a superior customer service experience.
13-2

Empower people Engage customers

By 2023, 30 percent By 2025, digital

60 30 50
of customer service adoption solutions will

% % %
organizations will be white-labeled in
deliver proactive 50 percent of customer-
customer services by facing software as a
using AI-enabled service (SaaS)
process orchestration applications, increasing
and continuous customer satisfaction
intelligence. and loyalty.
By 2022, 60 percent
of organizations
will utilize
packaged AI to
automate processes

70
30
in multiple %
functional areas. %
60 %

By 2022, 70 percent of By 2023, 30 percent

50
organizations will of organizations will
Transform products Optimize operations
%
rigorously track data exceed their data and
quality levels via analytics return on

(and services) metrics, increasing data

quality by 60 percent,
investment (ROI) by
governing the least
and significantly amount of data that
reducing operational matters most to their
risks and costs. strategic goals.

By 2025, 50 percent
of all enterprise
business-to-business
(B2B) sales technolo-
gy implementations
50 %
5X 90 %
30 %
30 %

By 2023, cloud-based By 2023, 90 percent of By 2023, 30 percent of By 2024, 30 percent

will use customer- By 2025, more than 50 percent of AI will have increased the world’s top 500 organizations will of organizations will
engagement analytics equipment manufacturers will offer five times from 2019, companies will have harness the collective invest in data and
making AI one of the converged analytics intelligence of their analytics governance
to optimize their outcome-based service contracts that top cloud services. governance into broader analytics communities, platforms, thus
sales processes. rely on IoT-based connectivity (up data and analytics outperforming increasing the
governance initiatives. competitors who rely business impact of
from by
Implementation Guide: Success lessDesign:
than 15 percent in intelligence,
Business 2019). reporting, and analytics solely on centralized or 331
trusted insights and
self-service analytics. new efficiencies.

Source: “Predicts 2020: Data and Analytics Strategies – Invest, Influence and Impact.” Gartner.com. Gartner, Inc., December 6, 2019.
 Transform products (and services)
Products and services have become sources of data that provide
better insights into lifecycle-related information that can be used to
benefit organizations and their customers. By analyzing those in-
sights, organizations can in turn transform their products and services
(Figure 13-2) to take advantage of opportunities and expand into
new channels and markets. For example, a smart air conditioner that
allows you to set an optimal temperature via an app can also apply
what it learns about your household’s daily routine to make sure
the house is cool when everyone gets home. The air conditioner’s
manufacturer then uses data collected from all users to enhance the
product’s smart features, leading to a continuous improvement cycle.

Optimize operations
Cloud-based AI usage is an increasingly common investment area for
most organizations—and not just for customer-facing technology.
For example, Dynamics 365 for Field Service can use AI to anticipate
hardware failures on a manufacturing floor and automatically dispatch
technicians before the malfunction. Organizations that optimize their
operations (Figure 13-2) with augmented analytics, AI, and embedded
intelligence will be more competitive in the marketplace.

Components of the
modern data estate
The Dynamics 365 platform can be an essential part of your modern data
estate architecture. Your business data is securely stored within Dynamics
365 as a function of your day-to-day operations. In addition, Dynamics 365
can export data to or ingest data from various sources to be used in re-
porting, workflow, applications, or in any other way that is required by your
business.

You can also generate insights and analytics from data created and managed
inside each Dynamics 365 application. Apps using embedded intelligence,
such as the audience insights capability inside the Dynamics 365 Customer
Insights app, enrich your data and allow more informed decision-making.

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 332
 Common Data Model
Dynamics 365 data estate components can ingest, store, prepare,
model, and visualize data to produce insights that will support,
boost, and even transform operations (Figure 13-3). The Common
Data Model (CDM) is one of the key technologies enabling access to
many kinds of information from heterogeneous services and
data stores.

The CDM—a shared data language used by business and analytical

applications—offers a set of standardized, extensible data schemas
to enable consistency of data and its meaning across applications
and business processes.

The CDM specification defines out-of-the-box standard entities

representing common concepts such as people, budgets, and cam-
paigns. Having your data in a common format for entities with the
relevant metadata makes it possible to build apps and services that
light up based on the data. The CDM format also makes it easier to
consume data in the Microsoft Power Platform, which uses Power
BI to get insights from your data. Normalizing the data sets up your
organization to better identify opportunities and puts your data in
a format that could be utilized for future projects. Dynamics 365
Fig. uses Microsoft Dataverse—which is structured according to the
13-3

Components of a modern data estate

App 1 App 2 App N

Common
Data
Model

Data source 1 Data source 2 Data source M

The Common Data Model (CDM) The CDM enables quick app deployment
unifies data in a well-known and development, out-of-the-box
schema with semantic consistency. intelligence, and more.

There’s a growing collection of solutions that store data in CDM form:

Power BI Dynamics 365 Power Apps Azure

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 333
 CDM—to store and secure app data. The CDM structure is defined
in an extensible schema, as shown in Figure 13-4. This allows
organizations to build or extend apps by using Power Apps and
Dataverse directly against their business data.

Fig.
13-4

CDM Schema

Core
Customer relationship management (CRM)
Account
Account Sales Service Solutions
Activity
Appointment Competitor Case
Marketing
Contact
Campaign Discount Contact Account

Currency
Contact Invoice Entitlement Contact

Email
Lead Opportunity Resource Event

Goal
Marketing list Order Service Marketing email

Letter
Phone call Order product Schedule group Marketing page

Note
Social activity Quote Task

Owner

Organization

Position Web Talent Healthcare

Link clicks Department Patient Risk assessment Procedure
Task

Web interactions Job position Practitioner Encounter

Webpage Worker Device Episode of care

Care plan Medication

Working with indus-

try and subject
matter experts, the
CDM is expanding to Account Attributes
include additional Description: A business that represents a • accountNumber • creditLimit • territoryId
business enterprises customer or potential customer. • accountRatingCode • openDeals • hotelGroup
and concepts. • createdOn • openRevenue • ...

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 334
 Data unification components
Services and applications that ingest data from multiple sources
serve a vital role in a modern data estate. Aggregation from data
stores and services provides users with business-critical information
supplied in dashboards and reports. The resulting data and events
can also be used to trigger workflows and act as inputs to the apps
running on the Dataverse platform. Many data unification
components are built into Dynamics 365 applications—and
organizations can design their own integrations to fit their business needs.

Customer Insights as a customer data platform

Customer Insights is a real-time customer data platform that brings
together transactional, behavioral, and demographics data from
various sources to create a 360-degree view of your customers.

The Customer Insights solution (Figure 13-5) offers prepackaged

customer-analytics capabilities such as segmentation, churn
analysis, and product recommendations, with Power BI embedded
for easy data exploration and visualization. It also incorporates
AI and machine learning models, so data scientists can create
Fig.
custom models in the Azure Machine Learning platform.
13-5

Customer Insights solution

Internal External Process optimization Channels of
and automation engagement
Ingest Map and match
mographic
Contacts Partner data
De Mobile
Point of sale
Events
(POS)
Power BI Web
Cases Loyalty Analytics
io n al

Campaigns Geo
Beh

Social
act
av

Web Wi-Fi Power Automate

ra
ns

Tra
io

l AI and
Bots
Email Social machine
Conflate Enrich learning
Mobile LinkedIn Power Apps In-person

Data Unification Insights Action

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 335
 The addition of Azure Cognitive Services provides text, speech,
image, and video analysis, and enriches data via Microsoft Graph.

Dataverse applications
Building an app typically involves accessing data from more than one
source. Although it can sometimes be done at the application level,
there are cases where integrating this data into a common store creates
an easier app-building experience—and a single set of logic to maintain
and operate over the data. Dataverse allows data to be integrated from
multiple sources into a single store, which can then be used in Power
Apps, Power Automate, Power BI, and Power Virtual Agents, along with
data that’s already available from Dynamics 365 apps.

Data export components

A key characteristic of a data estate is the ability to export data to ser-
vices as needed. Because data in the Dataverse conforms to the CDM
standards, businesses have the flexibility to export data from Dynamics
365 to other cloud services and applications using an industry-
standard format, greatly simplifying the standardization of exported
and imported data. To facilitate this business need, Dynamics 365 has
several built-in export services.

Export to Azure Data Lake

The Export to Azure Data Lake service is a pipeline and the preferred
option to continuously export data from Dataverse and the Finance
and Supply Chain Management apps to Azure Data Lake Storage
Gen2. Designed for enterprise big data analytics by delivering high
availability with disaster recovery capabilities, Azure Data Lake Storage
Gen2 provides a scalable storage facility for data that can be used in
conjunction with AI services.

When connecting from Dataverse, data is stored in the CDM format

to provide semantic consistency across apps and deployments. When
connecting via the Finance and Supply Chain Management apps, the
data is stored in a native table format and as entity metadata that can
be used for further analysis.

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 336
 Data Export Service
The Data Export Service replicates data from the Dataverse database
to an external Azure SQL Database or an SQL Server on Azure virtual
machines. This service intelligently syncs all data initially, and thereaf-
ter syncs the data on a continuous basis as delta changes occur in the
system, enabling several analytics and reporting scenarios on top of
Azure data and analytics services.

Bring your own database (BYOD)

Making data available to all users is risky. The bring your own database (BYOD) feature lets organizations export
Security in reporting and analytics should
data from the Finance app into an external Azure SQL database. The
be aligned with your business solution.
Be sure to have your data governance in app’s default operational reports already take advantage of embedded
place, and periodically review who needs
access to which information. This may
Power BI and Entity store (an operational data warehouse), but there
be a time-consuming process, but your are scenarios where organizations may need to export data for unifica-
data must be protected, secured, and only
accessible to those who truly need it. tion or other purposes.

Embedded intelligence
Dynamics 365 apps with embedded intelligence, such as Sales
Insights and Customer Service Insights, allow organizations to
use AI without depending on highly skilled resources. These apps
continuously analyze your data and generate insights to help you
understand business relationships, evaluate activities and
interactions with customers, and determine actions based on
those insights.

Unified data results in a 360-degree view of customers to ensure

high-quality decision-making and predictions informed by the best,
most recent data.

Power Platform
The Power Platform (Figure 13-6) enables organizations to analyze, act
on, and automate the data to digitally transform their businesses. The
Power Platform today comprises four products: Power BI, Power Apps,
Power Automate, and Power Virtual Agents.

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 337
 Fig.
13-6
Microsoft Power Platform
The low-code platform that spans Office 365, Azure, Dynamics 365, and standalone applications

Power BI Power Apps Power Automate Power Virtual Agent

Business analytics Application development Process automation Intelligent virtual agents

Dataverse

Power BI
Power BI is both part of the Power Platform and stands on its own
by bridging the gap between data and decision-making. Power BI
lets business analysts, IT professionals, and data scientists collaborate
seamlessly, providing a single version of data truth that delivers in-
sights across an organization.

Power BI helps you analyze your entire data estate within the Dynamics
365 or Azure platforms, or external sources. Power BI can connect indi-
vidually with siloed sources to provide reporting and analytics, or it can
connect with data stores within or outside of Dynamics 365. As data can
come from multiple sources, organizations should analyze how Power BI
will connect with those sources as a part of their data estate pre-planning.

Power Apps, Power Automate, and Power Virtual Agents

Built to work seamlessly with data in the CDM format, Power Apps
allows you to consume, create, and complement information to
contribute to scale and efficiency for the users consuming that data.

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 338
 Organizations can also get accurate insights by adding low-code AI tools
to their process automation via Power Automate. Power Virtual Agents
help you create and manage powerful chatbots—without the need for
code or AI expertise—and monitor and improve chatbot performance
using AI and data-driven insights.

Microsoft Azure
With Dynamics 365 at the center of the data estate, Azure provides an
ideal platform (Figure 13-7) for hosting services for business workloads,
services, and applications that can easily interact with Dynamics 365.
Built-in services in Dynamics 365 let you export data as needed or
scheduled. Power BI can aggregate information from Dynamics 365
and Azure sources into an integrated view, and Power Apps can access
both Dynamics and Azure sources for low-code, custom applications
designed for business use.
Fig.
13-7

Azure features and services with modern data warehouse architecture

Ingest Model and serve Visualize
On-premises
data source
Azure Synapse Analytics

Other data
source
Prep and train
Azure Data Factory Power BI
Logs and files Code-free data transformation and integration Azure Databricks Leader in the Gartner Magic Quadrant for Analytics
from more than 90 built-in connectors and Business Intelligence Platforms
(unstructured)

Store

Dynamics 365 Azure Data Lake Storage

High-performance, low-cost data
lake available in 54 Azure regions

Azure cloud platform

For data estates that include Dynamics 365 and other platforms,
Azure can help with unification of data to generate insights. There
are multiple options within the Azure cloud platform to ingest,
store, prepare, model, and visualize data (Figure 13-8), as well as
build intelligence on top of that data—whether it’s siloed or unified.

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 339
 Fig.
13-8

Modern data platform reference architecture

Data Process Business user

Stream datasets and

real-time dashboard

Hotpath
Real-time analytics

Stream Load and ingest Azure Stream Analytics

V=Velocity
IoT devices, sensors,
gadgets Lambda architecture
(loosely typed)

Coldpath
History and
Nonstructured Azure Event Hubs trend analytics Azure Cognitive Services Power BI Premium Analytics
V=Variety Azure Machine Learning
images, video, audio, free
text
(no structure)

Semi-structured Azure Data Factory Azure Data Lake Storage Gen2 Azure Databricks Azure Cosmos DB Application
V=Volume
Scheduled/event-triggered Fast load data
images, video, audio, with PolyBase/ Integrated big data
data integration ParquetDirect scenarios with traditional
free text data warehouse
(loosely typed)

Enterprise-grade
semantic model

Azure Synapses Analytics Power BI Premium Analytics

Relational database
(strongly typed, structured)

Store Server

Azure Stack
Azure Stack is a portfolio of products that allows you to use
embedded intelligence to extend Azure services and capabilities to
your environment of choice—from the datacenter to edge locations
and remote offices. You can also use it to build and deploy hybrid
and edge computing applications, and run them consistently across
location boundaries.

Data load, ingestion, and pipeline orchestration

Azure Data Factory is a hybrid data integration service for creating,
scheduling, and orchestrating ETL and extract, load, and transform
(ELT) workflows. Organizations can use Azure Data Factory to
combine CDM data and other data sources, and process that data
for unique insights. Azure Event Hubs can ingest data streams
generated by a client application.

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 340
Data store
Azure Blob Storage offers massively scalable object storage for any type of
unstructured data—such as images, videos, audio, and documents—while
Azure Data Lake Storage eliminates data silos with a single and secured
storage platform.

Machine learning and AI

Azure Databricks provides a cloud-hosted Apache Spark cluster where
data engineers, data scientists, and business users can collaborate to train
their models and get analytics.

The Azure Machine Learning service gives developers and data scientists
a range of productive experiences to build, train, and deploy machine
learning models faster.

Azure Cognitive Services puts AI within reach of every developer—without

requiring machine-learning expertise. It only takes an API call to embed
the ability to see, hear, speak, search, understand, and accelerate deci-
sion-making in your apps.

Model and serve

Azure Analysis Services employs enterprise-grade analytics as a service
(AaaS) to govern, deploy, test, and deliver business intelligence solutions.

Azure Synapse Analytics is a fast, flexible, and trusted cloud data

warehouse that lets you scale, compute, and store data elastically and in-
dependently with a massively parallel processing architecture. It can scale
across proprietary SQL and open-source databases and manage analytics
workloads to provide fast, cost-effective power over any data—whether it
is structured, semi-structured, run in real time, or globally distributed.

Azure Cosmos DB is a fully managed NoSQL database service for modern

app development.

Synergy
Getting maximum value from your data requires a modern data estate
based on a data strategy that includes infrastructure, processes, and people.

341
 Your data can flow inside a cloud solution or via synergy with other
components and platforms to provide an infrastructure and processes
for analyzing data and producing actionable outcomes.

People are a big part of the process, and the data journey will often
start and finish with them. The insights and actionable outcomes will
allow them to make better decisions—for themselves and the business.

Conclusion
In this chapter, we discussed how organizations are becoming more
competitive, expanding their global reach to attract customers, and
using business intelligence solutions to make the most of their data.

While seeking data on how customers interact with their products

and services, organizations are acting on that data to give customers
access to more content, new purchasing channels, and brand options.
By deeply understanding customer behavior, organizations can engage
customers proactively and make predictions about their future behavior.

Successful organizations will use data to empower their employees,

intelligently engage their customers, transform their products and
services, and optimize their operations. They’ll also use data to reduce
operational risks and costs, and to respond to opportunities and mar-
Due to budget, time, or skills constraints,
some organizations decide to deliver a ket changes to meet their customers’ needs.
business solution as a first step, with a
plan to improve insights and analytics
capabilities at some point in the future. Business intelligence, reporting, and analytics should not be an after-
Insights and analytics should be addressed
in the early phases of a business solution, thought. Focusing on building a unified and modern data estate gives
even if it doesn’t yet include all scenarios.
organizations access to augmented analytics and embedded intelligence
Leaving these vital elements until later can
affect the business and the user experience, solutions, which will be the differentiator in the future. By changing the
eventually reducing adoption, increasing
costs, and giving a negative perception of
ways that they use data to understand and engage with their customers,
the value the solution has to offer. organizations can reach each customer at the right time, on the right
channel, with the right message, for the right outcome.

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 342
 Checklist

Reporting and Use customer data platform offerings such as customer

insights to unify the customer data from various siloed
analytics strategy data sources.
Map out the organizational data estate to develop a
Focus on not just delivering a report but actions for the
holistic view of different data sources, the type of data
users in the context of the application.
they hold, and the schema used.

Define your analytics strategy and the tools to support

it. Ensure the approach meets the current and future
reporting requirements while considering the data
volumes and different sources of data.

Create an intelligence strategy that considers the report-

ing needs at various levels, such as strategic business
reporting, operational and process reporting, end-user
intelligence, and administrative reporting.

Take advantage of the out-of-the-box embedded

intelligence capabilities in the app.

Align the organization insights and analytics roadmap to

the business capabilities being delivered.

Align to the Common Data Model to take advantage

of the standardized business applications schema for
better inoperability across systems.

Understand the security implications when exporting

data from the app to external analytical data stores.

Align the reporting and analytics to the broader master

data management strategy.

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 343
 Case study
Premier yacht brokerage
cruises into smarter
marketing and boosts
sales by 70 percent with
Dynamics 365
To differentiate itself from the competition, a large company that supplies
luxury vessels decided to invest in reporting and intelligence. In an
industry where relationship-building efforts must be incredibly precise
and personal, any insight into a customer’s mindset is invaluable.

The company implemented a solution based on sales, marketing, and

finance data, using Dynamics 365 and Power BI to help identify trigger
points that would help increase sales to charter clients, strengthen
customer relationships with improved marketing activities, and
seamlessly move customers through the sales cycle.

This solution allowed the company to build intelligence on top of a

centralized system running in the cloud, and facilitated a move from
legacy systems where data was siloed in different applications, channels,
and departments.

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 344
 Dynamics 365 apps made it possible to separate the features each
department would use while accessing the same data. The marketing
and sales teams started customizing their content and campaigns to
nurture leads toward a sale, and used Power BI reports to generate
insights that helped them identify the best prospects and create
winning proposals.

Integration with external applications brought in data from other

sources to improve the company’s bespoke offerings. For example, the
sales teams could collate even more information about their products
and send it to customers in just a few minutes—instead of spending
hours manually putting together the data.

After deploying the Dynamics 365 solution, the company experienced

a 70 percent increase in sales. They next brought in more departments
to create a central hub for all activities, simplifying every area of their
operation. For example, the sales team’s process integrated everything
that the accounting department needed to efficiently close a deal and
reduce the possibility of human error.

The company is also using Dynamics 365 and Power BI to uncover

market trends, as augmented analytics help the company learn about
their customers in ways that were not possible before—and build
authentic relationships that will last long after the initial sale.

Implementation Guide: Success by Design: Business intelligence, reporting, and analytics 345
14
Implementation Guide: Success by Design: Testing strategy
Guide
Testing
strategy
346
 Overview
During the implementation of a solution, one of the
fundamental objectives is to verify that the solution
meets the business needs and ensures that the
customer can operate their business successfully.
The solution needs to be tested by performing multiple business execu-
tion rehearsals before the operation begins and the system is taken live.

It is imperative that proper testing of the solution is completed before

deployment to production. This helps avoid costly challenges that
could delay the adoption, create negative perception of the final
product quality, and reduce the confidence to run the business on the
new solution.
Defining a comprehensive test strategy
for your implementation is a combination
of considering the project scope, testing
The goal of testing is to detect defects in the process in order to ready
planning with a test plan and test cycle
schedule in combination with the phases the solution, processes, and people to operate the business in produc-
and solution versions availability, and
selecting all the test types and how to
tion application efficiently and accurately as soon as possible. Testing
execute them. allows us to detect defects as early as possible in this process, in a
• Define scope based on your systematic way and at the recommended time per the test type.
processes and detailed test cases
• Create a test plan with the necessary
test cycles This chapter explores the components of:
• Select the adequate test types for
▪ The value of testing
your project
▪ Types of testing
We explore each of these areas in more
detail in this chapter.
▪ Testing at the right time with clear scope
▪ Defining the strategy

Implementation Guide: Success by Design: Testing strategy 347

 You can also find recommended practices for executing the testing,
how to deal with the outcomes, and finally, reaching a business sign off
that confirms the readiness of the solution.

Having a properly planned and executed test strategy eliminates

the practice of doing parallel testing in production since it is no
longer a recommended risk mitigation approach during customer
implementations.

We look forward to identifying the key components of your testing

strategy in this chapter.

Focus on quality
and keeping scope
Testing is a continuous task under application lifecycle management.
Not only is it critical during the implementation of the solution, but
also during the Operation phase. It is a continuous evolution that
keeps bringing fixes or extending the solution. In the beginning testing
is completed manually, but over time with automation we can make
the process far more efficient. Our objective is to ensure the quality of
the solution is always meeting customer expectations.

To align with the scope of the solution, testing needs to focus on

validating the business processes in it to stay on track. Microsoft
Dynamics 365 Business Applications are rich with features that can
distract testers easily during testing, so it is important to focus on the
features that add value to the business process. By only focusing on
those features and the requirements needed for them, we can get to
the starting line of initiating operations for the solution being built.

The approach for testing

Depending on the methodology you apply, agile, waterfall, or a
hybrid of the two, the testing frequency differs for each iteration.

Implementation Guide: Success by Design: Testing strategy 348

 Regardless which approach you follow, you need to look for a
high-quality testing approach to confirm readiness.

On top of your implementation methodology, you need to have a

clear strategy for testing, as illustrated in Figure 14-1. This strategy is
composed of:
▪ A clear testing scope defined by your processes and requirements
to validate.
▪ A comprehensive plan that brings clarity to when, why, who, how
much, and where to test.
▪ The different types of testing with descriptions of how the testing
is executed based on the complexity of the solution and minimum
quality gates allowing us to detect defects early in the process.
▪ A comprehensive tracking mechanism that allows you to monitor
overall progress, pass/fail rates, ownership, and issues tied to
specific test cases.
▪ A clearly defined process for adding or modifying test cases during
each individual test pass.

Explore each component to define the testing strategy for your

Dynamics 365 implementation.

Defining the
testing strategy
After observing thousands of Dynamics 365 deployments, we can see
that most customers do a reasonable level of testing before go live.
But the difference between each of those implemen-
Fig.
14-1 Test strategy tations is how thorough the implementation team is
at defining their strategy, and the quality and depth of
the testing.

Scope Plan Test type

In general, a good strategy does not change in
approach between implementations. What changes
What you When, who, where, How you
will test and how much you will test is the tailored artifacts used for each implementation,
will test
for example, the process to define the scope can

Implementation Guide: Success by Design: Testing strategy 349

 be the same across projects, but the scope itself differs. Investing in
a good strategy and making sure the implementation teams use it
consistently increases the quality of the project outcome.
Make sure that every implementation
you execute has a clear and solid
testing strategy. Scope of the testing
The scope of the testing is defined early in the project after defining
the solution, and it is refined as the solution blueprint takes shape.
What changes throughout the project is the type of testing that comes
into focus as you move through the implementation lifecycle and the
ability to actually test the solution.

During the implementation, the number of test cases in the testing

scope keeps increasing as you progress in building the solution. After
golive, the scope is focused on maintaining the quality as you update
the solution, but testing can increase if you continue to expand.

Consider the testing journey as an incremental process, as illustrated

in Figure 14-2. Regardless of the timeline, you need to keep two
things in mind:
▪ A clear view of the end-to-end business processes.
▪ A clear definition of the depth and detail of the test case design,
but also the variations of testing, for example edge cases and
negative testing.

Testing connected to processes

To progressively build your solution you need to determine the implemen-
Fig.
14-2 tation processes and the requirements connected to each of them. As we
rage of the tes
Cove t
build the solution and make it ready for validation, the scope of the testing
increases as we add more processes. During the project progression, we
go from a simple functional test to an end-to-end testing scope during
user acceptance testing (UAT). Always connect the test to the process
Processes being validated since this is how the end user confirms if they can perform
Scope of
testing
a connected action to execute the business process in place.
Test cases

This connection is primarily important for testing connected functional

requirements, there is testing that checks other aspects of the solution out

Dee
of the process, for example testing quality of data, testing latency in the
pness of the test
connectivity, etc.

Implementation Guide: Success by Design: Testing strategy 350

 Testing is not an artifact of the development process; it is an
artifact of the business process created when you implement business
applications. Plan your testing by connecting those tests to the processes
Connect the test to the process being
and include processes that are executed outside of the new application
validated. This facilitates tracking,
reporting, and final sign off on being implemented.
readiness by the business team.

Test case definition

An important part of the scope is defining the test. Proper test doc-
umentation is key, and test cases (or test scripts) are the artifact that
achieves it. The nature of a test case varies depending on the test type.
Clear and detailed test cases guide the tester to validate the expected
outcomes without distractions and align to the scope of the solution.

Writing a good test case requires a series of tools and tasks that help
you validate the outcome. Dynamics 365 Business Applications provides
some of the tools to build test cases faster, for example, the task recorder
in the Operations apps where you can dynamically capture the steps
needed to execute a task in the app.

Test cases should reflect the actual business execution in the system
and represent how the end user ultimately operates the system, as
illustrated in Figure 14-3.

Fig.
14-3 Components of the test case

+ + +
What Get ready How Quality check
Linked business Prerequisites Reproduction steps Expected outcome
process requirements

Implementation Guide: Success by Design: Testing strategy 351

 Test cases should be composed of, at minimum, the following:
▪ The process and requirements that the test case is covering.
▪ The prerequisite, or entry, criteria to execute the test, which can
be dependent on other tests or necessary data to produce the
expected outcome.
Azure DevOps is great tool for ▪ The reproduction steps.
documenting test cases and connecting
them to the solution development ▪ The expected outcome, or exit, criteria that helps to confirm
lifecycle. Having this artifact in Azure
the readiness.
DevOps allows for tracking any bug
discovered during testing and helps
to trigger the fix and plan for the next
testing cycle while also determining
Finally, when you design your test cases, design for what you expect to
the coverage of your test; for example, happen based on the process objective but also on what it should not do.
process, user story, requirement, design,
etc. This is depicted in Figure 14-4. Plan for tests that can break the process.

Planning
We have defined what is needed to determine the scope of the testing
using the processes as a foundation. With this valuable information, we
can determine how to execute the testing. This is where the planning
becomes important.

Planning for testing is a fundamental part of the testing strategy. The next
section describes the minimal components needed to define a recom-
mended testing plan. This strategy can be used to implement any of our
Fig.
14-4 business applications.

Process Requirement Prerequisites

Prospect to cash Create sales order based on servicing Customer with no credit line available
system part request Stock available

Description
Dealership request parts for recreational vehicles out of warranty

Test case ID Test steps Test data Expected results Actual results Pass/Fail
PC-20 1. Create sales order header Customer: Sales order cannot Sales order is Fail
2. Check credit line in order CU050 - Great Adventures be posted posted
3. Create order line Dealership
4. Validate and post sales order
Part:
P001 - All weather speakers

Qty: - 4 pieces
Tester notes P001 - All weather speakers

Credit check is not stopping the order to be posted. Customer setup with wrong credit limit. Data quality issue.

Implementation Guide: Success by Design: Testing strategy 352

During the planning stage of the testing, you need to be able to answer
and provide clarity for the following questions. This helps design and
define the frequency of the testing based on the necessary iterations or
test cycles to secure a high-quality solution.

▪ When do you need to start testing?

▪ How do you control the versions of the solution to be tested based
on the progressive readiness of it?
▪ How do you document the outcome of the testing?
▪ Who participates in the testing?
▪ Where does the testing happen in terms of environments and type?
▪ How is the ownership of the outcome defined?
▪ How deep you are going to test?
▪ What types of testing do you execute based on the project needs
and solution complexity?

This helps to plan the quality control portions of the solution and must
happen at the start of the Initiate phase of the project.

The plan for testing must be documented and signed off on by the
business team prior to its execution. This is important because it
leads into other types of planning, like environment planning, that
determines where the test is done.

Creating a test plan

The test plan is the blueprint for executing your tests. It establishes the guard
rails to stay on track in terms of quality assurance for the solution, schedule of
testing, and resources needed, and it contains the scope of the testing.

Depending on the complexity of the project, this artifact can be developed

by the same implementation team. In larger, more complex organizations,
this is prepared in conjunction with the quality assurance team to secure a
neutrality at tracking and reporting the outcome of the test, as illustrated
in Figure 14-5.

Always create a test plan for your project regardless of if your

implementation is simple, complex, or if you are implementing a
standard solution or extending the solution.

353
 Fig.
14-5 Components of the test plan

Test cycles Roles Resources

How many Different personas Environments,
cycles involved devices,
printers,
etc.

Schedule Version Ownership Documentation

When testing Solution version Who will own Test asset
happens or scope per the outcome library
cycle

The test plan brings transparency and helps keep your goals and
objectives on securing a quality solution. It contains important
information that guides the team on how to conduct the testing.

When does testing happen?

Always create a test plan at It’s simple: the testing should start as early as possible. It is never too
implementation and obtain sign off
early to start testing. Your testing cycles are incremental in terms of the
on it by business stakeholders.
scope. Your first test cycles could be to test the standard functionality
to validate how far are you from the overall desired outcome.

Do not wait too long, or until the end of the implementation when you
think you have built the entire solution, to start testing. Testing late
adds serious risk to your implementation by having issues and gaps
that are hard to fix in a short period of time. This becomes a constraint,
especially when you have reduced time available for time sensitive
projects tasks like preparing to go live. Always plan to avoid testing
close to your go live date since it leaves no time to resolve issues. Poor
planning or no testing can ruin a project.

Testing cycles or iterations

Testing cycles are the frequency of the testing aligned to the availability
of the solution in scope for the testing at a given period time in the
implementation. Test cycles act as comprehensive testing milestones.

Implementation Guide: Success by Design: Testing strategy 354

 The successful outcome of the test cycle confirms the readiness of the
solution to meet the requirements of the business. There are different
terms used to describe this testing event, for example conference room
pilots, testing iterations, testing cycle, etc. The important message here is
that the testing event is comprehensive.

To implement Dynamics 365 Business Applications, it is recommended

that the scope in each testing cycle is based on business processes.
Test cycles by application module, or organizational department in
isolation, risk losing the big picture and thereby limit the effectiveness
of the test.

Every testing cycle represents a key milestone in building the solution.

Consider every test cycle as a mini go live, where you are rehearsing
the business operation at every test.

Solution versions and alignment to them

We discussed the importance of planning the testing cycles process as
part of the scope, this scope also triggers the readiness of the code and
data that supports the test cases connected to those processes. After
determining the planned scope of the test cycle, you should plan your
development efforts to cover those processes and the readiness of the
required data in terms of master data and configurations.

Your solution version needs to be controlled at every test cycle so you

can have a clearly defined milestone. For every new milestone, the
solution keeps expanding in scope and complexity.

The solution version is a combination of the version of your:

▪ Code
▪ Configurations
▪ Master data
▪ Migrated data

Different teams need to coordinate and plan how to combine their

efforts to prepare for the testing cycle.

Each component of the solution version is very important, and all of

Implementation Guide: Success by Design: Testing strategy 355

 them need to come together as part of the solution. This is especially
important with data, the earliest you bring migrated data into the
mix the better. One of the most common causes of errors during the
Align your solution version to the
first days of operation is poorly migrated data.
testing cycles and coordinate the
readiness of all the dependent parties to
provide their own artifacts. Missing one
necessary artifact can cause delays in
Ownership definition
the implementation, introducing risk. Having clear expectations of who takes care of the testing outcome
is key. Testing can result in bugs, but also it can result in discovering
configuration issues, gaps, or new opportunities where improvements
can be applied in future cycles or solution releases. Based on the type
of outcome, we need to define who takes ownership of the fix and
what test plan is needed to account for that.

▪ Bugs go back to developers for resolution.

▪ Configuration issues go to the consultant connected to the
test case.
▪ Gaps go to the project manager for further discussion with the
stakeholders and implementation team.
One common pattern for implementations
of standard functionality is the ▪ Standard product issues go to Microsoft support or any other
project team challenging the need to third-party solution provider connected to the issue.
thoroughly test when the application
has not been modified. The reality is ▪ Conceptual design issues go to the architects on the
that implementation teams are testing
implementation team.
the specific business solution being
implemented, not the standard software. ▪ Process improvements go to the business SMEs.
You are not just testing software, you
are testing the people being trained, the
processes that represent the operation, Documenting the outcome of the test cycle and assigning clear
the data that keeps the business alive, and
finally the technology. ownership for the fix facilitates tracking and resolution. Plan for
testing thoroughly, even if you are implementing standard product.

The right environment for the right type of test

Where to test is dependent on the type of test being executed and this
definition impacts the environment planning discussed in Chapter 9,
“Environment strategy.” Refer to this chapter for the different types of
environments available and their overall purpose. Our focus is on the
environments where you can do proper testing based on the test type.

This environment can be a different variation of a development or

test environment, but it is important that you never complete or
plan for regular testing in production environments.

Implementation Guide: Success by Design: Testing strategy 356

 Production environments are meant to run the real business operations.
Doing tests, or making changes to it without being tested first, is a high
risk for the stability of the solution because of the unknown variables
it can bring. The only test type that can be executed in a production
environment is a mock cutover test, and this test happens for a very
limited period before the solution is live.

You can require testing be done in an environment different than

Plan for non-Dynamics 365 testing Dynamics 365 environments, like in the case of integrations. Plan for
environments where you have a
the availability of the testing version of third-party systems that will
dependency in the solution that
requires validation. integratewith Dynamics 365; integrations or other dependencies
will need to be validated. This is a commonly missed opportunity
during implementation.

Plan for additional hardware required to emulate the real operation

like scanners, printers, handhelds, etc. These resource needs are dictated
by the test case.

Documentation
Documenting testing has two angles, the test cases and tracking the
outcome of the test.

Having a model to document the outcome of a test case and test cycle
Planning to document the results of allows the implementation team to validate the performance of the
your testing cycles helps to highlight
test cycle. Preparing to document and report the outcome highlights
the wins, but also the bugs and
patterns. This report determines other challenges as well, though not necessarily ones connected to the
the next course of action and can
impact future milestones. It allows
test cases itself. These issues can be solution performance, connectivity
stakeholders to make decisions to issues, usability, continuity of the transaction, gaps in the product, etc.
correct the paths if necessary.

Use Azure DevOps to build Other important benefits of documenting the outcome are that it
dashboards that can help to report
progress and quality of the testing. tracks the progress of the testing and keeps the big picture in sight.
It allows us to detect and build the backlog for new non-high-
priority opportunities that are discovered during the testing. It can
also trigger a change of the statement of work when the newly
discovered requirements are critical. This is highlighted as project risk.

Implementation Guide: Success by Design: Testing strategy 357

The different roles involved in testing
The people involved in testing depends on the type of testing being
done. For example, performance testing types require more technical
roles and consultants familiar with that type of testing, while user
acceptance testing types require your end users.

Testers can be developers, functional consultants, customer subject

matter experts, quality assurance testers, or end users.

It is important to plan for the resources required for building your

solution and for the different types of tests that require special skills.
Consider the volume of test cases involved and the business areas
impacted so you have good representation of the business.

Roles for testing are determined by who to involve in tests depending

on the following variables:
▪ Test preparation (system, data, test cases, training, etc.)
▪ Test execution
▪ Test administrator or manager
▪ Test results triage
▪ Test reporting
▪ Test defect fixes

The business team involvement during testing is critical, they own the
solution and plan for enough time to allow for proper testing by the
key business users. One common pattern of failure is poor involvement
from this group of testers.

Test types
In the previous section, we covered the different roles and environment
types that may be needed for testing, but we mentioned the need to
consider the test type. Let’s now combine the concepts in this section so
you can see how to use different test types to implement Dynamics 365
under the Success by Design framework.

In Figure 14-6, we show when we recommend executing the most

common test types when implementing Dynamics 365 apps across

358
 the different Success by Design implementation phases. Note that
some test types are more relevant to specific phase than others.

The minimal recommended test types are:

▪ Unit tests ▪ Performance tests
▪ Functional tests ▪ User acceptance tests
▪ Process tests ▪ Regression tests
▪ End-to-end tests ▪ Mock cutover

Each test represents an incremental testing approach. Some are

required for any implementation and others depend on the specific
Test types need to be planned with implementation circumstances based on the complexity of the
consideration to the complexity of the
solution, performance, and security requirements. Some test types
solution, performance, and security
requirements, and should be executed are especially relevant for specific stages in the implementation and
during specific times within the
implementation.
others can run for the life of the solution. For all cases you need to
test with real customer data and migrated data as soon as possible.

Fig.
14-6 Testing throughout the solution lifecycle

Initiate Implement Prepare Operate

Get ready to start Design Build Exchange-to- User acceptance Deploy Business as usual and
exchange testing continuous update

Unit test

Functional test
* Go live

Testing with customer data

Process test

End-to-end test

Performance test

Security test

Regression test

Interface test

Data validation test

User acceptance test

Mock go-live

Implementation Guide: Success by Design: Testing strategy 359

 Unit test
See it in action This test type focuses on individual function, code, and configuration
testing. It is done by developers and is the lowest level component of
Let us assume we are building a solution
for a customer that ships goods and the testing. In this test type, the developer verifies the requirements,
uses a specialized transportation system
validates, improves the code design, and finds and fixes defects.
that is part of the final solution. For
this example, we use unit testing to
test specific validations during a sales
order confirmation and interface with
This testing happens in a development environment and is expected
the transportation system using Power to be done during the implementation mainly, or during any bug
Automate. The unit tests are written and
executed for each validation requirement fixing. It is a fast, iterative process. This test type is required if you
and interface in this scenario.
are customizing or extending your solution and is one of the first
opportunities to introduce a quality check in the solution. At this
point you should validate performance behavior of the individual
components in addition to keeping security in scope. Being diligent
from the start with this test type can save you important time during
implementation by avoiding rework or bug fixes for issues that
should be detected during this test.

Functional test
See it in action
Functional tests can be either a manual test or an automated one. They
Following the previous example are done by the functional consultants, customer SMEs, or testers. The
regarding the sales order validation, the purpose of functional tests is to verify the configurations of the solution
functional testing of the interface with
the transportation system focuses on the or any custom code being released by the developers. The primary
configuration and setup, like customer
objective is to validate the design as per the requirements. This is
data, products, pricing, warehouse setup,
and the dependencies with payment done in a test or developer environment during the Implement
processing and other module-related
configuration and parameters.
phase of the Success by Design framework. At this point, testing
automation can also be introduced.

This test type is the first test to be done by consultants and customer
SMEs. Consultants still need to do it, as it is important that the first
line of testing is done by the consultant’s previous customer testing
at least to verify stability of the feature.

At this point the consultants need to map the test to the requirements
and processes and the test case is detailed. The link with process is
agreed upon with the business so the test case is not focused only on
the gap, but also to fit in the whole picture of the solution.

Implementation Guide: Success by Design: Testing strategy 360

 The data used to test beyond this point requires it to keep evolving
continuously using customer data and should reflect the reality of
the operation.

Process tests
The solution continues to be built, with unit testing being done by
developers and functional testing by consultant and customer SMEs.
The work from the development team is validated, and bugs and
corrections are completed while the test is mapped to the process.

This is the point at which running connected test cases is ideal. The
point where we raise the bar in our testing approach by connecting a
collection of test cases that are all connected to our process.

Process tests can be manual or automated at this point. The business

has visibility and is actively participating in this testing. It is our first
opportunity to go beyond testing that focuses on just functionality or
modules, and it gives different people the opportunity to start playing
an important role in the test by handing over the outcome from one
test case to another. This should be your first comprehensive testing
cycle, starting during the Implement phase of Success by Design.

The objective is to verify the solution allows us to operate business

scenarios, and the testing needs to be done in a test environment.

See it in action
Tracking the outcome becomes more critical at this point, so bugs
At this point we know how our sales
fixes and other dependent improvements must be tracked. During
order function works, but now we want this time, tracking tools like Azure DevOps become crucial as part of
to test how the sales order works in
connection with test cases preceding your application lifecycle management.
and following the process. Let us call
this process prospect to cash. We want
to test the collection of money after the For this test type, role security becomes crucial. We want to make sure we
shipment of goods happen. We involve
other teams in the implementation team
have the right segregation of duties and that the different personas can
that handles that part of the process, operate as they should when they are running the solution in production.
and we validate that we can generate the
invoice and collect money while closing
the cycle. From the business point of view, This testing should look for unexpected outcomes of the process,
the customer can confirm they can sell,
ship, invoice, and collect the money so the commonly known as negative testing, and not just the happy path.
process prospect to cash works.

Performance of the solution starts to become more evident with this

Implementation Guide: Success by Design: Testing strategy 361

 test type since more testers are involved, but also because it touches
more areas of the solution so the solution can start to enter into stress
mode when executing parallel functions. Nonfunctional requirements
need to be observed; for example, proper batch processing jobs are
properly configured.

End-to-end tests
After validating all the individual processes, it is time to connect all of
them and increase their complexity with new process variations. This is
the first test cycle that looks like a complete operation.

This test is manually executed, but automation starts becoming more

relevant in preparation of future iterations, regression testing, and
system maintenance because of continuous solution updates even
after go live.

The test is done by functional consultants who are preparing the cycle and
guiding the team, but the overall cycle is mainly executed by customer
SMEs and testers.

The main objective of this test type is to validate all full business processes
in scope and it needs to be done in an integrated test environment since
now it is connecting to other systems that interact with the solution. It is
See it in action iterative and the prerequisite to being able to execute your user
acceptance test (UAT).
On previous test cycles, we were able to
collect the cash of our sale. Now we want
to connect other processes dependent on
this one; for example, accounting process
This test type starts in the Implement phase and goes through the
to report taxes, update and interact with Prepare phase as per the Success by Design framework.
other systems like the transportation
system, optimize inventory, and
introduce new products. This allows us to It is important to execute as many of test cycles with end-to-end
combine different Dynamics 365 apps to
work together. testing, doing only one and at the end of the build of the solution is
not recommended since it add risks to confirming readiness by having
less time to react to final fixes.

Another important aspect here is that you run this test not just with
Plan for this test by making sure customer real data but with migrated data coming from the legacy
you include real customer data and
migrated data.
solutions as soon as possible.

Implementation Guide: Success by Design: Testing strategy 362

This test is key to validating the entire solution works in conjunction
with other systems that are part of the business, and testing is done by
having role-based access control enabled so it validates a real end-to-
end test.

Performance tests
Successful testing is not complete until we not only make sure we
can run the business on top of the solution, but also that we can do it
at scale. We are implementing Dynamics both for today and for the
future, and we need a solution that lasts and performs.

Performance testing is required, especially if there is concern over

operational volumes, peaks, variety of integrated systems, usability
challenges on the UX design, etc. In this section we focus on some of
the considerations you need to keep in mind for performance tests,
but we have dedicated an entire chapter just for this in Chapter 17, “A
performing solution, beyond infrastructure.”

We put special emphasis on this test type since there are many
misconceptions defining whether this test needs to be executed.
Our experience has shown that performance is one of the most
common reasons for escalation since teams often miss this test
when it is needed.

In the end, a performing solution is a combination of data, code,

configuration, and infrastructure. Microsoft, partners, and customers
play important roles for this test type, but we look to the implementation
teams to play their important role on validating performance beyond
infrastructure and at the right time of the implementation—so plan for it.

This test is executed by developers, functional consultants, customer

SMEs and testers mainly.

The objective of this test is to ensure the solution performs while focusing
on critical processes that require scaling with load and growth. Not all the
processes are involved in this test.

This test is completed in a dedicated performance testing environment.

363
 The basic version of performance testing starts during unit testing
See it in action so developers can influence proactive improvement of performance.
Regular dedicated test environments can also be used depending on
In previous example we involved
invoicing and the integration with the the load to be tested.
transportation system. Now we want
to test if we can process the operation
peaks considering the seasonality of the Performance testing happens as soon as the critical processes are
business. Invoice posting operation is
crucial since we need to have almost available to be tested throughout the implementation.
real-time interaction with the
transportation system as per the business
needs, so we need to test a day in the life Performance testing needs to be visible at the start of the imple-
with relative volumes across key business
mentation and be part of the test plan. It requires agreement on the
processes, not just prospect to cash.
performance test objectives, and it can require proper environment
planning to provide the necessary resources.

Do not delay performance testing till the end of the implementation

or avoid the review of the need for performance testing. This test can
bring to surface important challenges that can require fundamental
architectural fixes, so doing it late in the implementation puts the
entire solution at risk of wasting important resources.

It is important to mention that the infrastructure required to execute this

test type, especially when the plan requires a higher spec environment,
does not need to be available during all the implementation. You bring
them a needed environment or repurpose ones for other test types
during the lifetime of the project. This does not mean you should not log
performance bugs observed during other test types at any time.

Remember, you are not just testing the standard product in this
test type; you are testing the solution that you are implementing in
User acceptance testing (UAT) can be combination with your business operation needs. Some very basic im-
a great opportunity to start building
automated testing for recording the tests
plementations can justify not executing a performance test, but before
so you can have repeatability on testing. deciding you can learn more about what you need to have a performing
This sets the path for regression testing,
optimizing the investments to build solution in Chapter 17, “A performing solution, beyond infrastructure.”
that automation.

UAT is the last comprehensive test before User acceptance tests

going live, and every test case emulating
We are now getting ready for prime time. Our solution works in unison,
real operation is executed so end users can
record their test as a proof of the test case it performs, and it is ready for final validation by the actual people that
validation, therefore making this the first
step toward automation.
execute the business, our end users. User acceptance tests are the final
calls for the business to declare readiness for go live.

Implementation Guide: Success by Design: Testing strategy 364

 This test is manually executed—never automated. It is executed with
customer data, including migrated data, and with the latest solution
version. This test is the closest to being like running live operations, it is
an actual business operation simulation.

The objective is obtaining business sign-off of the solution, collecting

end user feedback, helping to manage organizational change, and it is
done in a dedicated and integrated test environment.

UAT is done during the Prepare phase of the Success by Design

framework and it happens prior to preparing for go live, which is a
prerequisite to reach this milestone. For more details, see Chapter 18,
“Prepare for go live.”

See it in action This test type needs to be executed by users from the business, the
implementation team is just a facilitator. The business team is the main
The business team now brings in a select
group of people who run the operation. owner of this test, failing to fully participate or test thoroughly is a risk.
This group consists of order processors,
the accounting team, the accounts
receivable team, shippers, and others. The The business users must be trained prior to UAT, not just on the
selected group of people run a real-life
operation simulation. All the processes are solution but also on how the pending process works once the solution
executed in parallel and in-sync between
is deployed to production; otherwise this group causes false error
teams. For the prospect to cash process,
they involve a selection of different end reports because the lack of training.
users connected to this process to run the
test. The team tests all the variations of the
processes in scope. At the end of the successful test, the end user connects the new
solution to its reality of their daily tasks and confirms the readiness of
the solution, but also their own readiness at being familiar with the
new system after being trained. The new solution fulfills the business
operation need.

You are required to plan for the

The final iteration of this test type requires the business sign off on
UAT milestone and to have business acceptance. We describe the importance of this step in later sections in
stakeholders sign off on the acceptance of
the complete, satisfactory results of this this chapter.
test prior to final preparations to operate
(go live).
This is one of the most important milestones in terms of testing types.
This test happens at the end of the implementation and prior to the go
live. It is the entry criteria to prepare for operations. UAT is an end-to-
end testing approach where all the processes in scope are tested and it
requires sign off on the acceptance by the business stakeholders.

Implementation Guide: Success by Design: Testing strategy 365

 Regression tests
Now we need to look at how to scale our future testing cycles, but also
to make sure prior positive test results continue as you keep evolving
the build for the solution.

Regression testing is testing that requires you to repeat previously

executed tests, due to changes or updates, to the solution to ensure it
stays healthy. Keep in mind that the solution is dynamic, new features
are introduced or new processes are required. Running regression tests
manually works of course, but it is also important that you consider the
need to prepare to automate this type of testing as soon as possible. It
is an investment that pays off with time.

The objective is to ensure the solution still performs as expected after

a change and that it does so in a pre-production environment, test
environment, or development environment.

A regression test happens when you have change in the code, or any
configuration or new solution pattern that can impact different processes.
See it in action
This test type is done by testers, developers, and end users.
Microsoft has released new functionality
that enriches the order processing feature It is important to perform this test type before change is introduced to
in Dynamics 365 Finance and Operations.
When the new update becomes available, the production environment.
teams can execute regression testing
to key business processes connected to
this new feature. Testing the order and As we realize that this necessary quality gate is needed, there are tools
warehouse processes following a new
configuration change is done because it
available to help you to automate. You can find links to more information
can impact the picking process. Once the about these tools at the end of the chapter in the “References” section.
update is done in the test environment,
the team runs an automated test to
confirm the solution produces the There are different techniques you can follow to run your regression test.
expected outcomes.

You can opt to test almost 100 percent of the processes, which can
provide you comprehensive coverage, but it is expensive to run and
maintain especially if there is no automation.

You can prioritize the test based on business impact. This technique
ensures you have coverage of the processes that are mission critical for
the operation, it is also a more affordable approach and the only risk is
that you cannot guarantee a solution 100 percent free of regression.

Implementation Guide: Success by Design: Testing strategy 366

Another technique is to focus on the areas that you expect to be
impacted based on the change. This technique is targeted where the
change is happening and requires less effort, but it has the limitation
of not being able to confirm that you are free of regressions and
the impact cannot be visible on the direct change itself, but instead
downstream on other processes.

You can also combine all the previous techniques, making sure you
test critical business processes, and you can target more testing on
the features being changed.

Always keep in mind that bugs discovered late in the implementation

process have a lifecycle to get fixed but also require testing. Testing allows
you to detect those bugs early, and if the bugs are detected late in the
process, let’s say in the last test cycle, to think twice on the value of the fix
versus the risk it introduces with not being able to do proper regression
testing. The fix can be more costly at that point, being so close to go live,
than waiting and doing a proper regression testing or delaying the go live.

Every time you test successfully, you put money in a trust. Every time
you have a failed test or fail to properly test, you deduct money from
that trust leaving a debt in the solution completeness. At the end,
you need to reach the end goal with the trust full of money.

Again, automation is key, and you should plan for testing automation.
Your solution is alive and ever evolving so change is a constant, not just
coming from the application, but from the customer business needs.

Start building your automated testing progressively, focusing first

on key business processes then expanding further over time. Do it as
early as possible during the implementation process and always test
it after a change and prior to production deployment.

Regression is important but it can be costly, especially if you do not au-

tomate in the long term. Automation brings the benefit of testing faster,
having better testing coverage, and providing repeatability of the test.
Automation takes as much planning as was needed for the manual testing
planning, but it is a long-term investment that pays itself off with time.

367
 Finally, keep in mind that finding bugs during regression testing could
be due the change of the solution design, a standard product update,
or just the resolution of previous bugs redefining how the solution
works, which can require recreating the test case.

Mock cutover
This is a special test since it is executed in a special environment, the
production environment. This type of test is especially important when
you test Dynamics 365 for Finance and Operations apps.

This is a temporary use of the production environment to do testing

activities and it happens during the Cutover phase. To learn more about
the cutover activities refer to Chapter 18, “Prepare for go live.”

This test brings important value because it helps to validate aspects like
connectivity, stability of the access by users and data, integration end
points configuration, device connectivity, security, network, and many
of the test cases in your processes may be environment dependent.

During this test, you are able to validate confirmation of the

estimated times between environment preparation for production
for all the planned tasks during go live. It is a confirmation that all
See it in action
the planned activities run smoothly once you do the actual
Continuing with the previous examples
final execution.
in this chapter, the team has tested
the entire solution and they are ready
to move forward. The team wants to Once the testing is completed you rollback your data up to the
confirm the cutover plan in terms of
point that there are not transactions being created because of the
sequence of activities, timing, and
production environment stability. They testing, allowing you to finish the environment preparation and
execute the cutover plan by rehearsing
and documenting so they can adjust
start running your system for real. If the mock cutover fails, there is
it. After executing the mock cutover, a risk of delaying the go live.
they found the time to load data and
sequence it required adjustment due to
other conflicting cutover tasks. The team We recommend you always plan for a mock cutover test type.
adjusted that sequence and confirmed the
readiness to execute the updated Cutover
plan knowing it works without surprises.
Other types of testing
We have described the incremental approach for testing through
the different test types, scale needs, and continuous validation of
the solution. But there are other types of tests that can be unique
for every implementation based on the design of the solution.

Implementation Guide: Success by Design: Testing strategy 368

 These test types are below.
▪ Smoke tests  Where you test if basic functions work as expected
so you can confirm the component can continue to test further with
other test types.
▪ Data acceptance tests  You validate if the migrated data goes into
the target environment correctly and if it is usable for actual operation.
▪ Interface tests  Here you confirm if your interfaces can be utilized
as intended.
▪ Environment testing  In this test type, you validate if the new
environment is fit for the purpose.
▪ Reporting testing and business intelligence testing  Validates
that reporting and Business Intelligence can be executed and operate
properly with the expected outcome and performance.
▪ Device testing  Focused on validating devices are operational,
connect, and perform as expected, for example, RFID readers, scales,
and warehouse devices, etc.
▪ Network/infrastructure testing  Validation of relevant
underlying networking, firewalls configuration, Wi-Fi, printers, etc.

While some projects are too small to justify separate planning to execute
some of these types of testing, the concept and meaning of the tests
should be folded into other types of testing.

Test types and outcome ownership

We described the importance of defining the owner of the outcome at
testing, now we need to connect this ownership role to the test type.

When you define your test plan,

In other words, who takes accountability to drive the resolution of the
define ownership for the type of fixes
but also who is accountable to drive outcome derived from testing based on the test type. It is important
the fix overall by the test type.
we have clear ownership to avoid defects bouncing around with
unclear roles driving the actions, even though the fix can come from
different roles. As an example, unit testing outcomes tend to be owned
by partner technical architects or dev leads. User acceptance testing
outcomes tend to be owned by customer steering groups devolved to
lead SME or PM.

Implementation Guide: Success by Design: Testing strategy 369

Test types and test plans
Test plans are living documents that keep evolving since project needs can
change in terms of adding new test cycles or test types. This is due to
business needs changing or the need to increase the quality due to
unknown factors at the start of the project. For the most important
test types, like user acceptance testing or performance testing, you
need to create specific test plans or keep enhancing your master test
plan while always bringing enough details to provide clarity for each
of them.

The bottom line on

defining a test strategy
Defining a test strategy for your implementation is a combination of
considering the project scope, testing planning with a test plan and test
cycle schedule in combination with the phases and solution versions
availability, and selecting all the test types and how you execute them.

Testing is iterative and incremental; it grows as you progress in the

implementation and it is a permanent activity once you are live based
on the regression testing technique you select.

Always test under the umbrella of the processes, in the end the processes
are the language of the business and the testing is proof that the solution
“can speak” that language.

Executing testing
In a previous section, we focused on the importance of defining a testing
strategy. Now we are going to explore the minimum components you
need to keep in mind to execute the testing.

The prerequisites required to move to execution are having a test

plan, test cases, and clear testing cycles that are defined according to
the progressive readiness of the solution.

370
 Now we focus on the tactical components for prepping the execution,
what to keep in mind to communicate the scope, and how to track the
progress and outcome.
Part of a successful test cycle is to set the

Communicating the plan for the test cycle

expectations with the participants, so
everyone keeps the focus on the objective.
Consider the test cycle like a mini go live
for the scope in place for the cycle.
Aligning the teams prior to testing execution is important so that
everybody stays focused on the expected outcome of the test cycle
being executed. This is especially important when you test with
several people involved, like process testing or functional testing.

During the communication effort you share the test plan for the test cycle.
You need to communicate the following with the team during testing.
▪ Scope  Before you start testing, you describe the scope and purpose
of the testing cycle, and what processes, requirements, and tests
cases are included in that test cycle. Every new testing cycle requires
detail on how the iteration has been designed in terms of incremental
testing and what is expected to be tested. The scope of the test cycle
is aligned to the solution version being used.
▪ Schedule  The time expected to be used to run the test cycle.
▪ Roles  Who is testing and how are the test cases distributed?
How do they report test case execution so dependent teams can
continue the testing? Who is the orchestrator of the test? Who
resolves questions to test cases per area?
▪ Resolution process  One of the objectives to test is to identify
any defect in the solution. The communication plan needs to
specify how those bugs are reported but also how to document
the feedback from the tester.
▪ Progress  How is the progress recorded and communicated so
everybody can see the current state of the testing cycle?
▪ Resources  The communication needs to specify where the
testing happens and how to access the apps. It determines any
additional equipment needed for testing, for example printers, bar
scanners, network requirement, etc.
▪ Test sequence  Especially on process test, end-to-end test, and
user acceptance test types. You need define and align how the
different teams interacts.
▪ Test objectives  Here you explain the purpose of the testing cycle.

Implementation Guide: Success by Design: Testing strategy 371

 Tracking during testing
Detecting bugs and properly documenting them is key. It facilitates the
resolution. But before reporting bugs, the tester needs to understand
the scope and objective of what the test is. Building the solution is pro-
gressive and it is easy to report issues where the scope of the test does
not include a solution for that and is part of future iterations. Keep the
focus with the team testing.

Tracking captures the outcome and performance of the test cycle; it is

how the development team and consultants understand the quality of
the test cycle. Azure DevOps is a recommended tool for this tracking
since it integrates into the entire application lifecycle management
and any bug reported is visible to the team that needs to fix it, but also
shows what is or is not working in general.

Let us explore some tactical components of the tracking of issues

discovered during testing.
Making configuration changes can
invalidate previous test cases connected
to those changes. Those changes need
Keep track of changes in configuration
to be handled much like it is a production
change to make sure those changes do It is likely some tests result in tuning configurations as you test.
not negatively impact previous passes
test. Track and execute regression testing.
Adjustments to the solution in configurations can change the behavior
of the app, and those changes can happen during tests. This brings
flexibility into the resolution, but also raises the risk of the team not
documenting those changes. Keep in mind that for the next test cycle
the data can be reset and if you do not document and make sure those
changes are not reflected in your golden configurations, you lose the
important opportunity of what you fixed during testing. In the end,
you do not want to repeat the same research over and over, so docu-
ment the configuration update.

Keep record of the progress

At every test cycle you come with a scope. Depending on the type of
test you have some dependencies. Teams need to know when they
should start testing each type of testing. A testing heatmap can help
you to show the progress of testing by process, test cases, and its
statuses. For example, in the scenario used to explain the test types in
previous sections we used the analogy of testing a prospect to cash.

372
 A meaningful process test type makes sure the process being tested
has continuity, from the order to the invoice and shipment and later to
the collection. If you are testing a sequence of processes, you do not
go ahead of the game and straight to the collections part. If you do,
you lose the opportunity to properly test the larger process sequence.

Keep record of the bugs

You know how to report the bugs and that the reports need to be ac-
curate. Make sure you can report the issue, but also that the case being
tested is in scope for that test cycle. Not doing that creates rework in
the investigation for resolution. Also, be factual. Is the test passing or
not based on the test case documentation and the expected outcome
when following the exact steps?

