AMD in IS

biggest summary ever and sample exam

questions/concepts

based on the “Analysis, Modeling and Design in Information Systems”’s course

granted by Professor Christine Legner

Spring 2018

contributors : Ahmad Ben M’Rad, Camille Glatz, Charles Guérin, Floria Papadopoulos, Daniel Sbai
Table of Content

Sample Exam Questions 4

What is an enterprise architecture? Define the term and outline the typical structure and components
of an enterprise architecture. 4
Enterprise architecture design is often compared to city planning and house building. Explain why and
illustrate the analogies. 4
What are key problems in today’s IT environments that an enterprise architecture helps to address?
Please give three examples and explain, how a company can address them using an enterprise
architecture approach. 4
What are enterprise architecture layers in ArchiMate ? Explain them and provide examples of EA
components for each layer. 5
How should a company proceed in system identifying and documenting their enterprise architecture?
Suggest an enterprise architecture framework and explain how it supports the company. 5
How can you assess the quality (or correctness) of an architecture model? 6
Why does enterprise architecture modeling require an analysis of stakeholders and concerns? Explain
for two different architecture descriptions / models the stakeholders and concerns they address. 7
Explain the different phases in application portfolio management. Which architecture models and
visualization are useful in the different phases? 8
How do you systematically translate business requirement into an IS solution ? Give two examples that
take into account different IS acquisition options. 8
What are the different phases in packaged software implementation? 9
What are the challenges in criteria to analyze and evaluate different packaged software solution ? 10
What are the constituent of a data architecture ? Explain the different levels of data architecture and
the models associated with them for the example of customer data. 10
Please comment three out of the following statements: Clearly state whether you agree or not, and
argue why. 11
1. Zachman framework provides a comprehensive structure for EA; but it is not an EA framework. 11
2. UML and ArchiMate propose similar models on the IS layer and consequently can both be used for
EA modeling. 11
3. The most difficult task in implementing enterprise architecture management is the design of the
target enterprise architecture. Once it has been defined, only little work remains for the enterprise
architects. 11
4. Configuration of a SAP system for the requirements of a specific organization belongs to traditional
software development approach 11
5. Big data is a collection of structured data with little impact on the IS architecture. 11
What types of system requirements can you distinguish? Name and explain the different categories of
requirements. 12
Zachmann Framework 13

Archimate 13
Business layer 13
Application layer 13
Technology layer 14
1
TOGAF 14
Synthesis 16

Application Portfolio Management 17

Purpose 17
Applications Portfolio Management - Steps 17
Application Inventory (Understand and Document) 18
Analyse and Evaluate 18
Transform and Design 19
Monitor and Progress 20
Business Architecture and Business Context / Requirements 20

Conceptualize the business architecture 21

Four approaches to conceptualize the business architecture 21
Capabilities 21
How to create a Capability Map 22
How to use a Capability Map 22
Operating Models 22
Operating model by Ross and Bell (2005) 23

Business Analysis 23
4 types of requirements 23
Functional Requirements: 24
Non-functional requirements: 24
Development vs Package Software Implementation (COST) 25
Custom Development Software 25
Package Software Development 26
Selection 26
Trailoring 26
Data Catalog 27
Entreprise Business Assets 27
Main Constituents of a Data Architecture 28
Data Architecture is Defined by Models at Several Levels 28
Data-driven Enterprises - Data Documentation & Data Catalogs 29
How to support data users? 30
Data catalog: maintaining an inventory of data assets 30
Case Study 31
Assura Insurance 31
InnoBike 31
Incident Management of a software company 31
BriteLite 31
Starting with the as-is capability map of BriteLite: 31
2
Allied Insurance Inc. 32
Allied Insurance Inc. plans to define its data architecture in order to improve firm-wide data
management. What should the data architecture comprise? 32

3
Part I#RevisionIsTheKey
Sample Exam Questions

What is an enterprise architecture? ​Define the term and outline the typical structure and components
of an enterprise architecture.

The fundamental structure of entreprise, of an organization, either as a whole, or together with partners,
suppliers and / or customers (“extended enterprise”), or in part (e.g. a division, a department, etc.).
Describing (Modeling) at an Aggregated Level the essential components and relationships of an entreprise.
as well as the principles governing its design and evolution.

Typical structure and components

The typical structure with its components of an EA is:

● The strategy (customers, markets, products and services, business goals, capabilities)
● The processes (activities, roles, …) and organisation (organizational units and hierarchy,...)

