The Capstone Project:

Increasing of cost efficiency of software

development at creation of self-services
channels at commercial bank.
Phase I: Preliminary Analysis
and Approval of the Project Charter
Author: Igor Dmitriev

1
 Agenda

Topics of presentation:
• Project Background
• Project Charter
• Project Gantt-Chart
• Business Case
• Appendix

Our project is about

efficiency of processes and about people in these processes.
2
 Project Background – Mortgage Business

Sales Self-services
Worldwide Banking teams and channels
Group. France. Paris.

B2B B2C
Services for partners Services for
(Realtors, Builders, individual clients
Brokers, Aggregators)

Mortgage & Real

Estate Ecosystem web mobile Private web Bot
Business Line interface apps API interface at chat and
(IOS, Android) messengers
Igor Dmitriev (fb, t.me, vk)
Head of Products, Marketing, Digital
Business. Chief Digital Officer.

Agile Mortgage
Development Common mortgage process
Team
Documents Real Estate
Lead Consultation Application Pre-approval Final Approval
Collecting Choosing

Sales Self-services
teams:
Collateral Collateral Deal Gov.
B2B & B2C underwriting appraisal
Insurance
Preparing
Deal
registration
Service stage

3
 Project Background – Development process

1d 1d Business
1d System
Ideas validation
analysis Analysis
• Work-shop • Cost • Preparing a
about design, calculating prototype of
• Data analysis • Preparing of • Preparing of
goals, • Resource design
requirements • Users approval
Business
• Approval
technical
interview Requirements requirements
Document document
Creation of Ideas
Prototyping
ideas prioritization
1d 3d 3d

1d
5d 1d
Testing Roll-out Support
• Software • Deployment at • Monitoring of
development clients zone usage and
• Auto and • Roll-out for all • Minor
stage
manual testing • Testing at users
stability
improvements
some users and bugs-fix

Development Launching Monitoring

3d 1d permanent

4
The example task: additional mortgage program at product calculator. Total time to market: 21 days
 Capstone Project Charter
Project Name Increasing of cost efficiency of software development at creation of self-services channels at commercial bank.
Project Sponsor (Company)
Project Manager (you) Igor Dmitriev
Submission Date 13th Sep 2020 Approval Date & Initials
I. Project Description and statement •Risk Assessment
Goal of the project is increasing of cost efficiency and speed of "time to market" for processes Type of Risk Level of risk Risks mitigation
of product and software development at creation of self-services for clients and partners at Risk of defocus Middle Set of block-point at work
Digital B2C and Digital B2B teams of Mortgage and Real Estate Ecosystem Business Line schedule for participation
Risk of breaking of Middle Prioritization, Escalating to
I. Scope deadlines the project committee
Process Scope: Product and software development Risk of wrong estimation Middle Escalating to the project
Organizational Scope: Digital B2C Team, Digital B2B Team of cost and timeline committee
RUN/Change Processes: Analysis of both run and change processes
Dimension of analysis: Project frameworks, engineering practices, customer development
practices, product development practices, management practices. Alignment with Strategic Goals
Timeline of project: 6 months Strong. Mortgage Strategy for 2021-2025 require of total cost decreasing by 7% per year.
Project Tools: Lean 6 Sigma tools (including analysis of all types of wastes), Engineering
Management tools. Additional Information

Project Team:
I. Objectives & Goals / Expected Outcomes / Deliverables
Project Sponsor: Igor Dmitriev, Head of Mortgage Agile Team, Head of Products and Digital
1) 5% Increasing of cost efficiency of processes, equal to 150 000$ decreasing of total cost for 2 Project Manager: Igor Dmitriev, Head of Mortgage Agile Team, Head of Products and Digital
years, Project Participants:
2) 10% Increasing of speed of time to market = 2.1 days (in comparison with an etalon task 1) IT Area Leader
"Additional mortgage program at product calculator" with 21 days of total time to market) 2) Product Owner of B2B Team
3) IT Head of B2B Team
I. Required Resources 4) Product Owner of B2C Team
HR Resources: Around 15 000$ (HR Cost), 6 months 5) IT Head of B2C Team
- 10 workshops with teams members 6) Agile Coach:,
- Work of Leaders of projects 6 months part-time 7) Members of B2B and B2C teams.
- Field Research 2 weeks 5
 Project Background – Financial Review

Total Cost
for 2 teams for 2 years: 3 mln. $

Total Benefits
from 2 teams for 3 years: 4,21 mln. $

Goal of the Capstone Project:

Improvement of cost efficiency +5%
= 150 000$ decreasing of total cost
(for 2 years after implementing of the
project).

Cost of the project: 15 000$

HR + External Cost
ROI: 400% (at first year)
Net Income

Financial Result
Return on Normalized Equity
Data at table in mln
FTE = full time employee

6
Project Chart (Gantt)
 Appendix I. Project Background – Existing pains
Stage of Product Best practies Zones of development
Creation
Creation of ideas Our strong and weak side. We know our targets for next 5 years, We try to switch to data driven culture. We want to identify ideas not from our expert view, but from data
and we created a roadmap. We understand market and about consumer deviation of our clients
interesting fields on it.
Ideas validation tbd It's our weak point. After creation of idea, and accept of a manager we rush to implement it without proper
validation with clients and partners. That approach increases our speed but creates wastes for products
with weak demand.
Ideas prioritization We started to use methodology for prioritization of our ideas It's difficult to set to single measure different types of activities: 1) for Sales growth, 2) for clients loyalty,
based on ratio of cost and result. 3) for Optimization, 4) for Regulatory norms meeting.
Business analysis tbd For some products we need to provide resource demanding analysis of several dimensions: legal,
finance, risk and accept our product with stakeholders. It's very long process and we need to transform it
for increasing of out TTY speed.
Prototyping We try to implement to all teams step of design prototyping, for Part of teams make prototype at zone of development of real systems, because it easier to understand
example with usage of figma software. Because one extended how to do it for developers, but that process is more expensive (at least more than 2 times)
template of new system is more effective for understanding than
10 pages of description.
System Analysis Even with usage of out-staff developers, we transfer all analysis We need to improve an analysis of inter-relationship between different systems.
part to in-house for saving of know-how and growth of internal
expertise
Development We use mix of inhouse, outstaff and outsource models based on tbd
different tasks with focus on in-house
Testing tbd We need to improve our automatical forms of testing and regress forms of testing