Bugs versus gaps

One common issue during bug reporting is that we can mix a bug with
a required feature or gap. Stick to the test case, if you identify a gap,
great! Just make sure the feedback is properly classified. We do not
want to lose the opportunity to capture that great idea or finding, we
just need to funnel it properly to the right tracking system.

Dealing with the outcomes

Finishing the test cycle requires reporting the results and maintain-
ing the bugs discovered during the testing, but also funneling the
feedback received. In the end, all these outcomes add onto the top of
the next wave of processes or features to be introduced in future test
cycles, as shown in Figure 14-7.

Reporting the outcome measures the performance of the test cycle

Fig.
14-7 and helps to identify areas of
opportunity for the implementation
% Failed team since it triggers clear
Performance of Scope
actions to improve the quality,
the test cycle so the next test cycle provides
% Passed
better performance.

Implementation Guide: Success by Design: Testing strategy 373

 For the next test cycle, we start over, as even passing test cases during
this cycle need to be re-tested in the next one.

Solution acceptance
and operation
Finally, the last test cycle takes the main objective of why you started
implementing the solution to confirm the readiness to run the business
using the new solution.

We described the importance of reporting the outcome of the test

cycles for user acceptance test types, this report helps to determine
the readiness to prepare for go live and is the foundation of the go
or no-go decision by the stakeholders. The performance should be
high at this point, and if not? Then the solution is not ready.

This confirms the readiness of the solution but also the readiness of the
production environment if you run a mock cutover test. It is important
that for any final test the business team signs off and confirms they can
operate the business with the new solution.

Accepting the solution does not mean it is 100 percent free of bugs.
You need to assess the value to bring a fix, at this point if the issue
is very low risk for the operation it is often better to go live with
known nonblocking bugs and fix them later than to introduce risk
Sign off by the business team on the
by making an unnecessary rushed fix and not having time to
overall pass of the final test cycle is a re-test properly.
required step prior to executing cutover
preparation for go live. This creates
accountability on the acceptance of the
From here you move to the maintenance mode of the solution if
solution by business. Any non-blocker bug
needs to be document and they should be you are implementing only one phase, or you continue adding new
low risk for the operation at not bringing a
solution at going live without them.
workloads or expanding the solution. Keep the discipline for testing
and scale using the tools for automation. Testing is a continuous
practice, the difference is the frequency, scope, and tools used
to test.

Implementation Guide: Success by Design: Testing strategy 374

 Conclusion
Testing is about quality, it is used to confirm the effectiveness of
the path taken by the implementation team to solve the challenge
of how to operate the business with the new solution, it confirms
the initial vision by putting it into action. It is a required step in the
process to implement Dynamics 365 apps, regardless of how simple
or complex your solution is, or if you are implementing standard
functionality or extending or customizing it. If you run into a large
number of quality related issues after you are live, it is generally
attributed to ineffective testing.

Testing during implementation is the step that builds the trust for
the business to run their operation. Always test and do so as early as
possible. It is never too early to start testing.

References
Finance and Operations Customer Engagement
Test Strategy TechTalk FastTrack automated testing in a day workshop offering

Performance benchmark EasyRepro – Getting started

Performance patterns and antipatterns EasyRepro in 60 Seconds

Performance troubleshooting tools Automated and manual testing with Azure Test plan in
Azure DevOps
Performance testing approach

Regression suite automation tool

Implementation Guide: Success by Design: Testing strategy 375

 Checklist

Planning Carry out performance testing throughout the

Implement and Prepare phases on critical processes to
Plan the test case creation, forecast the resource re- ensure the solution performs while focusing on scaling
quirements (people, hardware, environments, devices), with realistic load and growth.
and implement the test cycles in accordance with the
Throughout the Implement and Prepare phases,
processes in scope.
perform iterative and cumulative end-to-end testing
Have a tracking mechanism that monitors the overall of all full business processes in scope and connected
progress, pass/fail rates, ownership, and issues tied integrations and solutions in a test environment with
to specific test cases, and have a process defined for increasingly “real” migrated data.
remediation, retesting, and deployment when errors are
Perform UAT in a test environment during the Prepare
found and corrected.
phase to test the preparedness of users, processes, and
the solution to operate the business.
Implementation
Consider automation for functional, process, end-to-
Ensure the SDLC includes unit testing by the end, performance, and regression testing.
developer in a development environment, with focus on
Establish a test communication plan with clear scope
function, code, and configuration of all extensions
and objectives, specifying the schedule, sequence, roles
and integrations.
involved, and issue resolution process.
Ensure the functional testing is performed by functional
Perform regression testing in a test environment, or
consultants or SMEs in a test or developer environment
development environment for updates, throughout the
with the primary objective to validate the design against
Implement and Prepare phases as well as post-go live to
the requirements.
ensure the solution performs as expected after a change.
Ensure process testing is performed by SMEs in a test
environment, running multiple processes consecutively,
Go live
focusing on testing whether the solution allows business
operations and monitoring for unintended outcomes. Plan a mock cutover in the new production environment
to validate confirmation of the estimated times between
environment preparation for production for all the
required tasks during go live.

Implementation Guide: Success by Design: Testing strategy 376

 Case study
The right environment
for the right test: A
missed opportunity
A customer in the manufacturing industry that manufactures
recreational vehicles is implementing Dynamics 365 for Finance and
Operations and has created a testing strategy that included four testing
cycles, including user acceptance testing, as part of their plan.

The team included the most common test types except for performance
testing under the assumption that the first rollout would not require it
since the extension of the system was low, and the customer will be imple-
menting just manufacturing business unit. Next rollouts will include the
servicing business, including the warranty management of recreational
vehicles plus other operations. The servicing operations will continue to be
executed as a third-party solution for now and their dealership network
will be using this system to order parts and honor warranties.

Very soon, during the first testing cycle, the team started to test across
the first wave of the solution where there were no integrations, so
completing testing was satisfactory in terms of the cycle performance.

As the team got ready for their second wave of testing, they started
to introduce some of the integrations as part of the scope for the test
cycle, but those integrations were emulated with dummy data and the
team introduced some migrated data. Integrations worked for their
purpose and the volume of migrated data was small.

Implementation Guide: Success by Design: Testing strategy 377

 For their third testing cycle, things looked great and they were ready
to do a more comprehensive testing by using end-to-end testing, this
time with real integrations and real volumes of data migration.

At combining the testing for having a larger volume of users for the
test and integrations running with similar volume of operations to what
they expected in production, the team hit the first major blocker. The
solution was not keeping up with the needs of the test cycle, and they
were not even in production. The first reaction from the team was an
underperforming infrastructure, which raised the concerns of the business
stakeholders upon learning the test cycle outcome.

The team was ready to prepare for UAT and decided to continue ex-
pecting that this would not be an issue in production due to it having
higher specs. They assumed the production environment would be
able to solve this performance challenge, so they decided to continue,
complete UAT, and move to production. The customer signed off and
preparation moved to the next stage to get ready for go live.

The big day came and production and all the departments switched to
new system. The first day was normal and everything was working great.
The team decided to turn on integrations for the servicing solution on
the second day. When the second day came, they were ready to go in-
terconnect with the service department, and integrations started to flow
into Dynamics 365. This is when they had their first business stopper: they
had a sudden decrease in performance, users were not able to transact in
Dynamics 365, service departments from the dealerships were not able
to connect effectively, shipments started to slowdown, and the shop floor
was having a hard time to trying to keep inventory moving to production.

The implementation team was providing hyper-focused care and

immediately started to check for what could be blocking the operation
and they found the problem. The integration was creating a bottleneck
by making important OData calls to check inventory, creating sales or-
ders, and invoicing while connected to the dealership’s partners using
the servicing solution. They stopped the integration and things went
back to normal, but it was clear that the integration was the cause.

Implementation Guide: Success by Design: Testing strategy 378

 The team found a solution after troubleshooting, but it will require
a change to the integration design pattern and fast movement to
solve it and keep operating. Part of the revenue from the recreational
vehicle business comes from servicing and providing parts to the
dealerships. It took days of manual work to keep things working
during this painful first week of go live, which impacted the
perception of the organization in the solution.

In retrospect, validating designs for integrations with a volume of data

was key to preventing this scenario and the issue they had during real
operation. The team’s morale was affected after working hard during the
implementation season, all because they missed one important test—
the performance test. The team thought emulating early and validating
in production would be OK, but the cost of that decision was high.

This team learned that including performance testing as early as possible,

with actual volumes in the right environments that perform at the same
volumes expected in production, would help to detect a design problem
and correct it in time. The team also learned that “testing” in production
for the first time, and with the real volumes, was not the best idea since the
time available to make any fixes was reduced due to the live operations
and correcting as they go introduced bigger risks. The assumption of the
team that they were implementing almost standard functionality led them
to think that the solution would perform fine, but the actual customer
challenge made the solution unique which requires extensive testing.

For the next phases, this customer included performance testing. They
dedicated an environment to stress test the solution with their own
business needs and they executed these tests earlier with additional
scenarios and UAT that included parallel processing of people testing
and integrations running. They were able to have second go live
to include their accessories manufacturing business and it was a
breeze in comparison.

It is never too early to start testing and do so according to the project

needs. The later the testing, the bigger the risk.

Implementation Guide: Success by Design: Testing strategy 379

15
Implementation Guide: Success by Design: Extend your solution
Guide
Extend your
solution
380
The power of making it your own.

Introduction
Business solutions offer a rich set of capabilities to
help drive business value.
Still, in some situations, you need to extend the solution and adjust off-
the-shelf functionality to accommodate organization or industry specific
business processes. These adjustments can change how a feature works or
bring additional capabilities to meet specific requirements.

While business solutions natively offer rich capabilities, they also offer
powerful options to customize and extend them. Extending the solution
can open even more opportunities to drive business value. It is important
to note, however, that extending should not compromise the fundamental
advantages of an evergreen cloud solution, such as usability, accessibility,
performance, security, and continuous updates. These are key to success
Complex business requirements lead to
highly advanced solutions with customi-
and adoption.
zations and extensions to applications.

Defining your extensibility

Advanced implementations bring an
increased risk that user experience
suffers because of the introduction of

strategy
performance, stability, maintainability,
supportability, and other issues.

In this chapter, we delve into key factors

to consider when determining the scope
of depth of solutions. When organizations shift toward software as a service (SaaS) cloud
solutions, one of the main reasons they do so is to enjoy the many

Implementation Guide: Success by Design: Extend your solution 381

 advantages, including out-of-the-box security, native integrations with
other cloud solutions, cost savings, and improved agility. The evergreen
cloud approach also enables the continued ability to evolve with the latest
modern features. In addition, it is often easier to use off-the-shelf solutions
with embedded modern processes and further extend it. This tailors a
solution to meet business needs, rather than having to build a custom
solution from scratch with all the associated hassles and expense to meet
the precise requirements of your organization.

What is extending?
When organizations implement a solution, there typically is some degree
of customization and/or extensibility, which we refer to as extending.
Extending can vary from minor changes to the user interface of a particular
feature to more complex scenarios like adding heavy calculations after
certain events. The depth of these extensions has important implications
on how much the off-the-shelf product needs to change to meet specific
business requirements.

We can also categorize these extensions into distinct types:

▪ Extending by modifying off-the-shelf functionalities.
▪ Extending by introducing new functionalities on top of a native one.
▪ Creating custom solutions on the platform.

Determine the need

It is important for organizations to be aware of the impact of extending
the solution. They want to strike the right balance between the features
required to empower a user or bring business value, as well as efficiency
and the value it brings compared to what it costs.

A key driver in the decision to extend is understanding the potential risks

of extension to key solution characteristics, like its robustness, reliability,
performance, usability, security, and ability to take updates. Consider the
example of extending an invoicing system that could pose risks to the
operations or could disrupt the business because of increased complexity;
such risks are best avoided. When opting to extend the solution,
organizations may also depend on a third-party to maintain and evolve it,

Implementation Guide: Success by Design: Extend your solution 382

and in some scenarios block the organization from adopting
new functionalities.

Extending to integrate with other solutions is a common scenario that we

cover in Chapter 16, “Integrate with other solutions.”

Do not replicate your legacy solution

When gathering solution requirements, organizations may want to
reproduce every piece of functionality from the legacy solution or even
try to mimic the user experience. This could come across as a strategy to
reduce user resistance and drive adoption, but it can also lead to a highly
customized solution that fails to leverage the strengths of the new
platform in favor of familiarity.

Legacy solutions may have taken years to develop and evolve and may
not use the latest and greatest functionality available on the market. As an
example, Dynamics 365 natively uses artificial intelligence and machine
learning to provide insights that help users make the best and most
informed decisions.

Major deviations from the native features or repurposing existing features

can limit the value of an evergreen platform by compromising flexibility
and the ability to evolve with changing business needs. That increases
technical debt and maintenance costs.

Understand the power of the platform

Another challenge arises when an organization does not fully know the
power of the new solution and opt to extend before even understanding
the impact of their decisions. The platform may have numerous ways to
achieve the same functionality as the legacy solution already. Reasonable
customization of the off-the-shelf solution might be required, but the
point is to find the right balance. Thus, each scenario should be evaluated,
explored, and in some cases piloted. And of course it’s important to
involve key business users in these decisions.

383
 Leveraging ISV solutions
Leveraging independent software vendor (ISV) solutions from the app
marketplace instead of extending the solution to achieve the same
results may save development cost and time, as well as testing and
maintenance resources. ISVs typically support and maintain the solution
at scale for multiple organizations. Their experience can be an advantage
for organizations that require additional functionalities that are already
provided by ISVs.

When deciding to use ISV solutions, organizations should be aware of the

ISV support model and its alignment with continuous updates as well as
protection clauses.

Extensibility scenarios
Extending off-the-shelf solutions occurs when functionality is changed to
fulfill an organization’s requirements.

Customizations and extensions performed to out-of-the-box applications

could stretch from simple setting and configuration updates to custom-
ized platform as a service (PaaS) solutions that extend their functionality.
In this section, we explore levels in which customization is possible and
discuss their relevance.

Some organizations prefer to stay as standard as possible. That does

not mean that they cannot make simple changes with configurations or
low-code/no-code customizations using modern tools. Staying standard
allows for slight changes, as long as those changes do not highly impact
key characteristics of a solution.

App configurations
Configurations are the out-of-the-box controls that allow makers and
admins to tailor the app to the needs of a user. These setting changes are
low effort, requiring no support from professional developers. They are a
powerful way to make the application your own, for example changing a

Implementation Guide: Success by Design: Extend your solution 384

theme to reflect business branding.

In some cases, additional due diligence might be required for

configurations, even though they are tested and shipped as part of the
core service. Examples include security model configuration, out-of-the-
box integration setup, and enabling embedded intelligence. Also consider
new capabilities delivered via the service updates that might be controlled
using the app settings.

App settings are the safest and the least impactful way of tailoring the
solution to your needs and should be the preferred approach before
exploring another extensibility technique.

Low-code/no-code customizations
A differentiator for Dynamics 365 and the latest generation SaaS products
is the powerful customization capabilities made available through “what
you see is what you get” (WYSIWYG) designers and descriptive expression
based languages. This paradigm helps significantly reduce the
implementation effort and enables businesses to get more involved
with design and configuration.

This low-code/no-code approach also guides makers to take advantage

of the optimized patterns and best practices, avoiding the potential pitfalls
and antipatterns that are observed in a complex code. Because the platform
takes care of lower-level details and automatically optimizes the logic,
testing is focused on functional components. The impact on performance,
usage pattern and service boundaries, however, still needs to be taken
into consideration.

Typically, customizations result in user interface modifications and changes

to the field settings, security model, business processes, visualization
components, etc. Changes implemented through customizations are
performed to ensure data quality, reduce repetitive tasks, and adjust
visualization components that determine how information is organized
and displayed to the users.

With customizations, applications can be tailored to more closely fit

385
 organization requirements that are specific to an industry or unique
business processes, including specific functionality focused on specific
roles or personas. This allows personalization that streamlines the user
experience so a user can focus on what is most important.

Low-code and no-code customizations are the preferred extensibility

approach when the requirements cannot be satisfied by app
configuration changes.

Extend using pro dev

Over the years the low-code/no-code base configuration capabilities
have advanced, reducing the dependency on professional developers for
application changes. Edge cases or scenarios that cannot be achieved using
the configuration-based approach can still exist. In that case, the powerful
custom code execution capabilities of Dynamics 365 can be leveraged.

Dynamics 365 apps support extension programming models that enable

organizations to leverage existing expertise within the organization
for advanced scenarios. These models, depending on the specific app
and functionality extended, generally supports .NET based languages,
Software Development Kits, and open-source libraries around JavaScript
and HTML5.

The key aspect to keep in mind when using code-based customization

is to understand the extensibility framework of Dynamics 365 and only
use the documented extension patterns. Use of unsupported techniques
breaches the service level agreements and can have a potentially severe
impact on the live service.

It is important that your developer teams have the latest software

development toolkits available to them. They should also know about
community based third-party tools and samples. These can drive productivity
even though Microsoft may not provide direct support for them.

Extending into PaaS

In some scenarios, organizations leverage PaaS components to extend

Implementation Guide: Success by Design: Extend your solution 386

 solutions, which adds powerful capabilities that help address complex
requirements. Dynamics 365 has a core design philosophy that allows
our SaaS applications to be extended by leveraging the underlying Azure
PaaS capabilities. This is referred to as the no-cliffs extension approach.
The approach enables businesses to start with SaaS, and then for the most
complex long-tail scenarios, extend into the Azure PaaS. Doing so
alleviates the fear of being limited by the platform.

This no-cliffs extension provides the best of both worlds. The SaaS
application provides the off-the-shelf functionalities as well as the
options and methods to extend them. The PaaS extensions further
enrich the solution architecture by providing rich and powerful
mechanisms that scale and allow heavy processing of operations outside
of the business solution.

An example of how this approach is natively used is the Internet of

Things (IoT) Intelligence add-in for Microsoft Dynamics 365 Supply Chain
Management. This add-in integrates IoT signals with data in Supply Chain
Management to produce actionable insights, as illustrated in Figure 15-1.

The same happens with Connected Field Service. Connected Field

Service for IoT Central provides a direct integration of Dynamics 365
Fig 21-3 Field Service with Microsoft Azure IoT Central, an add-on solution
Azure
Fig. IoT / Supply chain management
15-1

Dynamics 365 Dynamics 365

Supply Chain
Management
Azure IoT IoT Intelligence Add-in
Central
Workspaces

Machine Product Production

status quality delays Dashboards

Connected Data
technology ingestion
Alerts

Analytics Insights Actions

Automation

Implementation Guide: Success by Design: Extend your solution 387

 Fig 21-3
Connected field service that brings in Azure IoT PaaS on top of Dynamics 365 Field Service, as
Fig.
15-2
shown in Figure 15-2.

In both examples, organizations can leverage PaaS to further extend the

Azure IoT solution. The same can be applied to specific organization requirements
Central
that can use PaaS features as part of the extension’s architecture. This
approach can be particularly valuable because it reduces the maintenance
requirements for parts of the solution compared to a fully custom-
built extension.

Power Automate
One example of this approach is when organizations leverage Azure
Logic Apps. Logic Apps provide a serverless engine to build automated
workflows to integrate apps and data between cloud services and
on-premises systems. Logic Apps provide the ability to trigger
workflows based on events or timers and leverage connectors to inte-
grate applications and facilitate business-to-business (B2B)
communication. Logic Apps are integrated seamlessly with Azure
Dynamics 365
Field Service Functions, as illustrated in Figure 15-3.

Azure Functions is an event-driven serverless compute platform that can

Fig.
15-3 also solve complex orchestration problems, by using it you can move
some of the heavy computing
Structured query processes away from Dynamics 365.
Azure functions
language (cloud)

Service bus Office 365

With Azure Functions
Logic applications organizations can code their own
Cognitive Services Microsoft workflow definition and simplify
Dynamics 365
complex, stateful coordination
Machine learning DocuSign requirements programmatically in
event-driven applications, as well as
On-premises data gateway
taking advantage of numerous
native connectors.

A multitude of examples
demonstrate where PaaS can
be leveraged to further extend
SharePoint Structured query IBM DB2 BizTalk server the capabilities of the solution.
language
Leveraging a platform helps

Implementation Guide: Success by Design: Extend your solution 388

 eliminate the costs and complexity of configuring and managing the
required infrastructure, while empowering development teams to focus
on building the apps and services that drive value.

Considerations
Every piece of an extension should focus on bringing efficiency or
value to the organization. Inherent costs do exist when implementing
changes and can vary from the cost of building, testing, maintaining,
and support. These should be taken into consideration when planning
to extend a solution.

In this section, we delve into the key considerations and impacts that
extensions can have on a solution.

Usability and accessibility

User experience is a key characteristic of a business solution. Providers
invest heavily in the user experience to ensure that the users have a
pleasant experience when interacting with the solution, and that the
solution operates seamlessly between devices, such as laptops, tablets,
phones, and platforms like Outlook and Microsoft Teams.

We earlier discussed that the purpose of extending off-the-shelf

functionalities should be to add further value to the business. One
way to achieve this is to improve user experience. This is achieved by
considering and adapting user requirements and expectations.

While the main purpose of extending may not be to improve the user
experience, it should not negatively impact the user experience and
how the solution behaves across different devices and platforms. In
addition, extending should not negatively impact the responsiveness
and performance of the solution.

Security, privacy, and compliance

Security and compliance are important aspects of a solution. When

Implementation Guide: Success by Design: Extend your solution 389

 extending, it is crucial to honor the security and compliance requirements.

Access to data, either implemented by security roles or any other

feature, should not be bypassed when customizing or extending the
native functionality. Breaking the security model can cause a large
negative impact, and security breaches are not to be taken lightly.

The same happens with compliance, for example the General Data
Protection Regulation (GDPR) policies. Compliance functionality
implemented natively also need to be inherited in any customizations of
extensions. Failing to do so may have consequences for organizations that
do not comply with regulations.

GDPR is just one set of regulations. Regulation of privacy and data use
exists in many different forms across several markets. While there is a
great deal of overlap in terminology, privacy and security are not identical.
Security is about preventing unauthorized access to any data, while privacy
is ensuring, by design, proper acquisition, use, storage, and deletion of
data defined as private under local regional and global regulations.

While security is natively built into Dynamics 365 business applications

and highlighted as best practices, privacy requirements tend to have a
higher probability of being overlooked by developers when extending
native apps.

Performance
Although cloud solutions provide a high degree of scalability and
performance, when extending a solution it is important not to
compromise performance.

When extending the user interface and/or the business logic, additional
efforts are added to create, retrieve, update, or even delete data. Those
additional efforts may have an impact on the user experience, depending
on the amount of extended functionality added.

Service protection limits also exist to ensure consistent availability and

performance. These should also be considered when extending the

Implementation Guide: Success by Design: Extend your solution 390

solution to avoid random or unexpected surges in request volumes that
threaten the solution’s availability and performance characteristics.

Performance goals should be identified, agreed upon, signed-off on by

the business, and measured as part of the testing strategy to identify any
deviations from those goals. These goals are especially important when
the solution is extended, as those extensions can highly impact them.

Depending on the environment strategy, organizations may have globally

distributed solutions where factors like latency, firewalls, network traffic,
organization proxies, and routing by Internet Service Providers (ISPs) are
important to ensure a good user experience. When extending a solution,
especially integrations with external systems, the number of external calls,
round trips, and volumes, on top of all previously mentioned factors, are
crucial to ensure that the solution meets the performance goals.

Scalability
The scalability of a business solution is also a key consideration to determine
how you extend it. While the cloud platform includes scalable servers and
micro services, other aspects of the platform need to be considered to
determine the impact of your business solution architecture.

Parameters, such as the number of parallel connections, requests, and

concurrent users, can determine how well the business solution scales, for
example when new business units, countries, or regions are added.

It is better to assess the impact of the chosen architecture as use of the

solution grows, for example, when the scope is small or when the solution
is first released.

When extending a feature, it is also important to understand how the

functionality applies to different business units, countries, or regions, and
what is the level of adjustments needed for each of them. If each
business unit, country, or region requires its own level of
customization, the solution at the extension level is not scalable by
itself and needs further adjustments.

391
 Service protection and limits ensure consistent solution availability and
performance, as well as a level of protection from random and unexpected
surges in request volumes that threaten the availability and performance
characteristics of the platform or solution.

When extending the solution, it is important to understand how use grows

over time and the impact of the design on parameters like:
▪ How much storage is required by the extensions and how does it
grow over time?
▪ How many additional application programming interface (API) calls
do the features require?
▪ What are the limits on workflows, code execution, or API calls?

Impact on application lifecycle management

When extending the business solution, all code and customization must
be packaged and shipped from their various environments. This means
extensions need to be included in application lifecycle management
(ALM). Typically, custom code, customizations, and development created
from the user interface or pro dev tools need to be developed, tested,
approved, and released into production. In addition, for easier
maintenance, they also should be added to a repository where they can
be backed-up, merged, and properly handled.

This means that all extended functionalities are added on top of ALM
practices, which also increases the complexity. As an example, if the
extended features required different configurations, or are only applicable to
specific business units, countries or regions, there needs to be a separation at
the ALM process, which may just be that they are shipped in different
packages (or solutions).

Aspects of the packages should be considered, depending on the number

of extended functionalities added to the native solution. Examples include
the order these functionalities are introduced, how they are split, and how
the size affects the ALM operations.

Implementation Guide: Success by Design: Extend your solution 392

 Maintainability, supportability, and
future-proofing
In a cloud world where updates are continuous, and new features are
released with frequency, it is important to think about the impact and
costs of maintaining extended functionalities.

When extending a solution, additional maintenance requirements

are added to the business solution, so it is important to understand
and be aware of deprecations and roadmap. This helps avoid building
something that might soon be offered natively by the solution or
being forced to replace parts of the extended solution because of
deprecated functionality.

Supportability
Extending a business solution can also complicate the support requirements
of the solution. Normally, the first line of support is at the organization itself
or a vendor. Support resources must have expertise on the business solution
and extensions built on top of off-the-shelf capabilities. This requires a
specialization of resources to support the extended business solution.

The solution provider is typically the last line of support, so it is important

to use proper techniques when extending a solution. Not complying with
these rules can void the warranty or support terms of your contract.

Because some apps provide the ability to customize them to meet

user needs, these customizations can also impact the time required to
maintain them. For example, when a new user is onboarded, the time
it takes to onboard and whether existing security roles and apps can
simply be applied or if additional requirements must be added that
require more maintenance time.

As the solution is extended, the importance of having a good testing

strategy grows because it is necessary to validate how the solution
behaves with each update.

Update and future-proofing

In a cloud solution, organizations must embrace the change and

Implementation Guide: Success by Design: Extend your solution 393

 innovation that comes with frequent new features and capabilities.
These capabilities, in turn, add value to the business.

When it is necessary to extend functionality, it should always be done

in a way that makes it easy and safe for other developers to extend
them down the line.

It is good practice to take the following perspectives into account.

Firstly, having sound developer craftmanship. This applies to basically any

language or technology and dictates that when APIs and extension points
are created, they are well thought through, robust, and well defined. Also,
that the extension is made in a way that allows for other extensions to use
the same extension point or API side-by-side with ours.

Secondly, you want to be a “good citizen” in the new solution. That is

to say that by your example, you encourage other developers who may
expand your code later to follow the same principles. This includes writing
code that others can easily read, APIs and extension points that others can
extend without surprises, unexpected results, or side effects, and design
structures and frameworks that enable others to utilize your effort in a
clean and concise way.

Thirdly, it is important to establish a process that anticipates the future

and works to minimize potential negative effects extending the solution
can bring. This is because, while it is common for organizations to build
custom extensions, deviations from native functionalities can have an
impact on the ability to take advantage of future capabilities.

Product-specific guidance
In the following sections, we will look at Dynamics 365 Customer
Engagement and Finance and Supply Chain Management individually.

Finance and Supply Chain Management

In Dynamics 365 Finance and Supply Chain Management, the

Implementation Guide: Success by Design: Extend your solution 394

 extensibility model provides the ability to extend the out-of-box
functionality to fit additional requirements.

Visit the Extensibility home page for ref- The extension model itself, the toolset, ALM in Azure DevOps, and the
erence information or follow the learning SDK is well described in Microsoft Docs, Microsoft Learn, and in multiple
path at Introduction to developing with
Finance and Operations apps - Learn | community sources. There are two great entry points for learning about
Microsoft Docs.
extensions and development in Finance and Supply Chain Management
apps in the links below and a list of additional relevant links at the end of
the chapter.

Introduction
In this section, we first look at what we can do to enhance functionality
and UI without extending the application. Then we give a very brief
overview of the extension model, and finally highlight some of the tools
and practices that are available for professional developers when they
extend the application.

Let’s first look at what we can extend without changing

applications components in Figure 15-4.
Fig.
15-4
Extension example #1 Extension example #2 Extension example #3

Requirement: Create a custom Requirement: Add a table to hold a list of Requirement: Add code to automatically change
workspace with tiles and lists from food allergens. Add the allergen field to the the status of the new production order to “started”
multiple forms and queries across the sales line record for the user to indicate that a when the user firms a planned production order.
manufacturing module. product contains the specific allergen.

Tools and components: Use the Tools and components: Use Visual Studio in a Tools and components: Use Visual Studio in a
personalization feature and “add to developer environment to add the extend- developer environment to extend the X++ business
workspace” functionality in the UI to ed data types, the tables and fields and the logic, add appropriate event handlers, class, methods,
put together the desired workspace extensions to the sales line table and the Sales and X++ code to catch the event that the planned
and components. No change to code Order form. This change requires new and order is firmed, and execute the correct code pattern
or application components is needed. changed application components and must to bring the new production order to started status.
follow software development lifecycle (SDLC), This change requires new and changed application
build, and deployment guidelines. components and X++ code, and must follow SDLC,
build, and deployment guidelines. Considerations
about scalability and performance when large
numbers of planned orders are firmed using
appropriate developer tools should be considered.

This requires user level experience with Requires entry level professional developer Requires medium- to expert-level professional
navigation of the relevant module, experience in Dynamics 365 Finance and Supply developer experience in Dynamics 365 Finance and
navigation, and personalization. Chain Management and familiarity with Visual Operations and familiarity with Visual Studio, build
Studio, building procedures, and best practices. procedure, best practices, and, ideally, frameworks
like SysOperations Framework, Multithreading, and
performance-checking tools and patterns.

Personalization and no-code options

As we mentioned above, even though Dynamics 365 Finance and Supply

Implementation Guide: Success by Design: Extend your solution 395

 Chain Management offer a rich extension model with minimal impact on
future updates, it is always preferred not to extend the product in the first
place if possible. This lowers risk, maintenance, and project complexity.

Before deciding that a certain requirement must be solved through

an extension to the standard product, it is a best practice to consider if
one of the options for noninvasive personalization or configuration can
meet the required functionality. Dynamics 365 Finance and Supply Chain
Management offer wide options for personalizing the UI, creating user
specific experiences, and automating processes without having to do any
programming at all.

Figure 15-5 is a table of these options.

As the table below shows, many options are available to find alternative
approaches to extending the application. The below is not an
exhaustive showing. Additional options for low code and no code
options are mentioned in Chapter 16, “Integrate with other solutions.” The
Fig.
decision of whether to extend comes down to user efficiency and value for
15-5
your customers.

Tools What is it? Description Characteristics

The application remembers the While this is hardly considered a way of extending, it Personalizations can be shared
Restricted last settings for sorting, column does give the user the opportunity to store selections in by users or managed centrally
width, criterial values in queries dialogs, expansion of relevant fast tables, and aligning by an admin from the
personalization
and dialogs. the column width so that more columns are visible on a personalization page.
given form.

Personalization bar in forms and Personalization allows users or admins to add or hide Personalizations can be shared
Personalization workspaces. fields or sections, change labels, change the order by users or managed centrally
of forms/UI of columns, and edit the tab order by skipping fields by an admin from the
when pressing tab. personalization page.

Saved views is a combination of Saved views is a powerful tool that allows the user Saved views can be shared by
personalization of the UI for a the ability to quickly switch between tailored views of users or managed centrally by an
form and filtering and sorting of columns, filtering and sorting on the same screen admin from the
the form data source. depending on the specific task at hand. For example, Personalization page.
Saved views
a buyer in a pharmaceutical company may need a
simple view of the purchase order screen for May require the Saved views
nonregulated materials purchasing regulated materials feature to be turned on.
used for manufacturing

Users can use personalization This functionality allows users to tailor their experience Custom and personalized
and the “add to workspace” but- to their needs. Workspaces provide glanceable workspaces can be shared by
Custom ton to add tiles, views, or links to information about the most important measures, users or managed centrally by
Workspaces an existing or custom workspace. actionable items, and relevant links to other pages. an admin from the
personalization page.

Implementation Guide: Success by Design: Extend your solution 396

 Tools What is it? Description Characteristics

Users with certain security roles Finance and Supply Chain Management have a rich set Custom fields and the
can add up to 20 custom fields of features that apply across a wide set of industries. personalized forms to show the
to tables Some organizations require additional fields on certain fields can be shared by users or
tables; for example, the item or customer master or managed centrally by an admin
Custom fields the sales order header. This feature allows the user to from the personalization page.
create these fields. Note that these fields are specific to Deleting a custom field is
the environment that they are created in and cannot irreversible and results in the loss
be referenced by developer tools. of the data in the custom column.

The grid on forms in the system The grid offers the following capabilities. The admin can enable the
have some extended features New Grid Control feature from
▪ Calculating totals for columns in the grid footer
that may eliminate the need for feature management. Note that
an extension. ▪ Pasting from Excel there are certain limitations, see
▪ Calculating math expressions. For example, the reference at the bottom of
if the user enters 3*8 in a numeric field and the section.
Grid capabilities presses tab, the system calculates and enters the
result of 24

▪ Grouping tabular data in one or more levels in

expandable groups

▪ Freeze panes so important information does not

scroll out of view when scrolling horizontally

The user can add a canvas app to The ability to embed a canvas power app enables citizen Please see more about UI
a form or workspace as embed- developers to use low-code/no-code options for inter- integration for Finance and
ded into the UI or as a menu item acting with data in Dataverse directly from the UI in the Supply Chain Management in
Embedded that can pull up the app from the Finance and Supply Chain Management apps. Chapter 16, “Integrate with
power apps menu. other solutions.”
canvas apps
It is important to note that if the app must interact with
Finance and Supply Chain Management data, that inte-
gration to Dataverse must be in place and the app must
of course support the required actions.

Users can view, edit, and act on IT admins can build and publish mobile workspaces that Simple actions can be done from
business data, even if they have have been tailored to their organization. The app uses the mobile workspaces. Most
intermittent network connec- existing code assets. IT admins can easily design mobile more advanced actions require
Mobile tivity on an app for iPhone and workspaces by using the point-and-click workspace extension.
workspaces Android. designer that is included with the web client.

The application has 13 predefined workspaces out-of-

the-box that cover Finance, AR, AP, and HR functions.

Users can extract or edit data The Excel integration allows for a wide variety of With great power comes great
on most forms in the system by scenarios for entering, presenting, and interacting with responsibility. While it is easy to
clicking on the office icon. data in way that is not possible from the UI. In addition change a whole column of data
Excel to the export-to and open-in Excel option, the user can and publish that data into the
integration create workbooks and templates for specific purposes. system, it is equally easy to make
Excel has many features for presentation and offers a mistake.
data manipulation capabilities for larger datasets that
users cannot do in the system UI.

Extending the application

In this section, we provide a high-level introduction to the extension
model and discuss required skill sets and practices used when extending.

Implementation Guide: Success by Design: Extend your solution 397

 We also highlight some of the tools and techniques that we recommend
you use to ensure a clean, high-performing, and future-proof solution.

Figure 15-6 is an overview of the components, the consideration for

extensions, and the related characteristics.
Fig.
15-6

Component Editor Considerations Do Not

Graphical editor in Visual Studio Forms must adhere to patterns. The editor in Visual Deviate from the predefined
with preview. Studio has a rich modeler that will help the developer patterns and create monster all-
applying the right structure to the screens. This will in-one screen style forms. They
User interface
ensure performance when loading the form, adapt- are often a result of the designer
ability across different resolutions and screen sizes trying to replicate a legacy
and consistency across the UI. system experience.

Metadata editor in Visual Studio Follow best practices and frameworks. Apply indexes Create redundancy or replicate
Data model for Tables, Fields, Extended Data and define delete actions. Normalize. Use effective poor modeling from legacy apps.
and queries Types, Enums Queries and Views date framework when applicable. Keep performance
in mind. Use field lists when performance is a risk.

X++ Editor in Visual Studio. Adjust compiler setting to alert about best practices, Ignore best practices, write long
code with the goal of zero deviations. Use code patterns methods or over comment
and frameworks. Practice good developer the code.
Business logic
citizenship. Write clean easily readable code.
Run CAR report and use compatibility checker.
Unit test the code.

SSRS report editor in Visual SSRS reports are good for certain precision Do not reach for the SSRS report
Studio designs and tabular lists. See Chapter 13, option if there is a better way.
“Business intelligence, reporting, and analytics,”
Reporting
for more information.

Metadata editor in Visual Studio The out-of-the-box data entities are general purpose Do not create data source
entities are built to support a wide variety of features proliferation. See Chapter 16,
surrounding business entity. In scenarios where a high “Integrate with other solutions,”
volume, low latency interface is required it is recom- for more information.
Data entities mended to build custom data entities with targeted
and specific features needed to support high volume
interfaces in the implementation.

Development architecture
The architecture of the development environment, as shown in Figure
15-7, includes the software development kit (SDK), which consists of Visual
Studio development tools and other components. Source control through
Azure DevOps allows multi-developer scenarios, where each developer
uses a separate development environment. Deployable packages are

Implementation Guide: Success by Design: Extend your solution 398

 compiled and packaged in a build environment or a build service and
deployed to Dynamics Lifecycle Services (LCS) for further deployment to
nonproduction environments. Deployment to the production
environment happens through a service request, after proper testing has
been done in the UAT environment and users/stakeholders have signed
off as described in the SDLC.

See Chapter 11, “Application lifecycle management,” for more information

about the SDLC.

Microsoft Power Platform integration with Finance and Supply

Chain Management
Finance and Supply Chain Management is a virtual data source in
Dataverse that enables full create, read, update, and delete (CRUD)
operations from Dataverse and Microsoft Power Platform.

Fig.
15-7

Visual Studio Debug/Run (F5)

Local Runtime

Batch
Storage
Project system Cloud manager
service
X++ code editor instance service
Designers
Application explorer
Design-time meta-model Deploy
Application object server Service
Best practice integration
hosted by Internet information service

Package
(metadata, binaries, and data)

Business
database

Build
Model binaries Runtime
(file system)

Model store
Metadata API
(file system)

Development environment

Implementation Guide: Success by Design: Extend your solution 399

 Fig.
15-8

Implementation Guide: Success by Design: Extend your solution 400

 Virtual tables provide a mechanism to use Microsoft Power Platform with
Finance and Supply Chain Management without having to physically
copy data to Dataverse. Virtual tables and dual-write are complementary
technologies, as illustrated in Figure 15-8. That means they can be used
together if required, depending on the different scenarios.

Dual-write, shown in Figure 15-9, is a feature that provides near real-time

interactions between customer engagement apps and Finance and Supply
Chain Management apps for specific integration scenarios. It provides
tightly coupled, bidirectional integration between Finance and Supply
Chain Management apps and Dataverse.

By bringing data from Finance and Supply Chain Management apps

into Dataverse, it is possible to leverage the Power Platform and use
low-code/no-code developer models on Finance and Supply Chain
Management data. Either as embedded power apps or PowerBI reports
Fig.
15-9 or dashboards, in standalone apps, or integrated with other Dynamics 365
Business Applications, without making changes through extensions to the
application and data layer.
Dual-write

Advanced practices
Finance and
Model-driven Tightly-coupled, real-time, and
Supply Chain
applications in bidirectional integration for Management
Dynamics 365 documents, master, and
reference data.
applications and tools
Common
As we mentioned earlier, it is
data always sound advice to be good
service
citizen developers. In Dynamics
365 for Finance and Supply
Chain Management specifically, you can ensure that you follow that
principle by understanding, learning, and using the following tools
and best practices as shown in Figure 15-10.

For more on these topics, see the “Reference links” section later in this
chapter.

Additional components
It is important to note that the extended product architecture contains
several additional components. Along with that are multiple avenues
and tiers for approaching a requirement for extension. It depends on the

Implementation Guide: Success by Design: Extend your solution 401

 Fig.
15-10

Tools What is it? Description Characteristics

Best practice checks for X++ and Developers should strive for zero deviations. The best You can stop developers from
Best practice application components are built practices are made to ensure and updatable, performing checking in code with best
check into Visual Studio. They can be and user friendly solution. practice deviations.
errors, warnings, or informational.

The compatibility checker tool can The compatibility checker tool is available as one of Not all breaking changes can
detect metadata breaking changes the dev tools in Platform update 34 and forward. be detected by the tool. See the
Compatibility
against a specified baseline release You can use it to ensure that your solutions are Compatibility checker docs page
report or update. backward-compatible with earlier releases before for specifics.
you install or push updates.

The users can take a trace of You can use the trace parser to consume traces and The trace parser tool can be
runtime execution directly from analyze performance in your deployment. The trace found in the PerfSDK folder in
Traces and
the UI. The trace parser to read parser can find and diagnose various types of errors. your development environments.
trace parser the trace. You can also use the tool to visualize execution of X++
methods, as well as the execution call tree.

Performance timer is a tool in To open the Performance timer, open your web page The tool itself has a performance
the web client that can help you with the added parameter debug=develop. You can impact.
to determine why your system's see counters for client time and server time, and the
Performance performance might act slow. total time. Additionally, you can see a set of performance
timer counters, a list of expensive server calls, how many SQL
queries were triggered by this individual call and which
SQL query was the most expensive.

Under Environment Monitoring The logs provide for example: The tools under Environment
in LCS there a comprehensive Monitoring are very effective at
▪ Activity Monitoring: A visualization of the activity
collection of tools and information diagnosing potential or growing
that has happened in the environment for given
that you can use to analyze and performance issues. Keeping
timeline in terms of user load, interaction and
diagnose your cloud environment. an eye on these metrics can
LCS Logs activity.
help pinpoint problems with
▪ SQL Insights: Logs that include advanced SQL extensions.
troubleshooting.

▪ Information about slow queries, deadlocks,

crashes etc.

The CAR report is an advanced The CAR report can be run by command line in a A clean CAR report is a require-
Customization best practice check tool. development environment. The output is an Excel ment for the go-live readiness
Analysis Report workbook with recommendations issues review prior to enabling
(CAR Report) and warnings. production.

A breaking change is a change that Breaking changes are, for example, changes to data That although the application is
can break the code consumers of model and extended data types, changes to access massive, we tend to only extend
your code and components make. modifiers on classes and methods and many others. the same relatively small subset
of elements. It is not as unlikely
Understand and
This is especially important in heavily extended solutions, that you may think that other
avoid breaking
if you are making a basis for a global rollout, if you developers use your components
changes
are making an ISV, or if you have multiple developers or code.
sharing common custom APIs or constructs but should
always be considered.

Implementation Guide: Success by Design: Extend your solution 402

 Tools What is it? Description Characteristics

If you find a need for an exten- Extensibility requests are logged to a backlog. Micro- Extensibility requests are
sion point that is currently not soft engineers prioritize all requests, and then work following the same cadence as
Log extensibility available, log the request early. on them in priority order. Please note that Microsoft the platform updates.
requests early This is done via an extensibility is not guaranteeing that all requests will be fulfilled.
request in LCS. Requests that are intrusive by nature will not be sup-
ported, as they will prevent seamless upgrade.

Sometimes developers put a Developers are the first line of defense against bugs, It is a lot easier and cost effective
lot of effort into building the performance issues and semantic issues that may for the developer to find and fix
extension, but little effort into exist in the specification. By simply going through the a problem before it is checked in,
unit testing it before determining intended functionality from perspectives such as: built and deployed.
Proper unit whether it is ready to deliver. ▪ Will this code scale with high volumes?
testing ▪ Does it do what I expect?
▪ Can I do things I am not supposed to?
▪ Could the user accidently do something unwanted?
▪ Does the requirement make sense or force me to
break best practices or patterns?

specific area and nature of the requirement.

In addition to the components in the Finance and Operation

application stack, there are components for Commerce and WMS that
require different technologies as well as developer toolsets and skillsets.
These include the Point of Sale (POS), The Commerce Scale Unit (CSU), the
eCommerce components, the Warehouse Management Mobile App, and
the Finance and Supply Chain Management Mobile App.

That means separate requirements for extensions may have to be applied

to different components. Slightly different approaches may also be need-
ed, and the complexity and skill sets required may vary. For all of these, it
is critical that developers and administrators familiarize themselves with
each of the specific SDKs, guidelines, application lifecycles, builds, and
deployment methods and the skill sets required before deciding to extend
the component.

Reference links
▪ Extensibility home page - Finance and Supply Chain Management

▪ Application stack and server architecture - Finance and Supply Chain Management

▪ Microsoft Power Platform integration with Finance and Supply Chain Management

▪ Microsoft AppSource – destination for business apps

▪ Commerce for IT pros and developers - Commerce

Implementation Guide: Success by Design: Extend your solution 403

 ▪ Write extensible code - Finance and Supply Chain Management

▪ Breaking changes - Finance and Supply Chain Management

▪ Extensibility requests - Finance and Supply Chain Management

▪ Grid capabilities - Finance and Supply Chain Management

▪ Extensibility requests - Finance and Supply Chain Management

▪ Mobile app home page - Finance and Supply Chain Management

▪ Take traces by using Trace parser - Finance and Supply Chain Management

▪ Testing and validations - Finance and Supply Chain Management

Customer Engagement
The Power Platform, shown in Figure 15-11, provides the ability to use
configurations, low-code/no-code customizations, and still allows devel-
opers to extend programmatically first-party customer
engagement apps like Dynamics 365 Sales. Custom business
applications can also be created. You can even mix the two approaches to
provide applications adjusted to specific organizational needs.

Fig.
15-11
Microsoft Power Platform
The low-code platform that spans Office 365, Azure, Dynamics 365, and standalone applications

Power BI Power Apps Power Automate Power Virtual Agent

Business analytics Application development Process automation Intelligent virtual agents

Data connectors AI builder Dataverse

Implementation Guide: Success by Design: Extend your solution 404

 Until recently, creating custom apps would require professional
developers. But by using Power Apps, organizations can quickly build
custom business applications that provide rich business logic and
workflow capabilities to transform manual business processes into
digital, automated processes.

Power Apps democratizes custom business app building experience

by enabling non-professional developers to build feature rich, custom
business apps without writing code.

When offering this ability to users, it is important to have a central

governance entity to determine guidelines for certain aspects of the
solution, like security, data model, etc. This prevents redundancy, gaps
in the security model, and bad practices/architecture that impact how
apps are maintained, secured, scaled, and shipped.

The Microsoft Power Platform Center of Excellence (CoE) starter kit

provides a collection of components and tools that are designed to
help organizations develop a strategy for adopting and supporting
Microsoft Power Platform. This helps organizations implement a CoE
to drive innovations and improvement while also providing standards,
consistency, and governance.

Power Apps also provide an extensible platform that lets professional

developers programmatically interact with data and metadata, apply
business logic, create custom connectors, and integrate with external data.

Model the user experience

Power Apps is a high-productivity development platform for business
apps and has four major components:

Canvas apps are intended to build a business app from a canvas in

Microsoft Power Apps without requiring professional developers to create
code. The apps can be built by using drag and drop elements onto a
canvas. It also provides the ability to use connectors to integrate data
from, and to, a variety of sources.

Model-driven apps are a component-focused approach to app

Implementation Guide: Success by Design: Extend your solution 405

development, which also does not require code. These apps can be simple
or complex. With Model-driven apps much of the layout is determined
and designated by the components that are added to the app. The
Model-driven apps can be completely custom or from first-party apps
such as Dynamics 365 Sale.

Both types of apps are built to be responsive by adjusting and being

accessible from different devices.

Power Apps makers can also create external facing Portals that allow users
outside their organization to create and view data in Microsoft Dataverse.
The Portals experience reuses components in other apps to determine the
experience. It is also possible to customize the Portals experience, similar
to how other apps are customizable.

Microsoft Dataverse securely stores and manages data that is used by

business applications.

Business logic and validation within the platform

Business logic and validation to ensure data quality and reduce
repetitive tasks can be applied within the Power Platform by using the
following approaches:
▪ Business rules validate data insertion and provide validation messages
within the user interface. As an example, with business rules it is
possible to display a message when a certain field does not have data
or does not have data in the expected format.
▪ Business process flows guide users to ensure they enter data and use
the solution in a consistent way. An example would be displaying the
information that is required to complete a business process, like lead
to invoice, this provides a consistent approach for all users while it
also improves usability of the solution by making the process more
efficient and improving data quality.
▪ Workflow allows business processes to be automated within the
process designer to apply a condition or perform a new action.
For example, when closing a support case to send a survey to
the customer.
▪ Business logic with code supports advances developer scenarios
to extend the application with code, for example using java script

406
 code for form events or plug-ins that apply business logic to data
transaction.

Dynamics 365 applications, such as Dynamics 365 Sales or Dynamics 365

Customer Service, use Dataverse to store and secure the data they use.
This enables organizations to build or extend apps by using Power Apps
and Dataverse directly against the business data.

Dataverse data is structured according to the Common Data Model,

which is a shared data language for business and analytical applications
to use. By using Dataverse, makers can simply extend the data model
of first-party solutions such as Sales, Customer Service, Field Service or
others, and jump-start app development by using the Common Data
Model with business logic, security, and integration already built in.

Dynamics 365 Customer Engagement and the Power Platform provide

a powerful customization and app making capabilities by using several
components that provide the ability to build the appearance and
functionality of an app.

These components are distributed in four categories.

Data
The entity designer and option set designer determine what data the app
is based on and allow changes to the data model by adding additional
tables and fields as well as relationships and components that use
predetermined options for users to select.

User interface (UI)

User interface components, like the app designer, site map designer,
form designer, and view designer, determine how users interacts with
the app and allow changes to be made to the components that display
in the app UI.

Logic
The business process flow designer, workflow designer, process designer,
and business rule designer, determine the business processes, rules, and
automation of the app.

Implementation Guide: Success by Design: Extend your solution 407

 Visualizations
These determine what type of data visualization and reporting the
app includes, like charts, dashboards, and reports based on SQL Server
Reporting Services.

You can create a range of apps with Power Apps, either by using canvas or
model drive apps to solve business problems and infuse digital
transformation into manual and outdated processes.

Solution analyzers
Solution checker  The solution checker can perform a rich static
analysis of the solutions against a set of best practice rules to quickly
identify patterns that may cause problems. After the check is completed,
a detailed report is generated that lists the issues identified, the
components and code affected, and links to documentation that
describe how to resolve each issue.

It is important to include solution checker with the solution release cycle,

and after releasing an updated version in a lower environment (sand-
box), to make sure that any deprecations or issues related to extensions
are identified.

Portal checker  The portal checker is a self-service diagnostic tool that

can identify common issues with a portal by looking at various
configuration parameters and providing suggestions on how to fix them.

Power Apps checker web API  The Power Apps checker web API
provides a mechanism to run static analysis checks against customizations
and extensions to the Microsoft Dataverse platform. It is available for
makers and developers to perform rich static analysis checks on their
solutions against a set of best practice rules to quickly identify prob-
lematic patterns.

Conclusion
Cloud-based solutions offer ready-to-use applications solutions
that can be easily delivered, reducing the time required for an

Implementation Guide: Success by Design: Extend your solution 408

 organization to start taking advantage of it. SaaS solutions typically
reside in cloud environments that are scalable and offer native
integrations with other SaaS offerings. These solutions benefit from
continuous upgrades that add innovation and new capabilities
multiple times per year. This reduces costs and effort and eliminates
the downtime associated with upgrades in a traditional
on-premises model.

This means that organizations can start using the solution as-is. Still,
in some scenarios, additional requirements are needed to add value
to the business or to empower users and drive adoption of the new
solution. Thus, modern SaaS solutions also provide rich capabilities
to further extend it. These can range from simple configurations us-
ing a low-code/no-code approach or an extension by using custom
code by professional developers.

It is important to consider that the level and depth of these extensions

can impact key characteristics, like performance, or increase the cost
of supporting and maintaining the solution, but they should hinder
the ability to take advantage of innovative functionalities that are
delivered as part of the continuous update strategy. Normally, when
deciding to extend a solution, a trade-off is necessary between new
functionalities and the impact of adding them. For example, if adding
too much information to a form it impacts the time required to load
the information included in it.

In this chapter we have discussed how these key characteristics can

be impacted and the consequences of overextending the solution.

With Success by Design, during the Solution Blueprint workshop,

such topics are discussed to establish a clear understanding of
the impact that the extensions have on the solution, as well as to
identify risks and issues that come with these decisions. One of
the most important factors to consider when deciding to extend
a solution is how performance is affected. During the Solution
Performance implementation workshop, the FastTrack solution
architect works along with the implementation partner and the
customer to review how the extensions can impact the overall

Implementation Guide: Success by Design: Extend your solution 409

 performance of the solution. Typical risks and issues are identified
related to overextending forms, synchronous events that impact
the user experience, impact on capacity such as service protection
limits, resources, and allocation limits.

References
Business Apps | Microsoft Power Apps

Build Apps – Canvas Apps or Model-driven Apps | Microsoft Power Apps

Microsoft AppSource – destination for business apps

Implementation Guides: Success by Design: Extend your solution 410

 Checklist

Define your Considerations

extensibility strategy Validate that extensions don’t negatively impact the user
Check that requirements to extend the solution are driv- experience, responsiveness, performance, and how the
en by the need to empower users and bring additional solution behaves across different devices and platforms
value to the business, or industrialize the solution. as part of the organization test strategy.

Ensure the solution doesn’t mimic the ways of achieving Ensure the extensions honor the security mechanism,
the same results as the legacy solution or the system privacy, and compliance requirements.
being replaced.
Ensure extensions are scalable, tested for high volume,
Understand the platform capabilities and use its and capable of handling peaks like holiday seasons.
strength to simplify and optimize the overall process to
Align extensions with ALM automated processes to
get the most of the out-of-the-box experiences.
build and deploy them in an efficient and fast-paced
Review if any potential ISVs were considered before approach.
deciding to extend the solution. The AppSource market-
Ensure code and customizations follow only the
place contains ISV-managed solutions that may replace
documented supported techniques, and don’t use
the need to create a fully custom solution.
deprecated features and techniques.

Implementation Guide: Success by Design: Extend your solution 411

 Case study
The power of making it
your own
A wealth management services company that delivers personalized
services had the need to standardize, simplify processes, drive efficiencies,
expose more data, and increase visibility and interoperability to their users.

After evaluating solutions, the company decided to build a modern,

completely customized, cloud-based customer engagement platform on
Microsoft Dynamics 365.

The off-the-shelf functionalities gave the ability for the users to achieve
a single view of the customer lifecycle across all products. Through
customizations and further extensibility of the solution, the company
added special wealth management services and fully customized the
user experience to better serve their customers.

This helped the company streamline processes, increase revenue,

reduce costs, enhance transparency, and manage regulatory com-
pliance, and allowed the users to better understand each customer’s
story, goals, and capacity for risk. The company was able to deliver
appropriate, innovative, high-value services and products to maintain
successful end-to-end customer journeys across every account.

The power of using off-the-shelf functionalities and being able to further

customize and extend the experience for their users set the foundation
to improve their internal processes and empowered their users to deliver
better services in a more efficient matter.

Implementation Guide: Success by Design: Extend your solution 412

16 Guide
Integrate
with other
solutions
 Introduction
Together, Dynamics 365 Business Applications
provide a rich and mature feature set across a wide
variety of industries.
However, there are situations in which you might want to go beyond
the boundaries of the application suite and extend processes or
incorporate data assets from other sources.

In this chapter, we examine how to integrate Business Applications.

We discuss the importance of keeping business goals at the forefront

This chapter covers the
while aligning integration projects with your organization’s overall
following topics:
solution strategy.
Defining business goals

Choosing a platform
In line with the design pillars, we look at how the choice of integration
Choosing a design platform should fit into the architectural landscape.

Choosing a pattern
We also look at how to choose an integration design that offers users
Challenges in integration
the capabilities they desire in the short and long term.
Product-specific guidance

We discuss integration patterns available with Business Applications

Implementation Guide: Success by Design: Integrate with other solutions 414

 and consider important characteristics and factors.

And finally, before diving into the product specifics, we walk through
some of the common challenges people face when integrating systems.

Defining business goals

Defining To ensure that your integration work aligns with the overall direction of
business goals the business, it’s important to match each requirement of cross-system

Choosing a platform processes against the overall goals of the project and the business. To
accomplish this, begin your integration work by defining goals that
Choosing a design map to the business perspective.
Choosing a pattern

Challenges in integration Reasons for integrating

Product-specific guidance You might have many reasons to want to integrate systems in your
solution. Given the array of tools available and the vast ecosystem provided
by Microsoft Power Platform, integration across processes and systems is
often readily available and a scalable and reliable choice. Here are some
examples of when you might want to unify systems by using integration.

Integration to legacy system

You might have a highly specialized system—perhaps a custom-built or
heavily customized system that fits a specific set of requirements in your
business—that you want to keep in place or phase out over time. With
this approach, you would integrate Business Applications into the legacy
system to get the benefits of both and allow for business processes to
span both systems.

Multi-phased implementation
You might be implementing Business Applications in a planned
multi-phased approach, starting with a geographic location, a single
division, or a single Dynamics 365 business application—Finance, for
example. In this scenario, some level of integration with the parts of
the legacy solution that will be implemented in future phases could
be necessary.

Implementation Guide: Success by Design: Integrate with other solutions 415

 Regulatory requirements
Regulatory, government, and industry data exchange or reporting
are often standard practice or required by law in some industries and
regions. It’s often reported electronically, but more complex exchanges
of data could require a more traditional integration approach.

Financial consolidation
Perhaps your organization is a subsidiary of a larger operation that
requires data to be consolidated and reported in a corporate parent
entity. This often requires extracts of data to be transformed and
loaded into the corporate consolidation system. In some cases, it’s
the other way around: your organization might expect integration
of consolidation and other data from your subsidiaries into your
new system.

Multiple system architecture

Sometimes the Dynamics 365 business application is a link in a bigger
chain of specialized line-of-business systems. For example, the Prospect-
to-cash process might start in one external system where the opportunity
is discovered, then go through quote and order management in Dynamics
365 Sales App, and then integrate to other systems for warranty tracking,
billing, and financial reporting.

Extract to reporting solutions

Perhaps you might have invested in building a specialized analytics
solution that includes a data warehouse with many years of historical
data and tailored reporting on top of it. In that scenario, you might
want to extract data from Business Applications for transformation and
load it into the existing data warehouse.

Many more scenarios are not described here. Some might even be
combinations of several of the examples.

When multiple systems “talk” to each other—systems that perform

distinct functions in the business—it’s not only about making it work.
Integration can enable business processes, reporting, and analytics that
would not have been possible otherwise.

Implementation Guide: Success by Design: Integrate with other solutions 416

 Goals and benefits
Organizations that integrate systems when building their business
solution can achieve several key goals and benefits, including:
▪ Process automation  Integration enables the automation of
repetitive tasks in the workflow.
▪ Data availability at the right time  Communication between
different departments is key. If the sales department depends on
information from accounting, integration can help the data to flow
at the time that the sales department needs it. Whether the data
arrives in real time or in batches, you can program it to arrive at a
certain time. The right pattern of communication will help satisfy
the different business units.
▪ Reduced human errors  An error in the data like a typo, an
incorrect number, or a button pressed at the wrong time could
significantly affect your process, or even worse, degrade the
customer experience.
▪ Increased productivity  The users become more focused on the
processes that add value, since they won’t need to reach for or
search other systems to retrieve the information they need.
▪ Process optimization  Integration simplifies your processes so
that you spend less time executing tasks, thereby adding more
value to your business.
▪ Increased security  By automating processes and data movement
through integrations, organizations implement better controls for
data access and reduce the need for users to directly work with
sensitive data.
▪ Regulatory compliance  Automated integrations could help
meet some of the controls needed to meet regulatory needs like
HIPAA (Health Insurance Portability and Accountability Act) data
or 21 CFR (Code of Federal Regulations) Part 11 processes.
▪ Reduced cost of operations  Integrations might help reduce
repetitive manual activities, human errors, and training activities,
which could lead to an overall decrease in cost of operations.
▪ Avoiding data silos  Integrations break down data silos and
improve the value of the organization’s data. With AI, machine
learning, and IoT usage on the rise, this data might help drive better
insights, data science, and predictive analytics in your company.