The information systems:

● Data (data objects, attributes, database…),

● Applications (software, functionality,...)
● Integration (interfaces, protocols, …)
● Infrastructure (System software, devices, network, …)

Enterprise architecture design is often compared to city planning and house building. Explain why
and illustrate the analogies.

Enterprise architecture applies the same methodology as the “common” architecture. Both need to
understand the foundation of the system that they are creating, and they both are dealing with different
stakeholders that have different interests, needs and requirement for the project. For instance, you can
consider a house from the perspective of his future owner, or from the perspective of the city planner. The
same can be established with the enterprise architecture: the different systems of an enterprise are
described through multiple viewpoints and different kind of relationships.

Also, it develops an AS-IS/TO-BE methodology for documenting and planning purpose.

What are key problems in today’s IT environments that an enterprise architecture helps to
address? Please give three examples and explain, how a company can address them using an
enterprise architecture approach.

- How IT can assess and establish a competitive advantage

Understand where the value creation comes from, leverage it and accelerate its efficiency, identify white
spots and red spots

- How to proceed to early important changes inside the business

prioritize project, eliminate redundant software by defining an architecture vision.

4
- How to assess the digital transformation
By defining the architecture as it’is for what needs to be achieved or reached.

- Help to communicate.
As explained in the previous question there is many stakeholders involved in a company from differents
backgrounds such as Management (business) with no deep IT understanding.

What are enterprise architecture layers in ArchiMate ? Explain them and provide examples of EA
components for each layer.

Each of the layer describes a specific view of the architecture of the entreprise.

- Business Layer ​develops the main processes, services, functions and events of the business units.
main concepts developed here are the ​business actors, roles, objects, processes, services​ and the
different relationships between them.

- Application Layer ​develops the software applications that support the components in the business
with application services. The concepts developed here are the ​application component, application
function, application interface and the data object​. ​More generally, this layer explain how behaves
part of the software and the data, what is automated and how the application service is made
available to a user or another component.

- Technology Layer​ deals with the infrastructure (hardware) that support the application layer
(software). Without the infrastructure, the software cannot be run. Concepts developed in this
slides are the ​node​ (ressource that stores artifacts), the ​device​ (hardware resource such as mobile),
the ​system software​, the ​network​, i​nfrastructure service​ and ​artifacts​ (physical piece of data used
or produced in software that is useless we never used that shit).

For more information on ArchiMate cf “Archimate” chapter of this summary.

How should a company proceed in system identifying and documenting their enterprise
architecture? Suggest an enterprise architecture framework and explain how it supports the
company.

System identifying and documenting the enterprise architecture refers to the first part of the design of an
enterprise architecture. In the best cases, the enterprise store and documents processes they do and you
can use this base to derive the business architecture, application and infrastructure. But it happens a lot of
times that this data simply does not exist. In this case, you have to retrieve it by yourself by whatever
means you can (observation, discussion, investigation..)

TOGAF ​is one of the main framework used to define requirements, plan and implement a system. It’s
divided in 8 steps.

TOGAF support the company by defining first the requirements, identifying the stakeholders, concerns in
what is called ​architecture vision​. ​Then, processes are the first source of concern to answer business needs
such as develop the baseline ​business architecture description. From there, the ​information systems
architecture are derived and backed up by the ​technology architecture​. After that, TOGAF allows the
managers to use the project management tool to plan and execute the implementation of the system such
as the solution choice, migration plan, Governance and Change Management.

5
How can you assess the quality (or correctness) of an architecture model?
It has to met three main requirements :

Firmitas ​(firmness, structural soundness). That means that the architecture should be “strong” and
consistent.

Utilitas​ ​(commodity, usefulness). Means that what has been modelled is useful/relevant for the business.

Venustas​ ​(grace, elegance). Means that it is more or less readable.

In a more ​formal​ aspect:

1. Domain (Purpose and content of the model (business correctness))

2. Audience interpretation ( stakeholders and concerns ( viewpoint selected and why))
3. Language ( Modeling notation (UML BPMN Archimate principles ))

Lindtland Framework

Syntactic quality

- A business/application/infrastructure is used by.. (not assigned)

- Application services are realized by application functions (not by components)
- Application interface is used by an application component or composes an application component
- Communication paths are realized by a network (and not associated with)

Network is associated with a device or a node and realizes a communication path.