Launching tbd -tbd

Roll-out We provide trainings for our staff, even with gamification forms tbd

Monitoring tbd Our weak point. We need to improve our post-launch monitoring of usage of all our functionality. For
example we can miss some changes of business processes and obsolescence of our functionality.
8
 Appendix 2: Agile Mortgage Development Team
Team Name IT Stack FOCUS Team Project Framework Interrelationship
Development of new 1 Product-owner (PO), 3
mortgage products without product-managers, 1
Products and MS Dynamics CRM
question of their delivery business analysts (BA), 1
(Conveyor), Automated Kanban With ALL teams
Marketing at channels, it's about Core syst. analyst (SA) + 1
Core Banking System (ABS)
Banking system, legal and developer (DEV) conveyor
risks aspects 1 SA + 1 DEV ABS
Credit Stream - internal 1 PO
MS Dynamics CRM
processes at Mortgage 4 BA
Core (Conveyor), Automated
conveyor, workplaces for 4 SA + 10 DEV Conveyor
Kanban, Agile With ALL teams
Core Banking System
internal employees 4 SA + 10 DEV ABS
1 PO
Automated Core Banking Internal processes and 4 BA
System, Digital Servicing Web/Mobile application External team of
Servicing Web (.Net) and Application for post-deal steps and Development
Agile Core, Products
(JAVA) servicing of current loan Service of Core team for
ABS dev.
.NET, C#, Angular, JS, web
app.
digital tools (web, mobile, 1 PO, 5 BA
MS Dynamics CRM
B2B Digital Sales (Conveyor)
integration) as a self- 1 UX UI analytic, 1 tester SCRUM Core, Products
service for partners 3 SA, 8 DEV
Flatter (mobile app)
Focus .Net, C#, Rest/SOAP API
of the
Capstone 1 PO
.NET, C#, Angular, JS, web
project 3 BA
app.
digital tools (web and bot) 2 SA
B2C Digital Sales MS Dynamics CRM
as a self-service for clients 6 DEV
Kanban, Agile Core, Products
(Conveyor)
No UX/UI analytics, no
Bot-platform
testers
 Appendix 3: RUN / CHANGE Processes
Interrelationship

Type of Process Example Team Problems

We always spend our
resources for tasks which
generates us new volumes
New Functional of Loyalty of business, and it’s always
Program for partners. a problem to spend ➢ Single resources
PO, UX/UI expert, BA, SA,
CHANGE Launching of new stage of
Developers, Testers
resources for refactoring of
credit process (for example processes for decreasing of
Insurance stage) level of errors and establish
➢ Conflict of priorities
a foundation for more
efficient development at ➢ Interrelationship at
the future.
process of development
BA, SA, Developers, Lack of global
Bugs fixing, corrections of
testers, prioritization (do we need
RUN fields, minor changes at
NO PO and UX/UI to do it?), lack of analysis
processes
experts of root cause of errors.

10
 Appendix 4: Online customer journey in mortgage

BANK’S OR PARTNERS’ SIDE CORE BANKING PROCESS

CUSTOMER SEARCH FOR MORTGAGE PERSONAL ONLINE INSURANCE PROPERTY

REAL PROVIDER DATA AND APPROVAL SELECTION MAINTENANCE
ESTATE SELECTION, INCOME
APPLICATION CONFIRM. COLLETERAL
APPRAISAL
API
DEAL

INTEGRATION

API API API API API

Real Estate Government ID • Federal Tax • Credit 30+ insurance renovation,

Aggregators usage for data Service bureaus companies moving,
filling • Pension Fund +100 appraisal furniture
companies companies

11
12

Thank You!

12
The Capstone Project:
Increasing of cost efficiency of software
development at creation of self-services
channels at commercial bank.
Phase II: Solution Exploration & Proposal
Based on project at Worldwide Banking group,
Author: Igor Dmitriev

13
 Agenda

Topics of presentation:
• Analysis of expectations of stakeholders
• Analysis of teams’ workshops
• Analysis of wastes
• Analysis of the data
• Proposed solutions
• Reflections
and further steps expectations

Second Phase of the project is about focus on details...