Implementation Guide: Success by Design: Integrate with other solutions 417

 Integration planning
To properly incorporate business goals into the entire project lifecycle,
we recommend you plan the integration architecture in the initial
For supplemental information, read the
Success by Design Integration architecture
stages of the implementation project.
guidance on Microsoft Learn.

When planning integrations, it’s easy to miss a critical point or

underestimate volume or complexity. To prevent this, we highly
recommend you create a blueprint of the design before starting
any specification work and do some calculations of loads and
transaction flow.

Conceptualizing
Creating a blueprint and thoroughly planning will also help you
Success by Design highly recommends that
formulate the testing and performance testing of the solution later in
you approach integration work the same the implementation. That’s why integration architecture is a key part
way you would an extension project, by
following a defined software development of the Success by Design solution blueprint.
lifecycle (SDLC) that incorporates business
stakeholders and collects their buy-in.
The SDLC should include requirements, To create a blueprint, you can leverage several types of diagrams,
design, development, and user acceptance
testing, as well as performance testing,
which we describe here.
deployment, and application lifecycle ▪ Application integration diagram  The application integration
management (ALM). The work-to-define
requirements should be business driven diagram is often a high-level representation of solution architecture.
and process focused. For more information, An example of one is shown in Figure 16-1. Many styles exist, but in
refer to Chapter 7, “Process-focused
solution,” and Chapter 11, “Application its basic form it should provide an overview of which systems in the
lifecycle management.”
solution need to be integrated and, ideally, what data is exchanged
and the direction of the data flow. Once the overview is established,
details can be added about the specific interface touchpoints,
frequency, and volume information, and perhaps a link to ALM
information, such as a functional design document (or FDD)
number or link.
Fig.
16-1

Dynamics 365 Customer Dynamics 365 Customer

Engagement Apps Engagement Apps

System 1 Data App 1 Data App 3

App 2 App 4

Implementation Guide: Success by Design: Integrate with other solutions 418

 ▪ Interaction diagram  The interaction diagram defines system
and user trigger points, data and process flows, and the sequence
of events. It can also uncover unforeseen complexity. By following
the flow, even nontechnical team members can understand and
validate the design before any work has started. The diagram in
Figure 16-2 illustrates a simple three-system interaction diagram
Fig.
16-2 for a simulated process flow.

System 1 System 2 System 3

Process start

User trigger event System event

Message Request

Message Response
User action User trigger event

▪ Process documentation, mapping, and volume calculations 

Once the high-level overviews are defined, we recommend you
Fig. document the cross-systems processes in more detail, for example
16-3
by defining user stories, before starting the system-to-system
mapping. Mapping can be done as part of a functional or technical
System 1 System 2
Sales SCM design document. (An example of mapping is shown in Figure 16-
3.) The design document, along with the process definition, should
AccountNumber CustomerAccount
be detailed enough to ensure that everyone agrees on the scope,
Name OrganizationName process, desired outcome, test criteria, and benefits of the
Description SalesMemo integration. Often such a document also contains information

Telephone1 PrimaryContactPhone
about the expected effort and duration. Calculating the volume of
transactions that flow through the interface is important because
EmailAddress1 CustomerGroupID
it helps you decide what patterns to use, and ultimately the size
CreditLimit PrimaryContactEmail
of the different platform components and services needed for
CreditLimit the integration.

Implementation Guide: Success by Design: Integrate with other solutions 419

 Choosing a platform
Defining business goals In this section we discuss what you should consider when choosing a
platform and, if needed, a middleware solution to achieve the previously
Choosing mentioned business goals.
a platform
Choosing a design Data storage has increased exponentially in recent years, and it will
Choosing a pattern continue to intensify in this direction. The amount of data that’s
generated and transferred today is practically unimaginable. In
Challenges in integration
addition to the new data that’s being created, historical data
Product-specific guidance might need to be preserved. The platform you choose will need
to incorporate all of these realities, plus the fact that the amount
of generated data varies from business to business and project to
project. The platform you choose must be able to reliably handle
the storage and transfer of vast amounts of data.

Cloud, on-premises, and hybrid platforms

Your organization’s size and maturity level are factors when
considering various platform scenarios. Most organizations find
themselves using some variation of one of the following types:
▪ Simple cloud  In the scenario depicted on the left side of Figure
16-4, the organization might have acquired the first cloud software
as a service (SaaS) application. Overall solution complexity is low.
As part of the implementation project, they want to integrate with
another cloud service.
▪ Simple hybrid  The scenario on the right side of Figure 16-4 is
similar to the previous one in that the organization might have
acquired the first cloud SaaS application. Solution complexity is also
low. But in its implementation project, the organization wants to
integrate to a system on-premises, adding the need for a strategy
for crossing the cloud to on-premises boundary.

When implementing a simple cloud or simple hybrid system, be

sure to use the capabilities that Microsoft Azure and Power Platform
provide. In Business Applications, out-of-the-box low-code/no-code
options are available for using Pull designs and RESTful service

Implementation Guide: Success by Design: Integrate with other solutions 420

 endpoints. Such an approach will build a scalable foundation,
allowing expansion for years to come.

Fig.
16-4 Simple cloud Simple hybrid

System 1
System 1 System 2

Cloud
Cloud
On-premises

On-premises
System 2

Whether in cloud or hybrid, be sure to not fall for the temptation of building
tightly coupled point-to-point, custom synchronous service integrations.
They will likely not scale well, and future expansion will likely make the
system more fragile. It’s good practice to lay a solid foundation early.
Organizations that choose not to do that often find themselves having to
make tough and expensive adjustments to make the architecture scalable
as the business grows.
▪ Primarily on-premises  In the scenario depicted on the left side of
Figure 16-5, the organization has in place a rather complex on-premises
solution with multiple line-of-business systems. The organization is
using a middleware platform that’s based on-premises to connect
systems and might be integrating its first SaaS cloud application.
▪ Hybrid  The hybrid example on the right side of Figure 16-5
represents large, diversified enterprises or organizations that are
halfway through their cloud journey. The organization is using
connected cloud and on-premises middleware platforms that can
integrate in hybrid scenarios, and it might add an additional cloud
application to the solution architecture.

Implementation Guide: Success by Design: Integrate with other solutions 421

 Fig.
16-5 Primarily on-premises Hybrid

System 1 System 2 System 3 Storage/DW

Cloud

Cloud
System 5

Middle-
ware

Middle- Middle-
ware ware
On-premises

On-premises
System 1 System 2 System 3 System 4 System 4 System 5 System 6 Storage/DW

When planning an incremental migration to the cloud, consider

the best options for where and how to achieve a reliable hybrid
middleware-to-cloud platform. Consider whether the current
middleware platforms and on-premises components are scalable and
provide opportunities for leveraging modern technologies, such as
low-code/no-code development options, or whether it would make
sense to leverage the possibilities within Azure and Power Platform to
migrate the current line-of-business systems to the cloud to achieve
scalability, elasticity, and extendibility.

Be sure to not silo data assets in on-premises line-of-business systems

and disjointed data stores that are difficult for citizen developers,
makers, and modern frameworks to access.
▪ Hybrid without middleware  In the scenario shown on the
left side of Figure 16-6, the organization has likely evolved the
integration architecture over time. Multiple systems are integrated

Implementation Guide: Success by Design: Integrate with other solutions 422

 using different technologies, and it seems System 4 is acting as
a hub. No middleware platform is applied, and strategy is likely
somewhat loosely defined. The perceived cost of applying an
When implementing a system that’s
integration platform and middleware is making it necessary to use
primarily cloud, compare the cost of
implementing a clear strategy and point-to-point integration when new systems are introduced.
using a platform and middleware such
as Azure and Power Platform with the
▪ Primarily cloud  In the scenario shown on the right side of
cost of operating and maintaining the Figure 16-6, the organization has a higher level of cloud maturity,
existing integrations.
and platform and middleware are platform as a service (PaaS)
cloud services. The organization has integrated a few on-premises
components that are necessary, and additional line-of-business
systems are all in the cloud.

Fig.
16-6 Hybrid, no middleware Primarily cloud

System 1 System 2 System 3 Storage/DW

System 1 System 2 System 3 Storage/DW
Cloud

Cloud
Middle-
ware
On-premises

On-premises
System 4
System 4 System 5 System 6 Storage/DW

It’s important to not continue using resources on maintaining an

integration architecture that spans multiple technologies and doesn’t
have a centralized mechanism for error handling, notification, high
availability (HA), and disaster recovery (DR).

Implementation Guide: Success by Design: Integrate with other solutions 423

 Be sure to continue using the cloud-based strategy and the toolset
provided by Azure and Power Platform to integrate to on-premises
components when necessary.

Middleware
Integration middleware is software or services that enable communication
Middleware provides specialized and data management for distributed applications. Middleware often
capabilities to enable communication,
transformation, connectivity, orchestration,
provides messaging services on technologies such as SOAP, REST, and
and other messaging-related functionality. JSON. Some middleware offers queues, transaction management,
Dynamics 365 Finance and Dynamics 365
Supply Chain Management are business monitoring, and error logging and handling. Different middleware
applications that provide integration
platforms can support on-premises, cloud-based, or hybrid scenarios.
capabilities to support interfaces, but they
are not designed to replace middleware. The following are characteristics you should consider when choosing a
middleware platform.

Key characteristics
When deciding whether to integrate to an existing system, you should
consider several important characteristics of the solution, the system,
and its context.
▪ Scalability and performance  The planned platform, middleware,
Business Applications and Power Platform
and supporting architecture must be able to handle your
allow users to design, configure, and organization’s expected persistent and peak transaction
customize an application to meet a
customer’s business needs. In doing so, it’s volumes in the present, the short term, and the long term.
important to consider performance and
▪ Security  Authentication defines how each system confirms a user’s
scalability. Ensuring a performant system is
a shared responsibility among customers, identity, and you should consider how that will work across
partners, ISV providers, and Microsoft.
systems and with middleware. Authorization specifies how each
system grants or denies access to endpoints, business logic, and
data. It’s important to ensure that an integration platform and
middleware are compatible with system security and fit into the
solution architecture landscape.
▪ Reliable messaging  Middleware typically provides messaging
services. It’s important that the messaging platform supports the
architecture of the integrated systems and provides a reliable
mechanism or technology to ensure that messaging across systems
is accurately sent, acknowledged, received, and confirmed. This is
especially important in situations in which a system or part of
the supporting architecture is unavailable. Error-handling concepts

Implementation Guide: Success by Design: Integrate with other solutions 424

 related to messaging, such as idempotency and transient versus
persistent errors, are discussed in the “Mind the errors” section
later in the chapter.
▪ HA and DR  The middleware platform supports the expected
level of stability and availability across the interface in line with the
cross-system requirements, for example, HA for mission-critical
scenarios. Requirements for DR is another consideration for the
platform and middleware. For example, if the integrated systems
or the supporting infrastructure experiences an outage, it’s
important to consider what might happen to shared or
synchronized data when the outage is resolved or if either system
has to roll back to the latest backup.
▪ Monitoring and alerting  The platform, middleware, and
supporting infrastructure must support the requirements for
monitoring activity and alerting users and administrators if
there’s a problem. In addition to ensuring that the platform can
manage adequate levels and venues of alerts and notifications, it’s
equally important to consider who in your organization is responsible
for responding to those alerts and notifications and whether the
necessary workload is feasible.
▪ Diagnostics and audit logs  It’s important to also consider the
requirements for monitoring cross-system activity and for audit
trails and logging. You should consider whether the planned
platform and middleware support that.
▪ Extensibility and maintenance  It’s important to consider
the impact of additional business requirements in the long term.
Considerations include the following: the level of effort required to
extend the integration beyond the first implementation; whether the
integration requires expertise in multiple technologies and changes
to both systems or whether there is a single low-code/no-code point
of extensibility; how and how often updates are applied; and what
level of technical and functional support is available.
▪ ALM  To keep costs as low as possible, consider how you can ensure
effective lifecycle management, version control, and documentation
of the integrated solution. You should verify that all components
and parts of the solution can use ALM tools such as Azure
DevOps. For more information about ALM, refer to Chapter 11,
“Application lifecycle management.”

Implementation Guide: Success by Design: Integrate with other solutions 425

 Data consistency across multiple systems is important, and for that to occur
you need to ensure data quality. Keep in mind the saying that “if garbage
comes in, garbage comes out”—it’s important to verify that your data
flows correctly and in a timely manner across all the integrated systems.

The following section discusses the preferred platform to use to integrate

with Business Applications on Azure.

Power Platform and Dataverse

Power Platform enables integration between your organization’s systems.
It also provides opportunities for your organization to use low-code/no-
code development models as well as more complex developer platforms
to integrate solutions. In Business Applications, this avenue for integration
and much more becomes available when you use Microsoft Dataverse.

Dataverse provides a data foundation for applications in Dynamics

365 and Power Platform. When your solution components live in
Dataverse, you can natively connect and extend the solution using
Power BI, Power Apps, Power Automate, and Power Virtual Agents,
Fig. as shown in Figure 16-7.
16-7

Microsoft Power Platform

The low-code platform that spans Office 365, Azure, Dynamics 365, and standalone applications

Power BI Power Apps Power Automate Power Virtual Agent

Business analytics Application development Process automation Intelligent virtual agents

Dataverse

Implementation Guide: Success by Design: Integrate with other solutions 426

 The primary benefits of using an integration pattern based on
Dataverse are as follows.

Works with any type of app

▪ Dataverse works with a broad set of data sources through
connectors or Microsoft Power Query.
▪ Dataverse has built-in connectors through Power Automate that
offer configurable, deep integration to not only Microsoft cloud
services such as SaaS, Business Applications, and Azure, but also to
popular applications such as Adobe, Mailchimp, and Twitter.
▪ Dataverse provides a modern REST-based API, a vast developer
toolkit, and a growing list of samples that link two or more
connectors, which means that you often don’t need to start
your integration project from a blank slate.

Works with any kind of data

▪ Dataverse is integrated into data-centric tools and services
such as Microsoft Excel, Outlook, Power BI, Power Query, and
AI Builder that are traditionally used by knowledge workers and
integration engineers.
▪ The Dataverse fabric has built-in analytics, reporting, and AI that
you can use to provide insights into your organization and support
decision making. You can obtain analytics and reporting data
using low-code/no-code options or by using the full capability set
of Azure Data Factory, Power BI, and Azure Databricks. For more
information about analytics and reporting, refer to Chapter 13,
“Business intelligence, reporting, and analytics.”

Let’s take a look now at how you can use Power Platform and
Dataverse with Business Applications.

Dynamics 365 Customer Engagement apps

Customer Engagement apps data is securely stored and managed
in Dataverse. This means that the Customer Engagement apps can
natively use Dataverse capabilities to integrate to other applications.

Implementation Guide: Success by Design: Integrate with other solutions 427

 Integrate Finance and Supply Chain
Management apps and Dataverse
A data warehouse provides capabilities
Finance and Supply Chain Management apps are built on a specialized
to store, process, aggregate, analyze,
data mine, and report on both current cloud SQL schema, optimized for online transactional processing (OLTP).
and historical data sets typically with
data aggregated from a varied set of cloud
However, you can integrate into Dataverse through several design patterns,
and line-of-business (LOB) applications. A some of which are seamless and synchronous or near real time. When you
data warehouse has specific architecture
optimized for analytical functions. An connect Dataverse and Finance and Supply Chain Management apps,
Enterprise Resource Planning (ERP) system you’re using a rich set of predefined denormalized endpoints in Finance and
such as Dynamics 365 for Finance and
Supply Chain Management maintains a Supply Chain Management called data entities. For more information about
highly normalized data set optimized for
how to connect Finance and Supply Chain Management apps to Dataverse,
transactional processing and business
functionality. It’s important to select the refer to the “Integrate business applications” section later in this chapter.
appropriate data warehouse platform to
meet your organization’s data warehouse
and analytical needs.

For more information on reporting and

analytics solutions, refer to Chapter 13,
Choosing a design
“Business intelligence, reporting,
and analytics.” Many factors influence the choice of patterns for your integration.
Integration scenarios can be grouped roughly into three kinds: UI,
data, and process integration. While there is some degree of gray area
Defining business goals and overlap between them, there are also some distinct characteristics
Choosing a platform and expected behaviors.

Choosing
a design UI integration
Choosing a pattern In UI integration, the primary point of integration is centered around an

Challenges in integration
action that’s performed on the UI. The integration might or might not
trigger business logic or cause anything to be written to the system. UI
Product-specific guidance integration creates a seamless user experience even though the data
and process might exist in separate systems, as shown in the example in
Figure 16-8.

In this figure, an order taker at a company works in the Sales system

to take an order from a customer, and in that process the order taker
looks at real-time information from an ERP system.

The UI integration allows the order taker to answer the customer’s

inquiries accurately and immediately without switching systems. The
information might or might not be used for the order-taking process.

Implementation Guide: Success by Design: Integrate with other solutions 428

 Fig.
The simplest form of UI integration is when a
16-8
widget, canvas app, or view from one system is
embedded into the UI of another system
System 1 System 2
Sales Enterprise without ever touching the business logic or data
Resource Planning of that system. This enables users to focus on their
tasks—for example, scheduling interviews with job
Create quote applicants or providing services to customers—
without the need to switch between several LOB
Enter item
systems. Sometimes values can be passed from the
Look up price hosting system to the hosted system to provide
Get price
Return price
session context from one system to the other.

The following are additional examples of UI integration:

Look up availability
Get date ▪ Commodity price ticker; an email view
Return price
▪ On-hand inventory
Finalize quote
▪ Embedded Power BI dashboard showing the
current day’s production quality issues in a
Send quote
manufacturing plant

Data integration
Data integration, shown in Figure 16-9, is integration between systems
that takes place at the data layer, and data is exchanged or shared
between systems. Data integration is different from process integration
in that both systems work with a representation of the same data,
Fig.
whereas in process integration the process starts in one system and
16-9
continues in the other system.

We use data integration patterns in the

following scenarios.
System 1 System 2
▪ Master or transactional data that isn’t the
central part of a single, continuous business

UI Layer UI Layer process is synchronized between a process in

one system to a process in another system.
Process / Process / ▪ Data is shared or exchanged between systems
Business logic Business logic when needed for calculations, classifications,
or references.
Data integration
▪ Data is shared or exchanged between systems

Implementation Guide: Success by Design: Integrate with other solutions 429

 so that actions that happen in one system are reflected in the
other system.
▪ Aggregate data from a system with a detailed level of data is
A key characteristic of UI integration
exchanged to a system with a higher level representation of data.
design is that it’s embedded.

The benefit of UI integration is that a

user can retrieve and provide real-time
The following are examples of data integration:
information from multiple sources without ▪ Aggregate marketing data assets
switching between systems, thereby saving
on training, user licenses, and more ▪ Bank account reconciliation
importantly, time when interacting ▪ Order status updates
with customers.
▪ Exchange rate updates
▪ Postal code and city list synchronization
▪ Accounts and customer data synchronization

Keep in mind that often data doesn’t originate from a system within
your own organization; it can come from an external source to upload.
Conversely, data might be extracted from your system to be sent to an
auditor, a regulatory body, or an industry data exchange.

In the example in Figure 16-10, orders are synchronized between

Sales and an ERP system. When the order is created in Sales, the
ERP system sends regular updates of the order data to Sales, which
Fig.
enables users to find accurate order status information.
16-10

Data integration is also a useful integration type because

System 1 System 2
Sales Enterprise multiple features within Business Applications enable
Resource Planning
administrators or users to configure many of these
Create order integration points out of the box, for example,
Create order
Microsoft Teams, journal uploads for expenses or bank
Ship order
account statements, exchange rate synchronization, and
Update status features for regulatory reporting and extracts.

Look up status
Whenorder
Invoice designing data integration, you should consider which system
Update status will be the system of record, or owner, of the information. There are
scenarios in which this is clear cut, for example, uploading worker data
from Dynamics 365 Human Resources into Finance and Supply Chain
Order paid

Update status Management apps (the former is the system of record). But there are also
scenarios in which each system owns a separate part of the overall entity,
Look up status for example, the integration of Accounts to Customers between Sales

Implementation Guide: Success by Design: Integrate with other solutions 430

 and Finance and Supply Chain Management. In this scenario, Sales
is the owner of the basic information, but that information is just a
small subset of the fields in the customer entity in Finance and Supply
Chain Management.

Another type of data integration, shown in Figure 16-11, is when two

systems share the data layer, so updates to one system are instantly
reflected in the other system. This is possible with Business Applications
Fig. in Dataverse.
16-11

First-party model-driven applications such as Sales,

System 1 System 2 Dynamics 365 Field Service, and custom Power Apps
Sales Enterprise all share the same data layer, even if applications are
Resource Planning
built and deployed separately.

Create order With Dataverse it’s also possible to integrate

Customer Engagement and Finance and Supply Chain
Ship order Management Apps so that a similar seamless experience
can exist between Sales and the ERP system.

Look up status
Systems may have separate processes and even different
business logic, but if the underlying data layer is the
Invoice order same, the need for transfers of data, synchronization,
and programming to transform the data is completely
Look up status eliminated. This kind of integration is now possible
because the development of shared data stores such as
Dataverse set standards for how a certain kind of data
is defined in the data layer.

Process integration
Process integration refers to when a business process is designed to span
A key characteristic of process integration multiple systems. There are many varieties of process integrations, such
design is that it’s event driven. as from a plan-to-produce workflow in which production forecasting or
The benefits of process integration are block scheduling occurs in a Production Scheduling System and the rest of
accuracy, efficient timing of information
the manufacturing and supply chain management process (production,
and activities in the organization, and
reduction of manual errors. order management, and fulfillment) and the billing process occur in an
ERP system. Figure 16-12 shows this type of integration.

Implementation Guide: Success by Design: Integrate with other solutions 431

 Fig.
In this scenario, the business process spans two
16-12
systems. The user in the Production Scheduling
System creates and maintains the production
System 1 System 2
Production Enterprise schedule, thereby initiating the business logic to
scheduling Resource Planning
procure, produce, and reserve the order in the ERP
system for later picking and shipping. User interaction
Plan production orders
might be required in the ERP system, but the point is
Firm production order Create order that the event of creating the order in the Production
Create
production Scheduling System triggers certain business logic in
order Procure raw materials
the ERP system, and the rest of the business process
Update Consume raw materials steps are handled either automatically or by the
production
order appropriate users in the ERP system.
Update production order Update order

Report finished Process integration can span multiple systems and

often requires batched, scheduled, and sometimes
Pick, pack, and ship order
near real-time integration.

In the example shown in Figure 16-12, without integration, orders

would have to be manually created twice, increasing time and risking
errors such as typos and missed updates.

Design options on Power Platform

Let’s now discuss some of the common services available across
Power Platform and Azure that can help in designing a dynamic,
scalable solution.

With Dataverse, you can leverage Power Platform to integrate apps. You
should consider several components and features of Azure and Power
Platform when designing an overall solution as well as individual integration.

Power Automate
Power Automate provides low-code/no-code solutions to help you
automate workflows and build integration between various applications.
Power Automate automates repetitive and manual tasks and seamlessly
integrates business applications inside and outside Power Platform.

Implementation Guide: Success by Design: Integrate with other solutions 432

 In fact, Power Automate has connectors to more than 300 specific
applications and technologies and a vast selection of integration
samples to start from, all with the goal of helping users to be more
productive. Power Automate is available to Office 365 users who have
the appropriate license and access. For more information, read about
Power Automate and “What is Microsoft Dataverse?”

Azure integration frameworks

Azure provides designers with a rich, scalable feature set and
user-friendly design elements as well as the option to build more
complex integration solutions. Developers can use tools including
Visual Studio, Visual Studio Code, and Azure DevOps.

Power Platform and Azure together can handle integration requirements

from simple scenarios to enterprise architecture scenarios.

For more information, visit “Azure Logic Apps” and “Azure

Integration Services.”

Choosing a pattern
Defining business goals Choosing the right pattern is a critical part of successfully implementing
integration between systems. When choosing an integration pattern, you
Choosing a platform
should consider factors such as what its main functionality is and how it’s
Choosing a design built, including platform, language, user interface, and the connectivity
type it handles. We recommend that you also consider what type of
Choosing a
actions you need the integration to perform, such as the following:
pattern
▪ Data types and formats  What types of data are you sending—
Challenges in integration
transactional, text, HTML?
Product-specific guidance ▪ Data availability  When do you want the data to be ready, from
source to target? Is it needed in real time, or do you just need to
collect all the data at the end of the day and send it in a scheduled
batch to its target?
▪ Service protection and throttling  When you use certain
integration patterns, service protection might be built in so
that there’s a maximum number of records allowed because the
performance decreases with quantity. Sometimes the provider

Implementation Guide: Success by Design: Integrate with other solutions 433

 places this limitation on you, but in other cases, if you’re offering
integration points to other systems, you want to impose a lim-
itation or throttling to ensure that the general performance of your
system is not impacted by traffic from external requests.
▪ Transformation of data  Do you want to convert or aggregate
transactional data into analytical data? Are you changing .txt
to HTML?
▪ Triggers  What action will trigger sending data from the source
to the target?
▪ Trigger actions  Do you want specific actions to be automated
after the data arrives at the target?
▪ Process error handling  What type of monitoring will you put in
place to detect any issues with the interfaces? What type of alerts
do you want to handle the patterns that you’ll be using?
▪ Flow direction  Is the flow bidirectional or unidirectional?
▪ UI  The UI determines how the data is going to be displayed once
it arrives at the target system. The UI determines the user
experience—how the user will interact with and handle
the information.
▪ Scalability  How will the interface pattern handle the expected
transaction volumes in the present, short term, and long term?
What would happen if you exceeded those expectations?

Pattern directions
Let’s take a closer look at some common patterns for individual
integrations and the pros and cons for each. This table is generalized; for more
information, refer to the “Product-specific guidance” section in this chapter.

Pattern Mechanism Trigger Considerations Use when

Real time or Data is exchanged User action or Pros: Fast request and response Generally, not a best
synchronous synchronously, invoking system event. round trip. Real-time values and practice to use because of
actions via services. information. the risk of processes getting

Cons: Small payloads. Action- stuck and solutions that

reaction can cause tight coupling of are fragile.

systems. Risk of ripple effect from Use when real-time

minor outages. Sensitive to latency. information is critical.

Implementation Guide: Success by Design: Integrate with other solutions 434

 Pattern Mechanism Trigger Considerations Use when
Asynchronous Data is exchanged Scheduled and user Pros: Loose coupling of systems Most recommended
unattended on a initiated. Can wait for off makes the solution robust. Load integration patterns and
periodic schedule or as a hours or idle time. balancing over time and resources. technologies supported are
trickle feed using Can be very close to real time. asynchronous, although
messaging patterns. Timely error handling. they can be near real time.

Cons: Delay in response and

visibility to changes across systems.

Push One system puts Originating system user or Pros: If technical expertise lies For reactive scenarios.
(pushes) data into system event. within the pushing system, the
Receiving system provides
another. Information custom effort lies here. Good for
a turnkey API and
flows from the originator reactive scenarios.
organization’s developer
to the receiver.
Cons: Pushing system might not skillset is with the
have visibility into availability and originating system.
load and idle times in the
receiving system.

Pull Receiving system requests Receiving system request Pros: If technical expertise lies For proactive scenarios.
data from the originator— based on schedule. within the pulling system, the
We might not have the
a subtle but significant custom effort lies here. Good for
option to add triggers or
difference from the proactive scenarios. Clear visibility
events in the originating
Push pattern. into availability and load and idle
system. Originating system
times in the receiving system.
provides a turnkey API and
Cons: Originating system might not organization’s developer
have the APIs needed to pull from. skillset is with the
receiving system.

One-way sync Data from one system is Data state, system, or Pros: Establishes a clear system of One system is owner or
synchronized to another by user event. record. Simple conflict resolution. system of record and
one or more trigger events. other systems consume
Cons: Sometimes the receiving
system or non–system of record that data.

doesn’t have built-in capability to Consider scenarios in which

make the data read only, which can the originating table is a
confuse users. subset of a similar table in
the target.

Bidirectional sync Data from two or more Data state, system, Pros: Data is kept in sync across For dual-write
systems are synchronized. or user event. applications. Acquired divisions on integration patterns.
multiple platforms can continue to
Scenarios in which there
use their existing systems. Users can
isn’t a clear system
use their system to make changes.
of record.
Cons: Complex conflict
Data from best-of-
resolution. Redundant data is
breed systems should be
replicated for each system.
available in Dataverse for
Synchronized data might be a
Power Apps and other
subset of data in systems. The rest
tools and services.
must be automatically given values
or manually updated later.
Implementation Guide: Success by Design: Integrate with other solutions 435
 Pattern Mechanism Trigger Considerations Use when
Aggregation Data from a specialized Any. Pros: Detailed data is kept in the Aggregates are needed for
system is integrated to system where it’s used. Aggregation calculating or processing,
another system on an can derive a small dataset from a for example, on-hand
aggregated level for large one, thus limiting traffic inventory by warehouse,
processing or reporting. across platforms. revenue by invoice header,

Cons: Users often expect to be able or revenue by customer by

to drill down to the detailed level. day, or operational data

While this could be done with for posting in a finance

embedding, it does require system.

additional integration complexity or

users operating in two systems.

Embedding Information from one User event. Pros: Simple because the data A mix of information from
system is seamlessly remains in the originating system. first-party applications (for
integrated into the UI of example, Bing, Power BI,
Cons: Difficult to use the data for
another system. and Exchange), third-party
calculations for processing.
components, canvas apps,
or other information
embedded in the UI of
an application.

This is common practice in

Customer Engagement.

Batching Batching is the practice of Any. Pros: Great for use with messaging Whenever it isn’t
gathering and transporting services and other asynchronous necessary to transmit
a set of messages or integration patterns. Fewer individual records.
records in a batch to limit individual packages and less
chatter and overhead. message traffic.

Cons: Data freshness is lower. Load

in the receiving system can be
affected if business logic is executed
on message arrival.

Implementation Guide: Success by Design: Integrate with other solutions 436

 Additional pattern considerations
When choosing a specific pattern, it’s important to take a broader look
Let’s take a moment to discuss
at the integration context.
synchronous integration patterns and
near real-time patterns because these are

Messaging patterns, middleware,

often mischaracterized and mistakenly
used interchangeably.

An example of a synchronous integration and service bus

pattern is a process in which a user performs
a web service–based lookup of current
A service bus refers to asynchronous messaging on a well-equipped
on-hand values from one system to another. messaging platform that includes a service-based channel and
In this example, the user waits for a response
before continuing the process, thus making endpoint connectors. This is a common, modern way of managing
the process both latency and availability complex integration architecture across distributed systems. The
dependent. This is known as coupling and it
should generally be avoided. messaging platform is often referred to as the middleware, and it often
offers message routing and message orchestration, transformation,
A near real-time pattern for the same
design scenario provides a user’s system and even in some cases, load balancing message error handling,
with fresh on-hand inventory information
every one to three minutes, thereby
monitoring, logging, and notification services.
enabling the user to continue the process
even if there’s a transient outage in the
providing system or the supporting Messages are pieces of information or datasets that can be transferred
architecture. This pattern combined with
between systems. Each message can contain instructions such as an
using a messaging-based integration
architecture results in a robust solution. event fired or a delete request, or it can contain a payload like a record
or dataset. Some messaging platforms allow for prioritizing messages
based on type, origin, destination, or payload.

Modern messaging platforms are based on service endpoints, REST,

and JSON, but the platforms can offer many additional features.
Connections between endpoints are called channels and the message
paths are called routing.

The benefit of messaging patterns is that the messaging platform

handles the flow of information between systems and detects
individual systems or SaaS service endpoint transient outages and
handles the queueing of messages to the affected systems until
they’re back online. Figure 16-13 shows the typical components of
a messaging platform.

The asynchronous nature of the messaging patterns allows for a robust,

decoupled integration architecture with less manual intervention and
lower maintenance costs.

Implementation Guide: Success by Design: Integrate with other solutions 437

 Fig.
16-13
Monitoring Orchestration
and alerts and routing

Endpoint Channel Routing Channel Endpoint

Message

Data logging

When implementing a messaging pattern, consider the following:

▪ Frequency and volume  It’s critical to have a clear picture of the
frequency and transactional volume of data that will go through
the interface, how that load is distributed over time, and what the
requirements for these are in the longer term.
▪ Idempotency and out-of-order messages  Mechanisms are
available to ensure that the messages are idempotent, for example,
using a message ID to check whether the message is a duplicate or
an out-of-order message. Services such as Azure Service Bus support
this capability and allow you to implement it without the need for
additional code.
▪ Service-level agreement (SLA) or end-to-end cycle latency 
Consider whether there are minimum requirements for how fresh the
data in the receiving system must be. For example, orders taken on an
e-commerce site should reach the warehouse in less than five minutes.
▪ Triggers and actions end-to-end process  It’s also important to
look at the cross-system business process in the bigger picture:
▫ What happens before the integration steps?
▫ Is there potential risk of introducing semantic errors earlier in
the process?
▫ What event triggers the integration?
▫ Does the integration trigger action in the receiving system?
▫ What business process steps happen after the integration steps
are done?
▫ If an error is identified, logged, and notified, how is the
problem corrected? How often might errors occur, how long

Implementation Guide: Success by Design: Integrate with other solutions 438

 might it take to correct them, and who is responsible for
making the corrections?
▪ Batching  Batching messages or datasets enables less frequent
communication, or “chatter.” But it also typically makes messages
and payloads bigger.
▫ Do the requirements support a batched approach in which
datasets are consolidated into each message, or does the
integration require individual records or rows?
▫ If a batched approach is suitable, how many rows and records
are in each message?
▫ Are there service protection volume limits to consider?
▪ Topology and architecture  Topology and architecture are
important considerations for the short and long term as your
organization grows internationally or as you migrate systems to
the cloud and need to address new challenges and considerations.
▫ Are there integration components that will span the cloud and
on-premises boundary? Does the messaging service handle that?
▫ Where in the world is the data located, and does the inherent
latency between geographic regions support the requirements?
▫ Does any data exist in countries or regions of the world that
have local privacy policies and regulations?
▫ Are there service protection limitations in technology—protocol
or platform dependencies in any of the integrated systems?
▫ Does the data exchanged require transformation, reformatting,
or remapping as part of the integration?

For more information, read about “Messaging services on Azure.”

We highly recommend that when you design integration, you don’t

assume that everything will always work as intended. In fact, when
working with integration, there are many external factors that we often
cannot completely control. Therefore, we need to make sure that our
integrations are robust by design, which we discuss in the next section.

Mind the errors

Designing for failure is a crucial component of architecting any solution,
including integration components. As part of the integration architecture,

439
 it’s important to incorporate the following:
▪ Error logging  Select the platform (such as a log file or database) to
log the errors.
▪ Error monitoring and notifications
▫ Define the process and technical approach to monitor errors.
▫ Select an error notification approach, which is the process to
notify administrators and other stakeholders if there’s an error.
Business-critical errors might need a different notification
approach than non-critical errors.
▪ Business continuity  It’s important to plan for business
continuity with minimal disruption to business in the event of errors.
For more information, see Chapter 20, “Service the solution.”

The more integration touchpoints there are, the greater the potential for
errors. Therefore, error handling also needs to be planned in accordance
with the integration scenario. For example, in the scenario of a synchronous
integration pattern, an error might require a complete rollback of the
entire process, whereas in an asynchronous data integration scenario, it
might be acceptable to fix a data issue just by notifying the administrator.
(See Figure 16-14.)

Let’s now discuss error management for two key patterns of synchronous
and asynchronous integration.

Fig.
16-14
As a reminder, in a synchronous integration pattern, each step is
dependent on the completion of a preceding
System 1 System 2 step. In the event of an error, consider a retrial or
Quote Enterprise rollback—which option to use is dependent on
Resource Planning
whether the error is transient or persistent. In the
scenario of a transient error, the normal behavior of
Create quote
the application will resume after a few retries. Note
Enter item that retrial limits should be predefined to avoid a
situation in which all resources are blocked. Once
Look up price
Get price the retrial limit has been crossed, the entire process
will need to be rolled back, and appropriate error
?
Return price
messages should be logged.

For an asynchronous integration, the error

Implementation Guide: Success by Design: Integrate with other solutions 440

 management strategy might differ depending on your business
requirements. Consider creating a queue that can continue to
validate the message being exchanged and retry until the retrial
limit is reached. The message that fails from being exchanged can
also be stored in another queue for later intervention.

Transient errors such as network timeouts get fixed after a few retries.
However, persistent errors require intervention to be fixed.

An important consideration as you build your retrial mechanisms is

to ensure that your requests are idempotent. This will prevent data
duplication issues due to retries.

The following are some of the most common error scenarios in any
integration between Business Applications and other applications:
▪ System becomes unavailable
▪ Authorization and authentication errors
▪ Errors caused by platform limits
▪ Errors caused by service protection limits applied to ensure
service levels
▫ API limits (limits allowed based on the user license type)
▫ Throttling
▫ Process errors
▪ Runtime exceptions

It’s critical to plan error logging, monitoring, and communication as

Fig.
part of an overall error-handling approach. Logging errors with the
16-15
exact timestamp and message details
System 1 System 2 helps the implementation team and
Quote Enterprise
Resource Planning application owners troubleshoot the
errors. Error tracking also enables the
Create quote
application owners to identify potential
Enter item
areas of improvement in their process by
Look up price
Get price reviewing frequent errors, if any. Figure

Return price
16-15 shows a persistent blocking error. The
Notification platform supports logging and notifications
so that an administrator can be notified and
Log
resolve the issue.

Implementation Guide: Success by Design: Integrate with other solutions 441

 It’s also important to identify owners who should receive error details
through email as part of regular monitoring. The monitoring and
notification approach might differ for business-critical errors versus
non-critical errors. In the case of a mission-critical application
integration, some organizations might choose to send text notifi-
cations to the administrators to speed up the process.

Error messages displayed to end users through the UI should be clear,

precise, and short. They should also include information users can act
on, such as point of contact information or a link to an FAQ. This makes
the user experience a more positive one and should be considered a
best practice as part of error handling.

Consider using tools such as Azure Monitor that help maximize your
applications’ availability to collect, analyze, and act on telemetry.

For more information, visit “Azure Monitor overview.”

To summarize, error logging and monitoring have a significant impact on

the following key factors and therefore are considered essential:
▪ Application health including performance and stability
▪ Ease of troubleshooting and resolving errors
▪ Minimize business impacts when errors occur
▪ Increase user satisfaction

For more information about integration patterns and technologies,

refer to the “Product-specific guidance” section later in this chapter.

Defining business goals

Choosing a platform Challenges in integration

Choosing a design The diverse applications used in organizations make up the IT backbone
of those companies. Organizations use applications that might be a
Choosing a pattern
combination of on-premises, cloud, and third parties. These applications
Challenges in need to communicate with each other for different business needs.
integration Integration challenges during a project implementation can cause delays
Product-specific guidance or cost increases. A key focus of any IT department is to ensure that these

Implementation Guide: Success by Design: Integrate with other solutions 442

 integrations enable business productivity and not become a blocker to
business growth.

Integration challenges can be categorized into the following areas.

Business
Each integration scenario has a direct impact on the application you’re
integrating with and on its users. Any downstream applications might also
be indirectly impacted. For integration to be implemented in a successful
manner, you should address the following during the Initiate stage:
▪ Identification of application owners and stakeholders
Application owners need to identify downstream applications that
might be impacted in an integration. However, a common scenario
is to bring in these owners after the planning is complete. This often
results in misaligned timelines and scope and in turn creates project
delays and poor quality. Integration design needs to take every
impacted application into consideration.
▪ Alignment between business owners  Business stakeholders
have different needs for their organization’s various applications.
Unless there is a collective understanding of integration scenar-
ios and approaches, requirements might be mismatched among
the various owners. This in turn often results in delayed timelines,
cost overruns, and a lack of accountability. System integrators
should consider the following:
▫ Identify the key owners and bring them together to walk
through the scenarios.
▫ Differentiate between process, data, and UI integration to
simplify and streamline the integration scope.
▫ Outline the impact on business groups affected by the integration.
▫ Highlight issues and risks in the absence of following a
consistent approach.
A transparent conversation enables business stakeholders to
understand the underlying risks and benefits and thus build a
common perspective.
▪ Ambiguous and unrealistic expectations  Integration
requirements can sometimes be ambiguous or incorrectly

Implementation Guide: Success by Design: Integrate with other solutions 443

 perceived as mandatory. Common examples of the latter are
unnecessary synchronous integrations and replicating high
transactional data volumes into your business application to
make it a reporting application. Such architecture decisions
can result in solutions that are neither scalable nor performant.
Encouraging and facilitating conversations to help business
stakeholders define their requirements clearly and understand the
necessary tradeoffs as part of the architecture decisions is a critical
step in ensuring a smooth implementation.
▪ Lack of business continuity planning  This is an aspect of
implementing any IT project that most often is ignored. However,
no matter the type of integration—cloud to cloud, on-premises to
cloud —every business group must define operating procedures
that will be enforced as a part of any availability and disaster
recovery situation. As an example, consider a scenario in which the
on-premises application of a Dynamics 365 business application is
temporarily unavailable. This can impact integration components,
resulting in data or process issues.
▪ IT-driven requirements and design  Implementation teams
sometimes consider integration between systems as a primarily
technical exercise and overlook the critical role played by business
teams. However, successful integrations between business
applications often depend completely on a well-defined
cross-system business process, and if the specific process details
and broader requirements are not properly defined, the
integration project might take much longer or go through
several iterations of trial and error. For more information,
refer to Chapter 7, “Process-focused solution.”

Technology
Most enterprises have legacy applications with traditional, on-premises
architecture. The move to cloud applications requires consideration
of the patterns they support and the best practices when planning for
integration. A low-code/no-code pattern should now be at the forefront
of any integration architecture. Engaging in conversations about not
just the current setup but also about future planning for performance,
extensibility, and maintenance plays a key role in choosing the right

444
 technology. When choosing the appropriate technology, consider
the following.
▪ Does one size truly fit all?  Many enterprises have an enterprise
architecture approach that might or might not be aligned with
the modern approaches for cloud applications. Prior to investing
in a specific approach, evaluate whether the existing architecture
aligns with cloud patterns. Sometimes, a generic approach is
taken—this can result in inefficiencies in integration, unscalable
architecture, and poor user experience and adoption. Therefore,
it’s crucial to consider design paradigms such as the following:
▫ Definition of the integration approach based on
multiple parameters
▫ Benefit of a proof of concept to determine the pros and cons
of one approach over another
▫ Synchronous versus asynchronous integration
▫ Process, UI, and data integration
▫ Single record or batch
▫ Frequency and direction of the integration
▫ Message reliability and speed
▫ Data volume
▫ Time expectations (some scenarios require a batch integration
to be completed during a specific time window)
▫ Error management and retrials
▪ Will sensitive data be exposed?  System integrators must
understand IT and customer concerns around security, especially in
the context of integrating on-premises applications with Business
Applications. Categorizing security concerns as follows aids in
identifying who and what is required to help address them:
▫ Access control
▫ Data protection
▫ Compliance and regulatory requirements
▫ Transparency
For more information, refer to Chapter 12, “Security.”
▪ Storage costs and platform limits  To ensure service quality
and availability, Business Applications and Power Platform
enforces entitlement limits. These limits help protect service
quality and performance from interference by noisy behavior
that can create disruptions. System integrators must incorporate

Implementation Guide: Success by Design: Integrate with other solutions 445

 these limits as part of the overall architecture. If these aren’t
planned for, the service will be throttled, resulting in failure and
errors within the integration layer. Storage costs are also often
ignored. Although this might not have an impact initially, over
time, it can result in increased storage costs and therefore
should be planned for appropriately.
▪ Connectivity  Network latency can become a constraint, especially
in heavy data-load scenarios. System integrators must ensure that
they design payloads accordingly for the efficient use of network
resources without compromising performance.
▪ Anti-patterns  Implementation architects should follow the best
practices for Business Applications. Sometimes these architects don’t
take cloud patterns sufficiently into account in integration scenarios
with on-premises applications, resulting in poor performance. The
behaviors leading to such situations are referred to as anti-patterns.
Consider the following common anti-patterns:
▫ Are there repeated connections between on-premises
components and Business Applications that impact
performance? If so, consider sending data in batches.
▫ Is there latency between a customer’s on-premises applications
and a Dynamics 365 datacenter? If so, consider using a cloud
service such as Power Automate, Azure Functions, or Azure SQL
to reduce the latency impact.
▫ Is a lot of data being synchronized with Business Applications
for reporting purposes? Keep in mind that the database under
Business Applications isn’t intended to be a data warehouse for
all of the organization’s data assets. Consider using a dedicated
datastore for reporting purposes.
▪ Proprietary technology  Customers might be using
third-party technology within their IT landscape that doesn’t
provide interface details or adequate support to enable integration
easily. Often such issues are identified either toward the end of
design or during the development stage. This causes delays in the
project timeline, burdening the customer with time constraints
to mitigate such risks. System integrators must highlight such
dependencies in the planning stage to ensure adequate support
or an alternate approach.
▪ Readiness  With the increasing pace of transformations in the

446
 technology world, architects sometime choose an approach due
more to its familiarity than its applicability. Customers and system
integrators must evaluate whether to request additional resources
specialized in the specific technology who will be a better fit for
their current and future needs.

Project governance
The initial stage of a project should include a defined project
governance model. Integration between on-premises and Business
Applications can range from simple to complex, and the lack of
well-defined project governance areas results in gaps and issues in the
smooth implementation of a project. Following are common project
governance concerns specifically for the integration components:
▪ Has the impact of the integrations been identified for the end user,
process, and reporting? This might require planning for change
management activities, including communication and training.
▪ Making a solution performant should be at the forefront of
any design decisions made by the implementation team. This
applies equally to the integration layer and the integration layer.
Is performance testing planned and does it cover integration
components? Performance testing is another activity that tends to
be considered optional. However, architects and project managers
must consider embedding this in their Business Applications
implementations. This will help identify any performance
bottlenecks prior to deployment for end users.
▪ Are development and test environments available for all applications
for thorough system integration testing? Is a plan for stub-based
testing during the unit testing phase required?

Asking these questions during the initial stages of the project enables
both the implementation partner and customer to proactively identify
and plan for any dependencies and risks.

For more information, refer to Chapter 8, “Project governance.”

Implementation Guide: Success by Design: Integrate with other solutions 447

 Product-specific
guidance
Defining business goals The following sections discuss specific integration-related considerations
for Finance and Supply Chain Management; specific technologies and
Choosing a platform
platforms for integration from Customer Engagement; and how to
Choosing a design integrate between individual business applications.

Choosing a pattern

Challenges in integration Finance and Supply Chain Management

Many ways of integrating into and from Finance and Supply Chain
Product-specific
guidance Management use an abstraction layer with integration and data migration
capabilities that shield users, administrators, and developers from the
heavily normalized database. The abstraction layer consists of data entities.

Data entities
In Finance and Supply Chain Management, a data entity encapsulates a
business concept, for example, a customer or sales order line, in a format
that makes development and integration easier. It’s a denormalized view
in which each row contains all the data from a main table and its related
tables instead of the complex view of the normalized data model behind it.

Said another way, data entities provide conceptual abstraction and

encapsulation, a denormalized view of what’s in the underlying table
schema to represent key data concepts and functionality. There are
more than 2,000 out-of-the-box data entities.
The out-of-the-box data entities are
general purpose entities built to support
a wide variety of features surrounding a
business entity. For implementations in
Data entities help encompass all the table-related information and make
which a high-volume, low-latency interface the integration, import, and export of data possible without the need to
is required, it’s recommended to build
custom data entities with the necessary pay attention to the complex normalization or business logic behind the
targeted and specific features. scenes. For more information, see the “Data entities overview.”
You can replicate an out-of-the-box data en-
tity and remove any fields and logic around
Finance and Supply Chain Management integration patterns
features not used by your organization to
create a highly performant interface. The following table is a list of integration patterns with pattern
descriptions, pros and cons, and use cases.

Implementation Guide: Success by Design: Integrate with other solutions 448

 Pattern Mechanism Trigger Considerations Use when
OData Data is exchanged User actions and Pros: It allows developers and Use with messaging services
synchronously, invoking system events. service platforms to interact with and point-to-point integration.
actions using RESTful data by using RESTful web services. Inbound (Push) or outbound
web services and the It provides a simple and uniform (Pull) of datasets from entities
HTTP protocol stack. way to share data in a discoverable when the volume and payload
All data entities are manner. It enables broad integration are relatively low.
marked public, and across products.
CRUD operations are Cons: The pattern is synchronous
mapped. and subject to Service Protection
limitations and throttling. Cold start
might be slow.

Custom Services— A developer can create User actions and Pros: Easy for developers to add Used when invoking an action or
SOAP, REST, and external web services by system events. and expose service endpoints to use update, for example, invoicing a
JSON extending the application with integration platforms. sales order or returning a specific
with X++. Endpoints are value. We recommend using
Cons: Requires ALM and SDLC for
deployed for SOAP, REST, REST Custom Services in general
coding and deployment of
and JSON. because it’s optimized for
extensions into Finance and Supply
Chain Management. The payload is the web.

low compared to other patterns. REST is preferred for

high-volume integrations
because there’s reduced
overhead compared to other
stacks such as OData.

Consuming A developer can consume Scheduled and user Pros: Easy for developers to add Use when consuming services
web services external web services by initiated. Can wait for and expose service endpoints to use from other SaaS platforms or
adding a reference in X++. off hours or idle time. with integration platforms. products, for example, com-

Cons: Requires ALM and SDLC modity tickers and lookups of

for coding and deployment of real-time values.

references into Finance and Supply The recommended pattern is

Chain Management. The payload is to use Power Automate instead
low. Risk of hardcoding connection when possible.
strings and credentials in related
code. Maintenance workloads and
risks when the service is updated.

Data The REST API helps Originating system Pros: On-premises and large- Large-volume integrations.
Management integrate by using data users or system events. volume support. The only interface Scheduling and
Framework REST packages. that supports change tracking. transformations happen
Package API
Cons: Supports only data packages. outside Finance and Supply
(asynchronous, Chain Management.
batched, cloud, or
on-premises)

Implementation Guide: Success by Design: Integrate with other solutions 449

 Pattern Mechanism Trigger Considerations Use when
Recurring Data With the Data Management Receiving system requests Pros: On-premises and large- Large-volume integrations
Management REST API, you can schedule based on the schedule. volume support. Supports files and of files and packages.
integration REST the recurring integration packages. Supports recurrence Scheduling and
API (asynchronous, of files and packages. scheduled in Finance and transformations happen
cloud, or Supports SOAP and REST. transformations (XSLT) if the file is inside Finance and Supply
on-premises) in XML. Chain Management.

Cons: None.

Electronic Reporting A tool that configures Data state, system, Pros: Data extracts and imports are Electronic reporting to
formats for incoming or user events. configurable in Finance and Supply regulatory authorities and
and outgoing electronic Chain Management. It supports similar entities.
documents in accordance several local government formats
with the legal requirements out of the box. It can be scheduled
of countries or regions. for recurrence.

Cons: Comprehensive configuration

when used for messaging purposes.

Excel and Microsoft Office Data state, system, Pros: Out-of-the-box integration Extracts for ad hoc reporting
Office integration integration capabilities or user events. (export and edit) on almost every or calculations.
enable user productivity. screen in the product.
Fast editing of column
Cons: Performance decreases with values and entry of data
the size of the dataset. from manual sources.

Import of journal records

such as general journals,
expenses, and bank
statements and similar
transactional data.

Business events Business events provide a User or system events. Pros: Provides events that can be To integrate with Azure
mechanism that lets captured by Power Automate, Logic Event Grid, Power
external systems, Power Apps, and Azure Event Grid. Automate, or Logic Apps.
Automate, and Azure To notify of events
Cons: Extensions are needed to
messaging services receive inherently driven by a
add custom events.
notifications from Finance single data entity, for
and Supply Chain example, an update of a
Management. document or a pass or fail
of a quality order.

Implementation Guide: Success by Design: Integrate with other solutions 450

 Pattern Mechanism Trigger Considerations Use when
Embedded Power Finance and Supply Users. Pros: Seamlessly integrates Whenever the required
Apps (UI) Chain Management information from Dataverse without data exists in Dataverse
support integration with integrating the backend. Opens op- and is loosely coupled
Power Apps. Canvas apps portunities for low-code/no-code with Finance and Supply
can be embedded into options directly into the Finance Chain Management.
the Finance and Supply and Supply Chain Management UI
Using an app embedded
Chain Management UI to without the need for updates and
in the UI provides a
augment the product’s compatibility.
seamless experience
functionality seamlessly
Cons: Power Apps and related for users.
with Dataverse.
artifacts aren’t deployed with the
build and must be configured in the
environment directly. Separate ALM
stories for Power Apps.
Embedded Power Seamlessly integrates Users. Pros: By using graphics and visuals Whenever reports,
BI (UI) Power BI reports, supported by Power BI to present dashboards, or visuals
dashboards, and visuals data from any source, workspac- exist in Power BI.
with information from es can provide highly visual and
Dataverse or any other interactive experiences for users
source, without integrating without leaving the Finance and
the backend. Supply Chain Management UI.

Cons: Power BI artifacts are not

deployed with the build and
must be configured within the
environment directly. Separate ALM
stories for Power BI components and
Finance and Supply Chain
Management Apps.

IFrame (UI) The Website Host control Users. Pros: Seamlessly integrates UI from When information from
enables developers to other systems or apps without loosely coupled systems
embed third-party apps integrating the backend. can be displayed within
directly into Finance and Information goes directly into the the Finance and Supply
Supply Chain Management Finance and Supply Chain Chain Management UI.
inside an IFrame. Management UI without the need The experience is
for updates or compatibility. enhanced if the external

Cons: External systems might have system supports single

separate lifecycles, and updates to sign-on (SSO) and

those might affect user experience deep linking.

with little notice.

Implementation Guide: Success by Design: Integrate with other solutions 451

 Priority-based throttling
Service protection is important for ensuring system responsiveness,
availability, and performance. In Finance and Supply Chain Management
apps, service protection is enforced by throttling. Throttling affects
OData and custom service pattern integrations only. Administrators can
configure priorities for external services (registered applications) directly
in the application so that lower priority integrations are throttled before
high-priority integrations.

For more information, read about “Priority-based throttling.”

Requests that are throttled receive a response containing a retry-after

value, indicating when the integrating application can attempt a retry.

Microsoft Dynamics Lifecycle Services (LCS) monitors throttling activity.

Customer Engagement
In this section we talk about frameworks and platforms to use when
integrating from Customer Engagement.

IFrames
IFrame is a popular approach commonly used for hosting external URL-
based applications. Consider using the Restrict cross-frame scripting
options to ensure security.

Power Apps component framework

Power Apps component framework is a modern approach for
creating rich visual components that allow enhancement of the
user interface of model-driven apps. It leverages client frameworks
such as React, and enhancements can be reused across Business
Applications. It can also pull data from an external application and
display the data in a visual form.

Canvas apps
Canvas apps is a cloud service that enables citizen developers to
easily build business apps without the need to write any code. These
apps can use connectors from other cloud services and be

Implementation Guide: Success by Design: Integrate with other solutions 452

embedded in model-driven apps to present data from other
applications on the user interface.

HTML web resources

A precursor to Power Apps component framework, HTML web
resources were commonly used to render data in a visual form,
providing designers with more flexibility. They can be used to pull
data into external applications using the available endpoints.

Dynamics 365 Channel Integration Framework

Dynamics 365 Channel Integration Framework hosts the widgets
and triggers the events for scenarios such as telephony and chat
integration. There are multiple versions of Channel Integration
Framework for different scenarios, including multi-session and single-
session needs. Channel Integration Framework also integrates channels
such as voice, chat, and messaging.

Virtual tables
Virtual tables pull data on demand from external data sources. This
approach is implemented as tables within the Dataverse layer but doesn’t
replicate the data because the data is pulled real time on demand. For
more information, read about the limitations of virtual tables.

Azure Data Factory

Azure Data Factory provides a code-free, serverless data integration
service to seamlessly orchestrate ETL (extract, transform, and load) and
ELT (extract, load, and transform) operations with disparate big data
sources. It can ingest data from Business Applications using the web
API endpoints. We recommend that you consider platform and service
protection limits as well as Azure costs.

Webhooks
Commonly used for near real-time integration scenarios, webhooks
can be invoked to call an external application upon the trigger of a
server event. When choosing between the webhooks model and the
Azure Service Bus integration, consider the following:
▪ Azure Service Bus works for high-scale processing and provides
a full queueing mechanism if Dataverse pushes many events.

453
 ▪ Webhooks enable synchronous and asynchronous steps,
whereas Azure Service Bus enables only asynchronous steps.
▪ Both webhooks and Azure Service Bus can be invoked from
Power Automate or plug-ins.

For more information, visit “Use Webhooks to create external handlers

for server events.”

Azure Functions
Azure Functions uses serverless architecture and can be used to extend
business logic, including calling external applications. It runs in Azure
and operates at scale. Azure Functions can be called through Power
Automate and Azure Logic Apps. For more information, read about
“Azure Functions” and “Using Azure Functions in Power Apps.”

Integrate business applications

In this section we look at the integration between two or more
business applications.

Some of the patterns mentioned earlier in this chapter are specific

to when we want to integrate one of the business applications into a
third-party or custom-built system.

Four options are available for integrating the Customer Engagement

apps suite and Finance and Supply Chain Management apps. Note that
we’re actually integrating Finance and Supply Chain Management into
Dataverse. These options have distinct characteristics, different scenarios,
and different pros and cons.

Virtual tables
The virtual table option enables you to connect Dataverse to Finance and
Supply Chain Management Apps entities as virtual tables that offer the
same full CRUD (create, retrieve [or read], update, delete) capabilities as the
entity endpoint in the app. A benefit is that we can access the data in Finance
and Supply Chain Management Apps in a secure and consistent way that
looks and behaves the same as any other table or construct in Dataverse. We
can also use Power Automate to connect to almost anything.

Implementation Guide: Success by Design: Integrate with other solutions 454

 Another benefit is that the data doesn’t have to live in both the
transactional database underneath Finance and Supply Chain
Management Apps and Dataverse but is still seamlessly available as
tables and rows in Dataverse.

Dual-write
The second option is dual-write integration, shown in Figure 16-16.
Because Dataverse is cloud based, this Dual-write also provides synchronous, real-time integration between
direct integration option isn’t available Finance and Supply Chain Management Apps and applications in
for on-premises implementations of
Finance and Supply Chain Management Dataverse. Dual-write even offers offline capabilities.
Apps. You can use the Data Management
REST API instead.
A common scenario for the dual-write option is one in which both
For more information, read the “Finance
and Operations virtual tables FAQ.”
Customer Engagement and Finance and Supply Chain Management
Apps are working on data that is fundamentally the same, for example,
customer and account, product and item, and projects.

The data is shared to provide seamless integrated experiences, for

Fig.
example, for product mastering, prospect to cash, and project to cash
16-16
across systems.
Dual-write
The benefit of using the dual-write approach is that the two applications
Model-driven share the same dataset, so changes in one system are seamlessly and
apps in
instantly reflected in the other one.
Dynamics 365

Another benefit is that in some cases the functionality is out of the

box and configurable with minimal effort, but it’s also extendable by a
developer. Additionally, the full set of capabilities of Power Platform is
available for the shared datasets.

For more information, read about dual-write.

Dataverse
Dynamics 365 configurable integrations
The third option is configurable integrations. These provide a starting
point for integration between specific Customer Engagement Apps
Dynamics 365 and Finance and Supply Chain Management Apps through integration
Finance and artifacts that enable integration between Customer Engagement
Operations Apps Apps in Dataverse and Finance and Supply Chain Management Apps.
Configurable integrations are built on data entity integration to

Implementation Guide: Success by Design: Integrate with other solutions 455

 Dataverse and enable an administrator to relatively quickly set up the
supported predefined integration scenarios.