System software is not a standalone component. It has to be assigned to a node.
Do not confuses artifacts (=files that are deployed on infrastructure components) with data objects

More generally, the ​syntactic quality​ focus on the correctness of the function that you use to modelize.

Semantic quality

Completeness

6
 - Your models should comprise the key elements brought forward by the software vendor
- Don’t forget the required user components : web browser, GUI, etc.
Validity
- Is the business application an application component or a suite?
- Assign the collected information to the correct layer
- Infrastructure components should be consistent with the architectural paradigm

More generally, the semantic focus on the validity and completeness of your model regarding the way you
modelize it (comparing to the reality) through a “business view”.

Remark : availability of the information may differ dependending whether the service you modelize is an
OpenSource project/software or a business with IP.

Pragmatic quality

Focus on the way you present the modelization of the business.

- Is it too overloaded? → Create model hierarchies and views

- Are the viewpoint chosen relevant to the audience ? → Choose correct viewpoint for the a
stakeholder
- Is your model well organized ? → Group similar elements

Why does enterprise architecture modeling require an analysis of stakeholders and concerns?
Explain for two different architecture descriptions / models the stakeholders and concerns they
address.

A system is situated in an environment. The environment determines the totality of influences upon the
system. Stakeholders of a system are parties with interests in that system, in order to model an
architecture that relevant to them we need to take into account their interests. Their interests are
expressed as concerns such as flexibility, agility, compliance, privacy, etc…

For instance, let’s take the following stakeholders.

- C-level’​s concern is to maximize the stock value of the firm on the market or control (KPI)
- Customers ​are concerned by their privacy when using the firm’s service.
- Shareholders ​want to maximize profit
- Government ​is concerned by the compliance to regulation

The system described by an enterprise architecture is not a standalone system. This system is used by
certain people in a certain environment, which makes these people the stakeholders of the system. This
means that these people have to be considered when designing the system and when we make it evolved,
two activities which are performed under the scope of enterprise architecture. For example, the Business
layer is of great interest to business owners, as it describes the key processes which create value (business
services) and satisfy customers (business actors); the application layer is of great interest to system
designers, as they have to understand which application services need to be made available through
application functions in order to best serve the business processes.

7
Part II #NotKnowingTheAnswerIsHard
Explain the different phases in application portfolio management. Which architecture models and
visualization are useful in the different phases?

cf. Application Portfolio Management Steps in this summary,

How do you systematically translate business requirement into an IS solution ? Give two examples
that take into account different IS acquisition options.

- By converting the requirements into the design of a system blueprint. (cf. Business Analys)
- By defining the business capabilities / abilities (combination of physical, human, technology
resources) that the new IS solution will impact.

There are two main strategies to meet the requirements.

1. Package Software Implementation (buy);

2. Custom software development (make)

for more information about PCS or CSD cf the “Package Software Development” and “Custom Software
Implementation” parts of this summary.

What are the different phases in packaged software implementation?

What are the challenges in criteria to analyze and evaluate different packaged software solution ?

- Products are built on architectural assumptions that may vary across system components. Just as
products have built-in processes, they also have built-in architectural assumptions that could
conflict with the evolving system architecture.
- Customers have limited visibility into COTS product source code and behavior.
- Solution are driven by the marketplace not what the specific client A needs…
- Rights and licences are involved

All these facts limit the visibility of criteria evaluation.

8
To analyze and evaluate different packaged software solutions, we first have to know what are our needs.
We then have to compare our requirements and the IS solution offer in order to know if there is a fit or a
misfit. Fits and misfits can sprout from structure components (deep = data models, surface = UI), or from
context specificity (pervasive = industry requirements, country requirements, organization specific =
decisions made by the company).

The problem with COTS software solutions is that, often, there is no transparency about the source code,
which does not allow the company to evaluate compatibility. Also, COTS often make built-assumptions
about how the system will be used and also architectural assumptions, which could conflict with the
Enterprise system.

What are the constituent of a data architecture ? Explain the different levels of data architecture
and the models associated with them for the example of customer data.

Nowadays data are a key asset for companies, its a strategic resource that can leverage the Value Creation
of a company. Data is highly dispersed… without a control of its quality.