14
 Project Goals

Objectives of the Project:

1) 5%
Increasing of a cost efficiency of
processes

2) 10%
Increasing of speed of time to market

15
 Results of analysis:
Key expectations of stakeholders and management
Stakeholders group Expectations Not meted expectations both at B2C particularity B2B particularity
B2C and B2B
1. Maximizations of efficiency ratio (Business 1. Difficultness at analysis of business result
result / Required Resources) and usage of resources for each task. Lack
General 2. Understanding of full utilization of of single methodology.
- 1. Internal set of reports
Management = Sponsors resources 2. Difficultness of resources control for
3. Prioritization of tasks based on business change and run tasks
KPIs 3. Difficultness at process of prioritization
1. Decreasing of defects level
1. Defects level 37,5% average
2. Increasing of level of clients satisfaction
Head of Digital and Projects 2. Lack of monitoring of clients satisfaction
(based on Net Promoter Score NPS) 1. Defects level 41% 1. Defects level 34%
(Chief Product Owner) 3. Lack of seamless funnel of usage
3. Increasing of level of usage of functionality
4. Lack of entire monitoring of TTM
4. Time to Market (TTM) decreasing
1. ROI (Return on Investments) >1 at first 3 1. Lack of entire methodology for analysis of
Finance Department years + RONE > 20% at first 3 years each task
- -
(Controlling) 2. Transparency at resources utilization 2. Lack of information about resources
3. Clear effect of each task utilization
1. Lack of understanding of impact of each
1. Clear business KPI for each team team for achieving of general KPI 1. Shared KPI with department of
Teams Heads 2. Understanding of general context and 2. Lack of time of general management for 1. Shared KPI with team of partnership and business
(Heads of B2B and B2C) needs of business synchronization with each team. digital marketing development managers at
3. Transparency at work with each task 3. Lack of usage of single resources control branch’s network
tools.
1. Taking their opinion to consideration at 1. Excess focus to partners
Heads of Branch’s network business requirements preparing 1. Lack of communications 1. Lack of focus to seamless journey. Without focus to clients
(participants at credit 2. Analysis of not only digital but also 2. Excess focus of each team to their context journey of a client at each journey (clients who were
process) processes of others channels for building without looking to entire client’s journey channel acquired through partners
of seamless clients’ journey network)

Other Product 1. Alignment with general IT Architecture

Development Teams, IT 2. Taking to consideration of processes of 1. Lack of collaboration with other teams - -
Architects other teams at development 16
 Results of analysis: Workshops with teams-members
Efficiency of the Product Development Process

People 1 Procedure 3 Measurement 5

Lack of analysis of Poor level of specification of Lack of usage of common metrics
users expectations business requirements of development efficiency

High pressure of deadlines doesn't Excess informal communication Lack of measurement of personal
allow to make result better instead of usage of prototypes efficiency of each participant

Flexible methods of development have their

pro and cons. Lack of usage of common
Technical Interrelation policy make difficult to manage expectations Lack of alignment at building of
with other teams from results. omnichannel process Low Level
of Efficiency
Lack of collaboration with teams – of the Product
participants of the credit process Lack of standardization of job Lack of alignment between team for common
requirements for each role mortgage product and a single clients journey.
Development
Process
Environment 2 Policy 4 Place 6

17
 Results of analysis: Workshops with teams-members
Time-to-market speed

People 1 Procedure 3 Measurement 5

Lack of Testers Low quality of business requirements Lack of single methodology for
increase wastes at next stages measurement of time to market speed

Tasks are waiting for time of Lack of usage of engineering practices Lack of continuous measurement of of
specific employee (Continuous Delivery,
Delivery, Deployment & time to market speed
Integration)

Interrelations between teams cause Lack of formal deadlines for each step Banking Legal and Compliance policies
policies block
additional waiting of approval or process fast time to market speed Low Time to
Market Speed
Need of approval from
Lack of formal expectation for result IT Security rules require additional
stakeholders cause waiting
of each step of development steps of control

Environment 2 Policy 4 Place 6

18
31
 Product Development Process Map:
Common and differences for B2B and B2C teams.
Both: poor Only 0,5 hour for System analysts Poor prototype at Lack of
B2B channel ideas validation. approval and and developers B2C due to lack of collaboration
require more only 2 hours, calculation, but 1 were not satisfied UX/UI expert between b2b and
time due to it’s but 1 day of day of waiting with a quality of b2c team
complexity 1d waiting 1d BRD 1d
Business System
Ideas validation
analysis Analysis
• Work-shop • Cost • Preparing a
about design, calculating prototype of
• Data analysis • Preparing of • Preparing of
goals, • Resource design
requirements • Users approval
Business
• Approval
technical
interview Requirements requirements
Document document
Creation of Ideas
Prototyping
ideas prioritization
1d 3d 3d

Development stage bugs-fixing after

launching
1d
5d 1d
Testing Roll-out Support
• Software • Deployment at • Monitoring of
development clients zone usage and
• Auto and • Roll-out for all • Minor
stage
manual testing • Testing at users
stability
improvements
some users and bugs-fix

Development Launching Monitoring

Excess 3d 1d permanent
communication due Lack of testers at Lack of monitoring
to poor specification B2C caused high of usage of
of requirements level of errors Business users were not satisfied due to lack of asking functionality
about their feedback.
The example task: additional mortgage program at product calculator (program with government subsidy 19
for primary market). We need to do it both at B2C and B2B channels. Total time to market: 21 days
 Results of analysis:
Gemba-walk (field research)

COVID 19 dramatically changed working environment and principles of management.

At product & Software development process this year (from March 2020 to February 2021)
all employees are working at remote mode with regular working meeting.

I was able to analyze quality of work at meetings, and ask about principles of business and system
analysis, but it was difficult to emulate real working conditions (for example real-life conversations
between different analyst or developers in case of asking about ideas for solution)

I think new fully remote conditions is a challenge for management

and process improvements activities.