The supported scenarios are as follows:

▪ Prospect to cash integration  The Prospect to cash integration
is a process and data integration between Sales and Finance and
Supply Chain Management Apps. The process integration enables
users to perform sales and marketing activities in Sales, and other
users handle order fulfillment and billing in Finance and Supply
Chain Management. For more information, read “Prospect to cash.”
▪ Field Service integration  Field Service integration adds integration
points between the Field Service App and Finance and Supply Chain
Management Apps to enable process integration for work orders and
projects. For more information, read the “Integration with Microsoft
Dynamics 365 Field Service overview.”
▪ Project Service Automation  Similarly, the configurable integra-
tions provide integration between Dynamics 365 Project Operations
and Dynamics 365 Project Service Automation by synchronizing
project essentials such as contracts, tasks, milestones, time, fees,
expense forecasts, and actuals. For more information, see the
“Project Service Automation overview.”

Other integration patterns

The fourth option for integrating the Customer Engagement apps
suite and Finance and Supply Chain Management is either Application
Connector or point-to-point OData integration patterns with data
entities. For more information, read about the “Application Connector.”

Conclusion
Implementing Business Applications is often a building block into an
organization’s larger solution landscape. In that case, the organization can
benefit from automation to gain streamlined and effective cross-system
processes and avoid manual errors.

Integration is also key in closing the digital feedback loop and making
all the organization’s data available, not just for reporting and visibility

Implementation Guide: Success by Design: Integrate with other solutions 456

 but also to make it accessible to users in the organization. With Azure,
Power Platform, and Dataverse you’re well on your cross-system
journey toward reporting, analytics, and AI-assisted forecasts and
predictive analysis so that the right information is readily available for
the right people at the right time.

With Microsoft cloud technology and Power Platform, low-code/no-code

options, and the citizen developer concept, once the systems are
integrated and the data is available in Dataverse, that journey can start
without requiring massive investments upfront.

In this chapter, we highlighted the path from business goals, high-level

architecture, design, and pattern considerations—the work that should
be done when integrating systems and processes—toward a successful
solution. We introduced integration patterns and discussed the dos and
don’ts and common challenges that could unnecessarily prolong the
start of the journey if not addressed properly, specifically about business
ownership, technology, and implementation project governance.

Finally, we referenced some of the technologies that people are using

across the globe to successfully integrate their Business Applications
into the solution landscape.

References
REST vs CRUD: Explaining REST & CRUD Operations
Messaging Integration Patterns
Open Data Protocol (OData)
Consume external web services
Custom service development
Data management package REST API
Recurring integrations
Electronic messaging
Office integration overview
Business events overview
Embed canvas apps from Power Apps
Analytical Workspaces (using Power BI Embedded)
What’s new and changed in Platform update 31 for Finance and Operations apps (January 2020)

Implementation Guide: Success by Design: Integrate with other solutions 457

 Choose a design

Checklist
Align the designs of each integration with the overall
integration architecture.

Avoid unnecessary replication and transmission of data

into the solution in integration designs.

Define business goals Clearly state the options and benefits of each of the
following: UI, data, process integration, and Dataverse.
Document and define goals and expected benefits of
integrations being implemented in a business-centric way.
Choose a pattern
Align the planned integration’s purpose with short- and
long-term organization goals. Design integrations to favor robust, asynchronous
messaging-based patterns.
Ensure the overview of the integration architecture, systems,
and integration points is clear and understandable. Align patterns used for each integration with expectations
for volumes, frequency, and service protection limitations.
Ensure that stakeholders have a shared understanding of
the purpose and scope of the integrations that are being Set realistic estimates of the operating costs for services,
implemented. platforms, and storage involved and be aware of how
scaling affects them in the future.

Choose a platform
Project governance
Ensure the organization understands the concept of
cloud versus on-premises platforms and the boundary Plan each integration for user and performance testing
between them. under realistic loads, as well as the end-to-end process
leading up to the integration, across the system
Plan to use either an integration middleware or
boundary, and after the point of integration.
messaging service.
Plan for testing the end-to-end process patterns used
Ensure the integration architecture, platform, or
for each integration in line with recommendations for
middleware supports the expectations for monitoring,
volumes, frequency, and service protection limitations.
audit, notifications, and alerts.
Have change management activities related to integrations
Ensure the integration architecture supports the expected
that reflect and support overall business goals.
level of security, availability, and disaster recovery.
Complete the impact analysis on upstream and
Ensure all components of the integration architecture
downstream processes.
support ALM and version control.

Implementation Guide: Success by Design: Integrate with other solutions 458

 Case study
Public sector infrastructure
organization learns how to
choose the right solution
for integration
A public sector infrastructure service organization was implementing
the Dynamics 365 Customer Service app and required integration
with a public-facing website, several on-premises apps, and Office
365 services such as SharePoint Online. One of the implementation
objectives was to gather insights from the data that was scattered
across different business applications.

The organization was working with an IT vendor that had several years
of experience building complex integrations using technologies such
as IBM MQ and Microsoft BizTalk.

They decided to keep a uniform integration architecture and proceeded

with a microservices-based architecture to build small, reusable
components using service endpoints to gain the following benefits:
▪ Faster and independent development, deployment, and release cycles
▪ Simplicity of architecture to allow decoupled and reusable components
▪ Fault isolation

Implementation Guide: Success by Design: Integrate with other solutions 459

 The organization chose Azure API Management to abstract their APIs
and implement a secure integration layer.

As the team started building and testing the initial components, they
identified some challenges due to the architecture:
▪ They experienced slow performance with batch-integration scenarios
because they called the services as they would have in a point-to-
point integration.
▪ They couldn’t use standard functionalities that would have been
available with out-of-the-box approaches such as SharePoint
Online integration with Power Platform.
▪ For an aggregated view, they decided to replicate all data into
the Dynamics 365 Customer Service app, which led to additional
storage costs.
▪ They encountered throttling and API limits issues, which prevented
successful completion of the integration.

At this stage, the organization decided to re-evaluate their standardized

architecture principle and redefine their integration strategy. They divided
their needs into two categories: batch and point-to-point synchronizations.
They also adopted a “Configure-before-you-customize” approach.

With a reasonable mapping of different integration needs to the

patterns, they were able to redefine a suitable integration architecture
for these components.

For point-to-point integrations, the organization used the previously

defined microservices-based architecture and incorporated the necessary
changes for aligning with Power Platform API limits.

For batch synchronizations, the organization designed the architecture

to manage the overall ETL processes with the following approach:
▪ They extracted the data from their source application by creating
SQL Server Integration Services packages and storing that data in
a staging database to perform all transformations.
▪ They hosted their staging environment in Azure SQL, within the
same datacenter as the Customer Service app.

Implementation Guide: Success by Design: Integrate with other solutions 460

 ▪ They sent that data as batch data to the Customer Service app to
reduce the number of individual connections.
▪ They considered API and throttling limits and built retries as part
of the design.

For an aggregated view of the data, they realigned their approach to

use Azure Data Lake, as Microsoft recommended. The out-of-the-box
Export to Azure Data Lake feature required simple configurations to
export data from the Customer Service app into Data Lake and didn’t
involve writing any additional code.

These changes in their overall approach provided significant benefits:

▪ Their batch synchronization performance improved significantly,
and they cut the duration of the daily sync by more than 70 percent.
▪ Their “Configure-before-you-customize” approach let them
choose out-of-the-box options to reduce development and
maintenance costs.
▪ The retry mechanism, combined with monitoring, ensured that
any failure scenarios were well handled and negative impacts on
end users were avoided.

Even though the architecture realignment reset the organization’s

timeline, the benefits outweighed the risk because they were able
to align their design with product features and the recommended
practices. The organization already had a change-management plan
in place to take care of stakeholder communication and alignment,
and to plan trainings for all affected teams in a timely manner.

Implementation Guide: Success by Design: Integrate with other solutions 461

17
Guide
A performing
solution,
beyond
infrastructure
Implementation Guide: Success by Design: A performing solution, beyond infrastructure 462
Patience is such a waste of time.

Introduction
When a new Dynamics 365 implementation is
delivered, users typically expect an improvement in
system performance.
Good performance is often assumed as a given, a default experience. The
The following are performance reality is that although Dynamics 365 products are scalable and powerful,
factors to consider when
designing solutions: various factors are involved in achieving a high-performance solution.
• Solution performance is critical for user
These include defining and agreeing to performance metrics, testing
adoption, customer experience, and throughout the phases of a project, and taking design and build consider-
project success and necessary to enable
businesses to achieve their goals. ations into account, particularly for customizations and integrations.
• Clear goals and realistic expectations
are vital to developing a solution that
performs well. In this chapter, we explore various aspects of performance and how
• Scalable solutions begin with the correct
use of the right software.
they should be addressed throughout the stages of implementation. We
• Customizations increase the risk of discuss why performance is directly related to the success of a project and
performance issues, but these risks can
be mitigated with the right mindset and why it must be prioritized early. We also discuss how to align expectations
planning. with business users to enable meaningful performance discussions so that
• Performance testing needs to be realistic
to be meaningful. challenges are identified and translated into design requirements.
• Performance issues are complex to
resolve and therefore are better avoided
than fixed. Finally, we cover performance testing strategies to protect organiza-
• Ensuring performance is an iterative
process that involves incremental im-
tions from common performance risks and anti-patterns as well as how
provements throughout the to approach resolving any issues that do occur.
solution lifecycle.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 463

 Why performance matters
Performance can pose challenges as projects approach go live. These
challenges are typically due to a failure to properly prioritize perfor-
mance in earlier phases. Let’s back up a bit and consider first why good
performance is important.

Let’s consider city planning—specifically, the importance of street de-

sign to ensure livability. Streets enable residents and visitors to get into
and out of a city. A city could plan a basic two-lane road to accomplish
this goal. However, what happens when 1,000 cars are on that two-lane
road at the same time? What happens if an accident occurs? What
happens if you add 10,000 cars? Yes, eventually drivers will get to their
destinations, but it likely won’t be in the desired amount of time. These
are the types of questions to ask when looking at performance.

People expect superior response times and a smooth experience from the
organizations they choose to provide products and services. If customers
don’t get this, they tend to look elsewhere. Let’s go back to our street
design. Because of poor street design, people might decide not to live in
or visit that city. Residents might even elect different leadership because
of their poor experience. The takeaway is that system performance affects
people in numerous ways and can have serious consequences.

Customer experience is key

A poorly performing application solution can definitely have a negative
effect on your organization. Microsoft Dynamics 365 products support a
wide range of business functions related to customer interaction and are
an opportunity to either delight or disappoint customers.

Customers might use integrated applications with a direct depen-

dency on Dynamics 365 apps, for example, through a sales website. In
such a scenario, poor performance can quickly turn into lost sales as
customers abandon the website to shop elsewhere. No one enjoys a
frustrating and slow sales experience, and customers can quickly switch
to a different business to purchase goods and services.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 464

 Often, though, customers interact with systems indirectly through an
agent within an organization, for example, when a customer calls a call
center to discuss a faulty product or delayed order shipment. These
types of interactions can be challenging for a business to resolve in a
positive way and are only made more challenging if an agent cannot
access information in a timely manner. This illustrates again that system
performance is key to ensure a positive customer experience, regardless
of whether the customer interacts with the system directly or indirectly.

System success
As businesses evolve and optimize, the ambition remains to achieve
more with less. The need to sell more products or service more cus-
tomers is always accompanied by the need to reduce the expenditure
of money, effort, people, and time, at all levels of the organization.

This constant need to achieve more with less creates pressure on em-
ployees to work in the most efficient way possible. Time spent waiting
for a system to perform an action is wasted time, and employees who
rely on these systems to do their jobs quickly realize this.

Poor performance can have significant effects on a project. It’s usually

not well tolerated by business users and can be detrimental in terms of
user adoption and system reputation.

User adoption
User adoption is a critical factor in the success of any software project.
Any business case (and projected return on investment) depends on
the system being used as intended. Poor performance directly drives
user dissatisfaction and can make user adoption incredibly challenging.

Users are keen to adopt systems that increase their productivity, which
essentially means minimizing wasted time. Business users achieve their
goals when solution performance is optimal and as expected. A poorly
performing system wastes users’ time and therefore reduces productivity.

If the impact of poor performance on users’ time is significant enough,

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 465

they’ll likely find more efficient ways of working on their own. These
workarounds ultimately serve the interests of specific users rather than
the business and often result in data stored off-system (for example,
in spreadsheets). This might eventually lead to a situation in which the
system no longer serves its purpose, doesn’t deliver a return on invest-
ment, and ultimately fails.

System reputation
Even before go live, performance can help or hinder user adoption.
During the development phase, the application typically is presented
to a set of key users in different areas of the business to collect feed-
back. These users then talk to colleagues about the implementation.
In this way, the reputation of the application spreads throughout the
business long before most users touch the system. Keep in mind that
performance impressions tend to spread quickly. For example, if a
demonstration goes well, a wave of excitement might flow throughout
the company. This positivity can help increase user adoption because
of the anticipated improvement in productivity.

Performance issues discovered in the buildup to go live or shortly af-

terward cause a great deal of stress for the project team, which usually
works long hours to ensure a timely delivery of a functional system.
The realization that the system doesn’t perform well can quickly force a
change in priorities and derail progress. These issues are often visible at
the most senior levels and tend to escalate quickly.

But it’s in the cloud

Having established that performance matters, surely you can just
deploy the fastest computers available to run the solution, right? Well,
not quite.

Dynamics 365 is a cloud-based software as a service (SaaS) solution.

Cloud solutions entail different challenges from on-premises deploy-
ments. For example, several years ago, Dynamics 365 deployments
were commonly hosted on infrastructure physically located close to or
even in the same office or warehouse as users. Today, single deploy-
ments are used by users across the world on a range of devices with

466
 different network infrastructures. Project teams therefore need to build
solutions that accommodate users who have a variety of hardware,
increased network latency, and a range of network quality.

The SaaS nature of an application means that the infrastructure on

which the solution runs is housed and maintained by Microsoft. Other
than a few predefined configuration options, the infrastructure is
beyond the scope of a customer’s control, which offers benefits in
setup cost and ongoing maintenance responsibilities but does limit the
options to change performance.

Having said that, the performance of an application extends beyond

the infrastructure. Although a correlation exists between the perfor-
mance of software and the quality of its underlying infrastructure,
the on-premises mindset of fixing software problems with additional
hardware is expensive and usually only a temporary fix.

Using our street design example again, adding a privately operated

toll road or highway doesn’t necessarily mean drivers will arrive at their
destinations any faster.

We look at more examples later in the chapter, but for now it’s impor-
tant to be clear that most performance issues are best solved by cor-
rect implementation decisions and not by adding hardware. Moreover,
it’s crucial to acknowledge that performance is not guaranteed simply
because the software is running in the cloud. It’s still the responsibility
of the project team to deliver a well-performing solution.

Prioritize performance
Given that performance is so important to users, customers, and ulti-
mately the success of the overall system, let’s look at how performance
relates to project delivery.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 467

 Performance doesn’t happen unless
you make it happen
Performance issues in Dynamics 365 projects are commonly analyzed,
understood, and fixed at a solution level, often against tight dead-
lines and high visibility. It’s fair to say that the most common cause of
unexpected performance issues is a failure to properly prioritize per-
formance from the beginning of a project.

Performance in any software project is not guaranteed. It requires

careful planning, expectation management, open communication,
and dedicated time and resources. This is especially true for projects
that feature integrations, data migrations, and extensions on top of
the base product. These are all common in Dynamics 365 implementa-
tions. Anything added to the core Dynamics 365 app has the potential
to negatively impact performance, so the impact needs to be under-
stood and managed to ensure performance remains acceptable.

Performance influences decisions in many areas of the project, includ-

ing the following:

Data strategy
▪ Is the volume of data stored likely to cause performance issues
for users?

Integration strategy
▪ Are real-time integrations feasible given the performance expecta-
tions of the users?
▪ Can overnight batch integrations complete within a given timeframe?

Data modeling
▪ Do we need to denormalize for performance reasons?

Security modeling
▪ Will our security model work at scale?
▪ Are there bottlenecks?

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 468

Environment strategy
▪ Is a performance test environment available?
▪ Is our performance test environment representative of production?
▪ Have we budgeted for performance testing?

Design and implementation of customizations

▪ Are the developers building to meet specific performance goals?
▪ Do user expectations align with what’s technically feasible?
▪ Do we need to use patterns (such as asynchronous or synchro-
nous) or batch jobs (such as background processes)?

Testing design and approach

▪ What’s acceptable—what’s considered a pass or fail?
▪ Is our testing representative of user behavior?
▪ Are we testing based only on today's business requirements or
keeping future expansions in mind?

User acceptance and adoption

▪ Is performance measurable?
▪ Are user expectations realistic?

These aren’t the types of questions the delivery team should be asking
when a performance issue surfaces, especially when approaching the
go-live deadline. A successful project will have answers to these ques-
tions early on. At the least, the delivery team should identify risks and
seek possible resolutions in the early stages of delivery. This might lead
to proof-of-concept work to test performance.

Let’s put this into the context of our street design scenario and consid-
er the questions that need to be asked and answered to maximize the
design. For instance, how many residents currently live in the city? How
much traffic can we expect to be on the roads? What’s the projected
population growth and how long will the design support it? Will traffic
lights be installed? If so, how many and how will that affect traffic?
What’s the speed limit and are there risks associated with that limit?
Each of these questions helps determine the best street design before
we even start the project.

469
 Resources
Considering performance from the early stages also ensures that the
correct expectations are set in terms of time, money, effort, and peo-
ple. For example, for performance testing, a dedicated performance
test environment is needed as well as the people to do the testing.
Business stakeholders might need additional time with the delivery
team to understand, agree with, and document performance require-
ments. It might even be necessary to allocate more development time
and resources for code optimization.

The most successful projects from a performance standpoint dedicate

workers to performance-related activities. Performance testing is a
subject on its own, and expertise is required to design test cases and
execute test runs. Furthermore, a single owner within the project
should be identified to drive performance planning and remediations
across the development teams during all phases of the project.

Fig.
17-1 Go live

Initiate Implement Prepare Operate

Identify and fix performance issues in these phases Do not wait to fix issues in these phases

Prevention is better than cure

A common approach with Dynamics 365 projects is to implement the
system, get it to a point where it’s functionally correct, and only then
deal with the inevitable performance challenges (see Figure 17-1).
Although this approach might seem logical under the pressure of
delivery, the inevitable performance issues place additional stress on
the team during the most difficult parts of the project and add work
during the buildup to go live.

Dealing with performance issues after functional testing is completed

can create additional project risk because the changes needed to
resolve them are typically significant and therefore come with a high
risk of introducing bugs. This essentially invalidates any prior func-
tional testing. A regression testing cycle is then required to ensure

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 470

 performance fixes haven’t broken the previously tested functionality.
This wastes time at a critical stage in the project.

User confidence
Performance is always linked to perception. It’s important to be aware
of user feedback during the implementation of the project because
it can affect users’ engagement during testing and ultimately help or
hinder user adoption at launch. Projects that prioritize performance
early on tend to present better-performing solutions during imple-
mentation. This early planning helps reassure users that they’ll receive
a solution that enables them to become more productive—and this
leads to better engagement and adoption overall.

Why not prioritize performance?

With all this in mind, it’s perhaps surprising that many Dynamics 365
implementations don’t prioritize performance. There are several
reasons for this. In some cases, people lack experience with the specific
products. Performance issues can be complex, and it’s often difficult to
foresee them without extensive knowledge of the platforms and asso-
ciated patterns and anti-patterns gained over years of experience.

Sometimes, however, projects intentionally don’t include performance

testing in an attempt to save on budget. For some projects, performance
activities are deprioritized because of timing constraints. However,
these are shortsighted approaches because they’re an opportunity for
performance issues to surface and risk the success of the project. The
best way to avoid performance problems is to be proactive and seek out
and address risks early to minimize their impact. It’s a more productive
approach than reactively responding to problems as they occur.

Establish requirements
Acceptable performance is the goal of every project, but the definition
of “acceptable” is often vague, if defined at all. To successfully deliver
acceptable performance, we need to be clear on what that means and
then be able to track progress against our performance goals.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 471

When we look at the requirements for our street design scenario, we
know that drivers want to go from point A to point B quickly and safely.
However, each driver might have a different idea of what quickly and
safely means. To get somewhere safely, a driver might need to drive
slower to handle unexpected situations. But then some drivers might
become upset because it takes too long to reach their destination.
If you create a goal for drivers to get from point A to point B within
X amount of time and that time increases, it might indicate that you
need to change the design.

In the next section, we discuss how to define performance goals that

are achievable and acceptable to users.

Why do we need performance

requirements?
If you ask a user for their performance requirements, you’ll likely get
a response that the system needs to be fast—and not much else.
However, “fast” is an ambiguous term; people have different expec-
tations of what that looks like, which makes understanding whether
you’ve achieved acceptable performance nearly impossible.

A better approach is to establish a common understanding across

the implementation team and stakeholders on performance, which
enables us to:
▪ Align expectations of the definition of “acceptable.”
▪ Understand the business impact of not achieving acceptable.
▪ Understand constraints during the design process.
▪ Provide clear instructions to developers during implementation.
▪ Test in a meaningful way with behavior models representative of
expected usage.
▪ Determine whether acceptable has been achieved.
▪ Assess gaps between what’s been achieved and what’s acceptable.

This approach is the same for other system requirements. It’s vital that
performance be considered like other requirements gathered in the
initial stages of implementation.

472
 More specifically for performance, it’s also important to know when
to stop optimizing. For example, developers might apply perfor-
mance tweaks to a piece of code, optimizing with no goal in mind.
Any optimization of tested code comes with some risk of regression
issues; therefore, this practice should be performed as infrequently as
possible. However, without a clear understanding of when enough is
enough, it’s difficult to gauge what level of optimization is sufficient
and when further optimization is unnecessary.

Work with users

Any performance-related discussion should include representatives of
the system users. They often give meaningful insight into what is and
isn’t acceptable. Decisions made without input from users risk rejection
in the later stages of the project, which is what we want to avoid.

This doesn’t mean users are solely responsible for deciding perfor-
mance requirements. It’s important to be clear that performance
comes with an associated cost, and business stakeholders need to
assess requests coming from users within the context of the wider
project. Aggressive performance requirements might be achievable,
but they require additional development, testing effort, and people.
With this in mind, it’s important to understand the underlying need
for each performance requirement and for business stakeholders to
be prepared to consider a compromise where it makes sense to do
so. Performance for the sake of performance is expensive and unnec-
essary. Communicate this to users and take a pragmatic approach to
focus on what specific performance requirements are important.

Identify performance-critical tasks

Users will use the system under development to perform their own
set of tasks. Some of these tasks will be more performance critical
than others. For example, consider a call center agent attempting to
access customer data at the start of a telephone call with a customer.
It’s critical that the agent is able to access the data, and every second
counts. Time spent waiting for a form to load is a waste of time for the
agent and the customer.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 473

 In contrast, a user preparing month- or year-end financial re-
ports might expect to wait a little while for actions to complete.
Performance is still important, but for activities like this that occur
infrequently, the business case for investing time, effort, and money to
improve performance is more difficult to justify.

For example, business users might offer the justifications shown in

Figure 17-2 when asking for a performance improvement.

Spend time with users to understand the activities for which per-
formance plays a critical role. Agree on these areas with project
stakeholders and then focus performance-related work on these
activities to maximize the value of the efforts. Consider performance
testing for each system area, including the following:
▪ Functional processes
▪ Background operations (for example, batch and workflows)
▪ Integrations
▪ Data migration
▪ Reporting and analytics

Understand user expectations

For each critical task you identify, have discussions with users about
what they deem to be reasonable performance. It’s important to
discuss each task separately and focus on the specific task when dis-
cussing acceptability. This leads to different requirements for different
tasks, which is a more granular, actionable, and sensible outcome than
Fig.
17-2
having broad, systemwide requirements.

SL
A

“I need to load this record “I need to be able to load “I have to be able to do “This overnight process
quickly for a customer on the products into the this within a few minutes; needs to happen within the
the phone; otherwise, I delivery truck to meet my we have a service-level time window; otherwise, the
might lose the sale.” shipment schedule.” agreement to meet.” users won’t be able to work.”

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 474

 Also consider that a user with exclusive access to a system is likely to
have a different experience than someone who uses the system only
during peak hours, along with thousands of other concurrent users,
or while an intensive background process runs. We recommend that
you map out predicted system usage throughout the day, week, and
month to understand the volumes of users, data, and actions in these
time periods to better locate any potential performance issues.

The usual measurement of performance is time taken, so requirements

should be shaped around what is an acceptable time to perform a spe-
cific action, in a way that can be measured. However, it’s often easier to
ask users, “What does unacceptable performance look like?” and work
backwards from there. This approach tends to lead to more realistic
requirements, although they might be unnecessarily aggressive. Take
the time to understand the business impact and risk of unacceptable
performance in these scenarios.

Anticipate growth
When discussing performance requirements, consider the roadmap of
the business as well as design requirements for an increase in demand
on the system. Data and user volumes play an important part in how a
system performs, so it’s important to anticipate any growth expected
in the near future and design for that rather than focus on the current
requirements. Along with future growth, also plan for seasonality in the
system load, for example, during the end of the year.

Document requirements
It’s crucial that performance requirements be documented (the same
as for other system requirements). Documenting requirements pro-
vides visibility to all parties about the expectations of the software and
provides clear goals for the implementation team. Additionally, any
performance risks identified during discussions with users should be
documented in the project risk register to ensure they’re tracked and
mitigated as much as possible.

Assess feasibility
The implementation team should review performance requirements

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 475

along with other system requirements as early as possible to assess the
impact to the organization. Some performance requirements might
have a significant impact on design decisions, so you want to be aware
of these as soon as possible.

Occasionally there might be doubt about whether a particular require-

ment is even achievable with the available software. In this situation,
it’s advisable to run a small proof-of-concept pilot project to under-
stand feasibility as early as possible, thereby minimizing the risk of
accepting unachievable requirements. Discuss requirements deemed
unachievable with the business—it’s generally a more positive discus-
sion during the early stages of the project than during user acceptance
testing (UAT) or immediately before going live.

With clear performance criteria established, accepted, and docu-

mented, the project team can move confidently on to implementing a
system that will perform within expectations.

Design for performance

A common cause of performance challenges in Dynamics 365 projects
involves incorrect use of the applications—in other words, uses that
aren’t aligned with the intentions of the software or are due to poor
solution design. Poor decision making during the design phases of the
project can open the door to performance challenges further into
the project.

Looking at this from the perspective of our street design discussion, there
are different ways to tackle a situation like traffic. We could design our
roads to be 10 lanes; that would handle a lot of traffic. But this creates
other complications. Do we have land that will support that infrastructure?
What are the relative costs associated with that design? How easy will it be
for residents to cross the street? How long will a light need to be red for
them to cross it? Will the street need additional lighting?

In the next section, we discuss the architectural and design consid-

erations for the project design phases that ensure success from a
performance standpoint.

476
 Use the right tool for the job
Dynamics 365 products are flexible and can be changed to suit the
needs of many businesses. However, exercise restraint when consid-
ering incorporating this flexibility into your design. Just because we
can adapt the products to achieve something functionally, it doesn’t
guarantee they will do them well.

For example, the xRM concept, which became popular around the
release of Dynamics CRM 2011, spawned many systems to manage any
type of relationship in the CRM product. The ease of development for
basic data access, including a built-in user interface and security model,
combined with its rich extensibility features, made it a popular choice
to begin system development. Although this proved successful in
many situations, many projects ran into trouble because they used the
product in a way unsuited to its strengths. Dynamics 365 products are
designed for specific use within a business context. They’re designed
and optimized for users to access master and transactional business
data, not for keeping high-volume historical transactions.

Additionally, keep the Dynamics 365 roadmap in mind. It’s worthwhile

to explore newer features because they often bring performance
enhancements compared to older approaches.

Many areas of this book discuss what to consider when making system
design decisions. Often the consequence of making the wrong deci-
sions during these stages are performance issues. It’s important that
use of Dynamics 365 products is in line with their intended purpose.

Design for scalability

Whether you design a solution for the first time or redesign one due to
issues, keep performance in mind from the beginning. The following
are recommendations for building a highly scalable solution.

▪ Design for parallelism and multi-threading for high-volume scenarios.

▪ Use set-based operations—RetrieveMultiple, insert_record-
set, and update_recordset—as appropriate.
▪ Build multiple lightweight processes connected to each other.
▪ Design shorter transactions to reduce lock escalations and increase
user concurrency.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 477

 Design for responsiveness
Response time is important for the success of your solution. The fol-
lowing are recommendations for building a highly scalable solution:
▪ Consider whether to design for synchronous, asynchronous, or
background (batch) processing.
▪ Consider data growth over time because rapid data growth re-
quires a lot of tuning to keep the same response rate.
▪ Design for aggressive caching in Dynamics 365 Finance and
Operations as appropriate.

It’s important to note that users typically assess whether performance

is in line with their expectations, not within an arbitrary period. With
this in mind, we can consider performance solutions that don’t actually
involve speeding anything up.

For example, choosing to execute a task asynchronously doesn’t re-

duce the time to execute the task, but it does ensure the user interface
doesn’t prevent the user from working. As long as the user doesn’t
need the result of the task in real time, this could be an acceptable
performance solution.

Data migrations and integrations

Many performance issues related to data migration and integration
happen due to unreasonable expectations of the platform, specifically
around data volumes. A good data strategy often works around this by
ensuring that Dynamics 365 products contain only the data required
for users to complete their day-to-day operations.

When planning a data migration, it’s important to understand the

business requirements for the data being migrated. Implementation
teams sometimes decide to migrate far more data than required and
then run into performance issues due to the size of the data set. For
example, importing the last 10 years of unconverted leads into a busi-
ness’s sales system will likely take a substantial amount of time. The fix
here is unlikely to be to speed anything up; it’s more likely to involve
questioning the value of 10 years of data, much of which is probably
stale and unlikely to provide value.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 478

 In the case of high-volume integrations, consider when and how often
these will run. Heavy processing can place a strain on the application,
and if this happens at the same time that users are interacting with it,
issues are likely to arise. This is particularly important to consider for
global businesses, whose systems might be in almost continuous use.

It’s also important to be aware of platform limitations when designing any

high–data volume or processor-intensive operations, which are common
in data migration and integration scenarios. For example, Dynamics 365
Customer Engagement has documented throttling limits that need to be
accounted for within designs. Working within these limits helps ensure
that the performance of the underlying platform remains acceptable. As
another example, the Finance and Operations OData endpoint has a page
size limit of 10,000 records and isn’t designed for high-volume integra-
tions. Other approaches, such as Data Management Framework (DMF),
are better suited for these types of integrations.

Chapter 16, “Integrate with other solutions,” discusses how to design

integrations and integration patterns and anti-patterns in Dynamics 365.

Prepare for implementation

In this section, we discuss the design decisions the team should make
before proceeding with the implementation.

Environment planning
Chapter 9, “Environment strategy,” discusses environment planning in
detail. From a performance perspective, consider the following as the
team moves towards implementation:
▪ Performance testing typically occupies an environment for a
significant amount of time, so a separate environment is
usually advisable.
▪ Latency adds overhead to every operation. Minimize overhead as
much as possible by ensuring that applications are located as close
to each other as possible.
▪ Choose a performance test environment that’s representative of
the production environment whenever possible. For example,
for Finance and Operations, the implementation team should

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 479

 recommend the appropriate environment tier based on expected
load and volume.

Keep in mind that if performance issues arise within scaled-down

development or test environments, it’s inaccurate to assume perfor-
mance will improve in an environment with more resources. Without
understanding the root cause of a performance issue, it’s impossible
to determine whether additional infrastructure will help at all, so it’s
crucial to understand and address performance issues early in the de-
velopment process rather than assume they’ll go away in production.

User personas
The organization should have a sufficient number of user licenses to be
able to realistically model the anticipated behavior during performance
testing. Teams also need an understanding of the user personas for
testing expected usage. Keep the following in mind for user personas:
▪ User locations
▪ User security configurations
▪ Data typically associated with each type of user
▪ Expected number of concurrent users divided by persona

Proof-of-concept
development
Teams sometimes run an initial proof-of-
Customization and
concept project before implementing
the main project to prove out ideas and
gain understanding of product suitability.
performance
Custom functionality built during proof-
of-concept projects is often carried over
The extensibility of Dynamics 365 applications provides the powerful
into the main project.
ability to tailor software to meet individual business needs but also
Although this approach can significantly
accelerate development during the main
introduces risk—performance issues commonly involve product cus-
project, it’s worth bearing in mind that tomizations and integrations, particularly those involving code. This
there’s often little in the way of develop-
mental governance for proof-of-concept section discusses these issues and provides guidance on how to avoid
projects, which are usually undertaken common problems.
without any performance requirements
or considerations in place. These proj-
ects can therefore become a source of
Chapter 15, “Extend your solution,” discusses how to approach custom-
performance problems unless the team
takes the time to review the outcomes izing Dynamics 365.
along with any new requirements prior to
further implementation.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 480

 Just write fast code?
Many performance issues are due to custom code, and the fix inevitably
involves reworking this code to remove bottlenecks. Project teams often
question why such code wasn’t written correctly to begin with, but the
truth is that software development is difficult. Developers regularly wres-
tle with complex and abstract processes based on vague and changing
requirements along with competing priorities against tight deadlines.

The most productive developers tend to focus on what’s necessary

and disregard anything superfluous to requirements. Keep in mind
that if a requirement is missing key information, the developer won’t
be aware of this and will proceed with implementation based on the
given information. It’s common for requirements to lack performance
criteria, which isn’t typically top of mind for the developer or part of
sign-off criteria.

It’s also uncommon for a developer to deeply understand the business

context of the system for which they’re developing code. A developer
likely might not have experienced the time-pressured environment
of a call center or order fulfilment center and therefore won’t be able
to judge the need for performance. Therefore, specific performance
requirements should be mapped through to functional requirements
where appropriate to ensure that developers have visibility to perfor-
mance constraints on the functionality for which they’re responsible.

Bad design leads to bad code

Performance issues related to poorly performing code often aren’t
due to the code itself but rather to a flaw in the design decisions. This
can result in situations in which developers are unable to write code to
perform an operation efficiently, even when carefully following best
practices, because of the constraints placed on the solution by its design.

Retrofitted performance
Sometimes developers focus on getting code working correctly before
working quickly. Although this can be a reasonable approach, the
pressure of deadlines often means that the optimization planned for a
later date doesn’t get done, leading to technical debt and a need for
rework. A better approach is to be clear on any performance constraints
from the beginning and then implement the solution accordingly.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 481

 Connected components
Performance issues can occur even when developers write fast,
efficient code. This is because in larger projects involving several devel-
opers, the isolated nature of development might result in performance
issues not being discovered until components are put together as part
of a larger process. This risk should be identified during design activi-
ties, and it can be mitigated using a code review process during which
team members can consider the impact of the implemented pieces of
code running together.

Requirement evolution
A change in functional requirements can be another reason for cus-
tomization-related performance problems. A developer decides how
to implement code based on the given requirement at a point in time.
A change in requirements might invalidate some or all of these deci-
sions and cause the implementation to become unsuitable.

Common mistakes
Code can become suboptimal from a performance standpoint for a
number of reasons, and specific guidance is beyond the scope of this
chapter. However, the following factors are often involved in perfor-
mance challenges, so we recommend that you understand and avoid
them during implementation.

Chatty code
One of the most common causes of code-related performance issues is
excessive round trips. Whether between the client and server or between
the application and database, an excessive number of requests for an
operation can really slow it down. Every request carries latency and pro-
cessing time overhead, and it’s important to keep these to a minimum.

Round trip issues are sometimes due to poor control of variables; a

process might be so complex that it’s easier for a developer to retrieve
the same data multiple times than structure the code to minimize the calls.
We recommend that developers avoid this practice and that the imple-
mentation team identifies such problems during the code review process.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 482

 Round trip issues are also sometimes due to queries being executed
within a loop, or worse, a nested loop. The code structure works for
the developer, but the parameters of the loop can add a significant
Some of the content in this section dis-
cusses technical concepts. For those less
number of calls into the process, which in turn results in a significant
interested in the technical aspects, performance issue.
note that the details aren’t essential to un-
derstand. Also note that free tools are avail-
Consider the pseudocode example in Figure 17-3 as triggered manu-
able for less technical people who want to
learn about these concepts, including the ally by a button click in the user interface.
Dynamics 365 Customer Engagement Solu-
tion Checker and the Finance and Opera-
If ExecuteRequest() sends a single request to the server, the total
tions Customization Analysis Report (CAR).
number of requests will be the product of the sizes of collection1 and
collection2. Assume each request takes an average of 50 milliseconds.

It’s common for developers to write logic that iterates across collec-
tions of data that are dynamic in size, due to the data-driven nature
of code within Dynamics 365 implementations. It’s also common for
developers to work with low volumes of data within development en-
vironments. However, that means that these types of issues often aren’t
identified until the latter project stages, which include meaningful data
volumes. Low data collection can be avoided by prioritizing minimal
Fig.
data retrieval during code design, retrieving each piece of data only
17-3
once, and identifying any devia-
tions from these steps as part of a
foreach (var a in collection1)
code review process.
{
(foreach var b in collection2) Retrieving too much data
{ Another common perfor-
ExecuteRequest (); mance-related issue is retrieving
} more data than necessary, often in
} columns. For example, the practice
of selecting all the columns in
Dynamics 365 apps to avoid speci-
Size: Size: Total requests Total execution time fying individual columns can cause
collection1 collection2
performance issues. The number
2 2 4 0.2 seconds
of joins executed on the database
5 5 25 1.25 seconds server to provide lookup names
10 10 100 5 seconds and option set values can be a
significant overhead, particularly
50 50 2,500 2 minutes, 5 seconds
when querying large volumes.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 483

 The same applies for rows. Be mindful of the amount of data being
retrieved because it takes time to query, download, and process the
volume of records even before the code is able to interact with them.

Broadly speaking, only the data required by a process should be re-

trieved; anything more is simply a waste of resources.

Intercept synchronous events

Many events in Dynamics 365 products can be customized, but if
the customization executes synchronously, it adds overhead directly
to the user’s experience. For example, if a customization placed on
RetrieveMultiple in Customer Engagement executes, or an event
insert or pre-insert in Finance and Operations executes, the user has to
wait for the customization to execute while they watch a spinning icon.
This approach isn’t recommended, but when it’s necessary, exercise
extreme caution to ensure minimal performance impact.

Unintended execution
Performance issues sometimes happen because customizations are
executed accidentally or multiple times in error, for example, duplicate
plug-in steps or duplicate method calls. Project teams should ensure
that developers are aware of exactly how their customizations are trig-
gered and mindful of circumstances that might inadvertently trigger
their code. Background processes or batch jobs recurrence should be
set according to business needs.

Performance testing
approach
The project team and users need to be confident that requirements
identified earlier in the project are achieved when the system is
implemented. A performance testing strategy ensures that system
performance is measurable and provides a clear indication of whether
performance is acceptable.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 484

What makes a performance test good?
A common approach to performance testing is to try to push the
system to the breaking point to learn whether the system can handle
the demand. However, this approach typically provides little value,
particularly in Dynamics 365 apps. The best approach is to model or
simulate expected system usage including peak load and day-in-the-
life business scenarios to make sure the solution is capable of achieving
defined performance goals.

Realistic
Being realistic means understanding the quantity and personas of
users using the system at a given time and defining day-in-the-life
activity profiles for the personas to understand the actions they’ll
perform. If the performance test incorrectly assumes that all the users
will run the most complex processes in the system concurrently, the
projected demand placed on the system will be far higher than reality.
Strive to model the users in an accurate way to get the most meaning-
ful results from the test.

It’s important to understand the difference between testing the

functionality and testing as a user would use the system. Functional
testing is often achieved by invoking specific logic and measuring the
outcome, for example, an API call, whereas user testing is testing that
models the actual behavior of a user. Testers often create a suite of
functional tests pointed at a web service, execute them with a large
number of concurrent users, and call that a performance test—but this
provides misleading results. Typically, this approach pushes certain
functionality much more than the usage expected once the features
are live (creating false performance issues), and other functionality can
be bypassed entirely (for example, the user interface). The result is that
genuine performance issues can slip through testing unidentified.

Keep in mind that user interaction with the application is fairly slow.
For a given process, users don’t typically click buttons as quickly as they
can; they take time to think between actions, and this can be incorpo-
rated into a performance test as a think-time variable. This can vary
from user to user, but an average figure is sufficient to model behavior.

485
 The key point here is to develop a performance test that represents a
number of users working concurrently and place a realistic amount of
demand on the system.

Developing a reliable performance test also includes realistically

modeling client infrastructure. For example, there’s little value in
performance testing your Dynamics 365 Field Service or Warehousing
mobile application from the office when users will likely use the appli-
cation in areas of low mobile reception.

The configuration of the system should also be realistic and based on

the expected infrastructure and configuration of the production envi-
ronment, use real security configurations for the test users, and work
with volumes of data anticipated in production.

Isolated
A dedicated performance testing environment is generally recom-
mended for two main reasons.

Performance test results are meaningful only if testers are aware of the
activities occurring in the system during the test. A performance test is
worthless if an unaccounted-for process places additional demand on
the system during test execution.

Additionally, it’s common for performance testing to occur around the

same time as UAT. Performance testing can take a substantial amount
of time to execute and push the system to its limits, thereby disrupting
any manual user testing activities. A dedicated performance testing
environment ensures that performance testing can happen in parallel
with other testing.

Business data
Key to good performance testing is using business data such as setups,
configurations, masters, and transactions. It’s recommended to use
the configurations and data to be migrated that will ultimately go live
in production. Additionally, all the data preparation activities must be
ready in advance—for example, data from 100,000 customers or sales
orders should be available via an import file.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 486

 Repeatable with low effort
Ideally, we execute a performance test once and use the results to
ensure that the criteria specified in the requirements are achieved.
However, performance testing is rarely a one-time activity—often,
it’s an iterative process. This happens for various reasons; for exam-
ple, functional changes might occur during UAT that risk a change
in performance. Similarly, if performance testing demonstrates an
unacceptable level, a retest of optimizations will be needed once the
solution is implemented.

If a low-effort performance testing approach isn’t available, it can be

tempting to skip performance test iterations, which makes it difficult
to assess the impact of any changes made between runs. Conversely, a
fully automated process, potentially including environment provision-
ing and data setup, can be executed frequently with minimal effort,
providing a clear picture of variations in performance in response to
specific changes.

Functionally correct
A performance test result is meaningful only if the system functions
correctly. It’s tempting to focus on the performance metrics of success-
ful tests. However, if errors occur, they should be corrected and the
test should be executed again before any analysis is performed on the
results. Deviations in behavior between test runs can significantly skew
a performance test result and make any comparison meaningless.

Document results
The output of performance testing activities should be documented
clearly so that interpretation is straightforward. The results should
be mappable to performance testing criteria and enable the team to
quickly assess whether the requirements were achieved or whether
there are gaps between requirements and results. For example, page
load times can be captured for certain activities and compared to
acceptable page load time requirements agreed to by the business.
It should also be possible to identify when the performance test was
executed and against which version of code if applicable.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 487

 Finally, it’s important that the information in the documented report of
results remains consistent with the defined template; the team might need
to compare various performance test run results as changes are made.

Plan performance tests

Document your performance testing strategy and include the following:
▪ Performance testing scenarios including key user activities, back-
ground (batch) processes, integrations, and reports
▪ Normal and peak data volumes for each scenario
▪ Ideal target and maximum acceptable target response time for
each scenario
▪ Day-in-the-life workload to simulate in the most realistic manner,
especially at peak hour
▪ Data preparation activities for each scenario, for example, existing
master or transaction data in the system or creation of file to import
▪ Environment for performance testing
▪ Iterations of performance testing (plan for two or more rounds)
▪ Performance testing tools
▪ Test results documentation tool
▪ Quantity and profile of concurrent users including security
configuration
▪ Guidelines and workflow for investigating and fixing performance
issues
▪ Risks such as delays and budget related to performance testing

We recommend investing the time and effort to automate this process

as much as possible. For example, with relatively low effort, you can set
up performance tests to run through Azure DevOps release pipelines,
providing a way to execute test runs with a single button click or auto-
matically after a deployment. You can even run them overnight.

Network conditions during testing

Be aware of network conditions during testing. Performance tests are
often conducted in isolated environments with the best network condi-
tions, so they usually show favorable results. However, when the solution
is moved to real-world use, performance issues might arise due to actual
network conditions and the solution being implemented at scale.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 488

To benchmark what’s feasible for the solution when all network conditions
are favorable, ensure that the performance testing environment is located
in the same Azure region in which you plan to deploy production for
business users. As mentioned earlier, latency adds overhead to every oper-
ation, and minimizing testing latency can help identify potential location
network issues if the test results vary with the location testing.

Address performance issues

Analyzing performance problems is complex and situation specific and is
beyond the scope of this book. However, let’s discuss some broad consid-
erations that might help you structure an approach to resolving issues.

Deal in facts
When faced with a performance problem, implementation teams
often begin to frantically point fingers at certain parts of the system
and develop ideas to find the magical quick fix to solve the problem.
Unfortunately, this approach often causes more problems than it
solves because teams make significant changes based on instinct, often
degrading performance further or causing regression bugs as a result.

It’s crucial to understand the root cause of a performance issue in detail

before making any changes. This can be a time-consuming process
depending on the complexity of the issue, but it’s necessary to ensure
that any optimizations made are meaningful. Ideally, the team will
understand the issue in enough detail to have a reasonably accurate ex-
pectation of the differences in performance prior to testing the change.

Expectations
Keep in mind that performance issues are rarely due to a single issue.
Suboptimal performance is most commonly the result of a number of
factors working together, so a single fix is typically unrealistic. Generally,
performance is an iterative process of applying incremental improvements.

In addition, as discussed throughout this chapter, a performance

issue in a business’s Dynamics 365 solution doesn’t necessarily imply

489
 a performance issue in the Dynamics 365 application or its underlying
infrastructure. Microsoft support teams are available to help with
performance issues, but keep in mind that when a performance issue is
isolated to a single environment, investigations typically start by ex-
amining anything unique in that environment, such as customizations,
rather than the application code or infrastructure.

Knowledge is power
Performance issues can be complex and difficult to resolve, so it’s vital
that the implementation team has sufficient knowledge to be able to
ask the right questions and analyze issues meaningfully. The imple-
mentation team is often able to assist with performance issues, but
issues can surface after go live and expiration of any warranty period.
It’s therefore crucial to transfer knowledge to allow business as usual
(BAU) teams to resolve performance issues.

Identify problematic customizations

To identify the source of a performance issue, we recommend that you
first assess whether the issue occurs in a specific customization. Many
Dynamics 365 customization options can be enabled and disabled,
and strategically disabling customizations often helps identify whether
the issue is customization related, and if so, which customization is the
source of the issue.

Proceed with caution

When you identify the root cause of an issue, it’s important to take a
risk-averse approach to optimization. Performance tweaks can involve
complex code changes and have significant risk of regression issues, so
it’s crucial to proceed with caution.

Low-hanging fruit
It's advisable to identify smaller opportunities for performance gains,
rather than consider reworking large and complex areas of the system.
For example, for a poorly performing piece of code, there are usually
several options for optimizations with varying risk of causing issues and
varying performance gains. In some situations, it might be advisable to
make a number of low-risk tweaks; in other situations, it might be better

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 490

 to make a more complex change and manage the risk. Which approach
to take is entirely situation dependent, but it’s important that the team
agrees on the approach and that the changes are managed carefully.

One change at a time

It’s crucial to assess the impact of a change as soon as it’s made—if
multiple optimizations are made at the same time, it’s impossible to
understand the impact of each change after a performance test run.
For example, if two changes are made and one increases the perfor-
mance by 10 percent and the other decreases it by 10 percent, the net
result will be no performance change. By testing the same changes
individually, the team can use the performance test results to reject
one change and accept the other one.

Workshop strategy
FastTrack runs a solution performance workshop focused on solution
design that covers the impact of additional configuration and cus-
tomization on the overall performance and end-user experience. The
workshop emphasizes the importance of performance prioritization,
goals, and testing during the stages of a project.

Workshop scope
The workshop includes the following topics that address how to
incorporate performance activities into the overall delivery plan and
allocate sufficient performance expert resources to the project:
▪ Data volumes  Projected workload and integration volumes to
ensure expectations are within limits and aligned with intended
product usage
▪ Geolocation strategy  Physical locations of users and servers to
identify any network-related challenges
▪ Key business scenarios  Main areas of the business for which
performance is particularly important
▪ Extension performance  Planned customizations to understand
how implementation is aligned with recommended practices
▪ User experience performance  Modifications to the user expe-
rience in conjunction with best practices

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 491

 ▪ Performance testing  Performance-related goals and testing
strategy to ensure performance is measured

Timing
We recommend that you conduct the performance testing strategy
workshop before solution design or as soon after as the team is able
to provide detailed information about performance requirements and
the performance testing strategy. Scheduling a workshop later in the
implementation is risky because any findings and recommendations
from the workshop could cause significant rework.

Product-specific guidance
Operations
Following are recommendations for achieving optimal performance in
your Finance and Operations solutions:
▪ Use Tier-2 or higher environments based on business objectives.
Don’t use a Tier-1 environment.
▪ Keep the solution up to date with hotfixes, platform updates, and
quality updates.
▪ Identify and maintain a log of performance-related risks.
▪ Use DMF to import and export large volumes. Don’t use OData for
large volumes because it isn’t natively built to handle large payloads.
▪ Use set-based data entities and parallelism to import and export
large volumes.
▪ Build your own data entities to avoid potential standard entity
performance issues. Standard entities contain fields and tables that
you might not need for your implementation.
▪ Configure a batch framework including batch groups, priority,
and threads.
▪ Define batch groups and assign a batch server to each batch
group to balance batch load across AOS servers.
▪ Design heavy batch processes to run in parallel processing.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 492

 ▪ Use number sequence preallocation.
▪ Perform cleanup of routines and jobs regularly.
▪ Avoid record-by-record operations; use set-based operations such
as insert_recordset and update_recordset where applicable.
▪ Define business user roles and assign the appropriate level of
access needed, which makes for better performance. The admin-
istrator role, for instance, will have better performance than a user
with limited access.
▪ Use the Optimization advisor workspace to identify business
processes to be optimized.

Performance tools
▪ Trace Parser
▫ Diagnoses performance issues and various errors
▫ Visualizes execution of X++ methods as well as the execution
call tree
▪ Lifecycle Services Environment Monitoring
▫ Monitors server health metrics
▫ Monitors performance by using the SQL insights dashboard
▪ Query Store
▫ Reviews expensive SQL queries during defined intervals
▫ Analyzes the index used in queries
▪ PerfSDK and Visual Studio load testing
▫ Simulates single-user and multi-user loads
▫ Performs comprehensive performance benchmarking
▪ Performance timer
▫ Helps determine why a system is slow
▫ https://yoursite.cloudax.dynamics.
com/?cmp=USMF&debug=develop
▪ Optimization advisor
▫ Suggests best practices for module configuration
▫ Identifies obsolete or incorrect business data

Customer Engagement
The following are guidelines taken from the Fast Track Performance
Optimization Tech Talk for optimizing solution performance:

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 493

 ▪ Configure before customizing. Achieve a goal by configuring
standard application components if possible. Customizations
such as plug-ins and JavaScript are useful, but they can carry a
performance overhead and can also make future upgrades more
challenging.
▪ Avoid OptionSet attributes in a Quick Find search.
▪ Stay current on deprecation announcements to align with the
product roadmap.
▪ Reduce form load JavaScript for a better form load experience.
▪ Display the minimum required number of fields required in the
forms.
▪ Design forms and pages to display the most important informa-
tion at the top.
▪ Minimize the number of controls in the command bar or ribbon.
▪ Use collapsed tabs to defer loading content.
▪ Avoid unsupported customizations such as direct Document
Object Model (DOM) manipulation.

Use the Solution Checker to detect potential performance issues in

code and configuration.

Conclusion
This chapter discussed why performance is expected and critical for
user adoption, the customer experience, and project success. We noted
that although Dynamics 365 projects are built to perform well at scale,
their flexibility means it’s crucial that implementation teams consider
performance as an iterative process throughout the solution lifecycle.

Performance often involves tradeoffs, so project teams need to work

closely with users and business stakeholders to determine and align
performance expectations.

Good solution performance starts with good design. This means

using Dynamics 365 products for their intended purposes and fully
considering the performance implications of design decisions early on in
the process. Many performance issues occur due to poor customization,

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 494

 configuration, and design choices, so it’s important to proactively work to
avoid this by clarifying requirements and implementing best practices.

Performance testing is crucial to ensure performance expectations

are achieved during implementation. However, performance tests are
valuable only if they’re meaningful and provide a clear indication of
how the system will perform in the real world.

Performance issues are difficult to resolve and place unnecessary strain

on the implementation team during the most critical points of the
project. With the right prioritization and planning, performance can
be baked into the implementation strategy, leading to a much more
confident launch with greater project success.

References
Product-specific guidance (Operations)
Monitoring and diagnostics tools in Lifecycle Services (LCS)
Performance troubleshooting using tools in Lifecycle Services (LCS)
Work with performance and monitoring tools in Finance and Operations apps
Query cookbook
Take traces by using Trace parser
Diagnose issues and analyze performance by using Trace parser
Performance timer
Query Store Usage Scenarios
Monitoring performance by using the Query Store
Optimization advisor overview

Product-specific guidance (Customer Engagement)

Performance tuning and optimization
Optimize model-driven app form performance
Tips and best practices to improve canvas app performance
Introducing Monitor to debug apps and improve performance
Performance considerations with PowerApps

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 495

 Checklist

Performance focus
Establish that the responsibility to deliver a performant
solution on the SaaS platform is shared by the cloud
service provider and the implementor who is customiz-
ing and extending the out-of-the-box application.

Design, develop, and test for performance in accordance

with the service boundaries.

Include performance as part of the requirement con-

versation and ensure it is reflected in the design and the
development of the solution, with specific documented
goals that are tracked through future change requests.

Establish development, user experience, and Information

Architecture guidelines that create a positive perception
of performance, prioritizing application responsiveness
and user experience.

Understand that the end-user performance is accumu-

lative of several factors, including the network latency,
device performance, browser add-ins used, firewalls,
and proxies. Establish the performance baseline in the
context of the actual user environment and location.

Ensure the performance testing considers realistic loads

and actual usage scenarios.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 496

 Case study
Corporate travel company
learns how testing is critical
to measure and optimize
performance
A global corporate travel company was implementing Dynamics
365 Customer Service to drive a call center transformation. Several
process flows required integration with legacy systems—many with
high transaction volumes—and there was a business-mandated cus-
tomization that the project team agreed couldn’t be achieved using
out-of-the-box functionality. This complex solution was projected to
support 4,000 users at scale and needed careful planning to ensure
optimal performance.

Despite the complex customizations and integrations, the customer

had decided performance testing was out of scope due to budget
constraints, assuming that because Dynamics 365 is a cloud service, its
performance is owned by Microsoft.

While still in the design stage, the project team decided to use the
best practices from the Success by Design framework and their own
experiences to highlight the risks of leaving performance testing out
of scope. The project team pointed out potential negative outcomes if
performance wasn’t considered:

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 497

 ▪ Users wouldn’t adopt the system if its performance affected
user productivity.
▪ Costs would be higher if reactive measures were required to
address performance issues later.

The project architect worked with the company’s stakeholders to

explain that while Dynamics 365 is a cloud service, factors such as
additional customizations, data volumes, and network latencies each
play a role in driving the performance of the application. Their recom-
mendations were summarized as follows:
▪ Review the application design and code for scalability.
▪ Establish baselines and review the areas that need to be optimized,
including network configurations.
▫ Formulate a baseline for form load times to understand perfor-
mance and align a new solution.
▫ Ensure performance with integrated systems meets the com-
pany’s expectations.
▫ Agree on realistic key performance indicators (KPIs) for
the business.
▫ Assess current end-to-end connectivity to determine network
performance and latency.

Based on the project team’s observations, the project steering commit-

tee was advised to approve adding solution performance testing to the
test cycle, and agreed to do so.

The customer’s architect worked with the implementation partner’s

architect to plan for performance testing by:
▪ Defining the key criteria for testing, including integration scenarios.
▪ Assessing the concurrent user loads and peak usage.
▪ Examining the data volumes.
▪ Establishing the tools for testing.

Performance testing helped the team identify several gaps before

deploying the solution to production. In addition, the team:
▪ Optimized their network configuration.
▪ Identified sync calls that were blocking resources, and made the
necessary design changes.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 498

 ▪ Reinforced and incorporated basic patterns (such as retrieving
only the required attributes) across their code components.

As mentioned earlier in this chapter, “Prevention is better than cure.”

Performance testing ensured optimal performance and stakeholder
alignment before the system was made available for end users. For
overall success and higher levels of adoption, a performance plan is
fundamental throughout an application’s lifecycle.

Implementation Guide: Success by Design: A performing solution, beyond infrastructure 499

Section
Prepare
18 Prepare for go live
19 Training strategy

500
18
Implementation Guide: Success by Design: Prepare for go live
Guide
Prepare
for go live
501
The final countdown for
the start of a journey.

Introduction
Go live is the process through which a new solution
becomes operational. It’s a critical milestone during
the deployment of a new business solution.
This is the stage in which all the parts of the project come together and
are tested and validated to ensure everything works as expected, not
just on their own, but as a whole.

A good plan strives to ensure a successful go live by minimizing un-

foreseen issues and project risks not identified during previous phases.
A good plan also has processes in place to deal with times when the
The go-live readiness plan addresses all
the activities to minimize unforeseen
unexpected does happen, which makes for a more resilient project.
issues and project risks and build
processes to help address possible
unexpected issues. The go-live process goes beyond getting the production environment
ready and deploying the solution. It includes a variety of activities such
Preparing the solution for a successful go
live focuses on activities such as training, as executing data migration and communication plans, completing train-
preparing the production environment,
ing, finalizing testing, signing off on the solution, and onboarding users.
conducting user acceptance testing (UAT),
and all the other activities described in
previous chapters. Activities should be
planned well in advance and completed
For scenarios in which there are phased rollouts, the plan should also
and signed off on before the cutover starts. ensure that additional deployments are considered in advance. It’s
And finally, the cutover, the last sprint! important to keep refining the plan and include the learnings obtained
during each rollout to improve the plan and build resilience.

Implementation Guide: Success by Design: Prepare for go live 502

 Go-live readiness includes all the key activities identified in the initial
stages of the project and refined throughout the project to ensure a
smooth go live. These activities encompass getting all required re-
sources to get the solution ready for production and ensuring that end
users are trained, engaged, and part of the process to drive adoption
and stickiness to the new solution.

When going live with a new solution, there’s often a transition period,
also known as cutover. The cutover process involves several steps that
need to be planned, executed, and monitored to ensure the completion
of all the essential activities critical to the transition to the new solution.

Go-live readiness
All the tasks and efforts undertaken during an implementation project
are preparation for the biggest milestone of the project: go live. In
the Success by Design framework phases, completion of these tasks is
when the project reaches the Prepare phase.

At this point, if you followed all the previous guidance and recommend-
ed practices, the solution should have sufficient maturity for going live.
You should perform an evaluation on the preparedness of the people,
processes, and systems for the solution to ensure no critical details have
been overlooked for go live. While you’ll never be in a position of zero
risk for go live, the go-live readiness review is a qualitative method to
determine how fully prepared the new solution is to run your business.
When the go-live strategy is aligned with best practices, requirements
for going live have been successfully completed (including testing, code,
and configurations), and there’s a concise and agreed-to plan for the
next actions required for the transition to the new system (cutover), then
the project is ready to go live. Figure 18-1 shows the Success by Design
implementation phases and when go live occurs.

To get ready to go live, you need to verify completion of the most

important project tasks. Once confirmed, you can perform the cutover
activities. The next sections discuss these topics in detail.

Implementation Guide: Success by Design: Prepare for go live 503

 Fig.
18-1

Initiate Implement Prepare Operate

Get ready to start Design Build Deploy Live

Preparation for go live Go live

Assess go-live readiness

The implementation team should assess go-live readiness early and
frequently. This provides clarity on the status, readiness, completeness,
and maturity of each activity and helps project stakeholders and the
steering committee make a go/no-go decision.

