Written by Vernon Naidoo, Senior Consultant at Info.Blueprint
Master Data Management (MDM) implementations are inherently complex due to the multiple (often legacy) data systems that are involved. These implementations require deep alignment across business and IT functions, as different departments often have conflicting definitions, ownership, and quality standards for core data entities like customers or products.
Integrating disparate source systems, resolving data inconsistencies, and establishing golden record rules introduces technical and governance challenges. Additionally, implementing effective data stewardship processes and ensuring user adoption demands strong change management and clear communication. These complexities mean that MDM projects must be approached holistically and with a clear view of the end state goal.
In 2024, Info.Blueprint implemented an MDM solution for a large JSE-listed company. We used technology that is known for its ability to unify and govern data across diverse legacy and operating systems, ensuring consistency and reliability. By integrating this particular platform into its data architecture, our client was able to transform how they handle their critical data, enabling more agile and informed decision-making.
Our MDM implementation involved configuring the platform to manage master data, data integration into and out of the intelligent data hub, and enforcing governance policies across various data sources. Two data domains were selected for the first iteration – Customer and Product.
Key to this implementation was the use of a micro-deliverables project management approach. It is common knowledge that whenever new technologies are implemented in large organisations, “teething” issues are to be expected. Info.Blueprint decided to minimise the potential risk of implementation delays and cost overruns that were not scope extension related by taking a micro-deliverables approach.
The rationale for developing and adopting this approach was that we quickly realised that the design, build and testing was very dynamic in nature. This meant there was no way to upfront design and build the system without adopting an iterative process.
For example, in developing the data quality cleanup rules, we recognised that while the rules looked ‘good to go’ on paper, after building and testing on actual data, we always found tweaking was needed.With this in mind, we needed a methodology that could help the teams from both organisations cope while managing the various components of the project.
We split the work into two major parts, Analysis & Design and Build & Test – summarised below:
Analysis & Design
Build & Test
Analyse Source Data
Profile Source Data
Design Data Model
Configure Data Model
Map Data
Ingest Data
Design DQ Rules
Configure DQ Rules
Design Matching & Survivorship Rules
Configure Matching & Survivorship Rules
UI Design
UI Build
Dashboard Design
Dashboard Build
Security Design
Security Build
Workflow Design
Workflow Build
Outbound Files Transfer Design
Outbound Files Transfer Build
Our methodology involved breaking down the project into the smallest level of micro-deliverables. This provided us with the following benefits:
– We could share technical and other project deliverables on a regular basis instead of a large deliverable at the end. This helped our busy client to read through and provide meaningful reviews and feedback without the burden of reviewing a large document
– The client could easily relate to the extent of the work involved in the project and support adjustments to the project plan, when needed
– Delivering frequently served as a communication mechanism to ensure that the client and project teams were aligned at all times
– The approach served as a risk management strategy to ensure the whole team was keeping to the original scope and goals of the project
– We were able to show continuous delivery and value throughout the project life cycle
– We saw huge reputational benefits in terms of delivering a very complex project professionally, on time, and within budget
A key consideration in taking a micro-deliverables approach was the need for an effective document classification system for the project as all deliverables had to be documented as artefacts.
We adopted the following naming convention: Client – Technical deliverable/Project management deliverable – Report identifier (eg Build) – Sequential number (eg Build05)
We had the following categories of micro-deliverables:
Micro-Deliverable Category
Total Number
Project Management
4
Infrastructure Setup
3
Design
6
Data Analysis
9
Build
10
Test
4
Deploy
2
Change Management & Operations
1
Additionally, we developed a tracking system to update the micro-deliverables when changes were made after the document was submitted for review and accepted by client.
The nature of the business means that there were always changes to the design of the DQ Rules or Match, Merge and Survivorship rules to produce the Golden Record. There were also changes to the Solution Architecture when the client required files to be either ingested from different sources or transferred to the new specified locations. The system we designed tracked changes on a weekly basis so that the deliverable documents could be kept updated and the final documents circulated at the end of the project.
Overall, our approach was highly successful and enabled us to complete the project within brief, budget, time and to the required quality standard whilst minimising risks.
It must be noted, however, that while this approach worked very well for this particular project, it won’t be suited for every project. Various factors need to be considered when adopting a micro-deliverables project management approach including but not limited to, the extent of the scope, timelines for delivery and the size and skills levels of the team.
That said, we will be looking to improve and extend this approach to other MDM implementations, as appropriate. This will enable us to provide even greater value to our clients, which will ultimately lead to greater cost efficiencies and gains.