- Enterprise data model​, that represent an organization’s most relevant business objects.
- Logical data models​ (also called high-level data models), that represent an organization’s data
objects, organized in terms of entities and relationships between them
- Physical data models​, that represent the concrete realization of the data in systems / applications
and may focus on different aspects

Please comment three out of the following statements: Clearly state whether
you agree or not, and argue why.

1. Zachman framework provides a comprehensive structure for EA; but it is not an EA framework.

I can agree with this statement. Indeed, even though it’s called Zachman framework, it only provides a
basic structure that supports organisation, integration, development and management of a set of
architectural representations (models) ⇒ description. However, it does not specify models or notations
and doesn’t support the design process itself ⇒ doesn’t tell how to do EA.

2. UML and ArchiMate propose similar models on the IS layer and consequently can both be used
for EA modeling.

Yes they can both be used for EA modeling but they have differences and though different uses. Archimate
can give different viewpoints and large overview whereas UML doesn’t, it drill down processes. UML is
more detailed on data and Archimate take into account different stakeholders and business processes.

3. The most difficult task in implementing enterprise architecture management is the design of the
target enterprise architecture. Once it has been defined, only little work remains for the enterprise
architects.

This statement seems wrong. Enterprise Architect, according to the TOGAF framework will mainly focus on
the first step of the framework, which are specifically establishing an architecture vision, derive from there
a business architecture, the information system architecture (which is basically explaining how interact
9
applications and data objects) and last but not least developing an infrastructure. Once this steps are done,
there is still plenty of work to do, as establishing a roadmap of the different transition architecture. Even if
we agree that it is certainly one of the most important task, the maintaining of the architecture of the
enterprise and application portfolio management are as much as important as the rest, as it can restart the
process of redesigning an enterprise architecture.

4. Configuration of a SAP system for the requirements of a specific organization belongs to

traditional software development approach

I would disagree with this statement, as we know that SAP is a COTS software therefore configuration
would fit in “ tailoring” and not traditional software development approach, according to lecture 10 slide 8,
or page 24 of this summary.

5. Big data is a collection of structured data with little impact on the IS architecture.

Wrong. Big data is usually unstructured data, it’s large dataset that are difficult to analyze. They might have
biases, noises and abnormality, the data need to be cleaned. They have a big impact on the IS architecture
as they are used for business transformation.

For more information cf “Data Catalog” part of this summary.

What types of system requirements can you distinguish? Name and explain the different
categories of requirements.

cf. “Types of requirements”, from “Business Analyse” part of this summary.

10
 Framework and Tools
Zachman Framework

WHAT HOW WHERE WHO WHEN WHY

SCOPE

ENTERPRISE
MODEL
ARCHIMATE
SYSTEM MODEL

TECHNOLOGY
MODEL

DETAILED
REPRESENTATIO
N

FUNCTIONING

Advantages Disadvantages

All relevant aspects of any architecture planning No model types and modeling procedures defined
are included

Regarding methodology/procedures everything is No notation defined

kept open

Archimate ( fews viewpoints )

Business layer

Business process viewpoint (designing purpose) focus on behaviors by detailing the functions within each
business process of a company (high-level structure)

Target:​ operational managers and process architect

Concerns:​ the structure of business processes, consistency and completeness, responsibilities

11
Application layer

Application structure viewpoint (designing purpose) focus on the structure (no way) of one or more
applications components.

Target​: Enterprise, process, application, and domain architects

Concerns​: Application structure, consistency and completeness, reduction of complexity

Note: ​in the Application Layer, Data and Applications are not represented as computer system but as
logical group of capabilities that manage data objects.

There is no reference to a specific technology!

Technology layer

Infrastructure viewpoint (designing purpose) focus on the software and hardware elements supporting
the application layer, such as network, device, node , server , system software, etc...

target:​ Infrastructure architects, operational managers

Concerns:​ Stability, security, dependencies, costs of the infrastructure

TOGAF
- TOGAF Architecture Development Model
- Core of TOGAF
- Provides development phases with objectives being clearly defined for each phase

The TOGAF’s ADM is iterative by nature. Each iteration corresponds to new decisions, with new ideas or
details, level of precisions or time horizon.

12
Archimate can be combined with TOGAF, as it corresponds to phase B, C, D of the ADM. But what about
the other phases?

The Preliminary Phase​ is about defining the architecture principles, which is mainly defining the qualitative
statement of intent that has to be met with the architecture (rational and measure of importance).

Example of Principles Statement

One example of the architecture principle could be for instance ​to derive from the business goal a few
points that can be improved through architecture.