The assessment can be performed and reviewed multiple times in

distinct phases of the project. When the project moves to the Prepare
phase, the implementation team evaluates the preparedness of the
solution, which serves as a quality gate to deploy the new solution.
This evaluation is done with the go-live readiness review.

Go-live readiness review

To assess the go-live readiness of a project, Success by Design promotes
the go-live readiness review, a deep-dive review that can be performed
as a workshop. The review validates whether a solution meets an organi-
zation’s needs and expectations, both functionally and nonfunctionally.
It also validates the plan that has been established to ensure a smooth
transition in the available cutover window and to avoid any surprises
during and after go live. The go-live readiness review summarizes all the
tasks that have been completed and covered from previous reviews.

The review uncovers potential risks and issues that could imperil go live
and provides a set of recommendations and actions to mitigate them.

Start early to prepare for the go-live review. Schedule time to complete
the go-live checklist and conduct the review, accounting for time to
mitigate possible risks and issues, especially go-live blockers. Keep in

Implementation Guide: Success by Design: Prepare for go live 504

 mind that time is needed to prepare the production environment.
Starting the review just a few days away from the go-live date risks a
delay. However, starting the review too early can also be detrimental
to the process because there might not be sufficient information about
the project to complete the go-live review with reliable results to
properly identify potential risks of the solution.

To execute the go-live readiness review (shown in Figure 18-2), you

should complete all required activities for going live and have the business
sign off on them. The main requirements for going live are as follows:
▪ System integration testing (SIT), UAT, and performance testing are
completed and signed off on.
▪ Code is complete and deployable packages are prepared to move
to production.
▪ A data migration plan is prepared.
▪ A cutover plan is ready and signed off on.
The go-live readiness review is a holistic
view of the implementation project: ▪ End user training is complete with the correct security roles.
looking back on the completed and
▪ The correct number of licenses are available and assigned.
signed-off on activities required for going
live, looking at solution preparations, and ▪ A production support plan is established and signed off on.
looking forward to the plan to roll out and
maintain the new solution.
▪ Any critical open issues from the Solution Blueprint Review or
other previous implementation reviews are mitigated.

Check your readiness

for go live
Once the solution is considered ready and the go-live prerequisites are
fulfilled, it’s time to execute the go-live readiness review and evaluate
how well prepared the solution and the organization are for the next step.
Fig.
18-2

Initiate Implement Prepare Operate

Get ready to start Design Build Deploy Live

Go live readiness workshop Go live

Implementation Guide: Success by Design: Prepare for go live 505

 Use the go-live checklist to perform the review to help ensure a
smooth go live. Create a mitigation plan to address identified issues
and risks and identify an owner for each one.

Keep the review objectives in mind

When executing the review, make sure you understand the responses
on the checklist. If something isn’t clear, ask questions and validate
the responses.

Identify all the possible risks and issues and get the recommendations
on each key area. All issues and risks must have a mitigation plan.
Identify workarounds for any go-live blockers. Follow up on mitiga-
tions of issues identified for critical activities.

Stakeholders must be kept informed about the results of the go-live

review and resulting mitigation actions.
Go-live checklist
After the review is complete, the implementation team determines
Solution Scope
to be released whether the go-live date is feasible.

Acceptance of
the solution

SIT completion
The go-live checklist
The go-live checklist is a list of requirements for go live. Use it to assess
UAT completion
and validate the preparedness of your solution. The checklist includes
mandatory activities as well as recommended practices.
Performance testing
completion
In the next section we discuss in depth the main topics in the
Data migration readiness
and validation checklist. Refer to the “Product-specific guidance” section for specific
product requirements.
Confirm external
dependencies

Change management for The Solution Scope to be released

initial operations readiness
What you check
Operational support
The Solution Scope conceptualizes the final solution that will be ready
readiness
for go live. It includes all the requirements designed and developed

Implementation Guide: Success by Design: Prepare for go live 506

 during the Implement phase, ensuring their alignment with the busi-
ness needs to be able to operate once live.

Consider the following during the Solution Scope review:

▪ Business processes supported by the solution
▪ Different solutions and applications used
▪ Whether the solution contains customizations and independent
software vendor (ISV) solutions
▪ Type and volume of integrations
▪ Number of go lives or rollouts
▪ Number of users at go live and in future rollouts
▪ Solution Scope signed off on by the business

Why you check it

It helps the stakeholders understand the complexity and magnitude of
the solution. In addition, even though the Solution Scope was defined
at the beginning of the project and you’re running the project accord-
ing to it, reviewing the Solution Scope at this time in the project can
help determine additional actions and identify potential risks.

Consider the scenario of a big-bang go live containing a heavily

customized solution in which multiple integrations are delivered. The
more functionalities going live on the first day, the greater the risk
that issues will occur. If the risk is higher, be sure to assemble a good
support team and the right escalation path. The Solution Scope must
be representative of the scope of your go-live event. The recommen-
dations you follow for a big-bang go live might be different from those
for a smaller implementation.

Expected outcome
Solution Scope is aligned with the solution that’s going to be live. It has
been communicated, shared with stakeholders, and signed off on by
the business, agreeing that the expected scope is covered.

Risks of not reaching expected outcome

Unless a Solution Scope is clearly defined and shared, stakeholders
won’t have an understanding of what will be included in the solution.
In addition, the solution might not include all requirements or might

Implementation Guide: Success by Design: Prepare for go live 507

 include developments that don’t meet business expectations, which
might cause delays. A well-defined Solution Scope, on the other hand,
significantly increases the probability of project success.

Mitigation plan
Compare the Solution Scope with the solution delivered for go
live, share the Solution Scope and comparison results with the key
stakeholders, and determine whether the solution is complete or if
functionalities are missing. If any are missing, prioritize them and
assign level of risks, owners, and expected completion date.

Acceptance of the solution

Go-live checklist
What you check
Solution Scope Verify that the complete solution has been fully tested with all required
to be released types of testing and accepted by the business stakeholders.
Acceptance of
the solution Why you check it
The outcome, quality, and coverage of testing are reliable indicators
SIT completion
of the readiness level for go live, ensuring that unexpected scenarios
don’t come up after end users get their hands on the new system.
UAT completion

Performance testing As described in Chapter 14, “Testing strategy,” testing accurately gauges
completion the readiness of the solution. Testing measures a solution’s quality and
Data migration readiness effectiveness because testing simulates how the solution will operate
and validation in real life. As a result, testing builds a solid foundation on which to
Confirm external determine whether the solution is acceptable, enabling the business to
dependencies make a go/no-go decision. The business is required to sign off on the
solution when there’s objective and rigorous testing to prove that the
Change management for
initial operations readiness solution fulfills the business vision.

Operational support
readiness Expected outcome
Unit testing through end-to-end testing must be completed success-
fully and the results signed off on by the business. By doing so, the
business states that the solution as built meets their end-to-end pro-
cess needs and that those processes are ready to be executed on the
new solution.

Implementation Guide: Success by Design: Prepare for go live 508

 Risks of not reaching expected outcome
Going live with incomplete testing or without having sign-off from the
business means that you cannot validate you have a reliable solution
and therefore cannot be certain that it will work properly in produc-
tion. There’s a high risk of issues such as rollbacks, slow performance,
unstable integrations, security holes, last-minute developments, and
platform churns.

Mitigation plan
Following the best practices in Chapter 14, “Testing strategy,” you’ll
ensure the integrity and effectiveness of the solution, minimize any
go-live blockers, and avoid rushing into fixing unexpected and critical
bugs close to go live. You’ll complete the testing and ensure the integ-
rity and effectiveness of the solution.

Several types of testing are covered in the chapter about testing. Some
elements from these testing types are included in the go-live checklist
Go-live checklist and need to be validated. The main testing types are SIT, UAT, and
performance testing.
Solution Scope
to be released

Acceptance of
the solution
SIT completion
What you check
SIT completion Validate that SIT is successfully completed. It’s important to test on
expected peak volumes and have the business sign off on it.
UAT completion

Why you check it

Performance testing
completion SIT validates that the behavior of the solution, including integrations
between different applications and with external systems, works
Data migration readiness
and validation properly. During SIT it’s important not only to test and verify the inter-
actions between all parts of the solution but also to plan and prepare
Confirm external
dependencies for the unexpected.

Change management for

initial operations readiness Expected outcome
The testing should include simulations of external systems that are down,
Operational support
how those scenarios are communicated, and the impact on the user
readiness
experience. Chapter 14, “Testing strategy,” describes how to perform SIT.

Implementation Guide: Success by Design: Prepare for go live 509

 Risks of not reaching expected outcome
Don’t wait until UAT to perform SIT because this could cause significant
failures during UAT. If you execute SIT during or after UAT and discover
that the incorrect integration pattern was used and you need to rede-
sign your integrations, this will put your go live at risk.

Going live without fully testing the integrations might result in un-
steady and ineffective system connectivity, performance issues, flawed
and weak interfaces, and data flow issues such as data inconsistency
and data not being available in real time.

Mitigation plan
Complete SIT with peak volumes that are close to actual peak volumes
and get sign-off from the business that the integration testing strategy
works for the go-live. For guidance on which integration pattern to
choose, see Chapter 16, “Integrate with other solutions.”

UAT completion
Go-live checklist What you check
Verify that UAT is successfully completed (or almost completed if there
Solution Scope
to be released aren’t any outstanding significant cases) and signed off on by the business.

Acceptance of
the solution Why you check it
UAT helps validate the inclusion of all the expected functionalities. The
SIT completion solution might still have some minor bugs and need some fixes, but
you can ensure that no critical part of the solution is missing.
UAT completion

Often during UAT, edge cases that were overlooked might be discov-
Performance testing
completion ered, resulting in continuous fixes and improvements to the solution
during this period.
Data migration readiness
and validation
UAT enables users to use the system and understand how to perform
Confirm external
dependencies activities. It also provides an opportunity for users to bring up scenarios
that might have been missed.
Change management for
initial operations readiness
UAT should include different types of personas, a variety of geographic
Operational support
locations, and different environments (such as virtualized, physical, and
readiness
remote) and devices (such as laptops, mobile devices, and tablets).

Implementation Guide: Success by Design: Prepare for go live 510

 Expected outcome
UAT should include the following requirements:
▪ Test cases should cover the entire scope of the requirements—
happy path and edge scenarios.
▪ Use migrated data for testing because it validates how real data
will behave in real life.
▪ Perform UAT with the correct security roles assigned to users; don’t
use a general security role or assign the System Admin role to all users.
▪ The solution must comply with any company-specific, country- or
region-specific, and industry-specific regulatory requirements.
Document all features and testing results.
▪ UAT must be conducted in an environment in a Microsoft apps
subscription to ensure that the environment’s properties approxi-
mate your production environment as much as possible.
▪ Obtain approval and sign-off from the customer.

Risks of not reaching expected outcome

If UAT isn’t completed close to go live, the live operations might be
delayed due to various issues including unsuitable processes, perfor-
mance issues, and poor training.

If the decision is made to go live with an incomplete UAT, the project team
might discover during go live that the solution doesn’t work properly, and
recovery would be difficult, costly, and take a lot of time and effort.

Mitigation plan
Typically, UAT is considered complete when all the cases are covered and
there are no blocking issues and any high-impact bugs are identified.
Establish a mitigation plan to address open items identified during
UAT, with an owner and an estimated time of completion. Without a
mitigation plan, there’s a substantial risk of delaying the go-live date.
Without a
mitigation plan,
Performance testing completion
there’s a substantial What you check
risk of delaying the Validate that performance testing is successfully completed and signed
go-live date. off on by the business.

511
 Why you check it
Go-live checklist Overlooking performance testing might result in performance issues
post–go live. Implementations often fail because performance testing
Solution Scope
is conducted late in the project or not until the production environ-
to be released
ment is ready, when it’s used as a testing environment. A production
Acceptance of environment shouldn’t be used as a testing environment. It’s more
the solution
difficult to fix performance issues in a live environment than in a test
SIT completion environment and might also result in production downtime.

UAT completion Any open issues critical for production need to be addressed before
going live.
Performance testing
completion
Performance testing covers the following:
Data migration readiness
▪ Load times with peak volumes of transactions
and validation
▪ Data search performance
Confirm external ▪ End-to-end processes and specific business processes
dependencies
▪ Performance from different devices, browsers, and virtualized
Change management for environments
initial operations readiness
▪ Integrations
Operational support ▪ Environment for performance meets the standards to ensure it’s
readiness
reliable for testing

Expected outcome
The solution delivered is performant and meets the business perfor-
mance criteria: it supports the load of expected transactions and user
concurrency and usage, and the speed, system response time, stability,
and scalability are acceptable.

Risks of not reaching expected outcome

A common misconception is that a production environment is more
performant than a test environment, and this belief is sometimes used
to justify skipping performance testing. However, performance is not
only about infrastructure, as discussed in Chapter 17, “A performing
solution, beyond infrastructure.” Therefore, we highly recommend
that you conduct performance testing before going live in a test envi-
ronment. A production environment cannot be used to conduct core
testing; it’s used only for mock cutover, cutover, final validations, and
live operations.

Implementation Guide: Success by Design: Prepare for go live 512

It’s important If performance checks are performed just during UAT, that might not be
a good representation of the actual usage post–go live. For example, if
to identify the during UAT there’s good coverage of roles and personas, regional access,
root cause of devices, and environments but those are for a small percentage of
issues so as not users, UAT can validate the solution readiness in terms of functional and

to replicate them even end-to-end performance of each business process, but it doesn’t
represent the full load and concurrency of actual usage post–go live.
in the production Therefore, it’s important to have a separate performance testing strategy
environment. that can simulate stress on the solution and concurrency usage.

Mitigation plan
Execute performance testing in parallel with UAT. Establish a mitigation
plan to address issues and assign owners and expected completion
dates. It’s important to identify the root cause of issues so as not to
replicate them in the production environment.

Data migration readiness and validation

Go-live checklist What you check
It’s crucial to validate that there’s a well-designed data migration plan
Solution Scope
to be released for go live.

Acceptance of
the solution Why you check it
Data migration is a key activity during the cutover. You need accurate
SIT completion data on day one of your live operations.

UAT completion
Data quality is an important aspect of data migration; you don’t want
to import bad data from a system into a new system. It’s important to
Performance testing
completion have a plan to conduct data validation.

Data migration readiness

and validation Expected outcome
Data migration is tested several times prior to execution during the
Confirm external
dependencies cut­over to validate the strategy and processes, identify data corruption
or duplicated data issues and address them, and make sure performance
Change management for
initial operations readiness is measured, validated, and fits within the cutover time window.

Operational support
All scripts and processes planned for the cutover migration are tested
readiness
and signed off on by the business.

Implementation Guide: Success by Design: Prepare for go live 513

 Any migration issues and risks are logged and a mitigation plan estab-
lished. This information helps inform the go/no-go decision.

Risks of not reaching expected outcome

It’s important to carry out all the steps of a data migration plan.
Otherwise, there’ll be a lack of confidence in the information because
it might not reflect accurate data. In addition, concerns will arise about
the ability to complete the go-live migration in a timely manner,
especially if record quantities are exceptionally large.

Data quality is a key factor in the user experience. For example, if a

user’s leads are imported but a critical field is missing or has corrupted
information (for example, truncated, not completed, or without the
correct currency or value), this severely impacts how the user perceives
the new solution, reduces the trust level, and might even eventually
lower solution adoption.

Mitigation plan
Go-live checklist Execute dry runs of how to migrate the data and prepare the
migration plan. Follow more recommended practices to prepare your
Solution Scope
to be released data migration by reviewing Chapter 10, “Data management.”

Acceptance of
the solution
Confirm external dependencies
SIT completion What you check
Verify that external dependencies such as ISVs and third-party systems
UAT completion
and services are aligned with the timelines and scope for go live.

Performance testing
completion Why you check it
External dependencies are outside the direct control of the project team.
Data migration readiness
and validation This means it’s even more important for these dependencies to be ac-
counted for when building the project plan and managing the schedule.
Confirm external
dependencies
Expected outcome
Change management for
initial operations readiness If the scope of the solution includes external dependencies, the
implementation team should coordinate with them to ensure that
Operational support
expectations are aligned, requirements are met, and their solutions are
readiness
aligned with the roadmap timelines.

Implementation Guide: Success by Design: Prepare for go live 514

 All external dependencies should be documented and monitored.

Risks of not reaching expected outcome

Lack of coordination and communication with external dependencies
might result in a delay in go live.

Mitigation plan
It’s good practice to have regular meetings to review dependencies
status because problems can cause delays and other project issues.

Go-live checklist Change management for initial

operations readiness
Solution Scope
to be released What you check
It’s important to include change management at the beginning of the
Acceptance of
the solution project and assess the level of change readiness.

SIT completion Why you check it

When delivering a business solution, the success of the technical im-
UAT completion plementation and the alignment with the business needs and goals are
paramount but not sufficient to deliver the return of investment. The
Performance testing
completion true value of the business solution is realized when end users actively
adopt and engage with the processes implemented.
Data migration readiness
and validation
It’s natural to be resistant to change even if it’s positive and can sig-
Confirm external
dependencies nificantly improve one’s experience. It’s important to have a change
management plan that prepares users for changes. There are several
Change management for
resources and strategies available to achieve successful change and
initial operations readiness
adoption that include ways to communicate, reduce resistance, and
Operational support allow feedback.
readiness

Expected outcome
Our goal here is not to explore the plethora of change and adoption
strategies but to highlight the key principles and activities that are
critical before the rollout of a new business solution.

Implementation Guide: Success by Design: Prepare for go live 515

 It’s important to assess early in the project—and continually reassess—
the changes in an organization to ensure there’ll be a smooth adoption
of the system.

The following are key elements to drive change management close to

go live.

Training  For go live to be successful, it’s vital that users know how
to use the system on day one of operations. Therefore, successful user
training is key to go live.

Ensuring that users are trained helps them achieve the desired results from
the new solution and leads to higher productivity and adoption. Failure to
establish a training strategy can negatively impact usage and adoption.

The most common training strategy is a formal set of sessions delivered

in a classroom setting or online. Refer to Chapter 19, “Training strategy,”
for additional training strategies.

When possible, end-user training should be delivered before the

go-live date because this prevents delayed access to the new solution.
It can also contribute to higher adoption because the users feel better
prepared and a part of the program and had the opportunity to share
feedback before the new solution went into production.

User engagement strategy  A user-centric mindset helps address

not just the functional requirements but also the unspoken needs of
the user. Providing a story of empowerment to users—how the solu-
tion will help them be more efficient in their jobs, for example—is a
message that strongly resonates with end users.

For go live to be Your program must have an effective way to engage the end users of
successful, it’s vital the solution to help drive adoption and also eliminate the risk of build-
that users know ing a solution that doesn’t necessarily meet user requirements.

how to use the

A key activity to perform early in the project is the identification of the
system on day one super-users, or influencers, within the community who are willing to
of operations.
516
actively engage with the project team. Provide opportunities through-
out the project for these people to provide input and feedback.

Following are sample project milestones for which super-user feedback

is most beneficial:
▪ Early minimum viable product (MVP) or proof-of-concept
demonstration
▪ Sprint demonstrations during implementation to highlight busi-
ness functionality
▪ User experience and information architecture design
▪ Data mapping and validation that data migration activities are
involved in the project
▪ Baseline performance metrics for frequently accessed components
▪ Expectations for end-user devices and accessibility needs
▪ Review of the timing of key milestones such as go live and cutover
▪ Backlog prioritization and setting expectations for gaps, open
issues, and future functionality

There needs to be a well-structured nurture program for the business

users that focuses on effective communication to keep them up to date
and engaged in the program. Although most of this work happens
throughout the implementation project, the culmination of all this
planning at go live should be validated and signed off on as ready to go.

Business sponsorship  The business sponsor, or executive sponsor,

is a trusted and influential leader who enables cultural changes and
plays an essential role in championing transformation throughout the
organization by communicating the value of the new solution.

Creating an awareness of the key executive sponsors and bringing

them in front of the user community to articulate the value of the new
solution can positively influence adoption and usage.

As part of the go-live activities, the executive sponsor’s communication

strategy needs to be well planned. This includes the types of communi-
cation, the stakeholders, the timing, and recognizing the contribution
from end users.

517
 To help overcome adoption challenges, it’s crucial to get leadership
support to encourage innovative technology use. End users are less
resistant to adopt novel solutions when business sponsors serve as a
role model.

User support  Creating a community that involves the super-users

and providing resources such as videos, tutorials, knowledge base
articles, and business processes will help with user adoption post–go
live and encourage ongoing user engagement.

A well-defined channel to communicate to users about upcoming

changes, open issues, and service notifications in case of planned and
unplanned maintenance and outages should be established.

Risks of not reaching expected outcome

A poor change management plan might cause ineffective execution of
processes on the first day of go live. Users would therefore likely have a
lack of confidence in the new solution, which would be counterproduc-
Go-live checklist tive for a successful rollout.

Solution Scope
to be released Mitigation plan
Reassess the change management plan throughout the implementa-
Acceptance of
the solution tion project.

SIT completion
Operational support readiness
UAT completion
What you check
Verify that there’s a monitoring and maintenance plan for the solution
Performance testing
completion in the production environment as well as for transitioning the plan to
support teams.
Data migration readiness
and validation
Why you check it
Confirm external
dependencies Verification ensures the health and availability of the solution once it’s
live. Before go live, it’s important to plan the transition to the solution’s
Change management for
initial operations readiness support teams.

Operational support
Expected outcome
readiness
Support teams can be from the partner or customer organization, and

Implementation Guide: Success by Design: Prepare for go live 518

they need to have the necessary resources, tools, and access to sub-
ject matter experts, business analysts, process specialists, and change
management champions. Support teams should also be trained on the
solution; this ensures that tickets that come through the help desk are
properly addressed.

Notify stakeholders when all parties agree that the system is ready to
go into production and send them a list of end users who will use the
new system.

Users should be notified of any unresolved issues or required work-

arounds and provided with a support contact.

Include the following in the operational support and maintenance plan.

Continuous deployment plan  It’s important to go live in a service-

able system version in order to take advantage of latest updates, have
a better quality of the solution, and have the support available if a
hotfix is required.

It’s also important to understand the update cadence of your new

solution so that you can incorporate updates and new features and
replace deprecated ones quickly. This should be part of your mainte-
nance plan; it’s vital for the continuity of your operations in production.

For discussions of continuous deployment plans, refer to Chapter 20,

“Service the solution,” and Chapter 21, “Transition to support.”

Monitoring  Establish a monitoring strategy for performance issues,

errors, feature usage, and adoption. This makes the post–go live
process easier to manage and makes it easier to provide updates to
stakeholders about the new solution.

Hypercare  Consider including Hypercare after go live because it

supports and expedites the mitigation of issues, reducing time for
stabilization of your operations.

519
 For additional details about monitoring and Hypercare, visit Chapter
21, “Transition to support.”

Risks of not reaching expected outcome

If there isn’t a well-defined plan to support production or if the plan
isn’t properly communicated before go live, when issues start surfacing
during the operations there won’t be a way to communicate them and
assign resources and severity levels.

There’s also a risk that the new solution could go live in a system ver-
sion that’s out of service. In such a scenario, if an issue is uncovered in
production, you’ll need to update the solution to the latest version to
be able to apply the hotfix. In addition, automatic updates that install
unexpectedly might affect the deployed solution and cause outages,
unavailability, and blockings.

Mitigation plan
As discussed earlier in this section, it’s important before go live to plan
the transition to the teams who will support the solution. A support
plan enables support teams to be more proactive and preventive
rather than reactive.

Production
environment readiness
It’s critical to prepare the production environment before go live. There
are different approaches to achieve this, depending on the applications
in the scope of the deployment.

A support plan For some applications, the production environment can be created
enables support from an initial phase without any dependencies. For other applications,
teams to be more a formal process needs to be followed to create or deploy the produc-
tion environment. Additionally, several milestones must be completed
proactive and prior to deploying the production environment to help with the sizing
preventive rather of the environment and impact on factors such as performance and
than reactive. scalability of the environment.

520
 The “Product-specific guidance” section later in the chapter details
the steps for each application. For a discussion of production environ-
ments, see Chapter 9, “Environment strategy.”

Establish a plan to prepare the environments for production and

execute the following activities before the cutover. This prevents any
last-minute issues and provides more time to ensure that tasks are
completed successfully.

▪ Deploy applications and update them to the latest version.

▪ Ensure that users were created ahead of time and assigned appro-
priate security roles.
▪ Assign security groups to environments and add users to the
groups to ensure that users are created only in the environments
for which they should have access.
▪ Ensure sign-off on the security and compliance strategy including
penetration testing when applicable, compliance requirements,
and data sovereignty requirements.
▪ Tune the environment when applicable.
▪ Create, import, and validate configurations.
▪ Run and validate integrations.
▪ Ensure that any specific configurations for different applications
are completed ahead of time, especially configurations that take
time to complete. For example, setting an integration with an
email engine such as Outlook might depend on the size of a user
mailbox, and the bigger the mailbox, the longer the process takes
to complete.

Finally, keep in mind that production environment readiness is also a

big dependency for other activities during the cutover. For example,
the final data cannot be migrated and validated until there’s a produc-
tion environment to start those activities. And before data migration
can occur, other activities need to be completed, such as creating the
users and assigning the proper security roles.

Cutover, the last sprint

The last mile before going live is the cutover, which marks the transition
to the new solution. A project cutover entails the execution of a set of

Implementation Guide: Success by Design: Prepare for go live 521

 tasks required to successfully go live with a well-defined scope on a
production environment. In the past, it was common to do a big-bang
go live after a multi-year solution implementation period. Now, how-
ever, businesses frequently follow a more agile approach. For example,
a company might plan an initial go live in a small country or region or
just for the core financial processes and complete the rollout to other
countries or more complex functionality at a later point. A cutover
process needs to be defined and executed for each rollout.

In most businesses, only a brief period is available to complete the

cutover activities because during this time, users cannot use the old
business system any longer, and the new business system is not ready
yet. Often the cutover period is referred to as “the cutover weekend”
because it’s usually no longer than 48 hours. For businesses operating
all the time, this time window might be even shorter.

The cutover process, shown in Figure 18-3, involves several steps

that need to be planned, executed, and monitored to ensure that all
activities necessary to transition to the new solution are successfully
completed. The cutover process includes strategy, planning, execution,
monitoring, and control.

Fig.
18-3

Pre-cutover Cutover Go live

Strategy Plan Rehearse Cutover execution Post–go live

Timing Actions Mock cutover Runbook execution Support

Resources and Timelines Go/No-go

competencies
Owners
Communication plan
Rollback steps
Roles and
responsibilities Resources

Rehearse

Communicate

Go/No-go

Implementation Guide: Success by Design: Prepare for go live 522

 Cutover strategy
The cutover strategy begins in an early stage of project planning, is
refined throughout the project, and is completed before the cutover
plan is created.

The cutover strategy ensures alignment of the cutover plan with orga-
nizational objectives and includes the following aspects:
▪ Timing, market conditions, and other environmental aspects
necessary to go live with the new business solution
▪ Organizational readiness requirements in terms of available com-
petencies and resources
▪ Setting up the communication plan
▪ Defining roles and responsibilities

Time your go live

It’s challenging to establish a realistic and achievable go-live date be-
cause this date is set early in the implementation. It’s always important
to have some buffer time when planning the milestones dates, but this
is especially true for the go-live date.

To set the go-live date, consider the time required to complete testing
and resolve any issues that might arise, in addition to time for the
preparation of the production environment and the cutover activities.

Preferably, plan to go live when there’s a slower flow of activities in the

business. For example, in the case of a seasonal business, like for some
retailers, choose the season when sales are slow to lessen any negative
impact on the business. Also consider avoiding going live during any
busier times, for example, month or quarter end. In addition, take hol-
idays into account—they might mean low activity for some businesses
and for others might be the busiest times of the year. It’s also important
to make sure that all team members and users are fully available
Plan to go live
during that time.
when there’s a
slower flow of In addition to those busy times, consider system release dates. Microsoft
activities in the continually makes improvements in the system to deliver a better product
and has established release dates that usually entail a mandatory
business.
523
 update. You’ll need to decide whether to set the go-live date before
or after a release. This consideration also applies for any other external
applications you might be integrating with your new solution.
To determine a realistic go-live date,
consider the amount of time required for
each pre–go-live activity, including both External factors might also affect the timelines in a project, especially
the internal and external factors impacting
your project. Then add some buffer time,
in long projects during which unforeseen events such as a natural
bearing in mind that adjusting dates and disaster, political changes in the country of implementation, company
resources along the project is expected
because there might very well be delays, fraud, or a pandemic might occur. It’s important to create a list of
unforeseen events, loss and replacement possible mitigation plans for unforeseen events when planning your
of resources, and external or internal
dependencies that don’t complete on time. project. External factors are part of the risk management of your project.

Resources and competencies

During the process to define the cutover strategy, program stakehold-
ers should identify which roles and competencies are required for the
activities in the cutover plan and assign ownership for each activity.

The stakeholders should verify that all necessary resources are avail-
able not only for the requested duration of the cutover plan activities
but also to support post–go live.

These resources might need to be requested from external companies

if they’re not internally available for the project team.

Communications plan
The communications plan is an essential part of the cutover plan. This
plan identifies all the stakeholders involved in the cutover. The plan
Effective communication helps avoid
should include at least one communication channel and appropriate
uncertainty and provides visibility to
stakeholders about project status and the distribution lists, depending on the type of communication required.
results of each activity, which is important
for a successful cutover.
It’s important to identify and document the different communications
required for go live, who is responsible for triggering the communica-
tions, and the recipients list. Having a communication plan as part of
the cutover plan enables better visibility into who the stakeholders are,
at what point in time they should receive communications, and who
the points of contact are.

Roles and responsibility

Effective delegation of tasks is key for a successful cutover. Therefore,
it’s important to define which roles are needed, based on the required

Implementation Guide: Success by Design: Prepare for go live 524

 activities, skills, and responsibilities, and assign them to members
involved in the cutover. Also assign an owner to each activity and
ensure that all owners understand what’s required of them during the
Team members working together
execution of the cutover.
effectively, including during the
cutover as we’ve been discussing, is the
foundation of a successful project. All
those roles functioning together, doing
The necessary skills and responsibilities will help better define which
the right thing at the right time, will roles need to exist and to whom they’ll be assigned. It’s important
result in a successful cutover.
that those with a role assigned know what to do and how to do it—it’s
expected that each person contributes with their tasks, delivering them
on time and well executed.

Cutover plan
The cutover process is a critical and complex step that must be planned
and practiced in advance. The center of the cutover process is the cutover
plan, which lists in detail every step that needs to be executed and moni-
tored to prepare the production environment for business processes to be
executed once the cutover is complete. Cutover activities include system
configuration, data migration, data validation, and decommissioning of
legacy systems when applicable. These steps must be precisely orches-
trated to minimize disruption and ensure a successful deployment.

When creating the cutover plan, it’s important to consider and document
all dependencies, the timing of each task down to the minute, who’s
responsible and accountable for each task, instructions, verification steps,
and any additional documentation associated with each task. To arrive at
the correct timings for the cutover plan, activities must be practiced and
tested multiple times. It’s recommended to perform a “mock cutover,” or
dress rehearsal, simulating the activities of the real cutover in sandbox
environments. Depending on the complexity of the solution, it might be
necessary to conduct multiple rounds of mock cutover. The goal is that
no matter how complex the cutover procedures are, the project team has
high confidence that final execution will work flawlessly.

A successful cutover plan is when there are zero or few disruptions

during the execution of the cutover and a plan is in place to mitigate
any breakdowns.

Implementation Guide: Success by Design: Prepare for go live 525

 Components of a cutover plan
A successful cutover activity always begins with a detailed and well-de-
fined cutover plan. The cutover activity itself is performed as part of go
live, but planning for the cutover should start at an early phase of the
project, and the business should sign off on it. The cutover plan is a living
document that is updated throughout the project as the solution is
implemented and evolves.

All project stakeholders should be involved in each activity. Project

stakeholders should sign off on activities, with a commitment that all
necessary resources are available for the requested duration to support
the go-live activities.

A clearly defined plan including detailed actions, timelines, and owners

needs to be established. The plan should also include a rollback sce-
nario, which adds to the value of the cutover plan. Planning a rollback
is especially important to ensure that the business is not affected by
the outcome of the cutover plan if it isn’t successful and that there’s a
Fig.
18-4 backup plan to retry the cutover at a later point.

Cutover plan The implementation team should rehearse all the cutover plan steps
example to ensure that the cutover strategy and communication plan are ready.

Go-live date: 8/1/2021

Today: 6/28/2021

Estimated Estimated Validated Actual Actual %

Cutover step Area Environment System Countdown Status Duration Owner
start end by start end complete

Finalize configuration environment

Customization
10 IMP Config Dynamics 3D Planned 1D 7/1/2021 7/1/2021 Team 1 PM 0%
installation
Configurations
20 IMP Config Dynamics 5D Planned 2D 7/2/2021 7/3/2021 Team 2 PM 0%
validation
Run data sharing Member
30 IMP Config Dynamics 6D Planned 1D 7/4/2021 7/4/2021 PM 0%
policies 1
Member
40 Take backup IMP Config Dynamics 7D Planned 1D 7/5/2021 7/5/2021 Tech lead 0%
2
Move to sandbox
50 IMP Sandbox Dynamics 7D Planned 1D 7/5/2021 7/5/2021 Team 3 Tech lead 0%
(UAT)
Backup applied Member
60 IMP Check TIME Dynamics 7D Planned 1D 7/5/2021 7/5/2021 PM 0%
to UAT 1

70

80

Implementation Guide: Success by Design: Prepare for go live 526

 By doing so, the predictability of the outcome increases, backup plans
can be created to avoid dependency issues, the duration of each activity is
validated, and any identified issues with the plan can be corrected. Figure
18-4 is an example of a cutover plan and shows the list of activities, the
different elements for good traceability, and execution of the cutover.

All key stakeholders should attend go/no-go meetings to review any

open issues, risks, and unresolved defects to collectively decide wheth-
er the application is ready for production.

The decision should be made according to the success criteria defined

during the cutover plan. For example, one of the go/no-go decision
KPIs might be the number of blockers or high-impacting bugs identi-
fied from UAT.

The cutover plan should also specify whether there are workarounds or
additional plans that might prevent any issues from delaying go live,
for example, if a blocking issue was identified during UAT but there’s
a workaround that can be implemented until a permanent fix can be
applied. The workarounds should take into consideration the impact to
end users. After all, users judge a solution by how well they can use it,
regardless of whether there’s a workaround. This is another reason why
all key stakeholders need to participate not only to define the success
criteria but also to make the final go/no-go decision.

Cutover execution
The cutover execution is the series of activities carried out in adherence
to the cutover plan. Cutover execution includes the following steps:
▪ Communicate activities, milestones, and results to stakeholders in
accordance with the communication plan.
▪ Ensure that all activities are executed and completed and that
any unforeseen issues are addressed or can be addressed after go
live, have been communicated, acknowledged by stakeholders,
and documented.
▪ Execute the activities.
▪ Monitor the execution of the cutover activities.

Implementation Guide: Success by Design: Prepare for go live 527

Cutover entrance criteria
Complete the following activities before initiating the cutover execution:
▪ Complete test cycles (performance, SIT, and UAT) with exit
criteria satisfied.
▪ Build and test data migration strategy including configuration data.
▪ Identify all mock cutover and cutover resources.
▪ Complete internal and external communication plans.
▪ Ensure that the production environment is ready.
▪ Complete the final cutover plan.
▪ Ensure the final go-live support plan is in place.
▪ Ensure end-user training plan will be completed by go live.
▪ Complete data validation plans.

Cutover exit criteria

Execute the following activities to complete the cutover execution:
▪ Execute all cutover steps in sequence.
▪ Ensure the steps fit in the time window established for the cutover.
▪ Document all failures and issues and resolve or identify solutions
to resolve them.
▪ Execute data validation plans successfully.
▪ Obtain sign-off at cutover go/no-go checkpoint.

If you execute a mock cutover, it’s crucial to validate the duration time
of the different tasks and their sequences so that you can achieve a re-
alistic plan for the actual cutover. This validation also helps you identify
the need for mitigation plans and whether you need to run additional
mock cutovers.

Successful go live and transition to support

Once the cutover execution is completed, the new solution is released
to end users. At this point, the solution becomes operational, and
it’s crucial to ensure a smooth handover to the operations teams,
following the operational support plans discussed in the “Operational
support readiness” section in this chapter. For additional information
on support, visit Chapter 21, “Transition to support.”

528
 Product-specific guidance
Operations
Operations must follow the go-live readiness review process to go live.
This review process acts as the quality gate for releasing the produc-
tion environment, which will be deployed only if the go-live readiness
review has been completed.

For Operations, consider the following:

▪ Lifecycle Services (LCS)  LCS is the workspace for managing all
aspects of your operations project, including project milestones,
environments, updates, deployments, support, and project
administrators. As part of the go-live readiness review, the LCS
workspace needs to be prepared with the correct configuration
and information such as milestone dates and project users and ac-
counts. In addition, create the Subscription estimator for sizing the
production environment. Familiarity with the LCS functionalities
for go live, including deploying a package, requesting a database
refresh, and requesting your production environment, will help
you efficiently run all the preparations for the go live.
▫ LCS also contains tools such as LCS environment to monitor,
diagnose, and analyze the health of the environment and
troubleshoot issues if they occur.
▪ Continuous updates  Ensure that your environments comply with
the One Version policy. This policy specifies that an environment
must have a version no older than four monthly service updates
from the latest available one. Set your Update Settings in LCS for
continuous updates.
▫ Review the targeted release schedule to verify that the version
of your environments hasn’t reached the end of service. It
won’t be allowed to deploy production in an older version.
▫ Regression testing is needed when updating or deploying fixes or
making changes to the solution. The Microsoft Regression Suite
Automation Tool tests regressions due to continuous updates.
▪ Upgrades  For Operations, upgrades from Dynamics AX 2012 are
common. It’s important to perform an assessment for go live with

Implementation Guide: Success by Design: Prepare for go live 529

 the specific requirements when upgrading your solution, such as
the following:
▫ Use the collation (SQL_Latin_General_CP1_CI_AS) because
it’s the one supported by the Finance and Operations apps
kernel/runtime.
▫ Run and execute all pre-upgrade checklist steps on the source
system before beginning any database upgrade activity.
▫ Store documents and attachments in Microsoft Azure
Blob Storage.
▫ Freeze all code and application configuration changes in the
AX 2012 environment.
▫ If you’re using a mobile solution in the warehouse to interact with
Dynamics 365, whether you’re using the standard Warehouse
Mobile Devices Portal platform in AX 2012, you should move to
the Dynamics 365 warehouse mobile application.
▪ Cutover  Figure 18-5 shows the ideal cutover sequence for
bringing data into the production environment and validating the
data. This works for Operations.

Dynamics 365 Customer Engagement

To go live with Customer Engagement apps, key stakeholders should
perform a go-live readiness review to understand whether the plan
covers all key activities, activities that can be executed ahead of time,
and the plan to transition to operations.

Fig.
18-5

Opening
Move Master data
Configurations balances in Mock go live Cutover
configurations in production Live
complete production complete complete
to production complete
complete

Configured Back up and Additional data, Load opening Simulate real-life After running From this point,
manually or with restore DB in like master data, balances with operations for a the mock go no more database
data entities/ sandbox and is added on data entities small period to live, restore movements
packages. production. top of restored to reflect a confirm solution production DB allowed, just data
configurations real operation stability in to any previous increments using
manually, with scenario or as production. reusable data entities or
data packages or final load. milestone. manually created.
integrations.

Configurations
environment
Production environment

Implementation Guide: Success by Design: Prepare for go live 530

 Go-live readiness consists of understanding the plan that helps ensure
all recommendations and best practices are being followed to antici-
pate any unforeseen issues during the last sprint.

When engaging with the FastTrack team, this review is conducted in a

workshop format and includes all parties.

This section provides key information that enables customers and

partners to perform their own go-live readiness review.

Test and acceptance

In the Prepare phase, the functionalities should be completed (code
complete), integrations should be tested with all external systems (SIT),
and a high-level deployment execution plan should be in place. This
will trigger the solution test and acceptance activities. In this phase, the
nonfunctional requirements (NFR) of the solution that were previously
agreed to and signed off on—for example, form load times, search
performance, and integration performance under the realistic produc-
tion load—should also be validated.

Several tools are available to aid with building testing automation. One
such tool is Easy Repro, an open-source library intended to facilitate
automated UI testing.

Training and adoption

To help with in-app contextual training, guided help functionalities
are available to give users a custom, in-app help experience tailored to
specific usage and business processes.
Post–go live,
management of The Dynamics 365 adoption guide is an easy-to-follow how-to guide
on the best way to roll out Dynamics 365 to your organization.
your solution is
important when Capacity management
migration activities Management of your solution is important, especially post–go live,
when migration activities are complete and users are actively using the
are complete and system. Storage usage and growth need to be validated; using storage
users are actively above the limits has an impact on some features, such as on demand
using the system. and restoring backups.

531
 Monitor API limits because after go live, there might be integrations or
bulk operations that cause peaks in the usage, resulting in API limit errors.

Data migration
You can use several tools—out of the box, ISVs, and custom built—to
migrate data. It’s important to follow best practices and include them
as part of your migration strategy.

Various factors can impact data migration activities, for example, the
service protection API limits that ensure consistency, availability, and
performance for everyone. Keep these in mind when estimating the
throughput and performing testing and the final data migration activities.

Requests limits and allocations can also have an impact and should be
taken into consideration.

Production environment readiness

When using Customer Engagement apps, the production environments
can be initiated at any point in time. It’s important that apps in production
are the same version as those in the development and test environments;
otherwise, issues might occur that impact the release process.

Several tools help you prepare for go live, including the following:
▪ Power Apps checker helps validate model-driven apps as well as
canvas apps and should be used throughout the implementation.
▪ Microsoft Dataverse analytics enables access to metrics that help
you monitor the solution, especially during the operations phase.

Continuous updates
In the words of Mo Osborne, Corporate Vice President and Chief
Operating Officer of the Business Applications Group at Microsoft, “To
enable businesses everywhere to accelerate their digital transformation,
we are continuously enhancing Dynamics 365 with new capabilities.”

Service updates are delivered in two major releases per year, offering
new capabilities and functionalities. These updates are backward com-
patible so that apps and customizations continue to work post-update.

Implementation Guide: Success by Design: Prepare for go live 532

 Upcoming features and capabilities can be validated months in advance
with early access, before they’re enabled in production environments.

To support customers and partners throughout the release schedule,

the release plans and early access are documented and available well
in advance.

Conclusion
As mentioned at the beginning of this chapter, go live is a critical
milestone during a deployment—every plan and test is validated at
this stage.

This chapter isn’t intended to explain how go-live readiness and the
cutover plan should be built; there are other resources that cover these
topics. The aim of this chapter was to provide a summarized explana-
tion of the key activities for go-live readiness and their purpose and
the important takeaways, best practices, tips and tricks, and common
pitfalls that can happen during this phase and their impact on how the
new solution is received and adopted by end users.

We discussed the importance of training, adoption, and change man-

agement plans and how these plans contribute to the user experience
and adoption of the new solution.

Planning and preparing are key to avoiding delays that can impact
the deployment execution. All resources must be available in time for
activities to be executed and completed. Environments must be creat-
ed, configured, and operational on time.

Planning and Data migration can be complex. Realistic goals and good throughput are
important to ensure that there are no delays that might impact the go-
preparing are key
live date and that all tasks are executed according to the data migration
to avoiding delays plan. Ensure that the recommended practices to improve throughput are
that can impact in place. Validate that throttling limits aren’t being reached.
the deployment
There should be coordination between the implementation team and
execution. external dependencies to make sure that expectations are aligned and
533
 requirements met. Any external solutions must work as expected with
the current version of the solution and platform and be aligned to
the roadmap.

Support during cutover, go live, and post–go live is essential for all
activities to run in the expected time. It’s crucial to have resources
available to mitigate any issues so that tasks can complete on time
or with as slight a delay as possible. A poor support strategy for a
successful rollout will impact the success of the rollout over time. It’s
important to create a list of responsibilities for first, second, and third
lines of support as well as an escalation path and to transition knowl-
edge from the implementation team to the support team. Ensure that
there’s coverage outside traditional working hours and on weekends
for critical situations.

Ensure that application and platform monitoring is in place, and set

up appropriate alerts so that the operations team can take necessary
actions and notify users. A strong monitoring and communication plan
has major impact on adoption and confidence in the solution. All crit-
ical application components should have a logging mechanism with a
well-defined process to monitor the system and alert administrators.

A process needs to be in place post-deployment and post–go live to

manage changes (such as upcoming updates or new versions of ISV
solutions) and prepare users for changes and the adoption of new
features and capabilities.

References
GitHub - Microsoft/EasyRepro: Automated UI testing API for Dynamics 365 Dynamics 365 release schedule and early access

Use solution checker to validate your model-driven apps in Power Apps Prepare for go live

Microsoft Dataverse analytics Report production outage

Microsoft Dynamics 365 Targeted release schedule (dates subject to change)

Create guided help for your Unified Interface app One Version service updates FAQ

Dynamics 365 Adoption guide Monitoring and diagnostics tools in Lifecycle Services (LCS)

Accelerate Dynamics 365 Adoption checklist Evaluate go-live readiness for Dynamics 365 Customer Engagement apps

Build a champion program

Implementation Guide: Success by Design: Prepare for go live 534

 Checklist

Go-live readiness Ensure the production environment is prepared and

administrators and IT are familiar with monitoring,
Assess go-live readiness early and often through quality troubleshooting procedures, and avenues of support.
gate reviews to provide clarity on the status, readiness,
Create a cutover plan that considers all dependencies,
completeness, and maturity of each activity and to help
the timing of each activity, roles and responsibilities,
project stakeholders and the steering committee make
instructions, verification steps, and any additional
go/no go decisions.
documentation associated with each activity.
Ensure that the solution scope is aligned with the solution
that’s going to be live and that it has been communicated,
Cutover
shared with stakeholders, and signed off on by the busi-
ness, agreeing that the expected scope is covered. Implement the cutover plan by completing all activities
in sequence, satisfying the exit criteria, and including the
Ensure that the solution has completed all test cycles,
sign-off at the cutover go/no-go checkpoint.
which include system integration, performance, and
UAT, with exit criteria satisfied and signed off on by the
business stakeholders.

Ensure all scripts and processes planned for the

cutover migration are tested and signed off by the
business stakeholders.

Ensure external dependencies such as ISVs and

third-party systems and services are aligned with the
timelines and scope for go live.

Plan all change management activities such as training,

user support, user engagement, and communication,
and implement all pre-go-live change management
activities from Initiate through Prepare phases.

Have a plan for monitoring and maintenance routine for

production as well as for transitioning to support teams.

Implementation Guide: Success by Design: Prepare for go live 535

 Case study
The gift: go live
in a cloud solution
A toy company recently embarked on an implementation project and
migrated all their processes from on-premises to Dynamics 365. They
had integrations with Dynamics 365 Customer Engagement, Microsoft
Power Apps, and Power BI. The company plans to implement Dynamics
365 Commerce and Warehouse Management mobile app in Microsoft
Dynamics 365 Supply Chain Management in future rollouts, but they
needed to move all financial data and their CRM system to the cloud in
time for the end of the year, their busiest time.

At the end of October, the team was excited and a little nervous
about the approaching go live. The impact of changing from a system
in place for many years would be huge. The team was used to their
manual and inefficient processes but were ready to move to a system
that would automate and improve processes and increase revenue
growth and profitability along with scalability. Change management
was critical for success.

The implementation team needed to make sure that everything was

ready for going live. With go live four weeks away, it was time to start
reviewing their readiness for the implementation. SIT was completed,
and UAT and performance testing were almost completed. The system
seemed performant—although not optimal, it was considered ready
for production.

Implementation Guide: Success by Design: Prepare for go live 536

 The team was ready to start the go-live review, to switch to the Prepare
phase. They prepared the go-live checklist and asked themselves the
following: Everything was thoroughly tested, but was the system really
ready? Were the users ready? Was all the infrastructure ready? Were all
the systems up to date? Was the mock cutover planned? Were the ISVs
working well?

Concerns came up over risks and issues identified during the as-
sessment. For instance, they realized that the mail service and other
services were in another tenant, so they needed to perform a tenant
move to enable single sign-on for their users. In addition, Microsoft
was planning an automated version release by their go-live date, so
they needed to update their environments to the latest version. This
would necessitate conducting another smoke test and verifying that
the ISVs worked correctly with the version. They missed dates because
they were focused on UAT and making fixes and addressing new re-
quirements identified during the testing phase.

The team realized that open transactions weren’t ready to be closed in

the old system and therefore they created them in the new one. The
data migration plan wasn’t solid. In addition, performance was weak,
so the team needed to troubleshoot the code.

Business stakeholders made a go/no-go decision after this assessment.

It wasn’t an easy decision, but they decided not to go live. Despite all
the heavy lifting they had done, they didn’t have a mock cutover plan,
which they realized was necessary to complete the readiness process.

Would they need to change the go-live date that was so close to the
holiday season?

What could they have done better? They couldn’t have changed the
date because they needed to be ready for the holidays, but they could
have planned the go-live date earlier so that they had more time for
the ramp-up and to address any delays or issues. They could also have
had an earlier assessment review, with UAT almost complete. The time-
lines were tight, and they had the choice to go live with what they had,
which was risky because that might mean stopping operations in the

Implementation Guide: Success by Design: Prepare for go live 537

 middle of the busiest season, or move the go-live date to January—
December wasn’t an option. A third option was to hire more resources
and work more hours to try to deliver on time, which was risky because
the ramp-up would be during November. A big-bang go live wasn’t
optimal for this situation.

The team learned that it’s important to start the readiness review on
time and to schedule sufficient time and resources. Additionally, it’s
crucial to have a solid support plan for production. The importance
of the Prepare phase also shouldn’t be underestimated—plan with
enough time to mitigate any issues and risks.

Implementation Guide: Success by Design: Prepare for go live 538

19
Implementation Guide: Success by Design: Training strategy
Guide
Training
strategy
539
“There is no saturation
point in education.”
– Thomas J. Watson, founder of IBM Corporation

Introduction
At its core, the goal of a training strategy—and
training—is to ensure that all necessary users of
your system are educated on the new application so
that their knowledge of how to complete their work
results in successful user adoption following go live.
Training is not the only factor in meaningful user education and adop-
In this chapter, we define a training
strategy and determine at a high level
tion; empowering users to perform their necessary tasks in the system
what components you need to embark correctly and efficiently should be the ultimate “why” in the develop-
on a methodical approach to a successful
training execution for your Dynamics 365 ment of a training strategy. As part of user adoption, organizations
implementation. Each of these areas is
should strive to train in a way that gives users confidence in the appli-
covered in detail in this chapter.
cation and inspires a sense of delight when using the system.
For a successful training
strategy consider these
In this chapter, we cover several high-level objectives, as well as exam-
main areas:
• Training objectives
ples of more organization-specific objectives, that should be included
• Training plan in your organization’s training strategy. A proper training strategy
• Scope
• Audience should center around the creation and execution of a comprehensive
• Training schedule
training plan. Furthermore, the training plan should align to the broader
• Training material
• Delivery approach training strategy of your Microsoft Dynamics 365 implementation. We
• Assumptions, dependencies, and risks
• Training as an ongoing process
discuss how to create an appropriate scope for your organization’s
• Training best practices training as part of this plan, as well as how to confirm the different
groups of users that need to be trained.

Implementation Guide: Success by Design: Training strategy 540

We cover training topics that should be part of your organization’s
overall project plan, and how best to keep your training plan and project
plan coordinated to avoid significant issues. This chapter also dives into
guidance surrounding different types of training material and their ef-
fectiveness, as well as how to create a training delivery plan and conduct
effective training that can serve all your users (before and after go live).

Dynamics 365 is an ever-evolving and growing application. Because of

this, training is an ongoing process that requires constant adaptation
and updates, and we cover how to best prepare your organization for
this continual update process. We also recognize that training strategy
and guidance can vary depending on your choice of Dynamics 365
application. We provide suggestions on how best to successfully adopt
a training strategy, whether your organization is implementing appli-
cations built on Dynamics 365 Customer Engagement, Finance and
Supply Chain Management, or both.

Lastly, we discuss best practices by highlighting the behaviors and

strategies that we have seen organizations successfully employ when
developing a training strategy. We share our collective experience
regarding these best practices toward the end of the chapter, in the
hopes that you will carry them forward in your own projects.

By following the guidelines laid out in previous chapters of this book, your
organization should be well on its way to rolling out a successful Dynamics
365 application that fulfills the needs of your business and users. Successfully
training your users on using modern technology, as well as understanding
business processes that they will need to incorporate to do their jobs safely
and efficiently, is integral to any organization’s training strategy.

Training objectives
One of the first things that your organization should consider when
beginning to develop a strategy surrounding training users, and a
plan behind it, is the objectives. Defining the objectives of a successful
training strategy is key—and it can help shape the crafting of a training
plan as well as lead to a more successful rollout of training itself.

541
 At a high level, every Dynamics 365 application training strategy
should include some version of the following objectives. Many change
management objectives can also be addressed by ensuring that your
training strategy explicitly addresses these training-focused objectives:

▪ Employees will gain necessary knowledge required to use

Dynamics 365 applications as a part of their day-to-day jobs in a
safe and efficient manner.
▪ Training should be meaningful and not overwhelming, to provide
users with an efficient way to do their jobs.
▪ Training should help support business goals for the application
(in other words, employees understand the “why” behind training
materials and scenarios).
▪ Training should be inclusive of everyone (for example, experienced
users as well as new hires) so that all users of your application
receive the education necessary to do their jobs.
▪ Training should be an ongoing process that does not stop when
the system goes live.

Your Dynamics 365 implementation will have additional training

objectives that are specific to the applications you are developing,
and to your organization. It is important that you work to define these
objectives as they, along with the more general objectives of training
outlined earlier in this chapter, should formulate the “what” of your
training strategy goals. An effective way to evaluate your organiza-
tion’s training goals is to follow the SMART acronym—that is, specific,
measurable, achievable, realistic, and timely. While it might not be pos-
sible for every training objective to meet every one of these metrics,
we recommend that you create training objectives that are relevant to
your organization’s goal of user adoption and fit the outlined criteria.

What not to do
As a counterpoint to the recommended strategies, let’s spend a little
time discussing the training objectives that an unprepared organization
might create.

In this example, let’s say the organization creates the following objective:
“Objective: Dynamics 365 Field Service users should be trained on the

Implementation Guide: Success by Design: Training strategy 542

 mobile application.” While this objective captures at a basic level what
needs to be done, it falls short of accomplishing what a proper training
objective should. When rewritten using the previously mentioned
criteria, and separated for increased detail and specificity, the objec-
tives become much clearer:
▪ Objective: 100 percent of Field Service technicians attend the
in-person mobile training session or view the complete online
seminar series prior to system cutover.
▪ Objective: All Field Service technicians finalize the Work Order
completion process on their mobile devices without assistance,
and are able to complete the lead-generation process on their
mobile devices in under three minutes.

These rewritten objectives also serve as good examples of making sure

your training objectives specifically relate to the application being
deployed. A poorly written training plan often does not include
application-specific objectives; in addition to not following the SMART
philosophy, these objectives aren’t thorough enough to capture the
essence of the goals of training. Clearly defined and measurable training
objectives also help an organization determine the effectiveness of its
training. It is significantly easier to judge the success of training in your or-
ganization if you can track against objectives that have been well-written.
Later in this chapter we discuss specific ways to assess the effectiveness of
your training; this process starts with creating proper training objectives.

Ensuring that your users are comfortable with the most challenging
obstacles to user adoption is important to achieving training success.
It is significantly While your training objectives do not need to specifically reference
easier to judge tasks or challenging business processes directly (as these would be too

the success of specific and thus not useful), they should reflect knowledge of areas of
your application that may require additional attention from a learning
training in your perspective. The mobile objective in the previous paragraph is a good
organization if you example of this strategy because we recognize that Dynamics 365 mo-
can track against bile apps can represent a significant change in business processes for
users and, therefore, could require additional attention from a training
objectives that and user adoption perspective. Thus, a key objective should be that all
have been well- users receive proper training on this specific job function.
written.
543
 Creating a training plan
Proper training is critical to user adoption. Organizations must develop
a training plan at the start of the project and engage resources from
the beginning.

It’s important to remember that managing a training plan is an iterative

process. As you progress in the project and learn more about your busi-
ness requirements and training needs, you might have to update your
plan. This sample training plan charter gives you something to start with.

We recommend you put together a training plan that, at a high level,

consists of at least the following elements (as illustrated in Figure 19-1):

▪ Training objectives
▪ Scope
Fig.
19-1 Training plan elements ▪ Audience
▪ Training schedule and re-
source availability (project
Training
objectives planning)
Training
Scope
resources ▪ Delivery approach
▪ Validation of training
Training as success/objectives
an ongoing Audience
process ▪ Assumptions/dependencies
▪ Risks
▪ Training environment
Training
schedule and management
Training
materials Training plan resource
availability
(project ▪ Training materials
planning)
▪ Training as an ongoing
process
Training Delivery ▪ Training resources
environment approach
management

Just as the project plan is key to

Validation
of training
project timelines and execution,
Risks
success/
Assumptions/ objectives so too is a training plan key to
dependencies
training timelines and training ex-
ecution. End-user training is one
of the last tasks on a project with

Implementation Guide: Success by Design: Training strategy 544

 dependencies on several key items, such as a near-complete solution,
good quality representative business data, and reserved time from end
users. If a project is running even a little late, the final preparations for
training often get squeezed. A robust, well-thought-through training
plan that reduces the number of items in the critical path is essential to
avoid last-minute compromises and lower-quality training outcomes.

Scope
A crucial element of ensuring that your organization’s project has a
proper training strategy is defining an accurate scope for the training.
Here are different areas that cover the breadth of scope that you will
need to build into your training plan:

▪ Business processes completed by end users, derived from your

user stories and functional requirements
▫ Example: Users must be able to create and process an invoice
from end to end in Dynamics 365 for Finance and Operations.
▪ Any nonfunctional requirements that require training users for the
requirement to be met
▫ Example: Users should have the ability to provide feedback
directly from the application.
▪ Any administrative tasks that users might need to accomplish
while in the application, separate from their business-driving work
▫ Example: Users must be able to configure and use the
Advanced Find tool inside Dynamics 365.
▫ Example: Users should be able to upload templates using the
Dynamics 365 data management tool.

There could be other areas of your application that should fall into
the scope of your training. If there is a task or process that your users
need to accomplish while working directly or tangentially with your
Dynamics 365 application, it should probably fall into the scope of your
training, as you should assume there are users who need to be trained
on that topic. An example of this would be training project team mem-
bers during your implementation who will need to work inside the
application from a development or configuration perspective.

Implementation Guide: Success by Design: Training strategy 545

Once you have the full set of business processes, requirements, and
tasks, you can use them to create a true scope, consisting of actionable
training areas, for your training. There might not be a one-to-one rela-
tionship between them, as there may be overlap between processes
as well as additional training topics that need to be added. This is our
guidance on how to avoid overlap and how to add more training topics.

▪ Avoiding overlap  If you have 10 sets of business processes that

are different in terms of function but almost identical in terms of
system execution, and each takes 30 minutes for a user to com-
plete, consider setting your system training areas to represent the
similarity between completing these in your application, rather
than to represent the disparity between business processes. You
can still reference different business functions and even take time
to walk through any differences in execution, but it is important
that the scope (that is, each training area) is framed in such a way
to be efficient and not duplicate training effort.
▪ Additional training topics  If you require users to switch between
their mobile and desktop devices frequently, with minimal delay
between each, it might be important to cover this transition even
if it is not necessarily covered in any business processes or other
requirements or tasks.

The next step in setting the scope for your training is to categorize
and prioritize the list of training areas that you have created. As an
example, some business processes in your new application might
be different from the “as is” processes in your organization’s current
application. This is important for a few reasons:
▪ System training and business process training are two separate
efforts, even though training might address both types.
▪ Training surrounding changed business processes should be given
special attention in training sessions and in training material.

It’s common for business process training challenges to affect system

training effectiveness. As part of your scope of training, and combined
with your training materials and delivery, you will want to separate
system training on the new processes and application from topics that
veer more into organizational change management.