20
 Results of analysis:
8 types of Wastes at Product / Software Development
Type of Waste Type of Wastes Description /
Canonical at Software Development Specific problems
1. Development of features that users
1. Needs to support useless functionality
Overproduction don't need,
2. Obsolescence of tasks (and BRD)
2. Overproduction of task at backlog
Waiting Waiting / Delays Waiting between stages
Needs to delete useless code. Risk of
Transportation Task switching conflicts of code for new task and code
from not finished and not tested task.
Inventory Incomplete / Partial work done Conflicts between code of different tasks.
Rework due to poor preparation of Rework, retesting, due to lack of clear
Over/Incorrect Processing
requirements business and system requirements
Lack of detailed requirements leads to
Excess communication due to lack of
Unnecessary Movement excess communication, subjectivity in
requirements formalization
interpretation, difficultness at testing.
Defects Software Errors Inoperability or wrong functionality
Short deadline / lack of time for task
Generating the volume not a quality of
Unused employee creativity doesn’t allow to improve quality of
functionality
requirements or realization
21
focuses for increasing of cost efficiency focuses for increasing of time to market speed
 Results of analysis: Statistics analysis
Analysis of right setting of team roles

If we share an average workload at each stage and available roles and

resources, we able to see shortage of resources of Business Analysts and
System Analysts (both at B2C and B2B), testers for B2C. f

But it doesn’t mean that we need to fire developers, because they are main
production workforce. It mean that we see bottle neck at product development
conveyor. And if we improve these zones we improve overall speed and
efficiency of the overall process.
22
 Results of analysis: Statistics analysis. Analysis of
best practices at work of teams (based on the data)

Parameter B2B B2C As we can see based on the data work of B2B and B2C teams
have several differences:
Team Size (FTE) 19 12
Existence of internal yes no B2C has 1 single team, B2B has 4 sub-teams
groups
B2C uses bigger size of task for 1 employee, B2B uses more
Amount of internal groups 4 1
detailed separation of tasks.
Tasks for 6 months 840 168
Tasks for 1 team-member 7,4 2,3
per month

23
 Results of analysis:
The main categories of errors
Product Development Errors
= Did we create a correct product?
Equal to waste: Overproduction, Over/Incorrect Processing
- Development of features that users don't need
- Rework due to poor preparation of requirements

Technical Error
= Is our functionality working correctly?
Equal to waste: Defects
- Inoperability
- Wrong functionality

u
Product Development and Technical errors have I
different nature but common causes.

But what if technical errors meet to errors at

product development stage?

Cost of an Etalon Task = X

X * % (probability of prod. errors) * % (probability of
technical errors) = Х * 72,5% * 74,6% = 184,9% X
24
 Results of analysis:
Statistics analysis

4 1 + 2 +3

As we can see both for B2B and B2C – vague system requirements cause defects at
development stage. But it really interrelated with causes of errors at products
development process. System analysis based on results of business analysis and we
can see that it also poor.

It’s a strong insight for focusing of our research.

25
 Results of analysis:
Statistics analysis

If we stack results
of the main categories of
errors both for product and
software development
processes we again able to
see strong impact from
stages linked to business
and system analysis for
quality of the overall
process.

26
 Results of analysis:
Statistics analysis

It important not only identify main causes of wastes, but also weight it with average amount of wastes.

After analysis of results of multiplication of probability to value of rework we able to see the main causes for rework.

Based on the goal of ”5% increasing of the overall efficiency” I identified 4 key wastes as a focuses for improvement:
- Lack of approval of stakeholders;
- lack of analysis of needs of users and clients;
- Vague business requirements;
- Vague system requirements. 27
 Results of analysis:
Statistics analysis

One of the main insights of analysis stage was hidden at differences at practices of teams in question of separation of overall
workload between stages.

I analyses differences of approaches of each teams, and moreover differences of time spending and used roles.

28
Results of analysis:
Statistics analysis
As we can see 13% of
extra work at stage of
business analysis and
research of users and
clients expectations
+17,9% saves 18% of
+13,1% B2C development cost.
B2B
It’s not a focus of the
project, but If we take to
account cases where cost
of 1 developer higher
than cost of 1 business
analyst, we get one more
strong point for efficiency
improvement.

29
Results of analysis:
Statistics analysis
If we talk about
decreasing of time to
18% of overall market speed we need to
decreasing of look at types of wastes at
time-to-market each stage of the project.
speed
We able to see that
waiting and excess
communication are critical
at stages of system
analysis and
development.

Overall wastes at these

stages more than 18%.

30
Results of analysis:
Statistics analysis to be

Proposed solution is increase at 15%

amount of efforts at stages of:
- Ideas Validation
- Business Analysis
- System Analysis
- Prototyping
For 20% decreasing of work at stages:
- Development
- Testing
- Monitoring 31
 Stage 3: Improvement
Proposed Countermeasures
Proposed Solutions for increasing of efficiency level
Linked Cause of Wastes Countermeasures
Lack of approval of stakeholders 1. Usage of Poka-yoke (check-lists for interviews)
2. Formalization of approval process (formal deadlines and scope of responsible persons)
3. Usage of more effective ways for approval (less text more prototypes and process maps)
4. More active feedback usage at process of development
5. Including of stakeholders to working groups
Lack of analysis of needs of users and clients 1. Less experts opinions, more data driven decisions
2. Usage of customer-development practices (deep interviews, feedback collecting)
3. Usage of technical tools for analysis (f.e. clicks monitoring)
4. Usage of beta stages with loyal users
Vague business requirements 1. Usage of Poka-yoke (check-lists for analysis)
2. Increasing of usage of prototyping
3. Usage of special analysis frameworks (f.e. user stories)
4. Additional learning of business analysts for required data for system analysis and
development stage
Vague system requirements 1. Usage of Poka-yoke (check-lists for analysis)
2. Usage of required analysis frameworks (for example process maps, data maps)
3. Including of system analysts to stage of customer development

+ formalized expectations and roles descriptions for each participants 32

 Stage 3: Improvement
Proposed Countermeasures
Proposed Solutions for increasing of time to market speed
Stages of Process Type of wastes Countermeasures

System Analysis Waiting 1. Setting of deadlines for time of answer and approval
2. Setting of prioritization between analysis of new tasks and
+ answering for questions for existing tasks
3. Usage of more effective notification of start of waiting and arrival of
Development the answer
Excess Communications 1. Implementation of set of improvement for business and system
analysis (setting of requirement format of documents, usage of
prototypes and mock-ups, usage of schemas, preparing of data and
process maps)
2. Usage of more effective tools for collaboration

33
 Appendix: Phase 2 Reflections

Causes of deadlines shifting:

• Wrong estimation of required time resources
• Excess communications and approvals at bank
• Defocus due to pressure from other tasks

Key Insights and experience:

• ”Snow ball” effect. Difficultness of prioritization of findings.
• COVID 19 environment: all discussions at online mode
• Avoid of resolving before finishing of the analysis.