13
The Architecture Vision ​(Phase A) the high-level definition of baseline and targets environment, from a
business, information systems and technology perspective.

(Phase E) identifies the parameters of change, the major phases along the way, and the top-level projects
to be undertaken in moving from the current environment to the target. The output of Phase E will form
the basis of the Implementation Plan required to move to the Target Architecture. This phase also
attempts to identify new business opportunities arising from the architecture work in previous phases.

Migration Planning ​(Phase F) has as a purpose to create a migration plan in cooperation with the portfolio
and projects managers (Ch. Lecture on Portfolio Management). In this phase we evaluate the resources
that are required to proceed to the migration, prioritize the different projects and generate the
Architecture Implementation Roadmap and Migration Plan.

When the firm reaches this level she enters in what is called “transition phase”, with a “​transition
architecture​” that shows the each transition state. The transition architecture enhance the different
opportunities that can be used to model or represent concepts that lead the firm to the target point, from
baseline to target (as-is/to-be). The transition architecture shows the status of the firm at ​incremental
states.

With each of these nodes has as a content a stable representation of the enterprise architecture.
14
Note : the Architecture Roadmap ​lists individual increments of change and lays them out on a timeline to
show progression from the Baseline Architecture to the Target Architecture.

TRM → Technical Reference Model

Provides information about the taxonomy (descriptions, appellations).

Synthesis

A framework is a basic conceptual structure used to solve or address complex issues. A method is a
series of steps to achieve a goal.

An architecture framework establishes a common practice for creating, implementing, analyzing and
using architecture descriptions within a particular domain of application or stakeholder community

A framework
... defines what you should do.
... does not define how or with which tools you should do it. ... should be vendor-neutral.

- architecture viewpoints and models to structure the relevant aspect of a firm

- metamodels to describe the overall enterprise (the component of the models, their taxonomy)
- methods and procedure that describes in what order it should be (architecture development
model, ADM guidelines & techniques, EA framework capabilities)
- glossary of term
- references (potentially) that could be use to model

How do you evaluate the pertinence of a framework ? Below some criterias

- must be consistent and structured

- make the business-driven approach simple and natural
- multiple levels of abstraction
- define process for developing the architecture
- describe the artifacts that will be present

For instance, Zachman Framework is not really a model as it only provides architecture viewpoints. It is
considered as a taxonomy.

15
Application Portfolio Management
- Is the ongoing application of a systematic and decision-making processes
- Helps to evaluate an organization’s application portfolio, weights them against relevant criteria.
- Take actions to implement out to date app. and so on…

Purpose

● Reduce the complexity and redundancy of the application landscape,

● Mitigate eco and tech risks
● align business needs and IT

Applications Portfolio Management - Steps

Understand and Analyze and Evaluate Design and Transform Monitor and Progress
Document

Focus on identifying Focus on the Design and transform Monitor and Progress
applications and collect assessment of Define optimization
data (what application application (do they strategies and
does the organization cover the current and scenarios, target
possess or plan to future needs, how do application
possess they score in terms of architecture, scoring
business value) per application

Deliverables Deliverables Deliverables

Application fact sheets Application landscapes Application landscape

Application inventory Portfolio matrices (to-be)
EA models (App. Layer) Heat maps Roadmap

To evaluate an Application, the first step is to go through a functional fit and gaps analysis. To do so, we
use an application landscape map. It represent a landscape viewpoint. It is a tool designed for decision
making (managers, CEO, CIO), enterprise architect, application owner and the concern is about complexity
reduction, comparison of alternatives.

Application Inventory (Understand and Document)

- Basically a list of applications

- A set of data and information for each application
- A set of quantitative and qualitative characteristics for each application

Analyse and Evaluate

Example of Application Landscape. Focus on process and geographic location

16
Scoring criteria Example

Business value of strategic fit

application business value in terms of quantitative and qualitative benefits
application usage
user satisfaction

Technical Health or operational performance

Technology Maturity platform maturity or health
architectural fit
alignment with global it standards

Risks operational risk

risk of failure
compliance risk
system complexity

Costs Total cost of ownership

maintenance costs
depreciation

17
Transform and Design

The third phase is about shaping the to-be portfolio and planning the transformation. For this purpose,
actions​ are defined per application :

● Innovate and invest by creating a new application

● Maintain the application and invest in it
● Maintain the application but not invest in it
● Replace the application with a new solution or consolidate with existing ones
● Engineer to new language or platform

Monitor and Progress

At this point, a Roadmap is created.