About Vernon Naidoo
Vernon Naidoo is a Senior Project Management Consultant at Info.Blueprint with over 25 years of experience, including 15 years specialising in IT. His expertise spans a wide range of IT initiatives, including application development, digital transformation, data migration, technology implementation, data quality, and master data management. Vernon is dedicated to advancing project management methodologies and systems to drive efficiency and foster innovation. This commitment has earned him a CSIR Innovation Team Award and an All Africa Public Sector Innovation Award (AAPSIA). He holds an MSc in Chemical Engineering (cum laude) from the University of KwaZulu-Natal (UKZN).
Taking a micro-deliverables approach to MDM implementations
Written by Vernon Naidoo, Senior Consultant at Info.Blueprint
Master Data Management (MDM) implementations are inherently complex due to the multiple (often legacy) data systems that are involved. These implementations require deep alignment across business and IT functions, as different departments often have conflicting definitions, ownership, and quality standards for core data entities like customers or products.
Integrating disparate source systems, resolving data inconsistencies, and establishing golden record rules introduces technical and governance challenges. Additionally, implementing effective data stewardship processes and ensuring user adoption demands strong change management and clear communication. These complexities mean that MDM projects must be approached holistically and with a clear view of the end state goal.
In 2024, Info.Blueprint implemented an MDM solution for a large JSE-listed company. We used technology that is known for its ability to unify and govern data across diverse legacy and operating systems, ensuring consistency and reliability. By integrating this particular platform into its data architecture, our client was able to transform how they handle their critical data, enabling more agile and informed decision-making.
Our MDM implementation involved configuring the platform to manage master data, data integration into and out of the intelligent data hub, and enforcing governance policies across various data sources. Two data domains were selected for the first iteration – Customer and Product.
Key to this implementation was the use of a micro-deliverables project management approach. It is common knowledge that whenever new technologies are implemented in large organisations, “teething” issues are to be expected. Info.Blueprint decided to minimise the potential risk of implementation delays and cost overruns that were not scope extension related by taking a micro-deliverables approach.
The rationale for developing and adopting this approach was that we quickly realised that the design, build and testing was very dynamic in nature. This meant there was no way to upfront design and build the system without adopting an iterative process.
For example, in developing the data quality cleanup rules, we recognised that while the rules looked ‘good to go’ on paper, after building and testing on actual data, we always found tweaking was needed.With this in mind, we needed a methodology that could help the teams from both organisations cope while managing the various components of the project.
We split the work into two major parts, Analysis & Design and Build & Test – summarised below:
Our methodology involved breaking down the project into the smallest level of micro-deliverables. This provided us with the following benefits:
A key consideration in taking a micro-deliverables approach was the need for an effective document classification system for the project as all deliverables had to be documented as artefacts.
We adopted the following naming convention: Client – Technical deliverable/Project management deliverable – Report identifier (eg Build) – Sequential number (eg Build05)
We had the following categories of micro-deliverables:
Additionally, we developed a tracking system to update the micro-deliverables when changes were made after the document was submitted for review and accepted by client.
The nature of the business means that there were always changes to the design of the DQ Rules or Match, Merge and Survivorship rules to produce the Golden Record. There were also changes to the Solution Architecture when the client required files to be either ingested from different sources or transferred to the new specified locations. The system we designed tracked changes on a weekly basis so that the deliverable documents could be kept updated and the final documents circulated at the end of the project.
Overall, our approach was highly successful and enabled us to complete the project within brief, budget, time and to the required quality standard whilst minimising risks.
It must be noted, however, that while this approach worked very well for this particular project, it won’t be suited for every project. Various factors need to be considered when adopting a micro-deliverables project management approach including but not limited to, the extent of the scope, timelines for delivery and the size and skills levels of the team.
That said, we will be looking to improve and extend this approach to other MDM implementations, as appropriate. This will enable us to provide even greater value to our clients, which will ultimately lead to greater cost efficiencies and gains.
About Vernon Naidoo
Vernon Naidoo is a Senior Project Management Consultant at Info.Blueprint with over 25 years of experience, including 15 years specialising in IT. His expertise spans a wide range of IT initiatives, including application development, digital transformation, data migration, technology implementation, data quality, and master data management. Vernon is dedicated to advancing project management methodologies and systems to drive efficiency and foster innovation. This commitment has earned him a CSIR Innovation Team Award and an All Africa Public Sector Innovation Award (AAPSIA). He holds an MSc in Chemical Engineering (cum laude) from the University of KwaZulu-Natal (UKZN).
Categories