The main goal – looking for the best focuses of improvement

based on ratio: required resources / expected result. 34
 Appendix: Phase 3 Expectations

Revision of Phase 2 results:

• Insights at implementation stage able to
improve our understanding
• Validation of methodology of measurement
of results

Following to focuses:
• Focus on the more critical points of
process improvement
• Avoid to try to improve anything, it’s impossible
• Set the system of the quality maintenance.

35
The Project Chart

36
Thank You!

37
The Capstone Project:
Increasing of a cost efficiency of a software
development at creation of self-services
channels at a commercial bank.
Phase III: Implementation & Evaluation
Part 1: Tools Implementation
Based on project at Worldwide Banking group, Author: Igor Dmitriev

38
 Agenda

Topics of the presentation:

• Goals of Phase III / Part 1 of Phase III
• Preliminary results
• Implementation Process overview
• Phase 3 Part 1 retrospective analysis
• Phase 3 Part 2 and overall project
finalization next steps and expectations

39
 Short Overall Overview of The Project
Goal of The Project: Phases I, II Key Insights: Phase III Goals: Preliminary Results:
Goal of the project is increasing of cost Key zones of Wastes Reduction: 1st of March: first 1/2 of the final Challenges:
efficiency and speed of "time to market" Based on the goal of ”5% increasing of the presentation • We met resistance of part of team
for processes of product and software overall efficiency” I identified 4 key wastes - with parts about prepared tools (f.e. members against of rebalance of
development at creation of self-services for as focuses for improvement: lack of poka-yokas, process improvements, work-load.
clients and partners at Digital B2C and approval of stakeholders; lack of analysis of project frameworks) and results of their • We don’t’ sure about efficiency of
Digital B2B teams of Mortgage and Real needs of users and clients; vague business usage (some pro and cons of each tool, part of solutions, but we unable to
Estate Ecosystem Business Line requirements; vague system requirements. some primary qualitative and make a decision without analysis
quantitative results for each tool) of an overall effect to full value chain.
Objectives & Goals / Expected Outcomes / Key zones of an overall efficiency Quick Wins:
Deliverables: increasing: As we can see 13% of extra green – done • We see positive impact of several
1) 5% Increasing of cost efficiency of work at stage of business analysis and orange – partially done implemented tools
processes, equal to 150 000$ decreasing of research of users and clients expectations • We deeply understand problems of
total cost for 2 years, saves 18% of development cost. Proposed 1st of April: last 1/2 of the final our value chain and our next steps
2) 10% Increasing of speed of time to solution is increase at 15% amount of presentation. Next steps:
market = 2.1 days (in comparison with an efforts at analysis stages for 20% - qualitative and quantitative results of the • We need to finish our analysis of
etalon task "Additional mortgage program decreasing of further stages. implementation of the proposed process overall results of system of new
at product calculator" with 21 days of total - "to do" for the after-project stage solutions
time to market) Key Zones of Time to Market value - as my important focus - some approaches • We need to make a decision about
Decreasing: Waiting and excess for ensuring the sustainability of the final process.
communication are critical at stages of implemented improvements • We need to implement a solution for
system analysis and development. Overall - reflections about the overall project results maintenance.
wastes at these stages more than 18%.

40
 Changes at Overall Process

1d 1d Business
1d System
Ideas validation
analysis Analysis
• Work-shop • Cost • Preparing a
about design, calculating prototype of
goals, 1. Merge •• Data
Users
analysis
• Resource
• Preparing of
Business 1. Merge design • Preparing of
technical
requirements approval Requirements • Approval requirements
interview
Document document
Creation of Ideas
Prototyping
ideas prioritization
1d 3d 3d
+ Parallel mode of BRD and prototype
+ Usage of Tools of Ideas Assessment preparing. Increasing of skills at B2C team
+ Validation and Data First Approach before a Manager Approval
as a base for creation of ideas

1d
5d 1d
Testing Roll-out Support
• Software • Deployment at • Monitoring of
development clients zone usage and
• Auto and • Roll-out for all • Minor
stage
manual testing • Testing at users
stability
improvements
some users and bugs-fix

Development Launching Monitoring

3d 1d permanent

+ Usage of auto tests

at Developers stage 41
The example task: additional mortgage program at product calculator. Total time to market: 21 days
 Changes at Overall Process

2. steps removal
Xd Xd
Ideas
System Analysis
prioritization
• Data analysis • Preparing of Business
• Users Interview • Cost calculating
Requirements Document
• Preparing of technical
• Work-shop about design, • Preparing of prototype
• Resource approval desigh
requirements document
goals, requirements

Creation and Business analysis

validation of ideas and Prototyping
Xd Xd

3. workload
shifting
Xd
Xd Xd
Testing Roll-out Support
• Software • Deployment at • Monitoring of
development clients zone usage and
• Auto and • Roll-out for all • Minor
stage
manual testing • Testing at users
stability
improvements
some users and bugs-fix