It collates a set of actions, clustered on a time-scale basis, and supplies the organization with immediate,
short-term and long-term opportunities to improve the application landscape.

18
Business Architecture and Business Context / Requirements
In ArchiMate, the business architecture is represented in the business layer (seems logical).

Conceptualize the business architecture

Four approaches to conceptualize the business architecture

● Business Model (logic of value creation)

● Capability Map (alignment of resources with business strategy)
● Operating Model (operations strategy)
● Business Process Model / Value stream (value chain)

19
Capabilities

- Describes firm’s capacity

- The way the firm deploy its resources and are a main drivers of firm performance
- From an EA perspective, capabilities are abilities that an organization, person, or system possesses.

Example of a capabilities map for a bank

Can also be modeled in ArchiMate

20
How to create a Capability Map

- Visualization in network or tree structures

- Describe logical groupings and dependencies between capabilities
- 5-9 generic capabilities of the firm

Note : business capabilities are the building block of the business and represent stable business
functions.

How to use a Capability Map

Classification Purpose

- core vs non-core
- strategic vs operational vs supporting
- customer-facing vs internal
- innovating vs differentiating vs commodity
- Either charts can be used to visualize the capability analysis in more details

Decompose a capability map

- Hierarchy of capabilities
- Common three levels
- Procurement level choosing the good level of granularity

Operating Models

21
An Operating Model is the abstract representation of how an organization operates across process,
organization, technology domains in order to deliver value defined by the organization in scope. From the
operating model, IT leaders can derive the “to-be” architecture design.

Operating model by Ross and Bell (2005)

Abstract of how an organization operates, helps and drives the to-be architecture phase.

- Process​ ​standardization​ : defining exactly how a business process will be executed

- Process​ ​integration​ : links the efforts of organizational units through shared data

Business Analysis
The practice of enabling change in an enterprise by defining needs (requirements) and recommending
solutions (design) that are of value for stakeholders.

- Requirements : something that need to be done or changed

- Design : the solution to this problem (requirement).

4 types of requirements

- business requirements
- stakeholders requirements
- solution requirement
- transition requirement

Solutions to requirements can be either functional or non-functional

22
Functional Requirements:

● Process-oriented, examples :
○ Allow employees to place orders for new vehicles + recording addition of new vehicles to
inventory
○ Record customer deposit and customer payment and customer vehicle purchase
○
● Information-oriented, example :
○ Allow managers to view current vehicle inventory

Non-functional requirements:
● Operational, examples:
○ Prepare sale contract from existing templates
○ Run on portable devices
○ connect to printers wirelessly
● Performance, example:
○ System should support sales staff of 15 salespersons
● Security, example :
○ No sales person can access any other salesperson’s customer contracts

Note : a good indication of a bad design is that nobody is using the service. ​A good success factor is the
fitting between the requirements and the solution provided by the information system.

When dealing with the system deficit and the system excess, one approach could be to understand what is
the most costly decision : to change the system or switch the requirements.

23
How to redesign a business process? 4 components should be taken in account to evaluate the impact of a
new re-designed business process

- Cost
- Flexibility
- Time
- Quality

Ideally, a redesign of a business process decreases the time required to handle an order, decreases the
required cost of executing the business process, improves the quality of the service delivered and improves
the ability of the business process to react to variation.

Of course all of this almost never happen together. → Trade off.

Development vs Package Software Implementation (COST)

Custom Development Software

The solution adopted here is to make the software by your own means. In order to produce a solution that
fits with the criteria of the firm, we have to go through 4 steps. This is called the Rational Unified Process
(RUP), an iterative software development framework. Important ! It is use-case driven !

1. Inception: ​determine the vision, the scope of the system who, what, how is going to use it,
understand the ​risks
2. Elaboration: ​deep understanding of the requirements, design and validate the baseline architecture
3. Construction: ​Iterative development like SCRUM filing, completing, testing of the solution
4. Transition: ​prepare, train, communication, launch the beta test

24
Package Software Development

Selection
Selection of a Package Software Implementation is an important concern for the firm. Below you can find
some tools useful to help to choose the most fitted software.