546
 Also, you should prioritize, or at least highlight, any training related
to critical nonfunctional requirements that have impacts on legal
regulations or company policies. For example, if your organization is
legally required to fill out fields on your sales order form in a particular
manner, this section of training should be emphasized so that users are
aware of its importance.

One process that might help determine priority of training areas is

the creation of a change-impact assessment, whereby you detail the
changed areas, the topics and stakeholder groups that are affected,
and the impact that the changed areas will have. This will allow you to
have a consolidated list of training topics, an estimated list of people
that will need training on that topic, and the beginnings of a prioritiza-
tion around each.

Figure 19-2 shows an example of a change-impact assessment.

Audience
Now that the objectives for your organization’s training strategy have
been defined, along with the scope of content that training should
contain, we can define the next piece of the training puzzle: who to
train, and what type or types of training they should receive.
Fig.
19-2

Change impact Training topic Impact (1-5) Action

The purchase department’s Procure to pay 3 Trainers will hold a series of

legacy system will be replaced business process webinars detailing differences
with Dynamics 365 Supply Chain between the legacy system used
Management and business wants for purchasing and Dynamics 365
to use the standardized process in Supply Chain Management
Dynamics 365

Field technicians are moving from a Using the new mobile 4 Because of the impact this change
pen-and-paper-based work order application to complete has, multiple live training sessions
management system to a mobile a work order will be held for technicians to ensure
application proper education

Implementation Guide: Success by Design: Training strategy 547

 Who receives training?
It is critical to identify the different groups of people who need
training. Dynamics 365 offers excellent role-based access control
and application segregation as a means of providing access to your
users based on how they need to use your application. As a result, it
is important that all employees who will need to use the product are
trained in the correct way, on the correct topics.

The following section identifies the core groups of people who should
be trained on your Dynamics 365 application. Note that each group of
users can be broken into subgroups. Breaking down these users into
personas is important to successfully executing training; if personas are
incomplete or incorrect, your organization might not complete prop-
er training on the correct business processes or, worse, it might not
complete training at all for specific user groups.

System end users

System end users are the people who will use the system daily and drive
direct business value. Your application could have multiple types of end
users, separated by licensing, security role, or even the application itself
(Customer Engagement versus Finance and Operations users). Each of
these groups will require different training. This group of users might
cover multiple departments and job functions; as a result, the training
should both be general, as to provide high-level knowledge of your
application regarding key core entities that all users may need to access,
as well as specific to aid these users in their specific job functions.
Fig.
19-3
As part of your training plan, be sure to identify each business role that
will need to use your application differently and will require training. This
Salesperson Sales team lead is related to, but distinct from, creating groups of users based on security
• Creating records role. Certain personas might have identical security roles in Dynamics
• Updating records
• Sending emails
365 but different job functions; these groups of users will need to use
your application differently and will, as a result, require different training.

• Approving records
• Creating and As an example, consider a salesperson and a sales team leader (see Figure
viewing charts
• Dashboards 19-3). Even if your organization has defined that their required access in
your application is similar enough to warrant identical security roles, it

Implementation Guide: Success by Design: Training strategy 548

 is important that these groups of users are treated distinctly during this
portion of your training strategy and receive correct training. Salespeople
will need to be trained more on the “doing” functions of your application,
like creating records, updating records, and sending emails.

Sales team leaders, on the other hand, might work on similar tasks as
salespeople, and they might require some or all of the same training, but
they will also need additional types of training, such as reviewing and
approving records, and creating and viewing charts and dashboards.

Another benefit of being specific when defining training personas

In many organizations, certain end user based on business role rather than security role is the value it can
roles have higher turnover than other
provide for new hires. Having a new hire in a specific business role un-
personas receiving training, so your
training material should be structured in a dergo an exact set of training based on what was identified previously
way that can help staff onboard easily.
in this chapter can make onboarding significantly easier and reduce
time in trying to match job titles to security roles, and then back to
business processes in your application.

Trainers
Trainers are a special category of users that require training. They
should be knowledgeable in all areas of the system and are respon-
sible for developing training material and training other users. “Train
the trainer” is a common approach to onboarding this type of user
to your application. Depending on the size of your organization, you
might have several trainers. Since the goal of your trainers is to create
material and conduct training for your other users, it is important that
at the end of “train the trainer” sessions, your trainers are aware of and
able to perform their job function—training other users—effectively.
The goal of training these users is different from that of your end users;
trainers should be part of the training strategy process and will be
advocates of the new application.

Super users
Super users are an elite group of (usually) end users who act as cham-
pions of the application. This group is key to driving excitement and
adoption of the new system. These users are often involved in early
feedback cycles with the project team. As subject matter experts (SMEs),

Implementation Guide: Success by Design: Training strategy 549

 they become educated on the application early on so that by the time
end-user training begins, your super users are experts in various business
processes. As part of your training delivery plan, your organization can
use super users strategically (by having them attend training sessions
and serving as application advocates) to aid your trainers in their job
functions. Additionally, these super users can act as a “first line” to help
answer employee questions—keeping your support desk from being
overwhelmed and at the same time keeping morale high.

We have mentioned several times that meaningful user adoption is

critical to the success of an application and relies heavily on a
well-executed training strategy; super users are instrumental in this
process. If your organization is deploying several different applications,
it is recommended to have multiple super users, spread across each
application, and corresponding to their area of expertise.

Project stakeholders and executives

If your project stakeholders and executives need to be able to access
and use your application, they need to receive training as well. This
training should be tailored to their specific role.

Executive buy-in is important for the success of a new application, and

meaningful user adoption applies no differently to this group of users
than it does to your end users. To take advantage of the additional
effect that executive buy-in may have, you should consider “visibly”
training these users so that other members of your organization are
aware that training is ongoing. You can also notify these stakeholders
of training that other roles are expected to undergo, which can help
from a buy-in perspective.

Support desk and administrators

Another critical group of people that will need training will be your
Executive buy-in support and administrative users. While these users might need a
certain amount of training on the application and certain business
is important for processes, they will need to become familiar with the application from
the success of a a maintenance perspective. Training for these users will focus on tasks
new application. related to system maintenance, troubleshooting, and administration.

550
 Training scope
Once you have defined the distinct groups (and subgroups) that
should receive training, you should work on matching the different
training areas defined in the Scope section of this chapter with the
groups of users defined earlier. Each group of users will require a spe-
cific set of training (based on training areas). We recommend creating
a matrix in your training plan—with training role against training
subject area—that you can refer to when creating training materials,
planning training delivery, etc.

You should When determining which personas need which types of training, refer

structure your again to your user stories as a baseline. You should also consider which
groups need training in non-functional areas and administrative areas.
assignment of Note that certain groups of users might require different levels of
training topics training on identical subjects. For example, trainers might require only
to users in a basic training on Dynamics 365 administration (to assist other users
during training and to implement training tools), while administrators
way that is and support desk personnel might require advanced training on the
comprehensive but same topic. You should structure your assignment of training topics to
not superfluous. users in a way that is comprehensive but not superfluous.

Training schedule
Successful integration of a training strategy into the main project
effort requires coordination with project management resources. Your
organization’s training strategy does not exist independent of other
ongoing efforts in the project, and it is important that key milestones,
resource requirements, dates, and tasks are acknowledged and ac-
counted for as part of your training plan and are incorporated as tasks
or dependencies in your overall project plan.

Regardless of the size of your organization and complexity of the

application being deployed, you must consider the following
training-related factors. We have included here examples of common
questions project teams encounter as a part of their training initiatives.
If these questions are applicable to your project, you will need to

Implementation Guide: Success by Design: Training strategy 551

address them as part of your training strategy and project plan. Note
that this is not a comprehensive list of questions that your organization
will need to answer, but this should help you get started:

▪ Project team resources

▫ What is the availability of the trainers and super users?
▫ Who are the super users, and if not defined, when will that
decision be made?
▫ Are there external resources involved in this process that will
require onboarding or administrative setup before they can
begin their work?

▪ Trainees
▫ Who is necessary to take part in training activities, and when
will these activities happen?
▫ Are there any resource dependencies (e.g., Resource A is re-
quired before Resource B)?
▫ How will training, and attendees, be staggered so that there is
no gap in coverage for day-to-day business functions?

▪ Dates or key milestones

▫ When should training begin for distinct groups of users?
▫ What dependencies do these trainings have?
▫ Are any project-related activities dependent on training?
▫ Is training reliant on the creation of material or development
work, whose schedules could shift?
▫ What milestones should training include, and are there any
impacts of these milestones?

▪ Training plan
▫ When will the training plan be completed?
▫ Who will review it, and what will be the plan for updating
it if necessary?

▪ Training execution and feedback

▫ When, and where, will training occur?
▫ Will training have an impact on other project efforts?

552
 ▫ If there are subsequent trainings, how will feedback be evaluat-
Fig.
ed and what is the cycle for processing this feedback?
19-4

Question
As you can see, many questions related to a training strategy have an
impact on, or are reliant upon, activities that occur separate from tradi-
Are any project-related activities
tional training activities. In Figure 19-4, we examine one question from
dependent on training?
this list to underscore the importance of asking and getting answers
Answer for these questions as part of your project planning.

We often recommend conducting training

prior to your user acceptance testing (UAT). This example shows why it is crucial that your training plan and overall
The major benefits of this order are: training strategy be reflective of and connected to your organization’s
During UAT, users are already aware of how to project plan, and that updates to either plan are reflected in the other
navigate through the system and accomplish
so as to not create conflicts, which can lead to project delays.
tasks. And as a result, there should be fewer
items logged as bugs that should be listed as
training issues.

UAT serves as a measure to validate the

effectiveness of training. If a large number of
training-related issues are identified during
Training materials
UAT, it may be necessary to make updates to
There are multiple consumable types of content for Dynamics 365
your training material and even potentially
conduct future training on topics with which applications that your organization can create. Depending on your
users are struggling.
project, you could create multiple types for use in different trainings.
Here are a few common examples of the types of training materials
that an organization could create.

Documents  Your organization can create written documents that

guide users through the application and educate them on business pro-
cesses. These documents can also serve as reference material. This type
of material is the least interactive but is the easiest to create and update.

Videos  Trainers can create videos that explain key training areas in de-
tail and walk through the application, providing guidance on how users
can navigate the system and complete their job tasks. Videos can be
more helpful than documents for many users, since they “show” rather
than “tell” certain features of the application. Videos can also be record-
ed in small chunks that represent discrete processes in an application,
as opposed to creating one longer video that walks a user through an
end-to-end flow. This benefits trainers from a maintenance perspective;
instead of having to rerecord the entire video when changes are made
to your application (and create rework around having to record training

Implementation Guide: Success by Design: Training strategy 553

for concepts that have not changed), they can focus on updating specific
videos around new features and functionality.

Microsoft Learn  Microsoft has published several labs via the

Microsoft Learn portal that provide a wealth of knowledge on Dynamics
365 applications, as well as other applications on which your organization
might require its users to be educated. Microsoft Learn allows the user
to follow any number of predetermined paths specifically designed for
different user personas. Individuals can also create their own knowledge
path that is specifically tailored to their learning requirements. Users
can learn on their own schedule and can browse videos related to topics
when necessary and engage with other community members.

Guided help  Dynamics 365 includes the ability to create custom help
panes and guided tasks, out of the box, to assist your users in walking
through the application and completing basic tasks in the system.
Guided help is easy to set up and implement and does not require an
additional solution on top of your application. Additionally, Dynamics
365 applications can install the Guides application, which allows for
visual or holographic displays to show step-by-step instructions on
tasks that your users need to perform.

Virtual material via a digital adoption platform  Built on top of

Dynamics 365 and allowing users to directly interact with the applica-
tion while training, digital adoption platforms are the most interactive
form of training material. This training material can be easily updated,
and once published, it has an immediate effect for all users who need
to view it. It reduces the need to update physical documents or re-re-
cord parts of video material. Digital adoption platforms can come
in multiple flavors. One effective method is a tool that lets trainers
create multiple “scripts” that walk a user through a business process
in Dynamics 365. For example, say that all users must learn basic nav-
igation through an account record, but only system administrators
must know how to customize a price list. Earlier in this chapter, in the
“Scope” section, we talked about reducing overlap; using digital adop-
tion platforms is an excellent way to manage scope in the creation of
training material. Instead of having to explain the same material twice
(in the case of video trainings that cover similar subjects) or create

554
 multiple reference guides or a single complicated guide (in the case
of written training material), a trainer can create one set of scripts and
apply them to different groups of users.

What to consider when selecting

training material
Be thoughtful when choosing the type or types of training material to
create for your training audience. Consider the benefits and drawbacks
of each. For example, if your end-user base consists mostly of users
who do not have considerable experience working with web-based
cloud applications, you might want to consider a more hands-on ap-
proach—a digital adoption platform—to help train your users. On the
other hand, system administrators and support desk users might find
documents more helpful, as they would rather have the large volume
of written information that do not require the more hands-on training
that digital adoption platforms provide. When creating training con-
tent, keep in mind that training is ongoing. Your organization can start
to create training material and training early on in your project, and
then refine it later with minimal rework needed. In the world of cloud
applications, where software is constantly changing, it is important
that your training strategy prioritizes being adaptable, so that you can
update materials easily as a response to application changes.

You should also consider language and accessibility needs in your

organization. If you have a multi-geo or multi-language deployment,
any training material you create might need to be available in different
languages. This material must also comply with any federal, state, or
local regulations regarding accessibility so that all users have an equal
opportunity to consume effective training material.

Who will develop and review the content?

During the creation phases of your project plan and training plan,
you’ll need to identify the trainers, SMEs, and business process owners
who will assist in developing the training content. Trainers will usually
create content, which will subsequently be reviewed by the SMEs and/
or business process owners. Depending on the importance of the

Implementation Guide: Success by Design: Training strategy 555

 course, as discussed earlier, you might want to consider bringing in
multiple people to review the material.

How long will it take to develop the content?

Your organization must make sure to allocate time in the project
schedule and training plan to design, develop, and review content. The
length of time required to create training content will depend on the
complexity of your application and the type or types of content to be
developed. To estimate the time, we recommend that you identify the
scope of training areas and subsequent material that must be created
based on those areas, and the format or formats the training material
must take. You can then estimate total time to create training material
based on experience or an extrapolation calculated on how long it
takes a trainer to create a subset of that material.

When is the content needed?

Your project plan should list when content creation begins and your
training plan should specify when training material will be available to
certain groups of users. These two schedules should be coordinated to
ensure that content creation begins on time, and to ensure the mate-
rial is ready when needed. End-user training can begin roughly when
development is completed (that is, when the system is code complete
and no changes to the application are anticipated).

Where will the content be stored?

If more than one We recommend your organization employ a content storage control
person will be method, especially if multiple types of content are being created for
editing content, groups of users. Training content should exist in a single centralized
source instead of being passed around by email or shared links. This also
your organization ensures proper source control and the ability to institute access control
might need a for your user groups. Additionally, if more than one person will be
versioning strategy. editing content, your organization might need a versioning strategy.

Delivery approach
As a subset of your training plan, your organization should create a
training delivery document that contains specifics about the actual

Implementation Guide: Success by Design: Training strategy 556

execution of training itself. While many of these topics are covered in
the overarching training plan, it is important to have a specific plan
for training delivery to ensure that your organization is prepared to
execute training.

We’ll now discuss the topics that your training delivery document
should cover.

Who will be involved, and how?

We have defined the audience for training in the “Audience” section
earlier in this chapter, and explained the details of incorporating
resource scheduling into your project plan in the “Training schedule”
section. To help you create your training delivery document, we will
go a step further in organizing these resources into training sessions.
This is important if your organization will schedule multiple training
sessions at the same time. In addition to allocating your trainers and
trainees, you will also need to assign them to the correct training based
on the content to be covered.

What methods of training should the organization conduct?

Like training material, diverse types of training suit different situations
and users, so it’s important to consider which training or combination
of trainings best fit your organization. Consider the advantages and
disadvantages of different methods. Figure 19-5 shows common types of
training, and the anticipated advantages and disadvantages of each type.

When should training occur?

As a part of your training delivery plan, you should decide when dif-
ferent types of training will occur. In an earlier section of this chapter,
we discussed how super users and trainers should be trained in nec-
essary areas of your application earlier than your end users; this type
of consideration should be built into your training delivery plan. Plan
to conduct training for your end users late enough in the project—
post-development and pre-UAT—so your users don’t forget concepts
and tasks that they have been taught during training, but also be sure
to leave enough time after training occurs to accommodate any delays
in your project or training schedule.

557
 Fig.
19-5

Type Most effective when Advantages Disadvantages

Live training The content is critical for Better interaction Scheduling challenges.
(in person or business function and you between trainer and
want strict control over participants as well as In-person variants of this type of
virtual)
delivery. between participants. training require a physical room with
computers, since Dynamics 365 is a
Content contains non- Immediate feedback. web-based application.
functional requirements,
or business processes that Collaboration in business In-person variants of this type of
require specific devices. processes is significantly training are limited to the number of
easier (i.e., front office people per room, per session.
and back office need
to work together to
complete an order).

Interactive Content is moderate in Web-based content Trainings are less effective than in-
web-based difficulty. can be built once person, since there is less interaction
and consumed by an and the end user faces more
training
Business processes are unlimited number of distractions.
best learned through users.
repetition and ongoing Trainings are not usually real-time,
training, since users’ People can train on their meaning any questions or challenges
access to the application own. No scheduling that users face during training might
is necessary. requirements and no not be answered quickly.
physical location needed.
Web-based training, when built
with digital adoption tools, can be
costly and also require technical
maintenance.

Self-paced Content can be easily Written and video This is the least interactive form of
training explained in a user content can be updated training, meaning users who rely on
manual or video series. quickly as future releases interaction either with a trainer or
and updates occur on with a web-based application might
Content is best your Dynamics 365 struggle to learn when having to
consumed as written application. read/watch content.
reference material that
can be used as needed. Training content is easiest
to create.

Implementation Guide: Success by Design: Training strategy 558

 We believe that it’s never too early to begin training super users and
trainers. Since the main goal of training these users is to help them
become familiar with your application early in your project, you should
start conducting training with them (whether formal or informal)
during development and continue it as an ongoing process. These
early training sessions should familiarize super users and project team
members with basic application functions and serve as a backbone for
future train-the-trainer sessions.

Figure 19-6 is a high-level sample schedule that shows training ses-

sions for each important group that needs to be trained, alongside key
relevant project milestones:

Trainers and project team members should be aware of ways in which

training differs depending on the size of your organization. Larger
organizations with more users are likely to require multiple personas
and an increased number of business-use cases. Larger organizations,
especially international companies or companies with numerous
locations, might require different training sessions and venues, and
they might need to occur at different times in your project. In addition
to making sure training materials are available in multiple languages,
employing multiple trainers or multilingual trainers is important.
Fig.
19-6

January February March April May June

Super-user training session #1 conducted Train-the- Go live

trainer
Train-the-trainer session #1 conducted session #2 UAT Support desk
on first half of sprints conducted conducted training occurs
based on
Trainers begin work on training material remainder of
Stakeholder/executive training occurs
development
Design, development sprints begin work
End-user training
Project kick-off Code Super-user training session #2 conducted
Demonstrations/Initial training conducted for super users, complete
trainers, and project team members

Implementation Guide: Success by Design: Training strategy 559

 In the sample project plan seen in Figure 19-6, end-user training
takes place over the course of a single week, in a single location. If this
organization were international and multilingual, most of the items in
the plan become more complex. To illustrate the increased complexity,
we consider what end-user training might look like in a multinational
organization in Figure 19-7.

As this figure illustrates, training schedule and planning complexity

increase dramatically when the scope of the implementation increas-
es. Instead of taking roughly a week, this updated training schedule
now takes almost a month, in three different countries/regions, and
requires drastically increased coordination of resources. If additional
phases with different business process areas need to also be covered,
this would further increase the complexity of training. We hope these
examples are useful in demonstrating the value of planning the when,
where, and how of training and why it is important to understand that
your organization should be prepared to make additional investments
to train its users successfully.

How should your organization set up and maintain a

training environment?
If your organization has created a Dynamics 365 environment to train
users as part of a broader environment management strategy (cov-
ered in Chapter 9, “Environment strategy”), using application lifecycle
Fig.
19-7
management (ALM) best practices (covered in Chapter 11, “Application

April
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Trainers work to End-user training #1 End-user training #2 End-user training for field
develop/translate Conducted in the US, Conducted in the US, technicians
material into English, France and Spain France and Spain Conducted in France
French, and Spanish
Training material Environment cleanup Environment cleanup
(including translations)
are validated

Implementation Guide: Success by Design: Training strategy 560

 lifecycle management”), you will want to make sure this training
environment is deployed with code and data that is representative of
production use and is of adequate quality to meet your training objec-
tives. Both are extremely important points:
▪ If your organization’s training environment does not have up-to-
date code and application features, your users will be left confused
and frustrated, which could decrease the overall effectiveness of
your training. Your users might feel the training was unprofessional
or incomplete—especially if the quality of the training environ-
ment is insufficient to accomplish training objectives.
▪ If your organization’s training environment does not have suffi-
cient and relevant test data, your users might fixate on the quality
of the data and not on the training material. This will lead to
improperly trained users from multiple standpoints:
▫ Because users are distracted, overall training effectiveness will
decrease. This could have an impact on user adoption.

Poor data quality in your training environment might impact the

If your training environment contains training’s real-world relevance. One problem we see frequently in
only accounts named “Test 1” and
poorly conducted training sessions is that users complain the data is
“Test 2,” it’s unlikely that your
users will enter realistic data when not realistic. This in turn can have an impact on the quality of the data
progressing through their training
scenarios. If users don’t understand
that these users subsequently enter during training, which can affect
the implications of entering fictitious the quality of the data that these users enter once the system goes live.
data, this could have wide-reaching
impacts on production. If users
are entering fictitious data in
How to assess competency and effectiveness of training?
production, this could influence
downstream business processes, One of the most effective ways to improve training is to collect feedback
system logic, and reporting. To
ensure more realistic scenarios
and use it to make enhancements. Training is not meant to be a static
during training, we recommend your exercise in which material is created once and delivered the same way in
dataset be a subset of real data from
a legacy system. every iteration; it should be a work in progress that is updated frequently.
Once a training session is completed, there is still work to be done. Many
organizations require multiple training sessions, and all of them present
opportunities to revisit, evaluate, and improve future training sessions.

You can measure the effectiveness of your training through several dif-
ferent recommendations. The first is by assessing your users’ learning
at the end of, or shortly following, training—in particular, prioritizing
the processes that they will use most frequently in their jobs. It’s
much more important that sales users understand how to create and

561
 update an account record in Dynamics 365 without assistance if it’s
a process they will need to do every day. Such prioritization should
be emphasized during the creation of any assessment method you
choose. Earlier in the chapter, we discussed how creating solid training
objectives would help you assess the effectiveness of your training. We
can now take those training objectives and use them to form the basis
for how we will assess our users, so long as the training objectives we
created contain the most critical topics that we are training users on.

Many larger organizations create exams to assess knowledge of their

users or they might even create a certification program following
training to assess readiness at a user level, team level, or even larger
level, such as a business unit, subsidiary company, or geographical
region. Trainers can create metrics based on exam results and certifica-
tion statistics to evaluate training quality. If certain groups have scores
or statistics that are lower than desired, the organization can work to
improve future training or host follow-up trainings for these groups.

Accuracy is Day-in-the-life testing, which can be used to measure real-world

scenarios and the effectiveness or speed with which your users complete
important to ensure tasks, also measures training effectiveness and organizational readiness.
data quality, but it Accuracy is important to ensure data quality, but it is also important
is also important to know if certain business processes take your users 10 minutes or 2

to know if certain minutes to complete. This could be used to measure your organization’s
readiness for the new application. If certain business metrics must be
business processes met regardless of the application used, make sure these metrics are met
take your users 10 within the Dynamics 365 application. As an example, if your call center
minutes or 2 minutes staff is expected to process at least six calls an hour, it is important to
ensure that your staff has been trained to complete a single end-to-end
to complete. call process in your application within 10 minutes.

Another way to measure the effectiveness of your organization’s

training is by analyzing the results of your UAT. As we briefly discussed
in the “Training schedule” section, you can use UAT as a barometer to
gauge user competency of your application by looking at the number
and types of non-system-related bugs that are logged by users, and
by measuring (usually informally, since the goal of UAT is accuracy, not
speed) the amount of time users take to complete the processes they

Implementation Guide: Success by Design: Training strategy 562

 are testing. If you are finding that your testers are logging many items
that are not system bugs but rather due to knowledge gaps on the
function of your application, you might conclude that training in those
areas needs updating. It could also mean that additional user training
in those areas is required.

Similarly, after go live, you can use helpdesk statistics to determine the
Question
longer-term effectiveness of training. If users are submitting help desk
What if, during a training session, a tickets, or requesting assistance, on specific areas of your application,
critical bug is discovered that blocks
training from going any further?
it’s a good idea to review training or training documents related to
those features to see if they should be updated. Again, if necessary,
Recommendation additional user training might be needed.

We believe your training environment

should be treated similarly to a production Another common approach to collecting feedback on training is to
environment during a training session. conduct pilot training sessions with a limited audience. You can invite
Bug fixes should not be made directly in
production; likewise, bug fixes should almost super users to attend these pilot training courses and gather feedback to
never be made directly in a training instance
improve. A side benefit of conducting these pilot trainings is that it helps
while training is taking place. Depending on
the severity of the bug, a work item should super users in their process of becoming experts in your application.
be created for a developer in their respective
development or hotfix environment, and the
bug should be fixed and tested before being
pushed to downstream instances (including
training), as per standard ALM processes. Assumptions,
If the bug is critical enough to block entire
business scenarios and prevent trainees
from accomplishing any tasks, it might
dependencies, and risks
be necessary to fix these bugs directly in
the training environment. However, we As discussed in the “Training schedule” section of this chapter, suc-
recommend this only in situations where
the bug prevents any workarounds or other cessful training relies on many external factors that should be defined
sessions that could be taught. Fixing the in your training plan. These factors can be assumptions, which are
bug directly in a “live” environment could
cause additional, unforeseen issues to occur, statements assumed to be true of the project that are often critical to
which would further lower users’ confidence
training success. Here are a couple examples of assumptions:
in the application. Furthermore, any bug
fixes that happen in this instance would also ▪ Appropriate trainer resources will be allocated to create training
have to be fixed in the lowest instance where
development is occurring, and regression
material and train users.
testing would have to be completed and any ▪ The tenant management and development team will create and
potential conflicts would have to be resolved.
maintain a training environment for use during the project.

Your training plan and project plan should also include dependencies,
which can be represented in a process workflow. They effectively show
that certain tasks or activities must occur before others as part of your
training strategy. Figure 19-8 depicts a sample process workflow.

Implementation Guide: Success by Design: Training strategy 563

 This tree is an example of a straightforward dependency list for creat-
ing a specific type of training document. Your plan will contain more
We recommend making sure each training
complex dependencies that might rely on multiple, unrelated project
participant has a user account with the tasks. Defining and organizing these dependencies at the start of a
same level of access as their forthcoming
production user account. project will help reduce the amount of time your trainers or other
project members are waiting for required tasks to be completed, by
We do not recommend giving system
administrator access to training users. This scheduling upstream project tasks specifically to minimize this down-
is a common mistake we see organizations
make during training. They offer a variety
time. This must be done in coordination with the broader project team
of reasons for doing this, including the and project schedule.
desire to not run into access issues during
training and the security model having
not been finalized. Lastly, creating and maintaining an active training-risk log, as part of
If you are conducting multiple training the overall project risk log, is important to ensure proper attention is
sessions in a single training instance, it’s
paid to training risks that could impact the success of training or other
important to take these proper actions at
the conclusion of the training session: project activities reliant on successful user training.
• Any bugs that discovered during

Training as an
training are properly documented
(with screenshots, if necessary) and
sent for review or triage.

ongoing process
• If restoring from a Dynamics 365
backup, ensure that a backup for the
correct date is available. (Work with
your partner or IT staff, as well as
Microsoft if necessary, to ensure this As mentioned in the introduction of this chapter, your organization’s
is not a blocker.)
• If restoring from source control as training journey begins long before the execution of user training, and
part of your ALM process, ensure
its ending extends long after those users complete their first tasks in
all code and data are ready for an
environment refresh. your production application. Training is an ongoing and constantly
• Refresh your environment using your
chosen method.
evolving process that is reflective of changes in your organization and
• Validate user access and data via a application, as well as changes to the Microsoft Dynamics 365 ecosys-
smoke test, prior to the start of the
next training. tem. While developing a training strategy, it’s important to consider
the implications of both types of change.

Internal change
Fig.
Create training documentation on Internal change, for the purposes
19-8
lead/opportunity management of this section, refers to change
that, at a transactional level, is
unique to the organization ex-
Update Finalize
Create draft
Review 1 document Review 2 document periencing it. Each of these types
of document based on based on
of document of BPO
by SME feedback feedback of change could require training
from SME from BPO
material to be updated and sub-
sequently conducted on the same
or different groups of users.

Implementation Guide: Success by Design: Training strategy 564

It’s important that Personnel change
User turnover is inevitable—either when employees leave an organiza-
new users are tion and new ones take their place, or when employees change job roles
provided support within the company. All new users are expected to quickly learn their job
during and after roles and how to navigate and complete their work in the application;

onboarding, as they to do this, these users need to be trained. Note that it’s unlikely that
an organization would conduct multiple lengthy in-person training
might receive fewer sessions for a single user’s onboarding, and would instead rely more on
training resources self-paced or web-based training as a means of education. It’s important
than users who were that these new users are provided support during and after onboarding,
as they not only are initially behind more experienced users who have
present during the been using the application for a longer period of time, but they might
application’s initial receive fewer training resources than users who were present during the
rollout. application’s initial rollout.

Your organization’s training material for users who joined after go live
will be similar to that used by your end users before go live, but again,
be sure that these users are supported and given the assistance they
require to train on the application.

Application change
Many Dynamics 365 projects have adopted a phased rollout approach,
in which features of each application are released to production over
time, as opposed to a “big bang” approach, in which a single deploy-
ment to production prior to go live has all included functionality. While
the rest of this chapter covers the development and distribution of
training regarding the first release of application functionality, future
phases of the project must include updates to training material and
new training sessions.

Your organization might want to treat each phase of a rollout as a

distinct project from a training perspective, wherein each future phase
contains a distinct, more condensed training plan and training delivery
schedule to ensure proper user training is conducted at every stage.
These future phases will focus heavily on making updates to training
material based on new business processes, and training users—both
new and experienced—on the application.

Implementation Guide: Success by Design: Training strategy 565

 Consider a building-block approach when designing your training ma-
terial. Phase one training material and training can be the foundation
on which all users can learn to use the application. Future phases can
Phase two training might include
users who are already familiar with
build on that foundation; training material for subsequent phases can
the application, having used it after be updates of earlier material as well as newly created material.
the phase one go-live, as well as
employees who have never used or
been trained on your system. Any
External change
new and updated training material
should reflect the difference in In other chapters, we discuss Microsoft Dynamics 365 as a software as
experience and skill level of these
user groups. It should not be too
a service (SaaS) that provides continual updates to organizations, both
simple as to be irrelevant to the from a technical standpoint and a business standpoint. It’s important
experienced user, nor too complex
as to confuse the new user. that your organization prepares for these updates as part of its training
strategy and that they are acknowledged in the training plan and ex-
ecution. Note that we allow customers to activate updates to software
at their own pace.

As updates to the Dynamics 365 application are announced and begin

to be rolled out, your organization should set aside an instance used
for regression testing product functionality against your custom
application. As part of these activities, trainers should be involved and
assess the potential impact these updates have on business process.

Certain updates, especially improvements made on the software’s back

end, will result in little to no change in the way your users do their work
in your application. Thus, they should not require a significant update
to training material, nor additional training. Other updates that are
more focused on user experience or new features within the platform
might drastically affect certain end users and will necessitate updates
to training material as well as additional user training. Figure 19-9
contains examples of how software updates might or might not require
changes to your training offerings.

Training best practices

The authors of this book have experience with Microsoft Dynamics 365
implementations of all sizes and complexities. We have created and re-
viewed training plans, participated in and hosted training sessions, and
evaluated the outcomes of training on end-user adoption. From most

566
 of these projects, common themes have emerged. We share the most
important recommendations here as best practices—things to keep in
mind during your organization’s training process. These best practices
are by no means comprehensive, nor are they set in stone, but they are
meant to guide a project team to a more successful training outcome.

Earlier is better
Starting early (with a training plan) is important. Dynamics 365 projects
are often run on tight timelines with little to no leeway. It’s crucial that
a training plan, complete with staffing and resource availability, be
completed at the start of a project, to ensure no unexpected resource
requirements or shortages appear midway through implementation.
Give yourself sufficient time to prepare a training plan, develop con-
tent, and deliver training. Content development should be completed
Fig.
19-9 well before the start of training delivery. Additionally, project roles
and responsibilities in cloud solution deployments are rarely separate;
often a single IT resource will wear
Software update multiple hats, and one of these
We have made improvements on the schedule board that should result
in a 40% increase in performance. hats might be to create a training
Performance
Training needs improvements plan, review training material, or
Even though this update is relevant and beneficial to your users,
it would not require an update to training materials or necessitate maybe even conduct training.
additional training, since there is no underlying change
in how your users would conduct their work.
Making sure these individuals are
VS
aware at the outset of their con-
tributions to the training plan will
Software update
help them balance their workloads
We have made updates to the layout of the schedule board
Interface updates that should make navigation and creation of bookings easier
and complete certain activities
for your dispatchers.

Training needs earlier, which could prevent delays

This update is relevant and beneficial to your users and will require
updates to training materials (and could require follow-up training for
from occurring.
users) due to user interface changes in key business processes.

Keep an eye on release notes to leverage new features. Work with your

More is
implementation partner and your internal teams on your heatmap for
Dynamics 365 capabilities. Have a regular review of future projects,
incorporating new features delivered.

Join the D365UG User Group for Dynamics 365 to learn about member-
driven education, networking, and events. (usually) better
Create a Yammer group or Microsoft Teams channel to continue In Dynamics 365 applications, a
conversations on best practices and learnings.

Host engagement events like town halls" or "Lunch and Learns" to

user can often access the same
continue training after go live and drive end-user engagement.
record or complete the same
Share success stories about how people are using your application in
innovative and impactful ways. business process in several ways.
Thus, training that initially seems

Implementation Guide: Success by Design: Training strategy 567

 repetitive is useful in that it teaches multiple ways to accomplish a task
depending on the situation and context. While too much training can
mean a small amount of wasted effort, for an organization adopting a
new application, the cost of time is far preferable to having users who
are insufficiently trained on a sizable portion of their job. The latter
situation occurs in organizations far more frequently than the former
situation does, even for similar amounts of training.

It’s also important that your training material does not contain too
much duplicate content. While it’s certainly critical that both the web-
based and live content cover the same features (especially if users are
not consuming multiple types of content), too much overlap can result
in training fatigue, and overall apathy regarding training. An organi-
zation would certainly want to avoid a situation in which users skipped
valuable content because they considered it repetitive.

Train with a goal in mind

It’s important to Conducting a variety of training for all users using different methods
is important. But what is much more important is not losing sight of
not lose sight of the the end goal: successful user education and meaningful adoption of
end goal: successful your Dynamics 365 application. It does the organization no good if
user education and every technician understands how to use their new mobile application

meaningful adoption but none of them uses it correctly because of a failed training process.
Having comprehensive training material is but one step in this process.
of your Dynamics As stated earlier in this chapter, conducting effective training that leads
365 application. to excitement and high user adoption with accomplishment of critical
KPIs, in addition to the more obvious goals of high training partici-
pation and education, is the key that can separate adequate training
programs from excellent ones.

Consider your audience

In addition to having distinct job roles, your users come from diverse
backgrounds and probably different generations, too. Avoid alienating
groups of people with references, case scenarios, or even pictures or
stories that could confuse or even offend certain groups of people. Your
training should keep users engaged, but be careful not to be divisive.

Implementation Guide: Success by Design: Training strategy 568

For example, say you have a junior salesperson in their 20s and a senior
salesperson in their 60s attending the same introductory training. Your
application includes an embedded LinkedIn widget on your contact
entity. Trainers should not assume that both users (or, more divisively,
only one) are familiar with LinkedIn and its capabilities. A good rule of
thumb is to assume all training participants arrive with the same level
of background information and context of the application, plus any
technology knowledge required to use it effectively.

Accessibility concerns
Your organization must consider accessibility while creating training
material. All Office 365 products have an accessibility checker that can
help you spot potential issues in your content.

Identify ongoing challenges

As discussed earlier, feedback is an important part of any improvement
cycle. In addition to identifying areas of improvement for future train-
ings and training material, we recommend that the people conducting
training or collecting feedback from the training identify key areas
where users are struggling with new functionality. This information can
be used to create more focused training material that can be distribut-
ed to users, or to create webinars or other sessions specifically focused
on these problem areas.

Product-specific guidance
Up to this point in the chapter, our training guidance has applied
to Dynamics 365 Operations as well as Dynamics 365 Customer
Engagement application projects. While both applications live in the
Microsoft Dynamics 365 ecosystem and customers frequently adopt
both systems (often simultaneously), there are differences between the
two, which can mean differences in how each project should train its
users. In this section, we highlight some of these differences and how
to apply them to your training material.

569
 Dynamics 365 Operations
A number of resources provided in Dynamics 365 Finance, Supply Chain
Management, and Commerce can assist with product help and training.

Help on docs.microsoft.com
The Dynamics 365 documentation on Microsoft’s docs.microsoft.com
site is the primary source for product documentation for the previously
listed apps. This site offers the following features:
▪ Access to the most up-to-date content  The site gives
Microsoft a faster and more flexible way to create, deliver, and
update product documentation. Therefore, you have easy access
to the latest technical information.
▪ Content written by experts  Content on the site is open to con-
tributions by community members inside and outside of Microsoft.
▪ Content is personalized based on language selected  If, for
example, a user is set up for the German language, then any Help
You can find content on docs.
microsoft.com by using any search content they access will be provided in German.
engine. For the best results, use a site
search, such as site:docs.microsoft.
▪ Help is now on GitHub  Customers can copy and create person-
com Dynamics 365 “search term”. alized Help content and link it to their application. Customers can
create a personalized and contextualized Help resource for users
of their custom business processes.

In-product Help
In the Dynamics 356 client, new users can enter the Help system to
read articles that are pulled from the Docs site’s Dynamics 365 area
and task guides from the business process modeler (BPM) in Lifecycle
Services (LCS). The help is contextualized to the form that the user is in.
For example, if a user is in a sales orders screen and wants to know how
Consider the number of concurrent users to create sales orders, the Help system will show the Docs articles and
using the training environment as you
choose the right tier for the environment. task guides related to sales orders (see Figure 19-10).
Do not automatically select the default
tier 2 environment if the volume of
expected users exceeds the recommended Task guides
volume for the tier. For more information
about different environment tiers and
Other useful help and training features are called task recorder and
the volume of users each supports, read task guide. Task recorder allows you to record a user’s activity on the
this documentation on selecting the right
environment for your organization. UI. You can capture all the actions as well as any UI fields and controls
that were used. This recording can then be used in a task guide.

Implementation Guide: Success by Design: Training strategy 570

 A task guide is a controlled, guided, interactive experience that leads you
through the steps of a task or business process. You can open (or play) a
task guide from the Help pane. When you select a task guide, the Help
pane shows the step-by-step instructions for the task (see Figure 19-11).
Localized task guides are available, and you can create custom task
guides. The data you enter in the system is saved in the environment.

To begin the guided, interactive experience, select Start task guide at

the bottom of the Help pane. A black pointer shows you where to go
first. Follow the instructions that appear in the UI and enter data as
directed (Figure 19-12).

Dynamics 365 Customer Engagement

Dynamics 365 Customer Engagement has product-specific tools that
can aid in training users. Gamification and custom help pages are two
of the most common tools used; we describe them in more detail here.
Fig.
19-10

Implementation Guide: Success by Design: Training strategy 571

 Fig.
19-11

Fig.
19-12

Implementation Guide: Success by Design: Training strategy 572

 Gamification
Gamification is a tool available from AppSource that is attached to
Dynamics 365 Customer Engagement. Primarily used to increase the
productivity of users within your Sales, Service, and Field Service/
Project Service applications, Gamification can also be utilized as an
effective training tool when configured properly for users.

From a training perspective, an administrator can set up the

Gamification solution in an instance within your Dynamics 365 tenant
(usually Training, though games can also be created in Production as
a means of “learning while you work”). You can define players, KPIs,
and games to be hosted. Say your end-user training involves the
participants following a complex business process from start to finish
several times, and it concludes with a Case record being resolved. The
administrator could create a game out of this. A KPI would be the
successful completion of a Case by game participants (the trainees). To
help motivate trainees, Gamification allows the configuration of prizes
in Dynamics 365 Customer Engagement.

While Gamification and using games as part of training can be an

especially useful motivational tool for your employees, you will want to
ensure that competition is only so fierce as to foster a learning environ-
ment for your users. You do not want to create an adversarial culture in
which collaborative discussion is absent and the goal of training—
education and adoption—is lost.

References Numerous custom ISVs also offer similar gamification solutions that
Training plan template can be attached to Dynamics 365 Customer Engagement. Pick the one
Training plan charter template that addresses the needs and goals of your organization best.
TechTalk Series: Training Plans and Content for
your Finance & Operations Project (Microsoft
Custom help pages
Dynamics blog)
Dynamics 365 Customer Engagement lets you create guided help pag-
Dynamics 365 training offerings (Microsoft
Docs)
es that provide your users with in-product assistance. This assistance
can be customized to your application and user base and can include
In-product Help – Finance & Operations (Mic-
rosoft Docs) text, links, images, and video links. For more information about custom
help pages, read “Custom help panes and learning path.”

573
 Checklist

Implementation Ensure trainers and super users are trained well in

advance of end-user training, which should ideally occur
Have training objectives that are SMART (specific,
prior to UAT.
measurable, achievable, realistic, and timely), relevant
to your organization’s goal of user adoption, and fit the Ensure the training plan appropriately addresses
organization’s overall change management goals. accessibility.

Have a process to update the training plan incrementally Define a process for feedback and continuous improve-
as necessary to reflect scope, changes, risks, and depen- ments to training materials.
dencies to ensure adoption and engagement.
Identify a process to provide for continuous training in
Consider what to include regarding system process and alignment with updates and changes to the solution as
business process, so that the training provides the best well as changes in roles and responsibilities.
possible foundation to end users.

Clearly define the match between personas and roles,

and business processes and user level, for each training
session, so users receive relevant training.

Ensure the training schedule accounts for key project

milestones, resource requirements and availability,
dates, and tasks, and is in alignment with the overall
project plan.

Consider different types of content, the time to create

it, the effort to deliver and maintain it, storage, and the
skillset of the authors and trainers.

Set up the training environment with realistic data, user

profiles, and scenarios. Refresh and clean it after training
sessions to maintain quality.

Implementation Guide: Success by Design: Training strategy 574

 Case study
A proven path to
training success
This global fire-protection company has earned a reputation for
providing customers with maintenance and service of the highest
quality. In this mid-size company’s 75-year existence, they built their
business organically through natural growth, marketing, and geo-
graphic expansion, as well as through numerous acquisitions of smaller
fire-protection companies. They recently hired a new chief information
officer to replace the CIO who retired after 15 years in the job. The new
CIO had extensive experience working with Microsoft and SaaS prod-
ucts. The CIO’s immediate focus was to consolidate and modernize the
applications that employees were using to conduct their business.

In the training plan, the company included a mix of high-level and de-
tailed training objectives. The high-level objectives included these goals:
▪ To not allow access to the live system without a solid
training program
▪ To prepare a core team of trainers to help support the initiative
▪ To continue receiving feedback and improving the
training approach
▪ To develop the training materials early and schedule classes early
▪ To prepare all application users to efficiently use the applica-
tion (Dynamics 365 Finance and Dynamics 365 Supply Chain
Management or Dynamics 365 Customer Engagement), as well as ad-
dress any key business process changes required in their job function

Implementation Guide: Success by Design: Training strategy 575

 Specific objectives were centered on business processes like Prospect
to cash—for example, “All sales team members should be able to
execute the standard Prospect to cash flow on their own.” Their specific
training objectives, listed here, helped support business goals for
the application:
▪ Increase sales by 10 percent year over year
▪ Improve productivity by 25 percent by measuring Work Order
completion time
▪ Increase user satisfaction (covered in annual poll) by 10 percent
and user competency (measured by metrics within the application)
by 15 percent within the accounting business department

The team understood that for training to be successful and for mean-
ingful user adoption to be achieved, they needed to begin planning
early and set up a strong team to support it.

The team benefited from using a change-impact assessment matrix to

understand the changed business areas, topics, and user groups impacted.
The matrix helped them prepare and prioritize training sessions.

The company had experience developing and conducting trainings for

previous projects. They knew they wanted to incorporate a variety of
learning modalities—written documentation, videos, hands-on learn-
ing, and multiple training labs using real and recognizable data.

Their legacy application infrastructure was a reflection of the way

business had grown. The back-office system ran on an on-premises
application that had been in use since 1990. Front-office users (sales-
people and field dispatchers), as well as field technicians, did not
employ any application; they were still using pen-and-paper-based
methods. Due to the number of acquisitions they had made over the
years, these methods varied—sometimes by business unit, sometimes
by other organizational structure—and there was no overarching
governance model consolidating these different methods.

As a result, employees that were in identical roles used different

methods to accomplish their day-to-day work. This discrepancy in job

Implementation Guide: Success by Design: Training strategy 576

 execution versus job role could be attributed to the fact that many
of them joined the organization via acquisition and had never been
required to change. It was necessary to capture all “as is” processes and
assess the impacts of moving to the new application, from a technical
perspective and from a business process change perspective.

Given the numerous challenges of supporting a wide variety of

applications for all users, the company made use of the BPM, task
guide, and custom guided help features for content development
and user training. Having all business processes documented on one
platform gave the organization a single source of truth. Task guides not
only provided consistency to all the business processes in use across
the different departments, but they also made the employee orienta-
tion and training a much smoother experience. Task guides were also
embedded in the product help experience to provide a 24/7 interactive
help experience.

By clearly defining all these processes, as well as the group of users to be

trained, they were set up for success. Its subsequent training delivery was
a smooth process that included all user persona business processes.

Because of the evolving nature of Dynamics 365 applications and

organizations, as well as the fact that the project was being rolled out
in multiple phases, the company developed an ongoing training pro-
cess by which training was executed, reviewed, and updated as a cycle.

Formal feedback was recorded after the trainings and Microsoft Teams
channels were created for employees to continue providing feedback
to the team. Users were encouraged to share knowledge, ask questions,
and suggest improvements to the training materials. The team was
also able to collect feedback and create metrics using help desk tickets,
which helped them identify areas of the application that users found
particularly challenging.

The organization determined that mobile Field Service technicians

were logging more tickets than other application users. By learning the
types of tickets being created by their technicians, trainers were able to
pinpoint a specific area of the Work Order process that needed to be

Implementation Guide: Success by Design: Training strategy 577

 made clearer during training. Adjustments were made to the training
material, too.

The objective “To not allow access to the live system without a solid
training program” was a difficult objective to meet in the initial days,
but the company learned over time that making sure every user
received adequate training drove a significant reduction in business
process issues for the company.

In the first few months, an evaluation of key KPIs showed that the
organization was on track to meet all the detailed objectives set by
the team.

Implementation Guide: Success by Design: Training strategy 578

Section
Operate
20 Service the solution
21 Transition to support

579
20
Implementation Guide: Success by Design: Service the solution
Guide
Service the
solution
580
Continuing the business applications journey.

Introduction
Your solution has been deployed to production.
Users have been trained and onboarded. A support
process is in place.

Most implementation projects focus on building and deploying the

solution, but perhaps not nearly as much emphasis is put on educating
the owner of the new system in keeping it optimal and healthy.

But you might be thinking, “Dynamics 365 is a SaaS platform.

Isn’t Microsoft responsible for maintaining these environments?”
For successful project
governance, consider these Let’s take a scenario where you rent an apartment in a large building
main areas: complex. The owner of the building is responsible for cleanliness and
Monitor service health upkeep of the building and common areas. They make sure to have

Service updates
proper plumbing and electricity in each apartment. They may add or
improve amenities such as an exercise gym or a garden in the yard.
Environment maintenance
But it is each renter’s responsibility to keep their apartment in good
Continue the business condition. The renter should clean the apartment and throw out their
application journey
garbage. They will need to replace the lightbulbs in the fixtures if it

Implementation Guide: Success by Design: Service the solution 581

 goes out. If there are any issues with the plumbing, they need to con-
tact the building manager so that a repair person can be dispatched to
help. In a similar way, Microsoft, the owner of the software as a service
(Saas) system, will be responsible for the platform—the building and
property in our scenario. The customer is responsible for the upkeep of
the health of their solution—the apartment.

In this chapter, we first look at how to proactively monitor the health

of the solution. Then we cover the continuous updates provided by
Dynamics 365 and best practices for working this pattern into your
project plan to minimize impact to the changes in your environment.
We also address environment maintenance tasks to maintain a healthy
system. Finally, we challenge you to continue your education in the
ever-evolving Microsoft business application journey.

Monitor service health

A key principle for a successful onboarding experience to Dynamics
Monitor service
health 365 is knowing the health of your environments at all times. Your team
Service updates must be able to troubleshoot issues right away.

Environment maintenance Dynamics 365 and the Power Platform have gone through rigorous
testing, so why do customers need to check that it’s running smoothly?
Continue the business
It’s because of each organization’s unique usage patterns and the
application journey
extensibility of the platform.

Just about every customer makes changes to their Dynamics 365

environment. It may be as small as adding new fields and forms. Or
your business may require heavy customizations with complex logic
or performance-impacting integrations with other systems. Or your
organization may have more concurrent users than the typical use
case, so data management needs are also growing at a rapid pace.

Another reason to monitor your system is to understand usage

patterns for better business intelligence. Through this analysis,
business stakeholders can pinpoint where to invest and prioritize
improvements and focus areas for training.

Implementation Guide: Success by Design: Service the solution 582

Microsoft highly recommends monitoring service health and provides
you the tools to do so. That way, you know when irregularities are
found and action is needed.

As administrators and power users of Dynamics 365, you can

focus on several key areas when monitoring usage. These areas
impact performance, security, cost, and supportability, all central
to keeping the system performance at optimal levels.

This section addresses key areas for monitoring the solution:

▪ Performance monitoring
▪ Licensing and operational cost
▪ Storage allocation and usage
▪ API request and service protection
▪ Security and privacy
▪ User access and resource usage
▪ Application logs and errors
▪ Messages from Microsoft

Performance monitoring
Microsoft recommends that Dynamics 365 project teams consider adding
performance testing to the project’s test cycle. Performance testing is an
effective way to gauge the impact that your customizations may have on
the baseline Dynamics 365 solution.

Performance testing is typically conducted in a simulated environment.

This provides insight into changes in solution design or configuration prior
to production rollout at scale. By conducting performance testing early,
Dynamics 365 administrators have a baseline performance expectation of
the new system that can be used as comparison over time.

Many factors can influence how performance is affected in live

production scenarios:
▪ Integrations with other production systems have fluctuating
response times that are different than in the simulated tests
▪ Workflows such as back-office jobs could be running, which
stresses the systems

583
 ▪ The network traffic can vary throughout the day depending on an
organization’s usage patterns
▪ For remote workers, reliance on individual internet connections
Chapter 17, “A performing solution,
could cause different outcomes for each user
beyond infrastructure,” covers the
importance of having a performance
strategy that includes elements such as
defining performance requirements,
Ultimately, the responsiveness experienced by end users is caused by
establishing baseline metrics, and a mix of multiple factors that aren’t limited to the performance of the
ongoing performance testing.
Dynamics 365 application itself.

To understand the performance impact in these scenarios, proper

end-to-end logging and collection of telemetry is required. This
provides a complete picture of how long each step of the pipeline
takes to complete. This information can reduce the time required to
You can use many different tools to moni- triage and fix any issues.
tor performance. For Dynamics 365 Finance
and Dynamics 365 Supply Chain Manage-
ment, Lifecycle Services (LCS) is the first As this data is collected, monitor the performance of the solution and
place to go for performance monitoring.
set up notifications to warn you when performance of any aspect of
For Dynamics 365 Customer Engagement, the solution varies from a defined range.
Dataverse analytics is helpful. With it, you
can gauge and monitor performance from
within the Power Platform Admin Tool. Poor performing solution leads to low user adoption. But you can stay
You can also use Azure Application Insights ahead of this through proper monitoring and alerts.
to monitor applications like Dynamics 365
for custom telemetry needs.

Licensing and operational cost

As your solution matures and new users are onboarded to your
Dynamics 365 solution, keeping track of the rate of licenses being
consumed will be very important. Insight on the number of available
licenses will help you plan ahead and adjust your purchasing as needed.
Stay compliant with user licensing
requirements and view the independent
software vendor (ISV) license status for Another aspect is the operational cost. With solutions that include
Finance and Supply Chain Management cloud computing, usage volume heavily impacts this expense. Staying
for more information.
aware of any changes in your organization’s usage will help business
Review the business subscriptions and
and IT managers make key budget decisions.
billing documentation for Microsoft 365
and the Microsoft 365 admin center to
manage your subscription for Customer
Engagement and Power Platform.
Storage allocation and usage
The Dynamics 365 Licensing Guide pro-
vides details on licensing requirements. With a Dynamics 365 subscription, you start with a set minimum
amount of storage space for your environment.

Implementation Guide: Success by Design: Service the solution 584

 Depending on your needs, you can add more data to suit your
business size and expected growth.

Administrators must be mindful of the amount of storage that is

available for the organization’s subscription as usage grows. Different
data classifications (such as transactions, files, and logs) have different
allocations depending on your subscription.

Users and integrations aren’t the only cause of storage growth. Logs
from system jobs, indexes created to help performance, and additional
application data added from new modules also contribute to storage
Refer to Chapter 10, “Data manage-
ment,” for details on storage enti- growth. Another scenario that impacts storage allocation is in the copy
tlements, segmentation, and impact and restore operations of an environment. Depending on the type of
to allocations when backing up and
restoring instances. copy, the size of the database can be very different. As an administrator,
you need to be mindful of who can create new instances and what their
For information on storage capacity for
Finance and Supply Chain Management, true needs are to minimize impact on the storage allocation as these
see the Dynamics 365 Licensing Guide.
copies are being restored.
Note that Dataverse storage capacity
entitlements and usage changed in 2019.
Administrators should monitor the volume of storage that is currently
used as well as its growth rate. This information will help you budget
for any additional storage needs or look into data archiving and
deletion to free up space. Scheduling and cleaning up data from
time to time will help as well. This is covered in the “Environment
maintenance” section of this chapter.

API request and service protection

One of the advantages of choosing a cloud solution over on-premises
software is that your application is scalable. Among other things, you
For Finance and Supply Chain Manage-
ment, use LCS to monitor API usage. For can adjust the amount of processing and memory resources based on
more information, review priority-based
throttling and how to configure priorities
the usage. But to provide consistent availability and performance for
for special circumstances such everyone, Microsoft monitors and applies some limits to how APIs are
as integrations.
used. These limits are designed to detect when client applications are
For Customer Engagement, use the making extraordinary demands on server resources.
Dataverse analytics available in the
Power Platform Admin center. Refer to
the service protection API limits and the
The intent is to prevent over-utilization of resources to preserve the
ways to maximize throughput to avoid
hitting these limits. system’s responsiveness and performance for environments running
Dynamics 365. When an API limit is hit, the service may be throttled

Implementation Guide: Success by Design: Service the solution 585

 or even prevent API calls to be run. Error logging shows you when
these limits are exceeded.

Administrators can also pull historical telemetry reports to see if any areas
of the application are at risk of hitting these limits. Then you can work with
the implementation team to make appropriate design changes. Or better
yet, tools like Azure Application Insights allow you to set thresholds on
these API calls so that when they’re exceeded, the administrator is notified
and can mitigate the risk of throttling or being shut down.

Security and privacy

From a security perspective, you want to prevent unwanted access to
your system, applications, and data. The unified Microsoft 365 security
center combines protection, detection, investigation, and response to
threats for your Dynamics 365 system, all in a central portal.

At the solution level, you need strategic planning to create a robust

security policy that can be implemented during deployment. As the
administrator who services this solution, you must also monitor and act
upon other aspects of security.

Take a scenario in which an employee changes roles or leaves the

company. You need a process to change the security role and permissions
for that user. When a vendor is granted access to a system as a consultant,
they often continue to have access to it even when they’re no longer
working with the company. Microsoft recommends that you have a
process to address these scenarios.
Refer to Chapter 12, “Security,” for details
on security strategy and planning, and see
the security center overview and the compli- “Who has access to my customer data?” may seem like a simple
ance center overview for more details. The
question, but as the number of users and system usage grows, it can
Microsoft 365 security center provides the
ability to search through Dataverse activity be daunting to keep track. You need to be able to identify who is
logging for Customer Engagement.
accessing the system and what they’re doing with the data.
Your organization may be subject to
rules such as the General Data Protection
Regulation (GDPR) that give users specific Depending on the industry or the region of the organization, you
rights to their personal data. You may may also have regulatory or compliance-related requirements. You
need to respond to data subject requests
(DSRs) to delete a user’s personal data. might need to provide reports about customer data handling or
delete customer data upon request.

Implementation Guide: Success by Design: Service the solution 586

 Microsoft recommends that you have a proper auditing strategy to
capture the information needed to track user actions so that you
can satisfy these types of requests. Most common areas are covered
by Dynamics 365. Because auditing takes up more storage and
potentially impacts performance, administrators need to turn some
of these capabilities on where they may not be by default.

User access and resource usage

For any application or solution, it’s important to understand your
organization’s resource usage patterns. Business sponsors want to
know who is using (and not using) the system and the frequency of
business processes and use cases that are being run.

Monitoring and Diagnostic tools in LCS

You can clearly understand user adoption through telemetry on usage
provide telemetry for usage patterns for
Finance and Supply Chain Management. patterns, and you can take proven actions to improve it. For example,
Microsoft Dataverse analytics provides
if a certain capability isn’t being used much, you can query the users
user access and usage pattern teleme- to get their feedback on it. By using telemetry to identify areas with
try for Customer Engagement and the
Power Platform. the greatest impact to the user group, the project team can work with
business sponsors to prioritize features.

You can also use this information when estimating license needs. For
example, some licenses may be assigned to users who rarely access
the system; you can reassign these to someone new. Insights provided
by the telemetry are readily available in Dynamics 365. Use this data
to help improve business processes and allocate resources in the right
areas to maximize your investment.

The platform is constantly adding functionality to the standard usage

reports. If you have any gaps between your needs and these options,
you can tailor tools such as Azure Application Insights to your usage
pattern tracking requirements.

Application logs and errors

Typically, system errors are found by end users who report issues
through an IT service management (ITSM) process. But you may have
unreported issues that can only be found in the application logs.

Implementation Guide: Success by Design: Service the solution 587

 Turning to notifications and application logs to proactively look for
entries is a good way to find trouble spots and address them before
they impact users.

You can use tools such as Azure Application Insights for Dynamics
365 and other applications and services that are part of the overall IT
solution for your organization. Application Insights lets you collect
Monitoring and Diagnostic tools in LCS
allow administrators to monitor and query telemetry both in and outside of Dynamics 365.
logs for issues detected in the system for
Finance and Supply Chain Management.
For example, if a process is initiated by a user in Dynamics 365 that calls
Trace logging in Dataverse provides
plugin error information for Customer
on an integrated, but external system, Application Insights can still detect
Engagement and the Power Platform. performance and exceptions at different stages of the execution pipeline.
You can also use Microsoft 365 service You can see these issues whether they occur at the user interface level, in
health to identify service issues, and Dynamics 365, or the external system. This empowers the administrators
administrators can be notified via email
or through the mobile app. who monitor alerts to react quickly the moment the exceptions surface.
They also have immediate access to information on the source of the issue.

Messages from Microsoft

As your SaaS provider, Microsoft provides ample information about the
health of your service, upcoming release information, service exceptions
and interruptions, and other important details. The Microsoft 365 Message
center is the central location for technical and licensing communications
with our customers. As an administrator of your solution, this will be a
very important area to frequently check for new messages.