Development Launching Monitoring

Xd Xd permanent

The example task: additional mortgage program at product calculator. Total time to market: 21 days (previously) 42
Resulted time will be calculated at Part 2 of Phase III
 Changes at Overall Process
4. Tools
implementation
Xd Xd
Ideas
System Analysis
prioritization
• Data analysis • Preparing of Business
• Users Interview • Cost calculating
Requirements Document
• Preparing of technical
• Work-shop about design, • Preparing of prototype
• Resource approval desigh
requirements document
goals, requirements

Creation and Business analysis

validation of ideas and Prototyping
Xd Xd
+X% +X% +X% +X%

Xd
Xd Xd
Testing Roll-out Support
• Software • Deployment at • Monitoring of
development clients zone usage and
• Auto and • Roll-out for all • Minor
stage
manual testing • Testing at users
stability
improvements
some users and bugs-fix

Development Launching Monitoring

Xd Xd permanent
+X% +X% +X% +X% +X%

43
 Tools implementation at Product Development
Stages: Stage 1: Creation and validation of Ideas
Stage Check-List: Linked Wastes: Implemented Tools:
0. Do we have an idea or we need
to create an one?: 1. Overproduction 1. Required Quantitative (f.e. funnel metrics) and 6. Users Club
◻︎ yes, go through check-list Development of features that users don’t need Qualitative (f.e. clicks monitoring) Data Analysis We ask loyal external
◻︎ no, let’s create it based on data Allow to decrease impact of experts’ biases and internal users
2. Unnecessary movement join our users club for
Excess communication due to lack of ability to discuss
1. Requirements of our Strategy 2. Deep interviews and Feedback analysis ideas and be
◻︎ decreasing of the cost of process requirements formalization. Is it important for our users? Do we have they preferences?
involved to beta-test
◻︎ increasing of sales of new functionality.
◻︎ increasing of cross-sale level 3. Unused employee creativity 3. Strategy driven approach Also including of
◻︎ increasing of clients satisfaction It’s not required to ask developers and testers Which goal we try to achieve? internal stakeholders
◻︎ decreasing or risk level propose ideas, but it improves overall team’s to Users club
spirit and motivation. decreases risk of
4. Stage Check-list
Usage of check-list allow us not to miss important details rejecting at approval
2. Do we have an existent data Also we had unused clients and partners stage.
◻︎ if yes, let’s analyze creativity.
◻︎ if no, is it possible to collect it? 5. Template usage
- Preliminary expectations (cost, benefit) allow us to understand is our
idea important. We set minimum buy-back period of 2 years as a threshold
3. Do we have users feedback
- Required fields allow to avoid unnecessary communications.
◻︎ if yes, let’s analyze
◻︎ if no, is it possible to collect it? Indicator AS IS TO BE PRO CONS

Cost of stage (average, m/h) X Self-checking and a 1. Implemented tools

4. Filling of template
better formalization required of spending
◻︎ cost
Impact to total wastes decrease time of additional time
◻︎ benefit
spending at next 2. It’s not obviously
◻︎ auditory of users
Impact to total Time-to-Market stages for team, why their
◻︎ Is approval of stakeholders required?
need to spend
◻︎ short name
Impact to total cost additional time
◻︎ description 44
 Tools implementation at Product Development
Stages: Stage 2: Ideas Prioritization
Stage Check-List: Linked Wastes: Implemented Tools:
0. Use the scoring file
◻︎ Fill cost of task 1. Overproduction 1. Scoring File
◻︎ Fill expected amount of new sales (1 Development of features that not required by During the project we implemented best at our bank tool for ideas
year after implementation) business prioritization. If you put at calculator your expected cost and benefits,
◻︎ Fill expected size of cost decrease you will see buy-back period and resulted net result. It’s very useful if
(sum for 1 year after implementation) you need to compare different dimension of benefits (sales, cost
◻︎ Fill NPS (Net promoter score) decrease, clients satisfaction, cost of risk, cross-sale increase)
increase
◻︎ Fill Cost of Risk Decrease (1 year
after implementation)
◻︎ Fill amount of selling of cross-sale
products
◻︎ Compare net result and buy-back
period