1. The Weighted Scored Method

+ It is quite simple to understand how it works and to come up with a “rational” decision
- The priority and conformance is compromised by a certain form of subjectivity

2. AHP (Analytical Hierarchy Processing)

+ Goal clearly defined, provides more confidence in decision making as we enhance the opportunity
to choose a solution that could (fully) match with our criteria
- Complex to use

Trailoring (​for Packaged Software Implementation)

Trailoring focus on the specifications of a software for a company.

- Personalization
- Configuration
- Modification
- Add-ons/Bolt-on

One import regarding package software implementation is that you cannot for instance modify it by
yourself, or at least ask for modifications and expect this software to still be updatable.

25
Personalization and configuration regarding particular needs is still possible in some cases.
Add-ons refers to extension that are

Data Catalog
Data is obviously becoming a key asset in today’s economy and companies. However, the data’s potential
value is not fully appreciated as it is hardly managed as a strategic resource.

● “It’s frustrating that companies have a better sense of the value of their office furniture than their
information assets” - Douglas Laney, Tech Analyst at Gartner
● “Only 3% of companies’ data meets basic quality standards” - HBR, September 2017
● “80% of the work involved (in advanced data analytics) is acquiring and preparing data” - HBR,
December 2016

Indeed, because companies rely on more than one application to run their business, data is highly
dispersed and duplicated across systems, causing variations in quality, format and meaning.

The increasing use of packaged software applications is part of the problem, as each application comes
with its data requirements and formats (ie, it is not up to the user company to set formats). You therefore
have little control over the way high-level business data concepts are realized.

Entreprise Business Assets

Data Realm Structure Volume Description Example

Master Data Structured Low Enterprise-level data entities Customer, product,

that are of strategic value to an supplier, and
organization. Typically location/site
non-volatile and
non-transactional in nature

Transaction Structured & Medium-High Business transactions that are Purchase records,
Data semi-structu captured during business inquiries, and
red operations and processes payments

Analytical Data Structured Medium-High Derivations of the business Data residing in

operation and transaction data data-warehouses,
used to satisfy reporting and data marts and
analytical needs decision support
applications

Documents and Unstructured Medium-High Documents, digital images, Claim forms,

content geo-spatial data, and medical images,
multi-media files maps, video files

Reference Data Structured & Low-Medium Internally managed or externally Geo data and
Semi-structu sourced facts to support an market data
red organization’s ability to
effectively process transactions,
manage master data, and
provide decision support
26
 capabilities

Metadata Structured Low Defined as “data about the Data name, data
data”. Used as an abstraction dimensions or
layer for standardized units, definition of
descriptions and operations. E.g. a data entity, or a
integration, intelligence, services calculation formula
or metrics

Main Constituents of a Data Architecture

Regarding Archimate models, the data architecture is modelled through the business objects (business
layer), data objects (application layer) and artifacts (tech layer). Note that all of those are considered
passive elements.

In the context of the Zachman Framework, we are interested in the Data/What column:

View Data Description Stakeholder

Scope / Contextual List of things important to the business Planner

Enterprise Model / Enterprise data Defines the common terms Owner

Conceptual model and strategic business rules for
corporate entities without
technology constraints

System model (Logical) Logical data Is a representation of an Designer

model organization’s data, organized
in terms of entities and
relationships between them

Technology model Physical data Is a representation of a data Builder

(Physical) model design which takes into
account the facilities and
constraints of a given database
management system (DBMS)

Detailed representations Data definition e.g. DBDGEN Subcontractor

(Out-of-context) languages

Data Architecture is Defined by Models at Several Levels

● Enterprise data model​, that represent an organization’s most relevant business objects
● Logical data models (also called high-level data models), that represent an organization’s data
objects, organized in terms of entities and relationships between them
● Physical data models​, that represent the concrete realization of the data in systems / applications
and may focus on different aspects
- the realization of the data model in one or more applications (e.g. Realization overviews)
- the relationships between different realizations of the same data item (e.g. Source and Consumer
Models)
27
 - the transformation of data as it moves between systems (e.g. Transportation and Transformation
Models)

As a general rule, the logical data model will change only when there is a significant change in business
processes or business models, but physical models will exist in various versions representing the “as-is”
structure and one or more “to be” evolutions.