Typically, organizations have multiple people who manage Microsoft

services. You may want to provide appropriate team members with
access to these messages.

You can also configure your preferences on receiving emails. A mobile

app is available to manage these communications.

As described in this section, having the information needed to understand

the health of your system is key in knowing when to act. Multiple tools
are available to monitor your solution, which can feel daunting. But
with tools to alert administrators when services are running less than
optimal, the task will be much more manageable.

Implementation Guide: Success by Design: Service the solution 588

 Key alerts you can receive from Microsoft include the service updates to
your Dynamics 365 solution. In the next section, we discuss when and how
Dynamics 365 is updated and what you can do as an administrator to take
advantage of the predictable nature of the service updates.

Service updates
Monitor service health To enable businesses everywhere to accelerate their digital transfor-
mation, Microsoft is continuously enhancing Dynamics 365 with new
Service updates capabilities. We add product enhancements and performance improve-
ments at a rapid pace, so it’s important to make sure Dynamics 365
Environment maintenance
provides an optimized user and administrator experience. Our objective
Continue the business is to help you keep your computing environment current in a consis-
application journey tent, predictable, and seamless manner.

Service updates are continuous, touchless updates that provide new

features and functionality. They eliminate the need to do expensive
upgrades every few years. The service updates maintain backward-
compatibility, and contain application and platform changes that are
critical improvements to the service, including regulatory updates.

Microsoft’s push to One Version

One of the benefits of Dynamics 365 is that every customer runs on
the same version of the service. With market demand for increased
agility to gain incremental benefits of these cloud offerings at reduced
operating expense costs, Dynamics 365 took a bold step to refrain
from multiple releases. Now every security patch, bug fix, performance
enhancement, and functional improvement accrues to all implementa-
tions across the globe.

We call it One Version. This is also referred to as Run One, a solution

that is evergreen in the modern SaaS world.

One Version promises to bring predictable release management to

these areas:

Implementation Guide: Success by Design: Service the solution 589

 ▪ Predictable updates with continuous deployment
▫ Regularly scheduled intervals for major releases of performance
and reliability improvement updates throughout the year
Refer to “Modernizing the way we update
Dynamics 365” by Mo Osborne, Microsoft
▫ Using safe deployment practices and monitoring updates
COO Business Applications, for details on closely for any issues
the benefits of continuous deployment.
▫ Knowing the cadence up front so you can work it into your
project schedule
▪ Early visibility of upcoming changes
▫ Access to release notes well before the release for readiness
so you know what the new changes are and the dates to start
planning early
▫ Visibility into new features and updates for business sponsors
that can help their team be trained in new features
▪ Test new capabilities ahead of time
▫ Ability for customers to opt in for early access to the release
▫ Fully supported, production-ready updates that can be tested
in your sandbox environments
▫ Turning on features when you’re ready instead of having it
forced upon your environment
Fig.
20-1
▪ Opportunities to advance your knowledge
▫ Faster access to new features
▫ Ease in sharing ideas and collaborating with the greater
community because everyone is on the same version
Planning Scheduling
effectively to avoid conflict with
for updates project plan With One Version, Dynamics 365 addresses the tough challenges
faced by customers and partners and reduces common rollout
concerns (Figure 20-1). The solution automatically enhances the
Raising Handling
awareness predictability of product updates. For example, One Version
unsupported
of feature changes techniques reduces the effort and capacity needed to test the update. It
makes it easier to manage business change and facilitate
appropriate adoption.
Staying aware Encouraging
of what’s coming adoption
Although One Version greatly reduces the impact of deployments,
the solution owner and the implementation team are still responsible
for making sure certain tasks are addressed. These include planning
Understanding Aligning
effort and capacity the team with the for the update, assigning ownership of the tasks, raising awareness
for testing implementation plan
of coming changes, and supporting adoption. We cover these items
throughout this section.

Implementation Guide: Success by Design: Service the solution 590

 Release readiness
System administrators and others who have signed up for service notifica-
tions are alerted about upcoming releases, minor updates, and bug fixes.
Being properly prepared is the key to successfully managing your solu-
tion. This isn’t just up to the Dynamics 365 administrator. You need tight
coordination between systems administrators, the Dynamics 365 project
team that works on the platform, and the business user groups and subject
matter experts (SMEs) who are the end users and champions of the system.

These notifications provide the information you need to start your plan-
ning—dates of release availability, release notes, and the process to opt in
for early access.

Release notes and feature deprecation

Great planning starts with education—learning when and what is being
rolled out. Your notifications explain the upcoming changes with a link
to online documentation for more details. This information can help you
gauge the impact to your organization and determine when and how to
take action.

Administrators are notified of upcoming releases through the message

center and by email. Notes for major releases are made available months
before availability and provide several key pieces of information. First, the
notes list the features in the release, including a description of the feature
and the general availability (GA) date. The notes also show if the feature
will be available for public preview and early access. This information will
help in your readiness tasks.

Let’s break down the three types of release dates:

▪ Public preview  These capabilities are pre-release features
intended to collect early feedback and aren’t intended for produc-
tion use. For planning purposes, these public previews can give
you a good idea of what’s coming up in the short-term roadmap. Not
every feature is available through a public preview.
▪ Early access  You can apply these production-ready, fully supported
updates and features to your environment prior to the GA date.
See more details in the following section about the process of

Implementation Guide: Success by Design: Service the solution 591

 opting in early.
▪ General availability  This is the date that these features are
deployed to your environment if the administrator hasn’t opted in
for early access.

Each release wave includes features and functionalities that you can
enable for different types of users:
▪ Users, automatically  These features include changes to the user
experience for users and are enabled automatically.
▪ Administrators, makers, or analysts, automatically  These
features are meant to be used by administrators, makers, or
business analysts. They’re enabled automatically.
▪ Users by administrators, makers, or analysts  These features
must be enabled or configured by the administrators, makers, or
business analysts to be available for their users.

If you choose to opt in to early access updates, you get features that
are typically mandatory changes automatically enabled for users. Each
feature in the release notes indicates which category it falls under.

There may also be announcements about deprecated features—those

with a timeline to stop being supported. These announcements are
equally important because they allow you time to gauge the impact
and work with your business sponsor and technical team to address
any concerns.

Refer to the Message center Deprecation of features is an important part of a solution’s growth. As
for detailed information on
business and technology needs evolve, the solution needs to change
notifications, email preferences,
and recipients for service updates. to keep up with new requirements for security, compliance, and over-

For Finance and Supply Chain

all modernization. As a result, some features are deprecated and are
Management, refer to the One replaced by a better solution.
Version service updates overview
for details on release planning,
release notes, deprecations, and Deprecated features continue to work and are fully supported until
release cadence.
they’re officially removed. After removal, the feature or capability
For Customer Engagement and the
no longer works. The deprecation notes provide information on
Power Platform, refer to the release
notes, deprecation announcements, what features are being removed, when this will happen, why it’s
and release cadence.
happening, and what actions you can take to address the impact.
Just like when you’re getting ready for new features, organizations

592
 must plan and prepare well before the removal of features to avoid
negative impact from the deprecation.

Opt in for early access

At this point, you have the release notification and reviewed the
release notes to do an impact assessment. The next step is to work
with the project team to test the release with your solution. Some
features available through our public preview program may be of
interest for your organization.

Microsoft recommends creating a sandbox instance from a backup of

your test instances prior to update deployment. The environment should
be running your latest solution with enough data for meaningful testing.
If you already have automated tests, you can speed up this process by
validating that your solution functions as expected. If you need to fix any
areas, you can flag them and work them into your project plan.

If your solution has been developed using supported methods for

customization, the release in most cases will have little to no effect on
it. Scenarios regarding deprecation that weren’t addressed previously
may surface from testing and need to be fixed.

Another important area to cover for early access is to work with the
business sponsors to help them understand the impact to the end users.
As part of the release, some updates may impact user experience, such
as user interface (UI) navigation changes. Even small differences can
have a meaningful impact. Imagine users in a large call center scenario,
in which every additional second on a call with a customer can impact
their service goals. In such a case, business managers want to make sure
that the user group receives proper communication and takes time to
provide training if necessary.

After this due diligence, your organization can schedule and

apply the new release to the production environment. You
should time this task for when it’s most convenient to the users
and administrators. Your organization can determine the best
time to apply the release—when there will be the least amount of
disruption to end users, other technical dependencies, and impact

Implementation Guide: Success by Design: Service the solution 593

 to other projects. Don’t wait for the release to be automatically
applied to your environment.

Microsoft has deliberately planned the release cadence so that it

allows you enough time to prepare. Aligning your own internal
deployment planning with this schedule allows you to benefit from
the new features and bug fixes and make sure that any customizations
specific to your environment continue to function as expected.
Thorough planning is key so that there are no unexpected results
from the updates and you have control of when these releases are
applied to your environment.

Product-specific guidance: Service

updates for Finance and
Supply Chain Management
With continuous touchless service updates in Finance and Supply
Chain Management, the system maintains backward-compatibility,
so you don’t need to merge your code.

Safe deployment practices

Three new innovations provide additional enablement for the new
modern application lifecycle—safe deployment rings, automated
regression testing, and the customer Release Validation Program (RVP).
In this section, we explain these and how they apply to you.

Deployment rings are a technique used to decrease any risk associated with
rollouts for Azure and other Microsoft cloud services managed at global scale.
As Dynamics 365 updates are created each month, they progress
through a series of rings, with each ring providing broader exposure
and usage and validation through system telemetry (Figure 20-2).
Refer to the Dynamics 365 for
Finance and Operations Cloud The GA update benefits from extensive Microsoft testing as well as
Application Lifecycle for more
validation through each of the earlier rings.
information.

Update cadence
Customers are required to take a minimum of two service updates per
year, with a maximum of eight service updates per year (Figure 20-3).

594
 Fig.
20-2
Safe deployment practice for
Finance and Operations

Finance and Supply Targeted First Standard Servicing

Chain Management team release release release (GA) N and N-1

Finance and
Supply Chain Targeted First
Management team release release
▪ Extensive validation ▪ Preview early access program ▪ Select customers
▪ Compatibility checker ▪ Preview build ▪ Auto update
▪ Over 100 customer RVPs ▪ No production use ▪ Production ready

Standard release Servicing

▪ Predictability ▪ Hotfix support for critical issues
▪ Customer-controlled ▪ Available for N and N-1 updates
patching window ▪ Cumulative updates
▪ UAT followed by production
on consecutive weeks

Implementation Guide: Success by Design: Service the solution 595

 Fig.
You can choose to pause up to three consecutive updates at a time to
20-3
accommodate your project schedule.

8 updates
Pausing a service update can apply to the designated user acceptance
delivered testing (UAT) sandbox, the production environment, or both. If the pause
per year window ends and the customer hasn’t self-updated to a supported
service update, Microsoft automatically applies the latest update based
Jan
on the configuration selection available in LCS.
Feb
Mar System updates follow these guidelines:
Apr ▪ Updates are backward-compatible
▪ Updates are cumulative
May
▪ Customers can configure the update window
June
▪ Quality updates containing hotfixes are only released for the
July current version (N) or previous version (N-1)
Aug ▪ System updates contain new features that you can selectively
choose to enable
Sept
Oct
Release readiness
Nov We strongly recommend that you plan ahead and work the updates into
Dec your internal project release cadence. To do so, take the following steps:
▪ Step 1: Plan  Have a good understanding of the release schedule
and a plan to work this into your application lifecycle management
(ALM) strategy. Because you can pause updates up to three months,
you can set aside plenty of time for testing, impact analysis, and
developing a deployment plan. Use the impact analysis report from
LCS to identify areas of change that may affect your solution and
help determine the level of effort needed to remediate any impact.
▪ Step 2: Test  We recommend using a non-production
environment such as your UAT instance to opt in early and apply the
release. You can configure service updates through LCS and specify
how and when you receive service updates from Microsoft to your
environments. As part of the configuration, define the update
environment (production) and an alternate sandbox (UAT). Use the
Regression Suite Automation Tool (RSAT) to perform regression
testing to identify any issues. Work any fixes into your ALM cycle
and deployment plans.
▪ Step 3: Deploy  After you define environment and schedule for

Implementation Guide: Success by Design: Service the solution 596

 the service updates through LCS, the back-end tools automatically
update the system.

Product-specific guidance: Service

updates for Customer Engagement
Dynamics 365 is continually updated to provide valuable functionality,
fix any issues, and maintain the overall health of your solution. System
administrators receive communications and are alerted to when changes
will be made to their environments. Organizations need to plan to enable
these service updates in alignment with their internal ALM process.

Safe deployment practices

When rolling out the major and minor releases, Microsoft Dynamics
365 follows a series of progressive, staged deployments across the
world to adhere to safe deployment practices. At each stage, feedback
is gathered and quality is monitored for continuous improvement.

For example, for Customer Engagement apps and Power Platform

releases and updates, a station-based deployment plan is followed
(Figure 20-4).

After thorough internal integration tests and validations, Customer

Engagement and Power Platform updates are rolled out to Station 1.
Check the latest release of station map-
This first release station consists of select customers and production-like
ping and their corresponding regions. environments. This is set up for early validation testing of the service
Release updates to Station 1 through
Station 6 follow the dark hours defined update before it’s rolled out to customer production environments.
for each geography.
Based on the pass rate of the first release, the service update gets rolled
out to the customer environments by Stations 2–6.

Some organizations require instances to be in different regions, which may

place them into different stations than that of their primary geographies.
Therefore, as the deployment cycle of a new release commences, instances
in different stations can be on different versions of the solution. As your
project team is developing new features that are consumed into your ALM
deployment process, set up a version check for Dynamics 365 and make
sure that there is a match so that your project doesn’t encounter
incompatibility issues. When the version of the source environments

Implementation Guide: Success by Design: Service the solution 597

 matches the destination, you can safely deploy your solutions.

Types of service updates

There are four types of updates:
▪ Major release  These happen twice per year in April and
October. Major releases offer new capabilities and functionalities.
These updates are backward-compatible. New features with

Fig.
20-4 Safe deployment practice for
Customer Engagement
Customer First Standard release
Engagement Team release and servicing

Station 1
Station 2
Station 3
Station 4
Station 5
Station 6

Customer
Engagement Team Station 1: First release Station 2 JPN, SAM, CAN IND
▪ Extensive integration ▪ Production quality
testing and validation ▪ Early view of weekly release
Station 3 APJ, GBR, OCE
▪ Solution checker ▪ Select customers

Station 4 EUR

Standard release and servicing

Station 5 NAM
▪ Predictable weekly releases
▪ Hotfix support for critical issues
Station 6 SPECIAL

Implementation Guide: Success by Design: Service the solution 598

 changes that might be disruptive to the user experience are
turned off by default.
▪ Minor service updates  Minor service updates are deployed
on a weekly basis, region-by-region. They contain software
customization changes that support new features, product
improvements, and bug fixes.
▪ Planned maintenance  Planned maintenance includes
updates and changes to the service to provide increased
stability, reliability, and performance.
▪ Unplanned maintenance  Applications sometimes encounter
unexpected issues that require changes to maintain availability.
Microsoft strives to provide as much notification as possible
around these events.

Update cadence
Customers receive two major updates per year, in the April and
October GA releases (Figure 20-5). You can get early access and opt
Dynamics 365 apps have a different in months before the GA dates. These updates apply to both Power
cadence from the major releases. For
example, Dynamics 365 Marketing and Platform and Dynamics 365 apps. We encourage you to opt in early
Dynamics 365 Portals have monthly to test and apply the release. The releases are production-ready and
updates. Apps from ISVs from AppSource,
Microsoft’s app marketplace, may also fully supported even when applying prior to the GA date. Activation
have a different cadence. You should
for major updates is automatic through safe deployment processes
consult with the ISVs for any third-party
apps you’re using. for the region where the Dynamics 365 instance resides, on the
deployment dates specified for the region.

Release readiness
We strongly recommend that organizations work updates into their
internal project release cadence. To do so, take the following steps:
▪ Step 1: Opt in for early access  Before you apply the changes
to existing production or non-production environments (which
may disrupt users and developers), we recommend you create
a new instance. You can’t revert back to the previous version, so
all testing should be done on a new instance. Take a copy of the
test environment that has your latest solution and data and
Refer to the Dataverse storage capacity
create a new instance from it. Enable early access to apply the
guidance to understand the impact on your new release capabilities. After you opt in, some features are
storage allocation for your tenant when
creating new instances from a backup. turned on by default; others may require an administrator to
explicitly configure them. The details are documented in the

Implementation Guide: Success by Design: Service the solution 599

 release notes.
▪ Step 2: Test  Run regression testing to make sure that the
solution continues to function as expected. Early opt-in features
are production-ready and fully supported, so if you encounter
any errors, you can submit a service request to report and get help
with issues. Another important aspect of enabling the early access
capabilities is that some UI and navigation changes may impact
users. Work with your user group to do a side-by-side comparison
between the current and opt-in versions. Use the release notes to
identify the areas where UI or navigation changes have been made.
Document the change with screenshots to be shared with the users.
Depending on the significance of the changes, it may warrant some
level of end user training as well as communications out to the group.
▪ Step 3: Deploy  When the testing is complete, you can turn the
Fig.
20-5 Feature release
October October April April
public preview release public preview release

May Jun July Aug Sept Oct Dec Jan Feb Mar Apr May

Access to latest features (often weekly) Prep time latest release Latest release
(admin opt-in) (admin opt-in) (automatically applied)

F1
F2
F3 Features Features
F4 F1-F6 F1-F6
F5
F6 Access to latest features (often weekly) Prep time latest release Latest release
(automatically applied)
(admin opt-in) (admin opt-in)

F7
F8
F9
F10 Features Features
F11 F7-F12 F7-F12
F12

Continuous updates that can include feature code which has no UI impact (often weekly)

Opt-in to experience all UI features coming in the next scheduled update (preview)

All GA functionality is automatically deployed on a semi-annual schedule

Implementation Guide: Success by Design: Service the solution 600

 early access features on in other non-production instances. This
ensures that new features and customizations by the project team
are developed on the upcoming early release solution. You may
keep one or more instances without the early access features turned
on to support production bug fixes or anything that needs to be
deployed prior to enabling the early access features in the production
environment. When it’s time to deploy your solutions to production,
you enable the early access features on the production instance. You
should notify business users of the new features (within the product
as well as anything custom built by the project team) and any
changes to how end users will navigate through the system. Timing
the deployment is important. Microsoft doesn’t recommend opting
in at the same time you have a new project release in the production
environment. If you encounter deployment issues, it’s easier to
troubleshoot when you’re not deploying multiple solutions to
the environment.

Environment
maintenance
Monitor service health Protecting your solution and providing continuous availability of
service is your primary goal as the system administrator. In a cloud
Service updates environment, these maintenance jobs are automated, but it’s critical for
Environment an organization to have a strategy so that these routine jobs are
maintenance appropriately configured and scheduled. In some cases, you may need
Continue the business to perform these tasks manually but still in alignment with your overall
application journey planning and strategy.

User access and security

Members of business and technical teams fluctuate. New employees
You can manage users and licenses in
both Finance and Operations apps and are onboarded and existing ones may change roles or leave the
Customer Engagement.
company. As an administrator, you need to work with other teams to
You can also use Azure Active Directory manage user access, assign the appropriate licenses, and assign security
(Azure AD) groups to simplify user manage-
ment and provision and deprovision users.
roles that meet their business and technical needs. You may want to
hand over some of these responsibilities to business and IT managers

Implementation Guide: Success by Design: Service the solution 601

 who are more familiar with their resources’ user profiles and know if any
frequent changes in their roles require immediate changes.

Your organization’s data is likely one of the most important assets you’re
responsible for safeguarding as an administrator. The ability to build
apps and automation to use that data is a large part of your company’s
success. You can use Power Apps and Power Automate for rapid build
and rollout of these high-value apps so that users can measure and act
on the data in real time.

Apps and automation are becoming increasingly connected across

multiple data sources and multiple services. Some of these might be
external, third-party services, possibly even social networks. Users
generally have good intentions, but they can easily overlook the
potential for exposure from data leakage to services and audiences
that shouldn’t have access.

Microsoft recommends that you create data loss prevention (DLP)

policies to act as guardrails to help prevent users from unintentionally
exposing organizational data. You can scope DLP policies at the
environment level or tenant level, which provides the flexibility to
craft sensible policies with the right balance between protection and
productivity. For tenant-level policies, you can define the scope to be
all environments, selected environments, or all environments except
those you specifically exclude. You can define environment-level
policies one environment at a time.

Dynamics 365 environment management

Dynamics 365 provides point-in-time restore (PITR) capabilities for
Although the backup and recovery
operation is dependable, it could also be
databases. This means that all databases are backed up automatically
time-consuming depending on the size by the system and retained for a set number of days. In the event of
of the backup.
accidental corruption or deletion, administrators can choose to restore
In a Customer Engagement scenario the database from any of the backups taken. The automated backup
when solution imports fail, it’s often
better to fix the import issue instead system and PITR provides a zero-admin way to protect databases.
of restoring from a backup. Fixing the
import should take significantly less
time than restoring. If your organizations require a proactive approach to manually take
backups (such as before a deployment of a new release of your solution),

Implementation Guide: Success by Design: Service the solution 602

 the administrator may be called on to assist. You should perform these
tasks in line with your organization’s environment strategy.

Data management
Data is central to all applications. It drives business decisions through
analytics and artificial intelligence. It also reveals crucial information
about the overall health of the system and what administrators need
to do for maintenance. The theme of this chapter is to be proactive
in planning and strategizing around the upkeep of your system. Data
maintenance is no different. In this section, we discuss the service
aspects of data management. To explore the broader topic, refer to
Chapter 10, “Data management.”

Refer to Chapter 9, “Environment

strategy,” to explore the importance Data volume grows quickly. Dynamics 365 has routine jobs in place to
of having a strategy for creating and
maintaining environments.
manage data that maximizes system performance at the platform level.
You can also perform maintenance jobs like reindexing databases that
Learn more about database movement
operations in Finance and Supply Chain adjust to variations in transactional data. But more improvements can
Management and review environments
always be done to better fit how you use the services. Developing a
for more information about managing
instances for customer engagement. strategy to archive and delete data is an important step toward
maintaining a well-functioning system.

How fast is the data growing?

You first need to understand how fast data is growing in your environment.
The rate of growth is a key metric to monitor. Set alerts to let you know when
this growth rate exceeds a threshold determined by your organization.

You can find details on storage allocation The rate of growth can fluctuate depending on the number of users or
and purchasing additional storage for
even during certain times of the year if your business practice has special
Finance and Supply Chain Management in
the Dynamics 365 Licensing Guide. circumstances that may impact record creation. Monitoring storage
Read about the storage capacity model for
growth and using historical trends will help estimate data growth. This
Dataverse and how to check storage growth. information can help you determine how often the data archiving and
removal process should take place.

What data can be removed?

Planning starts with identifying the types of data that need to be stored
over particular timeframe. Besides cost, there are performance implications
of having a large dataset. Building a data removal and retention strategy
will help determine what to do when data is no longer useful.

Implementation Guide: Success by Design: Service the solution 603

 Transactional data may help you make key business decisions, assist
customers, and use AI to determine the next best action. But after years
of inactivity, the data may be taking up storage space and not providing
Review the cleanup routines for Finance
any value.
and Supply Chain Management to delete
historical logs and notifications. You
should only run these cleanup routines
after the business has completed a
You may have logging requirements like auditing customer records
detailed analysis and confirmed that the to understand which users have access to read and modify personally
data is no longer required.
identifiable information. But excessive data auditing leads to rapid
You can also free up storage space for
storage growth and can negatively impact performance. Storing large
Customer Engagement.
files and attachments to emails is also an area in which datasets can
quickly expand. Planning around what needs to be stored and for how
long should be done deliberately and conservatively.

You may have logs, notifications, and other system records that you can
delete with no impact to the business. You may also have other transactional
data that can be deleted. Because Dynamics 365 applications have heavy
parent-child relationships between records, pay careful attention to how
records are deleted and any impact to related records. Look for triggers
that run extension code or workflows when a record is modified or deleted.
A delete operation could potentially write a new entry in the audit log to
record the transaction. You must account for all these things when planning
for bulk deletion.

Archive and retention strategy

What happens if the database is growing too large, and you’re seeing the
impact of it, but deleting records isn’t an option? You may need to retain
the data due to organizational, compliance, or regulatory requirements. In
this case, you can archive the data instead.

We recommend the following strategy:

Your organization may be subject to rules ▪ Identify retention requirements  Get a full understanding of
such as the GDPR that give users specific
the requirements for the types of records that you need to retain
rights to their personal data. To comply
with privacy and security regulations, you and how long you need to do so.
may need to respond to DSRs to delete a
user’s personal data.
▪ Develop an archiving strategy  Once you have a clear
understanding of what data to retain, plan for how to archive these
We recommend reviewing the guidelines
for privacy and personal data for Microsoft records. Where will you store the archived data? How will you access
Dynamics 365. this data when needed? What is the process to periodically move
records from the Dynamics 365 systems into the archives?

Implementation Guide: Success by Design: Service the solution 604

 ▪ Remove the data  For any data that doesn’t meet the retention
criteria, plan out your process to remove it from the archives.

Continue the business

application journey
Monitor service health The importance of continuing your education on the business
application platform is often overlooked. In the on-premises software
Service updates world, you would go live with a version of your solution and it would
mostly remain the same until you decided to develop or purchase
Environment maintenance
add-on functionality. As mentioned earlier in the service update
section, the benefit of continuous updates in Dynamics 365 is that
Continue the
business application you always have access to the newest features to help your business.
journey This also means you have a constant stream of new and better ways
to meet your business and technical objectives through your solution.

For example, let’s look at integrating Dynamics 365 with external

systems. It wasn’t very long ago that you needed custom code
modules in order to efficiently pass data to and from Dynamics 365.
But with tools like Power Automate and Azure Logic Apps, you can
build very powerful integrations through configuration with little to
no code.

Reporting is another good example. The out-of-the-box experience

for reporting had limitations in Dynamics 365—you could only
build reports with data stored in Dynamics 365. Now, with tools
like Power BI and its capabilities to build reports from data in and
outside of Dynamics 365, you have much more flexibility to quickly
build out and embed powerful reports. Also, some advanced Azure
and Microsoft 365 services coexist with Dynamics 365, and are
going through the same type of evolution. Everything from business
functionality to ALM build and deploy tools are constantly seeing
incremental improvements.

The only way to truly understand these benefits and how to best
apply them in your solution for the most return on your IT investment

Implementation Guide: Success by Design: Service the solution 605

 is by continuing your education. Gain an understanding of where the
business application domain is headed. Strive to understand what is
in the Dynamics 365 roadmap, why it’s included, and how it can make
your solution and your organization more efficient and robust.

Each year, Microsoft creates many events and learning opportunities to

help developers, IT and business professionals, partners, and educators
extend their expertise. Here are some great resources for furthering
your education:
▪ Microsoft Events consolidate links to stay up to date on upcoming
events for all kinds of audiences—both business and technical focused.
▪ Microsoft Business Applications Summit helps you maximize
on using business applications through live and on-demand
recordings and discussions.
▪ Microsoft Ignite is for anyone who is motivated to be on the frontier
of innovation and tech. There are live and on-demand sessions
tailored for different audiences.
▪ Experience Dynamics 365 provides access to Dynamics 365 experts
and the Insider program to participate in preview features. You
can also provide feedback to Microsoft engineers and vote on new
product features.
▪ The Dynamics 365 YouTube channel delivers the latest product news,
announcements, events, demos, and customer success stories.

Conclusion
In summary, take steps to be aware of your solution performance, be
proactive in taking action, and be prepared with a solid strategy for
maintaining your environments. Keep up on new trends and tools that
can help improve your solution and your organization.

It all starts with visibility into the health of the system through proper
monitoring. Having the right telemetry and managing notifications
from the system as well as Microsoft will help you to prioritize and act
to address maintenance needs.

A large part of the overall maintenance is simply understanding

Implementation Guide: Success by Design: Service the solution 606

the continuous update model of the Dynamics 365 service. We
recommend that you be proactive in planning, communicating, and
deploying service updates to gain the benefits of new features and
enhancements made available throughout the year.

Appropriate maintenance of your environments and data will help protect

and optimize your solution. This is a team effort, not just the responsibility
of IT administrators. By working across the business and technology
resources, you can make sure that the solution performs well technically
and maximizes results for the user community.

Finally, just as your solution is continuously updated, organizations

also need continuous education in trends and tools related to your
business applications. With awareness of the product roadmap as
well as improvements in anciliary products, you can maximize your
organization’s investment in the Dynamics 365 platform.

607
 Case study
Fruit company learns the
importance of servicing
the solution
An agricultural business that grows fruit implemented Dynamics 365
Finance and Dynamics 365 Supply Chain Management as soon as the
cloud version of Dynamics 365 became available. The company has
been a global leader distributing fruit across different regions, and
warehouse operations and transportation are part of the company’s
core business.

Before implementing the Finance and Supply Chain Management

apps, the fruit company was using Dynamics AX 2012, and was familiar
with the ALM of that on-premises enterprise resource planning (ERP)
solution—but moving to the cloud required a mindset change.

Dynamics 365 has been evolving since its initial release, when the
application and the platform were released as separate
components. By the time continuous updates and a single, unified
version became the norm for Dynamics 365, the fruit producer’s
operations in the cloud were mature in taking updates under this
modality. The company was ready to adopt the modernized update
pattern and take advantage of the continuous innovations from
Microsoft. They also wanted to fulfill one of their expected returns

Implementation Guide: Success by Design: Service the solution 608

 on investment (ROIs) by entrusting Microsoft to bring new
functionality, instead of developing it on their own.

The fruit company’s solution involved standard Dynamics 365 apps,

some extensions, and an ISV. As the direct implementer of the entire
solution, the ISV created a strategic partnership with the company and
provided IT outsourcing services.

In the beginning, the incentive for the company to update Dynamics

365 with new releases was the continued support from Microsoft per the
software update policy. The app functioned properly and applying the
releases was manageable, so the company didn’t have hard requirements
to stay up to date with the release cadence.

But the company noticed that while the ISV kept the solution up to date
with the Dynamics 365 releases, the ISV always provided an update of their
solution using Microsoft’s last update ring—or missed it entirely. Because
of this, the company had to apply Microsoft’s updates late in the release
cycle or, in some cases, not at all.

To grow and distribute fruit effectively, the company relies heavily on

time-sensitive warehouse operations where every hour matters as fruit
is moved from the fields to climate-controlled warehouses to shipping
containers that are sent worldwide as quickly as possible to avoid
spoilage. They also were expanding into distribution of seasonal fruits
in new markets, which made management of warehouse operations
even more critical.

Then Microsoft notified the fruit company about new functionality for
advanced warehouse operations in an upcoming release of the Supply
Chain Management app. Because of the expansion of the company’s
operations and the complexity of their warehouse management, these
new features were crucial to increasing productivity and saving time
and resources while managing the expansion.

To adopt this new functionality, the company had to test the features
in advance to align their processes. They realized that being late on
updating standard features wasn’t workable, and they wanted to

Implementation Guide: Success by Design: Service the solution 609

 optimize their ALM and validate their solution in early update rings. To
be ready for their peak season, completion of testing was time-sensitive,
and the company would only have enough time to adopt new
functionality and do proper regression testing if all solution components
were in place and ready. So, they asked their ISV to provide software
earlier than the general availability of their new version.

The challenge came when the ISV wasn’t aligned to that timeline.
The ISV’s usual practice of adopting Dynamics 365 releases when they
became generally available—in the last ring—meant the fruit
company frequently was one version behind on Finance and Supply
Chain Management releases.

The ISV wanted to honor their partnership and be flexible in meeting

the timing requirements for updates, but due to dependencies on their
internal roadmap and ALM strategy, the ISV couldn’t make the software
available to meet the company’s timeline. As a result, the fruit company
couldn’t take advantage of the Supply Chain Management app’s
advanced warehouse operation functionality in time for their expansion.

Even though the company’s ALM strategy could adapt to Dynamics

365’s continuous update delivery model, reliance on the ISV’s software
prevented them from using the latest version of Dynamics 365.

After conversations with the fruit company, the ISV agreed to align
with Microsoft’s update rings and give the company—and their other
customers—an opportunity to test their entire solution when the
standard release was available.

Servicing the solution involves more than just Microsoft providing

continuous updates. Microsoft customers must synchronize updates to
their own extensions and any ISV solutions they’re using, so they can
benefit from the latest Microsoft innovations as soon as new features
are available.

Implementation Guide: Success by Design: Service the solution 610

21
Implementation Guide: Success by Design: Transition to support
Guide
Transition
to support
611
Poor service is always more
expensive than good service.

Introduction
When introducing a new system into an organization,
we need to think through the various areas that will
influence how the project transitions from project
mode to support mode.
In this section, we discuss how to construct a strategy to help you
prepare, define, and operate a support model.

Fig.
21-1 Organizations that spend the necessary time and energy to construct
a strategy that explicitly addresses how to create a fit-for-purpose
Support scope support organization for their Dynamics 365 application tend to have
▪ Enterprise architecture better user satisfaction, better adoption of the system, and therefore
▪ Business and IT policies
▪ Dynamics 365-specific considerations
higher-quality outcomes for the business.
▪ Business processes
▪ Business continuity
▪ System update cycles If this is the first Dynamics 365 business application for your company,
you may not have experience in setting up the support organization,
Support models support infrastructure, and support procedures for this application.

▪ Support model options

Establishing a support organization for a new system (for example,
▪ Support organization Dynamics 365) needs more due diligence and may require a different
perspective that better reflects the new technologies and new ways of
Support operations working that come with it.
▪ Transition
▪ Requirements management
This chapter is organized to reflect the journey of preparing, defining,
▪ Change management
▪ Hypercare and operating support through the categories of support scope, sup-
port models, and support operations (Figure 21-1).

Implementation Guide: Success by Design: Transition to support 612

 Support scope
A key aspect in preparing for production support is compiling the list
of topics, tasks, impacts, and responsibilities that supporting the new
Dynamics 365 application needs to cover. The support scope defines
the “what” of the support model.

Support models
The scope definition influences the decisions that need to be made to
identify the best support model and the constitution of the resulting
support organization. This helps define the “who” of the support model.

Support operations
Finally, we discuss the distinct support requirements that emerge from
the transition and hypercare project phases.

Support Support scope

scope Support scope activities should be started in the Initiate phase. It may
be tempting to defer defining the scope until much later in the project
Enterprise architecture because it’s not seen as urgent compared to getting the project off
the ground and creating the system solution. However, as this chapter
Business and IT policies
discusses, early attention to this topic enables the support team to be-
Dynamics 365-specific come engaged in the project from an early stage, which pays dividends
considerations during the project lifecycle.
Business processes

Business continuity
Enterprise architecture
System upgrade cycles
In most organizations, the Dynamics 365 system is embedded within
the wider enterprise system landscape. The Dynamics 365 application
has connections to multiple systems and to underlying and coexisting
technologies. When considering the strategy for defining the support
model, the focus is often solely on the Dynamics 365 application archi-
tecture. It’s worth accounting for the changes, constraints, procedures,
and influences from the surrounding architecture on the Dynamics 365

Implementation Guide: Success by Design: Transition to support 613

 Enterprise applications, and vice versa—how will the enterprise architecture and
architecture
environment be impacted by installing and operating the Dynamics
Business and IT policies 365 application?
Dynamics 365-specific
considerations Firstly, from the enterprise system architecture documents, confirm
Business processes that you have identified all the various systems, software, and data that
interact with Dynamics 365 applications. You should make a distinction
Business continuity between new systems being introduced into the enterprise architec-
ture and those that may be influenced by the new systems.
System upgrade cycles

During the project, the new Dynamics 365 system will probably be
implemented within a special sandbox (test environment) and not
necessarily be subject to all the influences and rules that the pro-
duction system is subject to. This also applies to the third-party test
systems that are part of the middleware or integrations. For example,
production environments have more limited access to the Dynamics
SQL database, and the process by which custom code is promoted to
production or how the Microsoft system updates are applied isn’t the
same. You should specifically examine the implications of the produc-
tion system environment on support operations and not rely solely on
the experiences of the test environment.

Identify the specific, new, or changed support requirements impacted

by the following:
▪ Enterprise-level architecture configurations in areas such as
Microsoft Azure, operating systems, any existing Microsoft 365
tenants, Azure storage, and Azure Active Directory
▪ Dynamics 365 Azure environments managed by Microsoft (for
example, the Dynamics 365 production environment)
▪ Identity and access management systems, such as operating
systems, browsers, mobile devices, and firewall configurations
▪ If this is the first significant cloud, software as a service (SaaS)
system in the enterprise, the related system architecture, and
servicing needs
▪ Systems to which the Dynamics 365 application interfaces directly
▪ Middleware systems to which the Dynamics 365 application
may connect

Implementation Guide: Success by Design: Transition to support 614

 ▪ Any systems that will be decommissioned but have residual tasks
that may need to be reabsorbed elsewhere by the support team
▪ Any document management systems such as Microsoft SharePoint
or third-party systems
▪ Any code configuration management systems such as Azure
DevOps or third-party systems
▪ Reporting systems to which the Dynamics 365 application inter-
faces, including Azure Data Lake or data warehouse

The impact of, and on, the surrounding architecture can be difficult
for the customer project team working on the Dynamics 365 business
application to fully identify. In almost all cases, you need the enterprise
architects from IT and the business to be involved in identifying the
changes. Some changes may also impact the roles of individuals formally
part of a support organization and those in peripheral organizations.

Business and IT policies

Enterprise architecture All new systems need to think about how they will operate within the
wider organization’s policies and standards. When Dynamics 365 is
Business and
IT policies introduced into the enterprise, it needs to follow the applicable poli-
cies already in place and may require new operating policies. In either
Dynamics 365-specific
considerations case, you need to review the existing policies and standards to deter-
Business processes mine which policies to add, which to review, and which to apply to the
support model (Figure 21-2).
Business continuity

In many cases, the policies apply not only to the creation of the support
System upgrade cycles
model and its scope of responsibility, but also to how it operates as an
organization and how it addresses the lifecycle of a support request.

Group and company policies

and standards
When evaluating group-level policies or standards, you may have to
review the impact of business policies and procedures on various levels:
▪ Contracting with and managing third-party vendors  The

Implementation Guide: Success by Design: Transition to support 615

 support model needs to include some type of contract with
Fig.
21-2 technology partners and Microsoft
▪ Operating hours  Current and future operating hours for the
Group and company policies and standards
business can directly impact the working hours for support
Dynamics 365 application-level policies
▪ Financial period end and seasonal deadlines  This may impact
the level of service expected at times when the business has critical
Dynamics 365 application-level security
and access management
business activities
Data classification and retention ▪ IT policies on change control on enterprise IT assets  This
Group and company-level security and may impact the access and approval chain for changes required
access management

for the Dynamics 365 application troubleshooting or servicing (for

Regulatory compliance
example, creating new Azure virtual machines, copying databases
between environments, and building IP allowlists)
Support scope
The overall support model on Dynamics 365 business applications
operates across many of the company’s business processes and is
therefore shaped by the relevant business policies.

Dynamics 365 application-level policies

Some policies are managed in the system by the business adminis-
trators, and some may be delegated to the support team. Even if the
administration of a policy is the responsibility of the business leads, the
monitoring and auditing of the business process is often the support
team’s responsibility.

You could set up some of these policies within the Dynamics 365 ap-
plication, such as new vendor approval policies, customer credit limits,
purchase order approval thresholds, and travel and expense policies.
The support team may need to help provide information on compli-
ance or enforce compliance as part of their duties.

Dynamics 365 application-level security

and access management
Administration and review of application-level security is usually an
area of responsibility of the support team. The support team needs to
prepare by gaining an understanding of the Dynamics 365 role-based
application security model and the related access policies.

Implementation Guide: Success by Design: Transition to support 616

 Group and company-level security and
access management
In addition to the security and access policies and procedures that
you need to define for Dynamics 365 applications, you may have to
consider enterprise-level security policies that intersect with the appli-
cation-level security requirements.

In larger organizations, enterprise-level security and access man-

agement may be handled by a separate IT organization. Some of the
support team members will need elevated access not just at the appli-
cation level, but possibly for other systems within the enterprise.

The Dynamics 365 support team needs to work with these other
enterprise IT and business teams to define the rules and procedures
for managing some of the security topics that may impact Dynamics
365 applications:
▪ Azure Active Directory groups and identities
▪ Single sign-on (SSO)
▪ Multifactor authentication
▪ Mobile device authentication and management
▪ Authentication and management for custom applications working
on Microsoft Dataverse (such as Power Platform apps), which
requires an understanding of Dataverse security concepts
▪ Application access for third parties (such as vendors and customers)
▪ Application access for third-party partner support organizations
(such as technology partners and Microsoft)
▪ Service account and administrator account use and management
▪ Password renewals and security certificate rotations
▪ Secure and encrypted communications within the enterprise and
Identify any outside the enterprise (such as those involved with integrations with
enterprise-level internal systems, or with external web services or banking systems)
policies that
The Microsoft Trust Center can help your organization consider overall
intersect with security and managing compliance in the cloud. Chapter 12, “Security,”
application-level provides a more detailed discussion of security for Dynamics 365
requirements. business applications.

617
 Data classification and retention
Consider how the organization’s data classification and retention pol-
icies reflect on and need to be expanded to include the new Dynamics
365 application:
▪ How is the support team expected to enable and enforce
these policies?
▪ What is the impact on the backup, restore, and archive process?
▪ What is the impact on creating and managing database copies?
▪ Do any data classification properties flow between systems, or do
they need to be recreated or audited by the support team?

Regulatory compliance and policies

In addition to any data retention policies, the organization’s businesses
may be subject to national and supranational regulations. Some may
impact business processes that are expected to be assisted by the
support team (such as e-invoicing and digital tax).

The support team may be expected to support, monitor, and audit

other regulations on a regular or ad-hoc basis, such as General Data
Protection Regulation (GDPR) or Health Insurance Portability and
Accountability Act (HIPAA).

All of these different areas of business and IT policies shape the nature,
size, and scope of the support organization. Early examination of these
factors will help the team be effective from the start.

Dynamics 365-specific
considerations
Enterprise architecture

Business and IT policies

In this section, we examine some topics specific to Dynamics 365 busi-
Dynamics 365-specific
considerations ness applications that we should consider when designing the support
model. These can be broadly grouped by operational and maintenance
Business processes
topics and business process topics.
Business continuity
Operational and maintenance topics
System upgrade cycles We can further divide operational and maintenance areas into the

Implementation Guide: Success by Design: Transition to support 618

 management and support of the following:

▪ Dynamics 365 environments

▪ Integrations
▪ System and data maintenance
▪ Performance

Dynamics 365 environments

When Dynamics 365 is in production, various aspects involved in sup-
porting the system environments need to be factored into the support
model preparation:
▪ Maintenance and operational tasks (both scheduled and ad hoc)
such as environment data refresh, database movements between
environments, configuration moves, and code promotion
▪ Support organization roles and resources needed
▪ Skills required to support the environments
▪ Resulting budget impact

Typically, you need to apply these tasks and considerations for the
following environments:
▪ Dynamics 365 application support environments, which are recent
copies of the production system
▪ Test environment for testing the next versions of the
application software
▪ Development environments
▪ Any related data migration, training, or integration environments
▪ Test environments for integrated systems

This is not an exhaustive list of environments, but should help prompt

the right discussions related to supporting the environments. For a
more detailed discussion on the Dynamics 365 environment strategy,
refer to Chapter 9, “Environment strategy.”

Integrations
In addition to defining the environments required for supported oper-
ations, it’s worth delving a little deeper into the support requirements

Implementation Guide: Success by Design: Transition to support 619

 specific to managing integrations. When defining the support require-
ments for all the integrations, consider what areas are in scope for the
support organization:
▪ Data management at either ends of the integration
▪ Security of the integration services
▪ Performance of the integration
▪ Auditing or monitoring of the integration
▪ General troubleshooting analysis and stopping and starting of
integration services

The level of skill and effort required to manage integrations depends

on their complexity, criticality, and robustness. Designs of custom
integrations that are tolerant of variations of data distribution and
variations in volumes, are resilient to errors, and have automated
self-healing mechanisms will require less support.

System maintenance requirements

As part of the definition of the system, maintainability requirements
is an area that doesn’t always get the attention it deserves. This can
manifest itself in unplanned activity for the support organization. So,
during the Initiate phase of the project, the current support organiza-
tion should look to derive the specific maintenance requirements for
managing the new Dynamics 365 implementation.

Because Dynamics 365 applications are cloud and SaaS-based, many

maintenance tasks and responsibilities that are common to on-prem-
The level of skill ises solutions are now managed by Microsoft. In general, the burden

and effort required of infrastructure provision and maintenance for production systems is
reduced, which leaves more time to focus on managing and improving
to manage business process performance.
integrations
depends on Define the system maintenance requirements and what is within the
scope of responsibilities for the support teams. Typically, these are in
their complexity, the following areas:
criticality, and ▪ Servicing the non-production environments, which can include:
robustness. ▫ Requesting and configuring new Dynamics 365 environments

620
 ▫ Requesting and configuring database copies and restores
between environments
▫ Managing any customer-managed, cloud-hosted
environments
▫ Performing specifically requested backups
▪ Managing system operations, which can include:
▫ Assigning users to security roles
▫ Reviewing and running periodic system cleanup jobs
▫ Managing system administration messages and notifications
▫ Batch calendar management
▫ System update calendar

Data maintenance requirements

Data maintenance is a significant part of operating a business system,
and it’s essential that the scope of data management is clearly de-
fined for the support team. Much of the data maintenance will be the
responsibility of data stewards from the operating business, but the
support team may be expected to manage some critical areas.

Typically, in a Dynamics 365 application, a support team may be man-

aging system administrator data, some types of primary data, security
and access data, data related to periodic processes, and so on; for
example, assigning a user to a security role, or adding or revising a new
customer attribute.

Performance management
Performance management for a Dynamics 365 business application is
a mix of tasks and responsibilities for Microsoft and for the customer. In
this section, we consider the implications on the support model.

The support team needs to have some way to proactively monitor and
respond to any questions from the users on performance. Therefore,
the support team needs to be able to do the following:
▪ Understand the impact of the reported performance issue
on the business
▪ Assign the relevant priority
▪ Diagnose the issue across the potential root causes from

Implementation Guide: Success by Design: Transition to support 621

 configuration, data, custom code, independent software vendor
(ISV) code, standard code, and so on to determine the root cause,
or to hand it off to a specialist team (internal or external) with a
meaningful starting point
▪ Reproduce the steps and circumstances to demonstrate the per-
formance issue (and have the necessary resources to do so)
▪ Communicate the problem to other more technical specialists and
work with them to resolve the issue

To better detect processes beginning to show poor performance, it’s

important to have a baseline definition of required performance for
critical processes, also expressed as performance goals. Having ac-
tionable performance goals allows for more focused monitoring and
proactive actions to stop any decline in performance quickly.

Because performance troubleshooting is a specialist skill, the internal

support team may not have all the skills and experience to examine
and correct the performance of custom development. It’s also a skill
that may only be needed sporadically, so you may need to consider
planning for access to external teams that can use performance tools,
interpret the results, and decide on corrective actions.

Business processes
As a business application, supporting the users in the day-to-day pro-
cesses is a key function of a Dynamics 365 support organization. The
Enterprise architecture support organization is expected to provide coverage across all the key
business process, or at a minimum, know where to direct questions and
Business and IT policies issues that were agreed as being outside their support scope.
Dynamics 365-specific
considerations Based on the definition of the key business processes in scope, consider
Business processes the following for each process:
▪ What is the level of expertise needed in the business process?
Business continuity ▪ What are the likely support tasks and queries related
to this process?
System upgrade cycles
▪ What is the type of usage, frequency, and volume?
▪ How critical is this business process to the business outcomes, and

Implementation Guide: Success by Design: Transition to support 622

 what degree of priority do the support issues need?
▪ What is the type and level of skills needed to support this process?
▪ What processes require resources aggregated at a department,
workstream, or value chain level?
▪ What are the interactions with other systems and communication
protocols required?
▪ Does this process require the support personnel to have specialist
security or compliance training or certification?
▪ Do any special requirements need support services outside of the
anticipated support hours (outside of business hours or during
holiday periods)?

An important factor when determining resources and escalation

procedures is the level of ownership and responsibility that the business
process leaders are expected to take from their own organization to
support the day-to-day operations for their processes.

These questions and more will help shape the operating model for the
support team.

Business continuity
Many organizations need to have a business continuity strategy and
exercise regular checks of business continuity in case of a system-down
Enterprise architecture disaster. This may be required by external regulations or by internal
policies. In any case, the support organization is probably expected to
Business and IT policies play a major role.
Dynamics 365-specific
considerations Depending on the size and complexity of the system landscape and
Business processes the types of disaster scenarios being exercised, this may need a signifi-
cant amount of preparation, coordination, and timely communication
Business continuity between multiple parties.

System upgrade cycles

As part of preparing for the establishment of the support organization,
you should consider the definition of the business continuity strategy
and what resources, skills, and tooling are required.

Implementation Guide: Success by Design: Transition to support 623

As a cloud-based SaaS service, Microsoft provides the production
infrastructure, platform, and application-level business continuity
services. For Dynamics 365 Finance and Dynamics 365 Supply Chain
Management applications, the operations as listed in the Finance and
Supply Chain Management service description provide details of the
services and service-level agreements (SLAs) from Microsoft and the
expected responsibilities and actions from the customer. You should
bear this in mind when you build your business continuity planning,
which should reflect your specific systems landscape and scenarios.

Of course, the Dynamics 365 business application may not be an

isolated system—it may be connected to other enterprise systems, and
your business continuity planning will need to consider your entire
enterprise architecture when defining your strategy and the related
support requirements.

Dynamics 365 Finance

and Supply Chain Management
Dynamics 365 Finance and Supply Chain Management applications
have specific high availability (HA) and disaster recovery (DR) features
for cloud deployment. To ensure service availability, all production
environments are protected by using default Azure HA features. HA
functionality provides ways to avoid downtime caused by the failure
of a single node within a datacenter, and DR features protect against
outages broadly impacting an entire datacenter. When planning
support requirements and procedures, consider the implications if the
disaster is severe enough for the primary production site to be forced
to switch over to the secondary site and the DR features are warranted.
In such extreme cases, you would experience downtime for production
access while access to the secondary site is being established, and the
customer team is responsible for several actions. For more information
about business continuity with HA and DR, see the Dynamics 365
Finance and Supply Chain Management service description. If the
production system is running on the secondary site (while the primary
site is being repaired), you may have restrictions (such as the inability
to perform package deployments) and limitations (such as perfor-
mance reduction) that the support team needs to be aware of and help

624
 mitigate to reduce the impact on the business. You may also need to
apply specific setups (such as IP allowlists) to the secondary site.

In summary, you should have a business continuity plan that includes

how the support team should plan, perform, and communicate during
a service outage, including during a severe disaster.

Dynamics 365 Customer Engagement

For Dynamics 365 Customer Engagement, HA provides mechanisms that
can reroute requests in case of failures, planned downtime (patching
and updates), and more. On the database side (relational storage), the
Recovery Time Objective (RTO) for failover is dictated by the timings
from Azure SQL Database. With diverse storage, the scheme extends to
other storage components. For DR, Power Platform brings the ability to
fail over and fail back in seconds. You can do this at the granularity of an
organization (environment). There is no data loss in the case of planned
DR failovers. In the extreme circumstances of an unplanned DR, well-de-
fined processes are applied by the administrators of the Dynamics 365
(online) datacenter to recover from a service interruption; however, the
support team needs to help with any potential data loss.

System update cycles

Microsoft Dynamics 365 application updates are one of the changes that
the support team probably needs to manage. From the perspective of
defining the scope of responsibilities for the support organization, you
Enterprise architecture must understand what is involved and plan for the updates.

Business and IT policies Creating a calendar for the Microsoft updates helps the support team
Dynamics 365-specific plan for the resources and effort associated with the following:
considerations ▪ Microsoft Dynamics 365 updates
Business processes ▪ ISV updates
▪ Updates for custom code
Business continuity ▪ Associated testing, including any regression testing and
release management
System upgrade cycles
Implementation Guide: Success by Design: Transition to support 625
 We saw in the previous section that the solution is unlikely to remain
static for long, and the solution functionality can change because of new
and backlog requirements for many different reasons. For details on
Microsoft system updates, refer to Chapter 20, “Service the solution.”

Support models
Defining the support model is ideally started in the Implement phase

Support so that it can be ready and exercised prior to user acceptance testing

models
(UAT), can help with the transition during the Prepare phase, and be
fully effective and working by the Operate phase.

Support model options

Support organization
Support model options
Thus far, we have discussed topics that help define the scope of re-
sponsibilities and actions for a support team. Now we look more at the
“who” and the “how,” starting with an examination of the spectrum of
support models.

One of the foundations of the support strategy is defining the oper-

ation model for the support organization. Some of the factors may
be driven by enterprise-level strategies, including on outsourcing
or existing commercial agreements. Some organizations reference
independent, formal IT support and service frameworks such as
Information Technology Infrastructure Library (ITIL) or ISO 20000, or
other formal standards and methodologies. These are typically ad-
opted at an enterprise level, rather than by system. Regardless of the
framework, you have some key options to choose from:
▪ Fully outsourced  This support model is more often seen in
smaller projects that don’t have many critical business processes
within their scope or have insufficient internal resources to create
a critical mass for any dedicated internal support members. This
model requires the outsourced support organization to be more
immersed in the business and understand the business well to
navigate the organization when looking for decisions, diagnostic
data, and approvals, as well as have a good understanding of the

Implementation Guide: Success by Design: Transition to support 626

 business priorities. To succeed, it needs the business and partner
to have a common language to communicate business needs and
technical application processes between each other so that expec-
tations remain aligned.
▪ Fully internal  A support model without partner support isn’t
seen very often because it’s a difficult model to sustain on an active
project without a larger internal team that can cope with peaks and
troughs of demand and has the necessary distribution of skills across
business workstreams, technical areas, and servicing, including any
specialist skills. If internal resources are constrained, you run the
risk that the business stops being ambitious and outward looking
in driving improvements and encouraging innovation. The solution
can become stagnant, where assumed stability can come at the
expense of extracting optimum value from the application.
▪ Mixed  This is probably the most common model. It has a bal-
anced blend of internal resources that develops expertise in the
application and retains a deep understanding of the business.
Partner organizations can offer strategic advice based on their
wider experience across multiple implementations and provide
support for any custom code, resources to cope with peaks, and
specialists when required.

Considerations per support level

In this section, we break down the business considerations by each
support level (Figure 21-3).

First level
Many customers have a concept of identifying and nominating super
users within a business unit. Typically, the super users gained their
deeper knowledge of the system and processes in their previous role as
subject matter experts (SMEs) during the project cycle.

Super users can provide the following benefits:

Super users help ▪ Provide immediate assistance to a user to unblock someone in
most simple cases
organizations scale ▪ Filter out the simple questions such as “How do I do this?” or “Is
out application this what I should expect?”
support. ▪ Provide ad-hoc training at a business unit level

627
 ▪ Triage and communicate issues in a way that makes it easier for
the helpdesk or resolving authority to understand the issue, repli-
cate it, and fix it rapidly
▪ Collate best practices, FAQs, and issues, and provide the core team
better data on areas of priority for them to tackle
▪ Provide early warning on issues that have the potential to escalate
and help with better adoption by the business

This model helps organizations scale out application support without

having to create a large central support team or pay external parties
to resolve day-to-day queries on internal business process or simple
application questions.

For larger business units, super users may be further divided by

functional area, such as Finance or Operations. These super users are
typically identified during the project phase and participate in formu-
lating and implementing testing (such as UAT), and may also become
the trainers for cascade training. This allows them to gain the necessary
experience and skills to provide first-line support.

Consider formalizing the super-user role with funding, and consider

formal titles or dedicated time to perform that role. Formal performance
objectives and compensation also help ensure this is not just one more
assigned task. Informal super-user roles don’t always succeed because
Fig.
21-3 day-to-day operational tasks tend to be given a higher priority.

Fully outsourced Fully internal Mixed

Super users per business unit Super users per business unit
1 or function
1 or function
1 Super users per business unit or function

Internal helpdesk to route to Internal helpdesk or dedicated Internal helpdesk or dedicated

2 the partner
2 Dynamics 365 support team
2 Dynamics 365 support team

Dynamics 365 CoE or Dynamics 365 Dynamics 365 CoE or Dynamics 365
3 Partner 3 project core team
3 project core team

Partner via support agreement,

4 Partner 4 Dynamics 365 CoE 4 or partner for custom code via
support agreement

5 Microsoft support or ISV support 5 Microsoft support or ISV support 5 Microsoft support or ISV support

Implementation Guide: Success by Design: Transition to support 628

 Second level
Almost all customers have some internal helpdesk function for IT
systems, and many also have them for business applications. The size
and nature of the helpdesk varies depending on the model adopted.

For a fully outsourced model, the internal helpdesk registers the issue
and assigns it to the resolving authority (internal or external). This fully
outsourced model is tougher to deliver for highly-customized business
systems compared to less customized ones. Determining the correct
resolving authority can be tricky in business systems—even assuming
the super-user roles have eliminated the cause as a business process
or training issue, you may have many different system layers, such as
an internal network, infrastructure software managed internally or by
Microsoft, standard Dynamics 365, custom code, ISV code, or integration
software at either end.

Determining how to prioritize and who should drive the issue to

resolution is a typical deficiency seen in this fully outsourced model,
because very often the issue can bounce around various parties. The
other consideration for business systems is managing what business
data and process steps should be made available to the next resolving
authority. Additionally, if the partner is driving this in a fully outsourced
model, you need a clear definition of the access the partner is given to
users, the business data, and what can be further passed on the next
resolving authority.

In the fully internal model, the internal helpdesk also routes the issue
to a small internal 365 Dynamics team that triages the issue and helps
determine the best resolving authority. This team is also responsible for
ensuring the issue is driven to resolution, regardless of the number of
parties (internal or external) that may need to be involved. The differ-
ence is that the next resolving authority may be the internal Dynamics
365 Center of Excellence (CoE).

For multi-company or multi-project organizations, many create a CoE

that works with group IT to ensure the following:
▪ Common standards on the Azure environment

Implementation Guide: Success by Design: Transition to support 629

 ▪ Operational and maintenance policies on backups, business conti-
nuity, and disaster recovery
▪ IT security and policies (such as password policies, MFA, SSO, and
firewall rules)
▪ Regulatory needs, including access and security, data retention

In the mixed model, which many customers adopt, the internal

helpdesk routes the issue to a small internal Dynamics 365 team that
triages the issue and helps determine the best resolving authority. This
team is also responsible for ensuring the issue is driven to resolution,
regardless of the number of parties (internal or external) that may need
to be involved.

Third level
In a fully outsourced model, the partner is responsible for triage and
resolution. In the mixed model, the most common scenario is for the
internal Dynamics 365 support team to determine if they can resolve
the issue; if not, they work with the Dynamics 365 CoE and the partner
to ensure it’s well understood and correctly prioritized.

In the fully internal and mixed model, the Dynamics 365 CoE includes
functional and technical experts (including developers) who can
resolve issues with custom code. If the issue is seen to be with stan-
dard Dynamics 365 or an ISV, the CoE logs the issue and works with
Microsoft or the ISV to get it resolved.

Fourth level
In the fully outsourced model, the partner manages the process and
the customer is involved as necessary. Most customers tend to have
some parts of the solution written or managed by their internal IT team
(such as integrations), so you still need a definition of how the partner
should work with the customer.

In the fully internal model, the Dynamics 365 CoE takes on the diagno-
sis and resolution if it’s within their control. They only involve external
parties if the issue lies with standard Dynamics 365, platform hosting,
or an ISV.

630
 In the mixed model, the internal Dynamics 365 CoE or core team typically
fixes the simpler issues, but involve a partner for issues that require deep-
er Dynamics 365 knowledge or for complex issues.

When the issue is resolved, the method to get the fix back into produc-
tion also requires a clear definition of the expected standards, testing
regimen, and deployment process. Even in a mostly outsourced model,
the partner drives this to a pre-production environment and the inter-
nal team deploys to production. Most customers don’t give partners
admin access to production environments.

Fifth level
Registering the issue with Microsoft support tends to be the common
escalation point, after the determination is made that the most likely
root cause is the standard Dynamics 365 software or service.