FYI: Methodology
Scoring file use financial model of
mortgage business and analyze impact
of different drivers to net result of
business. I use net income from new Indicator AS IS TO BE PRO CONS
loans (based on expected life-long
result), impact of Opex and COR Cost of stage (average, m/h) X Efficient tool for We currently unable
decrease to net income, amount of analysis of different to calculate impact of
income from cross-sale products. But we Impact to total wastes dimension of benefits clients satisfaction to
need to finish our research about impact at single measure. net income.
of clients satisfaction to net result (we Impact to total Time-to-Market
try to check impact to sales, cross-sale or
customer churn) Impact to total cost 45
 Tools implementation at Product Development
Stages: Stage 3: Business Analysis and Prototyping
Stage Check-List: Linked Wastes: Implemented Tools:
0. Usage of Template
◻︎ Fill all required fields, especially
- changed systems 1. Over/Incorrect Processing 1. Templates and User-Stories / Mock-Ups Usage
- group of processes (sales, service, credit Rework due to poor preparation of requirements. Usage of templates (f.e. BRD - Business Requirements Document)
conveyor) allows to analyst remember about all important points of analysis, and to
- new/changed functionality
2. Unnecessary movement user of document – easier navigate through it.
1. CJM or Process Map
◻︎ If new process include/change more than 3 Excess communication due to lack of
step of process and more than 2 points of requirements formalization. 2. CJM and Process Map Usage
decision usage of CJM or process map is Preparing of Client Journey Map and Process Map (for example in BPMN
mandatory notation) allows to easier understand a process and interrelationship
◻︎ All steps of process need to be described between it parts.
◻︎ If process map contains more than 2
participants – it’s mandatory to use BPMN 3. Matrix of Approvals
notation.
If we change several fields of process we need to approve it with
2. Matrix of Approval stakeholders.
◻︎ Use matrix of approval for checking of new
functionality with stakeholders
3. Requirements for mock-up 4. Stage check-list usage / Deadlines settings
◻︎ Visual effects Usage of check-list as a part of template allow to remember about
If it required to use any special visual effects – important details and deadlines. Feedback of System analysts and
please provide link or short video to sites with developers used for templates preparing.
the same functionality
◻︎ Elements of control Indicator AS IS TO BE PRO CONS
Please highlight elements of control and show
different conditions Cost of stage (average, m/h) X - Decrease of excess - Analysts resistant to
◻︎ Web / Mobile communications at new rules due to
Please check how proposed page looks at web
and mobile versions. Impact to total wastes next steps increase of level of
◻︎ Colors - It’s easier for their work.
In case of usage of new types of colors – please Impact to total Time-to-Market analysts to remember - Necessity of training
provide codes of colors about important for analysts for usage
◻︎ Alignment with BRD details of mock-ups tools (f.e.
In case of change of functionality please check it Impact to total cost 46
with author of BRD doc.
figma).
 Tools implementation at Product Development
Stages: Stage 4: System Analysis
Stage Check-List: Linked Wastes: Implemented Tools:

0. Data Model 1. Over/Incorrect Processing 1. Wiki for log of used modules and changes
◻︎ Description of changes and Rework due to poor preparation of requirements. Allow to different teams understand common changes
integration for fields
◻︎ Integration map 2. Unnecessary movement 2. Data Maps
If system requirements document (SRD) contains changes or
Excess communication due to lack of integration more than 5 fields.
1. Functional map requirements formalization.
◻︎ Extended description of systems
deviation 3. Templates and check-list for analysis
3. Inventory Usage of templates for providing of analysis of different systems.
Incomplete / Partial work done. Usage of required deadlines and priorities for analysis.
2. System map
◻︎ Description of required for use 4. Additional learning
system elements (plugins, services,..) Alignment of knowledge base between analysts (sql, databases, basic
and changes linked to it principles of development) for their ability to prepare more useful for
developers technical requirements.
3. Preliminary settings
◻︎ Our system (Dynamics CRM) allow to
system analysts create part of new
process steps and system objects Indicator AS IS TO BE PRO CONS
without development
Cost of stage (average, m/h) X Decrease of wastes Necessity of
4. Interrelationship analysis at further steps additional training for
◻︎ Analysis of common usage of system Impact to total wastes analysts
elements between different teams.
Impact to total Time-to-Market

Impact to total cost 47

 Tools implementation at Product Development
Stages: Stage 5: Development
Linked Wastes: Implemented Tools:
1. Inventory 1. Auto-tests
Incomplete / Partial work done Usage of auto-test for basic functions decrease time for overall testing
of work of the whole system.
2. Transportation
Task switching 2. Feedback Analysis
We included to our process flow system field for feedback of developers
3. Defects about lack of information at system requirements. It allow to system
Software Errors analysts correct their work.

! We identified some needs for implementing of software engineering

4. Unnecessary movement practices like continuous delivery and continuous integration, devops,
Excess communication due to lack of but implementation of tolls such these out of scope of the project. We
requirements formalization. will include that proposals to to do list for post-project stage.

Indicator AS IS TO BE PRO CONS

Cost of stage (average, m/h) X - Decreasing of time - Additional time for

of testing. building of auto-tests.
Impact to total wastes - Decreasing of level - Not all analysts are
of errors. happy to receive
Impact to total Time-to-Market feedback.

Impact to total cost 48

 Tools implementation at Product Development
Stages: Stage 6: Testing
Linked Wastes: Implemented Tools:

1. Unnecessary movement 1. Including of testers to stage of system requirements approval

Excess communication due to lack of Participation of testers at system requirements preparing allow to predict problems of
requirements formalization. testing and include requirement system objects for automatical testing.

2. Auto-test usage
Usage of automatical tests requires time for tests development, but decreases time
spending and time to market at future.

3. Beta Tests with User-Group

Proposal to members of User Group to participate at beta-test allow to find hidden bugs.

4. Learning for B2C team

We don’t’ have a tester at B2C team, but for decrease of impact of lack of that knowledge
at team we provided a training for business and system analysts. Also we will include
proposal of including of additional team member for consistency of team to to do plan for
post-project stage.

Indicator AS IS TO BE PRO CONS

Cost of stage (average, m/h) X - Decreasing of time - Additional time for

of testing. building of auto-tests.
Impact to total wastes

Impact to total Time-to-Market

Impact to total cost 49

 Tools implementation at Product Development
Stages: Stage 7: Launching
Linked Wastes: Implemented Tools:

1. Unnecessary movement 1. Preparing of learning materials

Excess communication due to lack of We analyzed that usually teams provide communications about changes
requirements formalization. without convenient learning materials. It’s cause additional
communications after implementation.

Low Result We were unable to prepare a template for learning materials due to lack
of resources.

Even with understanding of key steps of process for

improvement we try to improve all. As a result we
implemented too minor improvements at non-important
stages and spent time required for major improvements.