Data-driven Enterprises - Data Documentation & Data Catalogs

Data-driven enterprises are built upon three pillars/trends:

● DATA - Data democratization: it means that “everybody has access to data and there are no
gatekeepers that create bottleneck at the gateway to the data. [...] The goal is to have anybody use
data at any time to make decisions with no barriers to access or understanding”
● PEOPLE - Data citizenship: With data as an asset, every employee needs to be data-savvy - not only
data managers and data scientists. Consequently, all employees are to be empowered as data
citizens. Data citizenship includes both rights and responsibilities
● ORGANISATION - Data sharing culture: Instead of having data silos, data management needs to
establish a data sharing culture. Sharing data is a data citizen’s key responsibility and provides
benefits through additional uses of data in other areas of the company

Furthermore, more people are required to use data. Through this use, many questions arise, such as:

● Where can I find data?

● Who owns them?
● Who maintains them?
● Who can help me explain them?
● Are they still up-to-date?

By 2019, data and analytics organizations that p​rovide agile, curated internal and external datasets for a
range of content authors will realize ​twice the business benefits of those that do not. - Gartner, Market
Guide for Data Preparation, 2017.

Questions vary between business roles. For example:

28
How to support data users?

Data catalog: maintaining an inventory of data assets

29
“A ​data catalog maintains an inventory of data assets through the discovery, description and organization f
datasets. The catalog provides context to enable data analysts, data scientists, data stewards and other
data consumers to find and understand a relevant dataset for the purpose of extracting business value.” -
Gartner report: “Data Catalogs Are The New Black in Data Management and Analytics” (2017)

Case Study

BriteLite

Starting with the as-is capability map of BriteLite:

- Mark capabilities that already exist, but need to be augmented / modified.

- Add new capabilities that need to be developed.

Based on the BriteLite case description (see Lecture 5)

As part of its digital transformation strategy, BriteLite decided to introduce “Smart lighting” in the context
of the Internet of Things. Your task is to support the digital transformation program by identifying and
assessing the required future capabilities:

For each augmented /new capability, describe the capability realization by specifying the required people,
processes, information, and technology.

To be augmented:

STRATEGIC CAPABILITIES
Regulations management: this capability needs to be augmented, as using IoT calls for new regulations.
Data is collected, and this needs to be made clear to the customer. New contracts have to be signed for
data collection (!! RGPD). Required people: IoT law specialists. Processes: the starting process will be to
acquire enough knowledge and information about data management and law. After this, it will be a
continuing monitoring and correction process, as data laws quickly evolve. Information: ? Technology: tech
to anonymize data?

Marketing and sales management: this capability needs to be augmented as BriteLite is launching a new
product that does not belong to their traditional market. Required people: Mkt salespeople. Processes: ?

30
Information: information about if the current customers are interested. Otherwise, collect data about the
market. Technology: not to be augmented.

CORE CAPABILITIES
R&D Engineering
Logistics management +Production management
Partner management

SUPPORTING CAPABILITIES
IT management: the collection of data from the smartbulbs needs a supporting infrastructure. Required
people: engineers and technicians to augment the capacity of the IS. Processes: extract, transform and load
the data. Information: data objects coming from the smartbulbs. Technology: Databases, …

Allied Insurance Inc.

As a part of a data architecture initiative, you are asked to classify Allied Insurance data objects. Name 3
typical data types and explain the differences between them. Provide 3 examples for each data type in the
case.

Master Data​ = structured data which is generally low in volume. It has strategic value, is non-volatile and
non-transactional. In the example of Allied Insurance we could cite the ​“table des franchises”​ folder, which
is established by the insurance and then used throughout transactions.

Transaction Data ​= semi-structured data which is generally medium-high in volume. It is captured during
business operations and processes. In the case of Allied Insurance, we could cite for example the
“​sinistres” data object, ​which is opened and established once a claim process deposited by the customer
and accepted by the insurance.

Analytical Data = ​structured data, which is generally medium-high in volume. It serves reporting purposes,
and is derived from transaction data. In the case of Allied Insurance, we have for example yearly reports of
the main reasons cited in customer claims, for example the number one reason for car accidents, etc...

Allied Insurance Inc. plans to define its data architecture in order to improve firm-wide data
management. What should the data architecture comprise?

Enterprise data model =​ most relevant data objects, for example customer profile, customer orders,
sinistres…
Logical data model = ​how the data objects relate to each other, which dependencies exist
Physical data model = ​realization in systems/applications, how the data is transported and transformed,
for example where this information is stocked, for example in a data base that is filled by an operational
system, which uses an interface to communicate with the sellers.

31
32