Similarly, most customers have a support and maintenance contract

with their ISV suppliers. In the fully outsourced model and in some
mixed mode models, the partner is contractually responsible to work
with the ISV for resolution.

Support organization
As noted in the previous section, multiple operating models exist, and
within each model a spectrum of sub-options. Therefore, the specific
organization structure is influenced by the variant of the model that
Support model options is chosen. In this section, we explore some common patterns for roles
and structures.
Support organization

Support team responsibilities

Any organization looking to support a business application system of
a reasonably-sized footprint needs to make sure that they have the
necessary skills and resources—not just across the business process
areas and technical topics, but also to manage data quality, maintain
and improve user proficiency in the system, and help the organization
remain compliant with processes and procedures.

Implementation Guide: Success by Design: Transition to support 631

 Each of these areas needs to be defined in more detail to ensure the
planning for the support model is sufficient.

Support for business process areas

Support for business processes is usually defined by the main work-
streams and includes the following responsibilities:
▪ Liaising with process owners on the overall health of the process
▪ Keeping up to date with the current business priorities in this area
▪ Gaining and maintaining a deep understanding of the steps and
relative criticality of the various functions within a business process
▪ Gaining and maintaining a deep understanding of how the
Dynamics 365 application and related systems deliver this process,
so they can provide how-to advice and analyze potential problems
▪ Analyzing, triaging, and recording potential issues
▪ Communicating with the originator of an issue throughout its life-
cycle to maintain user satisfaction, promote business awareness,
and manage expectations
▪ Curating issues through multiple resolving authorities until resolu-
tion within agreed SLAs
▪ Liaising with internal and third-party resolving authorities (such as
any partner organizations, ISVs, or Microsoft)
▪ Monitoring the lifecycle of a defect resolution from registering to
promoting a fix

Support for technical areas

Support for technical areas is usually defined by the following categories:
▪ Service operations (database copies, new environment creation,
managing system downtime windows, maintaining Dynamics 365
system quality update calendar)
▪ System operations via services (integrations, batch processes,
reporting, business intelligence)
▪ System maintenance and monitoring (system health monitoring,
Responsibilities in security role maintenance, security certificate rotations, periodic
technical support system cleanup of log files)
▪ Dynamics 365 application management related to break or fix
roles may overlap, errors reported by users
but with important ▪ Dynamics 365 application management such as updating system
distinctions. administration parameters

632
 ▪ Dynamics 365 application system update management (code
branching, any software development, unit testing new changes)

Responsibilities may also overlap with those for business process support,
with the following distinctions for the expected areas of understanding:
▪ Gaining and maintaining a sufficient understanding of the relative
criticality of the various functions within a business process
▪ Gaining and maintaining a deep understanding of the underlying
technical processes involved in the system functions (especially
integrations, custom code, and critical processes)

Support for data quality maintenance

In many organizations, some of the business data maintenance respon-
sibilities are passed to the support team by the business data stewards
or owners. This type of support includes the following areas:
▪ Core business data and configuration in the system as well as IT
and system administration parameters
▪ Liaising with business data stewards to determine priorities
and actions
▪ Periodic audits and reports to confirm the data quality

Support for user proficiency in the system

In some organizations, the functional experts within the support team
act as a default team to provide user training both formally and on an
ad-hoc basis. This level of support includes the following:
▪ Monitoring user proficiency within the system by reviewing the
number of how-to queries and issues created by not following the
agreed business process in the system
▪ Creating training content and conducting formal end-to-end
training of new users
▪ Creating training content and conducting informal specific func-
Plan for the tional training of users on request or as needed
support team’s ▪ Conducting change management activities to help improve
involvement in user adoption

more than just the New features and developments

break/fix stage of In addition to managing the day-to-day and periodic maintenance
the lifecycle. tasks and addressing queries and issues in the preceding areas, the

633
 support team may also be expected to keep track of new features and
developments (positive and otherwise) that may impact the system.

Often, the planning for the break/fix part of a support team’s orga-
nization is well thought through, but the planning for other areas of
responsibility may be neglected. For example, many support team
members may also need to be involved in helping with the lifecycle of
the next system update. This may include the following duties:
▪ Helping with assessing the impact of proposed changes on the
current system
▪ Providing insights into areas that need reinforcement based on
statistical analysis of the areas with the most issues
▪ Testing the changes
▪ Updating documentation to reflect process and system changes
▪ Training in any new processes
▪ Sending communications of the system update
▪ Managing the next release through the development, test, and
production environments
▪ Refreshing the support environments with the latest
production environment

Dynamics 365 business applications are cloud-based SaaS applications,

so Microsoft handles a lot of the traditional effort (for on-premises
systems) of managing all the layers of the production infrastructure
hardware, operating system, and application software, and much of the
monitoring of the underlying system health.

The duties and responsibilities outlined in this section aren’t meant to

be exhaustive, but rather prompt you to make sure you have sufficient
definition of the coverage, skills, and effort needed to maintain and
support a business application. Later in this chapter, we discuss the
special considerations in the transition period and in the early days of a
new system implementation (sometimes called hypercare).

This section concentrated on the tasks and activities that your support
organization may need to cover. For details on operational consider-
ations, refer to Chapter 20, “Service the solution.”

Implementation Guide: Success by Design: Transition to support 634

 Support organization structures and roles
When considering the support organization, each business should consid-
er the level of effort that is implied by the range of duties expected as part
of support and determine how to distribute the various responsibilities.

The following is a typical set of roles (depending on the size and com-
plexity of the implementation, sometimes multiple roles may be filled
by a single individual):

▪ Business super user  As we discussed earlier, super users serve

as first-line support with some form of part-time role. The level of
formal recognition of the role varies from customer to customer.
▪ Business process expert  Also called SMEs or functional leads,
these experts in the support team usually have a full-time or near-
full-time role. A lead (technical architect) usually has oversight
on the full solution architecture across all the business process
workstreams.
▪ Technical architect  Also called system architect or solution
architect, this role within the support team ensures the enterprise
and Dynamics 365 technical standards are met. They have over-
sight across all the technical tasks and duties mentioned earlier in
this chapter. The technical experts, developer, and release manag-
er roles often report to the technical architect.
▪ Technical expert  This role has expertise in the technical servic-
ing and support duties that we discussed earlier.
▪ Developer  This role is responsible for bug fixes to custom code
and writing code for new custom functionality.
▪ Release manager  This role is responsible for confirming that a
release has been appropriately validated for release to production.
They’re also often responsible for other servicing and support tasks.
▪ Support manager  This manager is accountable for the support
services for Dynamics 365 applications (and often other
related systems).
▪ Business and IT stakeholders  You should consider these roles
significant because their sponsorship, direction, prioritization, and
decisions are key parts of establishing the right support organiza-
tion, defining fit-for-purpose processes for support, and meeting
business expectations.

Implementation Guide: Success by Design: Transition to support 635

Depending on the support model chosen (fully internal, fully external,
or a mix) and regardless of the organization responsible, you will likely
need these roles.

Organizations that are implementing across multiple business units or

companies with an incremental rollout plan should consider setting
up a Dynamics 365 CoE (Figure 21-4). This allows the organization
to bring together specialist knowledge and make it available across
different business, where each individual business or implementation
may not have been able to justify it. This model gives economies of
scale on constrained and specialist resources, and helps disseminate
good practice across implementations. In a CoE model, you may have
permanent internal roles or create sub-roles that have deep specialist
knowledge in business and IT processes. Because the Dynamics 365
CoE can employ its own specialists, it can also help manage other
specialist third-party organizations providing support and service.

A common approach is to outsource some of the Dynamics 365 appli-

cation functional and technical support to a Dynamics 365 partner. This
is especially true if the application has a lot of customizations that were
developed by that partner. In these circumstances, the customer may
still elect to keep a small internal support team of a technical archi-
tect, developer, and release manager to help manage the outsourced
services and maintain final control over the quality of the changes to
the production system.

The size and nature of the support organization should be determined

Fig.
21-4
• Provides executive guidance • Clears roadblocks
• Sets priorities • Secondary escalation point
Executive
• Authorizes budgets sponsors

• Provides goals and strategy

• Approves expenditures
• Provides day-to-day leadership Program lead
• Primary escalation point

• Day-to-day project management

• Responsible for deliverables and Planning/ Business Technical/ Architecture Change Support
own individual outcomes Release Development management
• Brings operational expertise

636
 by analyzing the scope of the requirements for the support organiza-
tion (as discussed earlier in this chapter) as well as the size, complexity,
functional spread, and geographical (time zone, language, local data
regulations) distribution of the implementation.

Standard Dynamics 365 platform and application support and ser-

vicing for production environments is also available (Figure 21-5).
Microsoft offers different levels of support for your Dynamics 365
products. Refer to Chapter 20, “Service the solution,” for more details.

Budgets, contracts, and agreements

When a new system is introduced into an existing enterprise architec-
ture, some changes can impact related systems, such as the changes
required to budgets and support agreements. A common issue that
can occur is that the budgets and agreements aren’t modified to reflect
the new system environment and processes. This can cause delays
when addressing issues that may span departments, or if the resolving
authorities or third parties have their own budgets based on previous
processes and agreements.
Fig.
21-5
After you have a reasonable draft of the new system solution blueprint,

Support provided by Microsoft

Infrastructure Application platform
• Storage and database capacity management • Diagnostics, patches, and updates
• High availability and disaster recovery • Network management
• Platform security • Application monitoring and
• Infrastructure capacity, scale up and down first-line support

• Infrastructure management and deployment

Implementation • Datacenter networking, power and cooling
roles and
responsibilities Customer leading, supported by
implementation partner
User/Data Application
• Security, identity configuration, and • Define and test business processes
management
• Develop and test customizations
• Monitor sandbox environments

Implementation Guide: Success by Design: Transition to support 637

 to-be enterprise architecture, and support organization, you should start
defining the budgets related to the new tasks and areas of responsibility.

Consider the full set of tasks and activities in scope for the support
organization (as discussed earlier) and map these to the various roles
and resolving authorities over the full lifecycle of a support job or request
so that no gaps appear in the flow. You can use this to make sure that the
specific responsibilities of all the roles and resolving authorities can be
mapped to agreements and service-level expectations.

For internal parties, this may mean defining budget and resource splits
and less formal service delivery agreements. For third parties, that may
mean formal contracts with formal SLAs. Mismatched expectations
between the customer and a third-party support provider in the mid-
dle of dealing with high-priority issues are not uncommon. A contract
created based on a deeper understanding of the expected tasks and
service expectations is much more likely to avoid misaligned expecta-
tions and provide a faster path through the process.

Support operations
The preparation and design work allows you to provide the level of
support expected by the business. There are many aspects to consider

Support when planning for the practicalities of operating the support services.
Preparation for support is necessary, but not sufficient. The support

operations team also needs to learn and practice to ensure they are ready for the
extra pressures of the period immediately following a go live. The next
section explores this in more detail.
Transition
Requirements management
Transition
Change management
Support organizations for business applications are rarely fully formed
Hypercare on the first day of go live. You normally have a transition period from
project mode to support mode. However, the transition period often
starts later than it ideally should, and the quality of the transition is
often not sufficient to provide the level of service that is expected. This
results in poor user experience, frustration from the support team, and

Implementation Guide: Success by Design: Transition to support 638

 a drag on the project team that is called into support requests long
after the transition is supposed to be complete.

What is a good time to start the transition?

Learning a new business application system, even one that is as user-
friendly as Dynamics 365, can take time, and actively using the system
is one of the best methods. For the business process-related support
roles, if the SMEs are expected to move into the support role at go live,
training happens naturally as part of the project, assuming the project
approach includes a significant amount of hands-on system involvement
by the SMEs. It’s important that the business process experts don’t just
become consultants to drive business requirements and have very little
exposure to how the system design meets the requirements.

If, however, the existing SMEs supporting the legacy system are
expected to support the new system, consider formalizing their in-
volvement in the project during the Implement phase, from the early
build steps and especially in shadowing the business functional leaders
at critical points in design and software playbacks.
If the SMEs aren’t being exposed regularly
and extensively to the system during its Furthermore, consider involving the support business process roles
build, a gap will likely appear in their
understanding and ability to support, in the formal testing phases as authors of some of the test cases or
which means increased reliance on the
in conducting the testing. Other opportunities to get the team to
partner functional consultants for support
(and testing). directly experience using the new system can come from creating task
recordings and task guides in Dynamics 365 Finance and Supply Chain
Management applications, or other forms of training documents.

You can apply a similar approach to the technical roles—they can be

involved in shadowing the project team at critical points and be given
project tasks to complete (under the supervision of the project team).

Consider setting intermediate targets for transition readiness through-

out the project so it doesn’t become a last-minute process. It may
help to exercise the whole support organization prior to go live and to
review the effectiveness of the readiness for supporting the production
system. Consider using the formal support processes during the later
stages of formal testing (such as UAT) to review readiness. This way, a
representative sample of issues raised during this test period can be

Implementation Guide: Success by Design: Transition to support 639

 recorded, triaged, and resolved through the support organization tools
and procedures. The project team will be expected to participate as
resolving authorities to help train the support team and ensure that
issues are addressed in a timely fashion. The learning from this type of
exercise can help make the transition more effective than last-minute
handover training sessions.

Most organizations run lean support teams, which doesn’t offer many
opportunities for the team to participate in the transition activities we
described without explicitly planning these activities so that the sup-
port team gets time away from their day-to-day operational duties to
participate fully during the project implementation.

Requirements
management
When the application is in production, it doesn’t mean that the sys-
tem solution can now be considered static for the next few years. The
solution will need to deal with new requirements for several reasons:
▪ Some projects implement a minimum viable product (MVP) on
Transition their first go live with the goal to incrementally add the lower-pri-
ority requirements over time
Requirements
▪ Some projects have planned multiple rollouts that may impact the
management
already live areas
Change management
▪ Some changes to Dynamics 365 are driven by changes in connect-
Hypercare ed systems in the enterprise
▪ Businesses need to react to the changing world around them
▪ In a modern cloud SaaS world, incremental functional and tech-
nical updates from Microsoft and ISVs help keep the customer’s
solution secure and updated

Some organizations will continue to have a project team that is respon-

sible for reviewing, prioritizing, and delivering the backlog, but many
won’t have a standing project team, or the project team will be purely
focused on the next rollout.

In such cases, the support organization needs to have the resources

Implementation Guide: Success by Design: Transition to support 640

 and means to keep the solution up to date for the users currently using
the live system.

In any case, the support teams are often at the front line of the feed-
back and requests for new requirements. At the very least, they need
a way to capture and channel the new and emerging requirements to
the right parties.

Change management
We saw in the previous section that the solution is unlikely to remain
static for long, and the solution functionality can change because of
new and backlog requirements.

Transition However, support teams may need to accommodate other sources of

change in the production system, such as the following:
Requirements management ▪ Core data changes (new customers, suppliers, items as part of
normal business operations)
Change management ▪ System parameter changes (such as in response to the growing
Hypercare volume of data)
▪ Configuration changes (such as in response to regulatory changes)
▪ Approval process changes in the system (such as in response to
internal policy changes)
▪ Bug fixes

Some of these changes can be made within the agreed change control
process enforced and facilitated by the system itself. In other cases,
proposed changes need to be reviewed, prioritized, and approved by
business or IT stakeholders and any other relevant parties.

This process should be considered as part of defining the scope of the

support teams’ responsibilities. It should also prompt the organization
to create the appropriate change management boards for changes to
the system in operational mode.

Hypercare
Most projects define a short period of time after go live in which

Implementation Guide: Success by Design: Transition to support 641

 additional resources and extraordinary attention are given to support-
ing the business. This hypercare period is necessary to ensure that the
Transition business can operate within acceptable thresholds while addressing any
initial growing pains with a new system and new business processes.
Requirements management
The role of the support team during this period is critical, and it’s
Change management
important that the expectations are clearly set out and agreed upon
with all parties:
Hypercare
▪ Define a clear exit strategy for the end of the hypercare period
based on meeting explicit criteria. For example, criteria might
include no open P1 issues without an acceptable action plan, key
business processes are operating at an acceptable efficiency rate,
over 90 percent SLA targets being met for P1 and P2 support re-
quests, or the support team can resolve over 95 percent of support
requests without using resources reserved for hypercare only.
▪ Decide on an end date for the hypercare period and have an
action plan in place to meet the exit criteria.
▪ Define expectations and an explicit plan for the project team (and
partner resources) during the hypercare period so you can get reli-
able help from the team. Otherwise, you may find that the project
team is committed to working on the next phase of the project
and will resist being drawn into support issues.
▪ Make sure the support team gets the necessary documentation
and top-up training from the project team and implementation
partner during the hypercare period.

Conclusion
During operational use of the Dynamics 365 system, the support
operations are expected to function efficiently and evolve alongside
the expanding use of the system. For that to happen smoothly, the
preparation and definition of the support operating model are
essential precursors.

Implementation Guide: Success by Design: Transition to support 642

 Scope of support operations
The sections in this chapter are designed to help define the potential
scope of support operations; you need to convert each of the areas
covered from the “what” to the “what in detail,” “who,” “when,” and “how.”

You should convert the enterprise architecture alongside the business

and IT policies and business continuity requirements into practical
reference architecture and detailed procedures with reliably repeatable
steps that the support team can implement day to day.

You should convert the Dynamics 365-specific areas and business

process requirements into the support organization charter or terms
of reference with a further detailed definition of the tasks, the roles
responsible, and the standards expected to keep the system operating
safely and efficiently.

Use the system update cycles section, including managing new re-
quirements and changes, to define the means by which the Dynamics
solution can stay updated and evolve with the additional and im-
proved features. This should deliver a detailed definition of the tasks,
the roles responsible, and the standards expected to continuously
improve the business value from the Dynamics 365 system. This is im-
portant to ensure that the support organization can keep the current
operational system functioning optimally while also working in parallel
on the next update.

Support operating model

Well-defined The sections discussing the support model and support organization
should help take the definition of the support operations scope and
scope and support use the chosen support model to create the shape and size of the
operations will support organization. This resulting design should be a good fit for the
yield an efficient requirements of support operations.

support model Finally, transition guidelines can help you make sure that the transition
that can evolve as from project to support happens incrementally and through practical
needed. experience of the system during implementation. We also encourage

643
 you to validate the readiness of your support processes and organization
prior to go live.

Operating considerations
The nature and details of the support tasks are influenced by the spe-
cifics of the business application because they are so embedded in the
core business. The details of the support operations are further affect-
ed by the details of the circumstances of the business being supported.
We discuss some of the circumstances next.

Routine maintenance and service operations

Consider defining a support and servicing shared calendar that lists the
regular, periodic tasks. This can help make sure those tasks are reliably
implemented and have the right visibility across the team.

Service hours, language, and geography

In addition to defining the normal hours of operation, consider how to
define out-of-hours service, especially for urgent issues.

Some larger organizations, especially multi-national ones, may need to

provide support in multiple languages and across multiple time zones
based on the geography of the users.

Furthermore, consider business-critical periods when support may

need to provide out-of-hours support; for example, during seasonal
peaks or period-end or one-off events such as company acquisitions.

All of these topics have a bearing on the operational patterns that the
team needs to support.

SLAs, priorities, and escalations

Support organizations need to provide some form of SLA for the dif-
ferent types of support request—if these terms are based on genuine
business criticality, they will help drive the right prioritization.

Establishing a formal escalation procedure with defined criteria will also

Implementation Guide: Success by Design: Transition to support 644

 help prevent frustration on slow-moving cases and help mitigate the
impact of those that escalate issues beyond their true business impact.

Tools and access

Most support organizations have some type of internal helpdesk soft-
ware that is used to manage support requests across all IT services. Most
continue using their existing helpdesk for Dynamics 365 applications
but need to consider new internal and external resolving authorities.
Third parties such as partner support organizations and ISVs have their
own helpdesk systems. When the issue is diagnosed as likely to be in
standard Dynamics 365, the recommended method of registering the
issue with Microsoft is to use the in-application process: Dynamics 365
Finance and Supply Chain Management Support and Dynamics 365
CE Power Platform Support. Tracing a support request across multiple
technologies and especially across third-party organizations requires
some deliberate planning and possible configuration so that the internal
helpdesk has a reference from start to finish of any ticket.

Policies on access for partner resources to the customer systems and

customer data need to be defined as part of the support organization
so that there are no delays in the middle of dealing with high-priority
issues while policy is reviewed and special approvals are obtained.

Avoid delays by
For example, if the policy only allows for partners to view anonymized
defining access customer data, having an automated way to copy the production
policies for partner system that includes anonymization will help reduce delays when
resources. troubleshooting or testing.

645
 Checklist

Support scope Support operational

Establish a practical reference architecture and detailed considerations
procedures with repeatable steps for the support team
Establish a transition strategy that drives the support
that can be operated day to day.
team to participate in the emerging solution and im-
Identify the impact of the business and IT policies on the prove their support readiness over the project lifecycle,
scope and the impact on the policies after the Dynamics rather than a limited handover close to go live.
365 application is in operation.
Confirm the readiness of the support team’s operating
Identify the Dynamics 365 application-specific functions model, resources, knowledge, and procedures during UAT.
and procedures as well as business processes that need
Have a defined strategy to manage the critical period
to be supported.
just after going live (hyper care) with explicit entry and
exit criteria.
Support model and
Define how to keep the solution up to date and evolving
the related support with additional and improved features and capabilities.
organization Determine a support and servicing calendar for regular
Define the type of support operating model to adopt and periodic maintenance activities.
and the nature of the role of the internal and external
Identify the normal hours of operation across different
support teams.
locations, as well as out-of-hours service windows,
Create a support organization charter with a detailed multi-language and time zone requirements, seasonal
definition of the tasks, roles, responsibilities, and stan- peaks, period-end, or one-off events.
dards expected to keep the system operating.
Establish formal escalation procedures with defined
Consider the implications of supporting a cloud and criteria, SLAs, and different tiers of support, including
SaaS application, and potentially including a Dynamics internal and external support.
365 CoE in the support model.
Determine the tooling that the support
organization requires.

Identify the policies to access data and environments,

either by internal support teams or by external
support teams.

Implementation Guide: Success by Design: Transition to support 646

 Case study
A story in two acts
Over the last few years, IT project implementation methodologies have
been changing to try to keep pace with the speed of innovation of
software in the cloud, especially Dynamics 365 applications. IT organi-
zations within businesses have similarly been evolving as the cloud is
changing the role and areas of responsibility for IT.

The following case study is an illustration of the dangers of assuming

that an existing, well-functioning support model will continue to
perform well in a different architecture and with changed business
processes in new applications, without deliberate planning and action.
This is doubly true in a cloud world.

Act I: The journey of learning

This is a story of a medium-sized organization (fewer than 1,000 em-
ployees) with multiple offices across three European countries and
their corporate headquarters in the UK. Their core business includes
professional services, field services, and bulk distribution from their
own warehouses.

The group was undergoing rapid business growth. To support and

accelerate this growth, they chose to replace several specialist local
applications and multiple on-premises ERP and CRM systems with
Dynamics 365 Finance, Dynamics 365 Supply Chain Management,

Implementation Guide: Success by Design: Transition to support 647

 and Dynamics 365 Customer Engagement across the enterprise. They
planned their first rollout to locations in two countries that had sig-
nificant interdependency (accounting for approximately 15 percent of
their business) and to the other businesses incrementally, as part of the
next phases of the project. Subsequent phases would also include addi-
tional functionality related to Human Resources, Payroll, the Customer
Service Center, and shared services on contract management.

The existing support teams were fragmented across various systems,

and sometimes also by business unit. These multiple support teams
were planned to be organized under a unitary structure in IT before
the transition to Dynamics 365. All support activities and teams were
planned to be managed by a director of support services, and who
would report to the CIO.

The project was planned to go live approximately 10 months after

project initiation. At the exit criteria review of the Initiate phase, they
identified the following:
▪ The to-be end-to-end business process flow diagrams and
description were adequate in most areas.
▪ The solution blueprint document was judged to be sufficiently
well-defined to confirm the scope of the overall design.
▪ Budgets and project resources to complete the project were
revised and approved in light of the latest solution blueprint and
project plan. The project scope and go-live date had been
reviewed and reconfirmed by the steering group.
▪ In readiness for the implementation of enterprise-wide Dynamics
365 applications, the ambition was for the support team to be
reorganized by main business areas: Finance, Sales and Marketing,
Operations, Field Service, Warehouse and Shipping, IT business
application, and IT servicing, rather than by legacy system.
However, in reality, the teams continued to be tied to the legacy
systems they had existing expertise in supporting.

There was an informal understanding that the support team would be

trained during the project. However, during the Implement phase, the
project team stated that they were behind schedule and too focused
on project delivery to help educate the support team. The mitigation

Implementation Guide: Success by Design: Transition to support 648

 plan was for the support team to be part of an extended, one-week
training and hand-off period after UAT was complete.

As the implementation progressed, the support organization contin-

ued to focus on the day-to-day needs of the currently operating legacy
systems. There was a general feeling in the project and in the support
organization leadership that the risk was low because the existing
support team was experienced and some of the project’s SMEs would
join the support team, thereby seeding knowledge into the team. The
underlying assumption was that the support team would be able to
pick up the new duties without too much fuss, given that they had
been working in a support role for a few years.

During the Prepare phase, when some members of the support team
were asked to assist with the testing, the team didn’t feel ready to par-
ticipate until they had been through the training, which was scheduled
after the testing. The project SMEs didn’t feel they could conduct the
internal training because they also had very little hands-on experience
in the working system.

When UAT was complete, the project team had other unplanned
activities that were considered to be higher priority, and so the training
and handover to the support team was reduced to two days.

The go live went ahead on schedule, but the support team struggled
to adequately support the business. The initial assumptions that the
support team would just pick up the new system support were found
to be mistaken, because the underlying processes had also undergone
significant changes. The support team hadn’t been informed about
all the business process decisions and weren’t familiar with the new
control and approval requirements configured in the Dynamics 365
application. The shortened training and handover were conducted
without reference to the significantly changed business processes or to
the new enterprise architecture, so didn’t provide the necessary context
to support a production system.

The support operating model had been revised, in theory, from the
distributed, system-level team structure to be more business process

Implementation Guide: Success by Design: Transition to support 649

 and enterprise IT-based, but it wasn’t exercised during the Implement
or Prepare phase; the gaps and deficiencies were only exposed once
the teams had to operate in production. The new operating model had
also not adequately considered the implications of changed require-
ments of the new business processes, architecture, and policies.

In the previous support operating model, all queries and issues came
directly (and with little structure) to an individual support team member,
and the full lifecycle of the issue was performed with little external visi-
bility. In the enterprise-level support, even with the first limited rollout,
the support team was receiving large numbers of administrative queries,
which reduced time to address the more business-critical tickets.

The team was also forced to revise IT servicing procedures to reflect

the new technology implications in a cloud-based SaaS model. They
had to revise policies related to data retention and backups because
processes based on the legacy, bespoke on-premises processes weren’t
reviewed and updated for the new cloud-based SaaS architecture.

The project team had planned to start working full-time on the next
phase of the project after only two weeks of hypercare. The budget
hours for the implementation partner hypercare support were planned
to last four to six weeks, but were exhausted within two weeks or so
because the support team needed far more of their time. Many SLAs
were missed in the early weeks, and the business was forced to formally
extend the hypercare period for both the project team and the imple-
mentation partner because trying to get the timely attention of the
implementation teams was difficult given that they had other planned
priorities for the next project phase.

After the P1 and P2 support tickets were resolved and the volume of
new tickets was down to a manageable level, the post-go live ret-
rospective review made the following key recommendations to be
enacted as soon as possible, to improve the support outcomes during
the next rollout phase:
▪ The scope of the activities, tasks, and actions expected from the
support team should be explicitly defined, in the context of the
new system, processes, people, and architecture.

Implementation Guide: Success by Design: Transition to support 650

 ▪ The support operating model needs to have a more formal defi-
nition of the different level of support and a clearer agreement
on the expectations from third parties. For example, using tools
such as Microsoft Teams channels where super users can help
address the higher-volume, simpler questions quickly. The super
users would also help to better formulate issues and questions that
needed to be addressed by the next support level.
▪ The support team needs to be formally allocated time to be in-
volved in the implementation project to ensure their learning is
sufficiently deep and is embedded in the new process and archi-
tectural context to be usable in production. The transition should
not rely solely on a last-minute, half-hearted handover.
▪ The transition should be formally planned within the project plan
with specific tasks (such as process design reviews and test case
preparation) so the team has sustained and multiple opportunities
to have hands-on experience of the developing system.

Act II: Fewer problems = shorter story

These recommendations were approved by the stakeholders and
enacted over the following months. Even though the next rollout (six
months later) had a much larger user base, it was a much smoother
operation with a shorter hypercare period, higher user satisfaction, and
markedly less cost than the first go live.

This story exemplifies the adage “Poor service is always more expensive
than good service.”

Implementation Guide: Success by Design: Transition to support 651

 Conclusion
Implementation Guide: Success by Design: Conclusion 652
 Conclusion
As we have seen in numerous examples
throughout this book, decisions taken during
the project implementation lifecycle can have
dramatic impact on overall project success.
For example, the case study in Chapter 14, “Testing strategy,” describes
the negative business impact resulting from deprioritizing perfor-
mance testing. What I love about this example is that it describes a
situation we see all too often. Perhaps it’s logical to think that perfor-
mance testing is optional because the first deployment is a limited go
live, but as the case study shows, it’s important to prioritize perfor-
mance testing before Phase 2 when more users are onboarded onto
the system.

Every decision taken during a project implementation ultimately comes

down to risk; the decision is either unknowingly adding risk or helping
intentionally manage risks during delivery. What we have attempted
to do with the Success by Design framework is provide you with a very
clear set of guidelines and practices that will help you manage risk and
ensure that you have approaches to mitigate all risks that show up as
part of a solution implementation. These practices and the prescriptive
approaches are based on years of experience gained by our team of over
200 Dynamics 365 solution architects, having overseen thousands of
complex implementations over the last many years.

Implementation Guide: Success by Design: Conclusion 653

At Microsoft, our passion for building the most innovative products
is rivaled only by our commitment to helping our customers and
partners bring these products to life so we can jointly realize the op-
portunities presented by digital transformation. The Success by Design
approach is a framework, and like all frameworks, it’s far from perfect.
We invite you to take this guidance and apply it to your projects—by
making it your own and adding to it what we may have missed. In
the spirit of jointly building the body of knowledge that drives our
Implementation Guide, we invite you to engage with this content and
let us know how it compares to your experiences in implementing
Dynamics 365 solutions.

Your feedback is critical to helping us evolve this guidance,

so please don’t hesitate to reach out. You can email us directly at
[email protected].

We’re immensely proud of the body of knowledge that our FastTrack

for Dynamics 365 team has formalized through Success by Design, and
even more excited to share this knowledge in the first iteration of this
book. What you’ve just read is only the beginning—we will ensure that
this guidance continues to reflect product truth and includes the latest
best practices gleaned from the thousands more implementations that
we will run in the coming years.

While we may never fully eliminate the nail-biting anticipation that

accompanies most go lives, I hope that our unique take on the fun-
damentals of successful implementations will bring more certainty,
predictability, and stability to your digital transformation journey.
Thank you for partnering with us on your journey. I look forward to
hearing from you as you put this guidance to practice.

Faisal Mohamood
General Manager, FastTrack for Dynamics 365

654
 Acknowledgments
This book celebrates Microsoft’s desire to share the collective thinking
and experience of our FastTrack for Dynamics 365 team, which is
currently made up of more than 140 solution architects who serve
Microsoft Dynamics 365 customers and partners around the world.

For being the kind of leader who purposely carves out time in your
busy schedule to think about, write about, and share your vision for
Dynamics 365 and the Microsoft Power Platform, thank you, James
Phillips. We hope that the way you’ve inspired us comes through on
every page of this book.

For your tireless efforts in making Dynamics 365 a better product and
for making us better solution architects, thank you, Muhammad Alam.
We’re also grateful for the time you gave us to round out the content
in Chapter 1.

For encouraging us to embrace James Phillips’s vision of making

Dynamics 365 Success by Design for all, and for supporting us
in bringing this book to life, we thank Faisal Mohamood, Swamy
Narayana, and Jason Kim. Successful leaders prefer details over

No one who executive summaries—the reason we know this is because of you.

Thank you.
achieves success
does so without Thank you to the book’s authors, including Ajit Kotwal, Alejandra
acknowledging Cabrales Orozco, Alejandro Gutierrez, Alok Singh, Aman Nain, Amreek

the help of others. Singh, Dave Burman, Jeremy Freid, Jesper Livbjerg, Matthew Bogan,
Pedro Sacramento, Rich Black, Richa Jain, Satish Panwar, Saurabh
The wise and Kuchhal, Seth Kircher, Tak Sato, Veselina Eneva, and Vidyasagar
the confident Chitchula. At the start of this journey, you were given the task to write
acknowledge this the truth into these chapters, to trust what you know, and to make sure
your message is valuable to the reader. That’s exactly what you all did,
help with gratitude.
in addition to your day jobs. Now please go share this book you wrote
-Alfred North Whitehead with those closest to you.

655
Also, a special thanks to Alok Singh and Pedro Sacramento for their
insistence that the checklists at the end of most chapters be included
in the first version of this book. You went above and beyond in your
efforts, and we’re certain that readers will find the checklists helpful
and provocative.

To the book’s many reviewers, including Ajay Kumar Singh, Dan Ogren,
Faisal Mohamood, Gokul Ramesh, Gregg Barker, Jason Kim, Jayme
Pechan, Matt Sheard, Paul Langowski, Praveen Kumar Srinivasan
Rajendiran, Swamy Narayana, Timo Gossen, and Umran Hasan—you
met us at the edge of exhaustion and pressed us to go on and to make
a better book. Thank you.

To Denise Francis for your effective project management skills—and

for keeping after us—throughout this effort. Thank you.

To Rishi Manocha for your critical eye and tireless assistance with all
things related to the marketing of this book. Thank you.

To Carol Dillingham, Anna Tatistcheff, and the entire Steyer Content

team for your highly competent and professional effort in advancing
the book’s content and design. Thank you.

Finally, a special thanks and debt of appreciation is due to all our

current and former Dynamics 365 customers and partners from whom
we learn and by whom we’re inspired with each encounter. We marvel
at what you do.

—Alejandro Gutierrez and Seth Kircher

656
 Appendix
Implementation Guide: Success by Design: Appendix 657
Phase
Initiate Regardless of which implementation methodology
you use, here are the typical implementation
deliverables, activities, and recommended roles that
need to be performed by the project team during
the Initiate phase of the project.

This table is for reference only; a missing deliverable,

activity/task, or role might not be an indication that
something is missing in your project.

Note that the names of deliverables, details

regarding activities and tasks, and roles and job titles
might change depending on your project.

When a deliverable or activity is application-specific,

it is specified as either:
▪ F&O for Finance and Operations apps; or
▪ CE for Customer Engagements apps

Implementation Guide: Success by Design: Appendix 658

Topic Deliverables Activities/Tasks People

Program strategy Program strategy/ Define: Customer project

charter ▪ Business vision sponsor, program
▪ Business goals/business director
drivers Project managers
▪ Business objectives
Customer business
▪ Identify key success
stakeholders
metrics or KPIs
▪ Project budget Customer IT director
▪ High-level project and Customer IT architect
solution scope Solution architect
Project plan Define the project plan, Functional consultant/
scoping all activities across architect
project phases. Change manager
Risk and issue Define the risk and issue Project management
register register (with risks, severity, office (PMO) team
mitigation, consequence, Resource manager
issues, escalation process).

Project Define the governance

governance controls through
establishment of a PMO
and steering committee to
manage the triple constraints
(schedule, budget, resources),
project risks, and change
management/change control
for the project.

Solution Define architecture review

architecture board (ARB) or design
management authority (DA), which governs
key architectural decisions
made during implementations.

Implementation Guide: Success by Design: Appendix: Initiate phase 659

Topic Deliverables Activities/Tasks People

Program strategy Project Install applications and tools Customer project

continued technology that will be used to deliver the sponsor, program
and tools project. These could be: director
▪ Lifecycle Services (for F&O Project managers
projects)
Customer business
▪ Azure DevOps
stakeholders
▪ SharePoint
▪ Microsoft Teams Customer IT director
▪ Conduct implementation Customer IT architect
team training on Solution architect
methodology and tools.
Functional consultant/
Familiarization Conduct initial familiarization architect
training training on the Dynamics Change manager
product(s) that will be
Project management
implemented, as well as
office (PMO) team
independent software vendors
(ISVs), add-ons, and every Resource manager
other software element.

Train the trainer Plan train-the-trainer strategy, Functional consultant

which includes content Technical writers
preparation for training;
Trainer SMEs
identifying the customer
trainers; number of training Customer trainers
sessions; plan of execution of Customer business
the training. consultant

End-user training Plan end-user training, which Customer IT team

includes number of users,
number of training sessions,
and the plan for executing the
training.

In-context help Plan in-context help

documentation document/manuals that
Dynamics 365 application will
provide for users.

Implementation Guide: Success by Design: Appendix: Initiate phase 660

Topic Deliverables Activities/Tasks People

Test strategy Test plan Plan the testing strategy. Test lead
Test consultant
Define test activities such as
unit testing, functional, user Functional consultant
experience (UX), integration, Solution architect
system integration testing Customer IT team
(SIT), performance, and time
Customer business
and motion study.
consultants

Provide high-level definitions UX consultant

of test cases, scenarios, Performance test lead
alternate scenarios, and Performance test
expected results. consultant

Test execution/ Establish KPIs for measuring

metrics test execution outcomes.

Performance/ Plan performance benchmark

scalability goals testing to assess the scalability,
concurrency loads, network
configurations (latencies,
bandwidths), and browser mix.

Usability Elicit accessibility,

compatibility, and ease-of-use
requirements.

Testing Confirm testing automation

automation tool tool that will be used for
testing, including regression
testing.

Business process Process Define process architecture, a Solution architect

strategy architecture commonly understood, shared Functional consultant
view of all business processes
Customer business
that an organization performs
consultant(s) (SMEs)
to deliver a product or service
to their customers. Customer IT architect

Implementation Guide: Success by Design: Appendix: Initiate phase 661

Topic Deliverables Activities/Tasks People

Business process Requirements Define requirements Solution architect

strategy traceability traceability matrix (RTM), a Functional consultant
continued management key deliverable and success
Customer business
measure. Every requirement
consultant(s) (SMEs)
needs to be linked to at
least one process. It can also Customer IT architect
depict the to-be processes as
requirements in the form of
user stories/use cases. The RTM
also includes nonfunctional
and negative requirements.

Fit-gap Assess which requirement

assessment and respective process can be
addressed by the system or
requires some customization.

Gaps evaluation Evaluate and estimate gaps with

and estimation the minimum information of:
▪ Related process
▪ Related integration
▪ Criticality
▪ Complexity
▪ Dependencies

Organizational Define organizational

structure structure based on feedback
from executives and senior
managers from functional
areas. Model how the business
operates. It might consist of:
▪ Geographical structure
▪ Organizational groupings
▪ Line of business
▪ Reporting lines
▪ Segregation of duties

Implementation Guide: Success by Design: Appendix: Initiate phase 662

Topic Deliverables Activities/Tasks People

Business process Country/region Validate country/region

strategy localization localization availability for
continued
availability each country/region in scope.

Application Solution Defines high-level view of Solution architect

strategy architecture the solution architecture, Technical lead(s)
blueprint which depicts how Dynamics
Customer IT architect
365 apps interact with other
systems and users. UX consultant
Customer business
ISV/add-ons Define need for considering
consultant(s)
any ISV solutions to meet the
business requirements.
Include all considerations in
the relevant strategies.

User experience Prototype high-level

(UX) wireframes from UX point of
view to derive the new user
experience.

Data strategy Data migration Define data that needs to be Customer IT architect
plan and migrated to Dynamics 365 Solution architect
execution apps, along with data volumes,
Functional consultant
strategy approach, and tools to be used.
Data migration
Define high-level plan of Lead/architect
data migration activities,
such as data extraction,
transformation, and load;
migration runs; initial and
incremental migrations;
data validations and data
verification; and production
migration.

Implementation Guide: Success by Design: Appendix: Initiate phase 663

Topic Deliverables Activities/Tasks People

Data strategy Data management Outline strategy, including: Customer IT architect

continued strategy ▪ Scope Solution architect
▪ Key requirements
Functional consultant
Approach, including:
Data migration
▪ Governance
▪ Compliance Lead/architect
▪ Migration
▪ Master data
▪ Configuration
▪ Sharing and promotion
Technology and tools

Configuration Define plan and approach for

management how to manage:
(F&O) ▪ Configurations
▪ Shared data
▪ “Golden” configuration,
including promotion
▪ “Golden” configuration
environment

Integration System Elicit the integration needs, Customer IT architect

strategy integration such as: Solution architect
requirements ▪ Identifying lines of
Functional consultant
business (LOBs) to be
integrated. Integration lead
▪ Inbound and outbound
integration requirements
from Dynamics 365 apps
▪ Integration types such as
Sync, Async, Batch
▪ ACID (atomicity,
consistency, integrity, and
durability) requirements
towards integration calls

Implementation Guide: Success by Design: Appendix: Initiate phase 664

Topic Deliverables Activities/Tasks People

Integration Integration Outline the strategy, including: Customer IT architect

strategy strategy ▪ Scope Solution architect
continued ▪ Integration approach
Functional consultant
▪ Landscape
▪ Technology and tools Integration lead

Integration Define the integration

interfaces mapping between Dynamics
identification 365 and other systems. Plan
the dependencies.

Dual-write Consider using the dual-write

infrastructure infrastructure if customer has
both Dynamics 365 CE apps
and F&O apps.

Intelligence BI and reporting Outline the strategy, including: Customer business

strategy strategy ▪ Scope consultant
▪ Reporting structure Reporting lead
▪ Key requirements
Functional consultant
▪ Roles and responsibilities
▪ Approach Solution architect
▪ Technology and tools AI SME

Printing strategy Outline the strategy, including: Customer IT architect

▪ Scope
▪ Key requirements
▪ Approach
▪ Technology and tools

Personal, ad-hoc Elicit the reports for personal

reports or ad-hoc needs and data
volumes

Operational Elicit the reports for managers

reports at various levels (e.g., team,
business group)

Implementation Guide: Success by Design: Appendix: Initiate phase 665

Topic Deliverables Activities/Tasks People

Intelligence Organizational Elicit the reports for broader Customer business

strategy and executive scope, consolidated data consultant
continued reporting across the enterprise systems, Reporting lead
and KPIs for executives (CxOs).
Functional consultant
AI/ML needs Elicit the requirements that Solution architect
are fulfilled through artificial
AI SME
intelligence, such as customer
churn analysis, and customer Customer IT architect
lifetime value.

EDW (enterprise Define the high-level view of

data warehouse) Dynamics 365 data integration
to customer’s EDW.

Common data Define the high-level

model (CDM) and considerations to use and
Azure Data Lake structure data using CDM.
Storage

Security strategy Security strategy Outline the strategy, including: Customer information
▪ Scope security group (ISG)
▪ Key requirements team
▪ Approach Solution architect
▪ Technology and tools
Identity SME
Federation, SSO Define the needs for single Customer
sign-on and federation with infrastructure SME
other identity providers.
Customer IT architect
Security roles Define all the roles required
and assign these roles to
business process activities so as
to map them to business roles/
personas.

Identity Elicit the requirements for

integration integration with other identity
providers.

Implementation Guide: Success by Design: Appendix: Initiate phase 666

Topic Deliverables Activities/Tasks People

Security strategy Azure Active Define the requirements for Customer information
continued Directory Active Directory or Azure security group (ISG)
(AAD) access Active Directory, identity team
management integration needs, and cloud Solution architect
identities.
Identity SME
Information Elicit security, privacy, and Customer
security compliance requirements; infrastructure SME
regulatory needs; and audit
Customer IT architect
events.

Mobile security Define BYOD requirements,

MDM, MAM, and/or
conditional access policies.

Compliance Define the privacy, regulatory,

management compliance, and security
procedures.

Run risk assessment.

ALM strategy ALM Define the required DevOps consultant

environments environments for ALM setup Customer IT architect
(e.g., dev, test, golden, pre-
Solution architect
prod, prod).
Technical lead
Solution Define the solution
Customer IT team
management segmentation strategy

Configuration Define the source control and

management branching strategy.

Build process Define the processes for build

management and automation.

Deployment Define the processes for

process deployment management
and automation between the
environments.

Implementation Guide: Success by Design: Appendix: Initiate phase 667

Topic Deliverables Activities/Tasks People

ALM strategy Release process Define the processes for DevOps consultant
continued
release management of the Customer IT architect
solution.
Solution architect
ISV ALM Define the process for Technical lead
managing the DevOps process
Customer IT team
for other components and
solutions (such as Azure/non-
Azure components) that make
up the solution.

Environment and Tenant, Define tenants/environment/ Customer IT architect

capacity strategy environment and instance needs and the Solution architect
geo strategy locations.
Technical lead(s)
Environment Define required environments Customer support
planning to be set up. team

Deploy required environments Technical lead(s)

to enable the team to start PMO
working. Customer information
Platform release Define the plan, maintenance Security group team
process and schedules for the platform Customer
updates release and upgrades. infrastructure team

Regulation and Define the regulatory and

compliance compliance checks that
must be adhered for the
implementation.

Service-level Plan the post-go-live

operations operational procedures, such
as application availability,
backup, archive, audit,
maintenance registers,
supportability, and service
updates.

Implementation Guide: Success by Design: Appendix: Initiate phase 668

Topic Deliverables Activities/Tasks People

Environment and API limits Define the capacity-planning Customer IT architect

capacity strategy requirements for integrations Solution architect
continued and user data access.
Technical lead(s)
Storage Define the capacity planning Customer support
requirements for storage. team
Deployment Plan the deployment strategy, Technical lead(s)
planning locations, required hardware PMO
and software prerequisites, etc.
Customer information
Rollout planning Create the high-level plan for Security group team
staggered rollouts. Customer
Continuance plan Create the continuous update infrastructure team
architecture and update policy.

Regression Confirm the regression

automation tool automation tool that will be
used.

Implementation Guide: Success by Design: Appendix: Initiate phase 669

Phase
Implement Regardless of which implementation methodology
you use, here are the typical implementation
deliverables, activities, and recommended roles that
need to be performed by the project team during
the Implement phase of the project.

This table is for reference only; a missing deliverable,

activity/task, or role might not be an indication that
something is missing in your project.

Note that the names of deliverables, details

regarding activities and tasks, and roles and job titles
might change depending on your project.

When a deliverable or activity is application-specific,

it is specified as either:
▪ F&O for Finance and Operations apps; or
▪ CE for Customer Engagements apps

Implementation Guide: Success by Design: Appendix 670

Topic Deliverables Activities/Tasks People

Program strategy Project scope and Following all related activities, Customer PMO
requirements list the project scope and RTM to Customer business
signoff be signed off on. stakeholders
Cutover strategy Complete this deliverable by Project manager
and plan the start of the build. Solution architect
Solution design Once design is documented, Functional consultant/
signoff it needs to be signed off on. architect
All changes that are required
follow review, approval, and
version control.

Train the trainer Prepare the training materials Functional consultant

for customer training teams, Tech writers
such as functional admins,
system admins, and identified
functional users.

Test strategy Develop unit test Build unit test cases for the Test lead
cases defined functionality to ensure Test consultant(s)
TDD (test-driven development)
DevOps consultant
practices are adhered to.
Technical consultant(s)
These unit test cases should Performance test
be executed as part of CI/CD consultant
pipelines.

Build verification Conform to DevOps practices

tests (BVT) and build the BVT scripts that
decide on the validity of the
build for accepting to the test
environment.

Execute test Execute test cases manually or

cases/test automatically.
scenarios

Implementation Guide: Success by Design: Appendix: Implement phase 671

Topic Deliverables Activities/Tasks People

Test strategy Automation of Build automation scripts for Test lead

continued test cases high-impact processes. Test consultant(s)
DevOps consultant
It is recommended to build
automation scripts for Technical consultant(s)
processes that are critical and Performance test
prone to changes. consultant

Performance Develop and unit test

testing performance test scripts and
execute the scripts, per the
load requirements.

Business process Financial structure Finalize financial structure Solution architect

strategy design with the components: Functional consultant
chart of accounts, dimensions,
Finance SMEs
and account structure.

Test cases/ Write test cases/scenarios Functional consultant

scenarios supporting the functionalities Test consultants
to be tested for various
identified test types, such as
functional testing and SIT.

Design the automation testing

needs for performance testing,
UI testing, and more.

Business Define business processes

processes and (linear and cross-functional
personas process flows) showing the
inputs, outputs, and activities
at a subprocess and process,
including intended personas.

Gaps solution Design solution, update process

design flow, and confirm approach.
Create development plan.

Implementation Guide: Success by Design: Appendix: Implement phase 672

Topic Deliverables Activities/Tasks People

Business process Low-level For all gaps identified in Functional consultant

strategy technical design the fit-gap assessment, it Test consultants
continued (CE) is recommended to have
low-level technical design
documentation (TDD) in place.

Focus on how the

requirements are achieved
using the Dynamics 365
platform constructs. This must
be detailed to the core.
▪ Define customizations,
such as business rules,
plug-ins, JavaScript,
workflows, and Power
Automate workflows.
▪ Depict a technical flow
on how a functional
requirement is fulfilled
using the combination
of client-side scripts and
server-side rules.

Application User experience Design the visualizations, such Functional consultant

strategy (UX) design as UI form definitions, UI views, UX consultant
UX (e.g., forms, and custom controls.
views) design

Solution design: Consider the design and Technical lead(s)

API limits/service- development impacts to API Solution architect
protection limits limits/service protection limits.
Customer IT architect
planning
Technical consultant(s)
Technical lead
Functional consultant
Technical consultant
continued on next page

Implementation Guide: Success by Design: Appendix: Implement phase 673

Topic Deliverables Activities/Tasks People

Application ISV/add-ons Include the installation of Functional consultant

strategy the ISV/add-ons early in the Technical lead(s)
continued solution design so that they
Data migration lead
are incorporated into the
processes, data, security, and Technical consultant(s)
integration design. Customer IT team

Solution design: Note that based on expected Customer data

storage capacity data migration activities and migration consultant(s)
capacity planning of the
solution, the design should
include storage capacity needs.

Build solution Develop the solution per the

development design documents.
code

Develop UX Build UX constructs (e.g.,

forms, views, controls, web
resources)

Data strategy Data model Confirm data model is

precise and complete. A clear
definition of the following
aspects of the data model
should be available as part of
defining the data model:
▪ Data types
▪ Data format
▪ Data length
▪ Rules (e.g., auditable,
change tracking)

Data migration Often, it is possible to write

design custom tools or leverage
ISV tools to execute data
migration needs. Whatever the
approach taken, it is important
to design: continued on next page

Implementation Guide: Success by Design: Appendix: Implement phase 674

Topic Deliverables Activities/Tasks People

Data strategy Data migration ▪ ETL (extract-transform- Functional consultant

continued design (continued) load) processes. Technical lead(s)
▪ Data dictionary: data
Data migration lead
migration mapping
between incumbent and Technical consultant(s)
Dynamics 365 system. Customer IT team
▪ Data migration scripts Customer data
design. migration consultant(s)
▪ Execution report.
For projects with heavy data
migration requirements, it is
recommended to maintain a
separate design document for
Data Migration activities.

Configuration Create data packages and/

management or golden configuration
(F&O) environment.

Configuration Store approved configuration

management and establish:
(F&O) ▪ Change management
procedures.
▪ Transfer process to and
from repository to the
various development and
test environments.
▪ Plan to move to
production.

Historical data Confirm requirements, scope,

migration and approach.

Evaluate migration in
Dynamics vs. data warehouse
storage.

Implementation Guide: Success by Design: Appendix: Implement phase 675

Topic Deliverables Activities/Tasks People

Integration Integration design It is recommended to maintain Integration lead

strategy a separate integration design Technical consultants
document if the integration
Azure architect/SMEs
needs are heavy. This should
(if required)
cover:
▪ Integration patterns to be Solution architect
considered
▪ Detailed integration
interfaces and mapping of
data attributes between
the systems.
▪ Transformation rules
▪ Integration response
processing
▪ Integration type (e.g., sync,
async, batch).
▪ Integration technologies
(e.g., logic apps, Power
Automate, third-party
ESBs)

Dual-write If dual-write is used, harmonize

infrastructure concepts such as company,
business unit, security,
etc., between CE and F&O.
Determine the plan to architect
the data properly between
apps.

Intelligence BI design If the project has a high BI lead

strategy number of reporting Technical consultants
requirements and the project
Functional consultant
needs custom reporting
beyond what Dynamics 365
offers out of the box, it is
recommended to build a
detailed design document
covering the following:
Implementation Guide: Success by Design: Appendix: Implement phase 676
Topic Deliverables Activities/Tasks People

Intelligence ▪ Sources for data beyond

strategy Dynamics 365 apps.
continued ▪ Visualizations using
reporting tools such as
Power BI or SSRS.
▪ Reporting queries.
▪ Logic to make up the
reporting interfaces.

Security strategy Security design Security implementation Technical lead

is one of the critical Solution architect
pillars in Dynamics 365
Functional consultant
implementations. Design
should be thorough and Customer IT architect
requires extensive analysis Customer information
of the current requirements security group (ISG)
before deciding on the
security constructs suiting the
Dynamics 365 apps. Example:
For Dynamics 365 CE apps
or Power Platform, design
of the following is of utmost
importance:
▪ Business units
▪ Security roles and access
levels
▪ Ownership, access, and
AAD teams
▪ Field-level security
▪ Custom security options,
if any

Implementation Guide: Success by Design: Appendix: Implement phase 677

Topic Deliverables Activities/Tasks People

Security strategy Security roles When a custom role needs to Technical lead
continued (F&O) be created, a good practice Solution architect
is to not modify the standard
Functional consultant
roles because it will impact
the continuous updates from Customer IT architect
Microsoft. Customer information
security group (ISG)
Apply strict change
management control.

Include security roles in the

golden configuration.

Segregation of Once the security roles are

duties (SoD) assigned to process activities,
identify SoD rules and baseline
the security testing.

Compliance Track, assign, and verify your

management organization’s regulatory
compliance activities related to
Microsoft Professional Services
and Microsoft cloud services,
such as Office 365, Dynamics
365, and Azure.

ALM strategy Develop CI/CD For continuous integration DevOps consultant

pipelines and continuous development Technical consultant(s)
needs, it is recommended to
build the routines for DevOps.

Environment and Solution design: Design considerations for Technical lead(s)

capacity strategy performance high-level customizations Technical consultant(s)
from the viewpoint of system
performance, such as UI form
loads, error handling, query
performance, solution health
checks, and more.

Implementation Guide: Success by Design: Appendix: Implement phase 678

Topic Deliverables Activities/Tasks People

Environment and Regression Build regression automation Technical lead(s)

capacity strategy automation test tool and test scripts to be able Technical consultant(s)
continued to frequently uptake Microsoft
service releases.

Implementation Guide: Success by Design: Appendix: Implement phase 679

Phase
Prepare Regardless of which implementation methodology
you use, here are the typical implementation
deliverables, activities, and recommended roles that
need to be performed by the project team during
the Prepare phase of the project.

This table is for reference only; a missing deliverable,

activity/task, or role might not be an indication that
something is missing in your project.

Note that the names of deliverables, details

regarding activities and tasks, and roles and job titles
might change depending on your project.

When a deliverable or activity is application-specific,

it is specified as either:
▪ F&O for Finance and Operations apps; or
▪ CE for Customer Engagements apps

Implementation Guide: Success by Design: Appendix 680

Topic Deliverables Activities/Tasks People

Program strategy Signoff The customer/partner has Customer information

documents gone through all internal security group/team
approvals, information security Customer IT architect
reviews, penetration testing,
Solution architect
and other steps to ensure
production readiness of the Functional consultant
system. PMO

Go-live readiness The cutover plan, go-live date, Customer project

document/ and go/no-go criteria are sponsor(s)
cutover checklists agreed upon with customer
business stakeholders and
the production deployment
runbook is created with all
tasks, owners, start time,
duration, and rollback plan
determined.

Train the trainer Complete training of the Functional consultant

customer trainers for Technical writers
functional and administration
Customer training
activities.
admins
End-user training Prepare and deliver the end- Customer business
manuals user training for effective use users
of Dynamics 365 Business
Apps.

Online help Revise online help guides

guides based on feedback from UAT,
if any.

Testing strategy UAT test cases Refine UAT test cases used for Customer business
validating the solution. users (UAT test users)

UAT test report Prepare and present UAT test Test lead
execution report that baselines Functional consultant
the UAT signoffs. Test consultant(s)

Implementation Guide: Success by Design: Appendix: Prepare phase 681

Topic Deliverables Activities/Tasks People

Testing strategy Performance These results help establish the Performance test lead
continued benchmark confidence to go live for the Solution architect
testing results anticipated concurrent loads,
Customer IT architect
and transactions.

UAT signoff Once the tests are completed, Program director

business signs off on user Test lead
acceptance testing.

Business Mock cutover Conduct mock cutover and Tech lead

processes strategy approval validate and approve outputs. Technical consultant(s)
Sign off.
Solution architect
Customer IT architect

Data strategy Data migration Team completes the last Tech lead
execution results/ cutover migration of the Technical consultant(s)
signoff data required following the
Solution architect
shutdown of the incumbent
systems before opening Customer IT architect
Dynamics 365 for go-live.

ALM strategy Escrow Escrow all the project artifacts Tech lead
collected during the Initiate, Technical consultant(s)
Implement, and Prepare
Solution architect
phases.
Customer IT architect
Knowledge Register the documents and
transition maintain all pre- and post-
documents deployment activities that the
support team should consider.

Maintenance Maintainability workshops are

register/ set of trainings/documents
maintainability that help the support
workshops team learn the operational
procedures of the production
solution.

Implementation Guide: Success by Design: Appendix: Prepare phase 682

Topic Deliverables Activities/Tasks People

Environment and Service-level Confirm and understand the Tech lead

capacity strategy agreement (SLA) SLA. Technical consultant(s)
Production Release and deploy Solution architect
environment production. Customer IT architect

Implementation Guide: Success by Design: Appendix: Prepare phase 683

Phase
Operate Regardless of which implementation methodology
you use, here are the typical implementation
deliverables, activities, and recommended roles that
need to be performed by the project team during
the Operate phase of the project.

This table is for reference only; a missing deliverable,

activity/task, or role might not be an indication that
something is missing in your project.

Note that the names of deliverables, details

regarding activities or tasks, and the roles or role
titles in the People column might change depending
on your project.

When a deliverable or activity is application-specific,

it is specified as either:
▪ F&O for Finance and Operations apps; or
▪ CE for Customer Engagements apps

Implementation Guide: Success by Design: Appendix 684

Topic Deliverables Activities/Tasks People

Project Post-go-live review Customer support

postmortem team
Customer IT architect
Production Operations Refine the operations guides/ PMO
operations guides/checklists checklists that help maintain
Solution architect
(maintenance) the production solution for
updates, changes, and solution Functional consultant
health.

Platform release Operations Update the cloud solution. Customer Support

process and guides/checklists Test customizations supporting Team
updates the new updates. Customer IT architect
PMO
Solution architect
Functional consultant
Technical consultant(s)
Test consultant(s)

Maintenance Operations This is a very important Customer support

guides/checklists practice to ensure necessary team
maintenance schedules are Customer IT architect
defined to undertake activities
PMO
such as:
▪ Platform updates.
▪ Custom solution updates.
▪ Batch data updates.
▪ Archival activities.

Periodical execution of archival

and data retention.

Implementation Guide: Success by Design: Appendix: Operate phase 685

Topic Deliverables Activities/Tasks People

Change Change request Documents that explain the Customer support

management Change impact change request (e.g., hotfix, team
assessment feature enhancement) to be Customer IT architect
Change control brought into the production
PMO
solution and the impact
assessment of the existing
components, as well as the
plan to implement and deploy
the change.

User adoption/ User adoption Document the solution usage Customer business
usage document (user patterns and user interviews, users
interviews, along with strategies to Customer IT architect
feedback loops) improve the adoption of the
PMO
system.

Implementation Guide: Success by Design: Appendix: Operate phase 686