Indicator AS IS TO BE PRO CONS

Cost of stage (average, m/h) X N/A N/A

Impact to total wastes

Impact to total Time-to-Market

Impact to total cost 50

 Tools implementation at Product Development
Stages: Stage 8: Roll-out
Linked Wastes: Implemented Tools:
1. Unnecessary movement 1. Usage of Change Management principles
Excess communication due to lack of We provided a training about basic change management principles.
requirements formalization
We were unable to prepare a template or check-list for that stage due to
lack of resources.

Low Result
Even with understanding of key steps of process for
improvement we try to improve all. As a result we
implemented too minor improvements at non-important
stages and spent time required for major improvements.

Indicator AS IS TO BE PRO CONS

Cost of stage (average, m/h) X N/A N/A

Impact to total wastes

Impact to total Time-to-Market

Impact to total cost 51

 Tools implementation at Product Development
Stages: Stage 9: Monitoring
Linked Wastes: Implemented Tools:

1. Overproduction 1. Training about post-implementation monitoring

Development of features that users don’t need We identified that users don’t use a part of functionality. We provided a
training for teams about principles of post-implementing monitoring.

We were unable to prepare data models and charts for monitoring due
Low Result to lack of resources.

Even with understanding of key steps of process for

improvement we try to improve all. As a result we
implemented too minor improvements at non-important
stages and spent time required for major improvements.

Indicator AS IS TO BE PRO CONS

Cost of stage (average, m/h) X N/A N/A

Impact to total wastes

Impact to total Time-to-Market

Impact to total cost 52

 Phase 3 Part 1 Reflections

Our mistakes:
• Even with understanding of key steps of process for improvement we try to improve all. As a
result we implemented too minor improvements at non-important stages and spent time required
for major improvements.

Lessons learned and External factors:

• We met resistance of several members of teams against of changes.
It’s a sign of our insufficient efforts of change management.
• Several tools were ineffective. At one hand it’s permissible at pilot stage.
At other hand – it’s a sign of our poor preliminary analysis.
• During our project team was involved to a process of political and structure changes at the Bank.
It’s was an obstacle for seamless implementing of tools and clear process of measurements.
• It’s difficult to use all Lean 6 Sigma tools if we analyze creative process. For example we were
unable to use multi-factors analysis and ask team to use tools at 2^K*R variations. Also each
creative task contains non-controlled variable (for example mood and overall condition of
employee (for example if any developer spent all night for computer gaming)).

53
 Phase 3 Part 2 Expectations

Data analysis:
• Merge all collected data to a single model with analysis of
4 dimensions at each step of process: cost of stage,
impact to overall wastes, impact to overall Time-to-
Market, impact to overall cost of a process

Purposes of the stage:

• Final analysis of all qualitative and quantitative data
• Final analysis of overall results and lessons learned
• Project Closing
• To Do for after-project stage including framework
of sustainability of implemented improvements
• Reflections about overall project

54
55

Thank You!

55
56

The Capstone Project:

Increasing of a cost efficiency of a
software development at creation of self-
services channels at a commercial bank.
Phase III: Implementation & Evaluation
Part 2: Evaluation and Overall Conclusion
Based on project at Worldwide Banking group,
Author: Igor Dmitriev

56
 Agenda

Topics of the presentation:

• Analysis of the Project’s results
• Proposals for after-project stages
• The project retrospective analysis

The third phase of the project is about

ensuring the sustainability of the implemented
improvements...
57
Analysis of the Project’s results
“AS IS” Model for the reference size of task

58
Analysis of the Project’s results
“TO BE” Model for the reference size of task

59
 Analysis of the Project’s results

1. Cost (Man/Hours) Weighted cost of reference

2. Weighted Wastes task (with average value of Data Model:
3. Time to market rework)
Steps of the process • Analysis of several dimensions of
data (cost, wastes, TTM) and
their overall results.
• Analysis of teams separately
• Specific value of weighted
wastes (probability of wastes * %
of total rework caused by errors
at stage).

60
 Analysis of the Project’s results

1. removal of stages Results:

• Goals of the Project achieved:
Cost reduction -6,87% / 5%
TTM reduction -12,5% / 10%
Also -50,53% reduction of wastes

2. Planned workload
shifting was lower
61
 Analysis of the Project’s results

Key Insights:
• Real workload shifting was
dramatically higher than planned
• Primary cost was decreased only at
1,43%. It’s was too low for teams
perception as a justification of
changes.
• Total wastes (causes for rework)
were decreased twice! It’s allow to
decrease total cost of work more
Decreasing of wastes was more efficient than general decreasing of than at 6,87% and decrease TTM
usual cost of each task... It seems obvious after calculation of results, from 21-22 days to 19 days!

but it was not perceived at separated stages. It’s a reason why support
62
of management is important and why wide looking is mandatory.
 Proposals for after-project stages

❑ Analysis of interrelationship with other teams (Core,

Products, Servicing)
❑ Implementing of post-launch monitoring
❑ Usage of Change Management principles
❑ Analysis of Change/Run process interrelationship
❑ Roll-our of best practice to other teams
❑ Analysis of waiting buffer between teams, implementing The best way of ensuring of sustainability
of Kanban principles of results – building of culture of
❑ Lean 6 Sigma Learning and Culture Implementing continuous improvements.

But we understand that we need to launch

improvements at step-by-step basis.

63
 Overall Project Reflections

❑ Limitations of implementing of tools of work

standardization at creative processes...
❑ But flexible agile and lean principles of process
improvement are effective.
❑ It difficult to see overall result at narrow stage.
It was the reason of resistance. But after presenting of
results for overall process team accepted changes.
It’s a reason why we need to use both helicopter view of
overall problem and deep mining of zones of
development.
❑ It’s was an insight that focus on wastes is really
effective, not lower than focus at just cost of stages.
64
Appendix: The Project Chart

65
66

Thank You!

66