Part 5 in our series on building a data Management program from scratch in Kazakhstan

Introduction

We continue our project here on Hub Designs Magazine, called DMFS – Data Management From Scratch. Any data governance initiative needs a functional framework to structure its activities and responsibilities at the top level, so this field report addresses the Data Management Framework we developed in the last year, as part of a reference business model.

Gizat Makhanov contributed sufficiently to developing this framework that I asked him to write about our approach and results. We also decided that our readers should be able to download a data management chart, which we hope you find useful. Our holding companies use this framework and chart as a transformation guide while they are developing new processes and organizational structures.

Yours sincerely, Dmitrii Kovalchuk

Data Management Framework

At the beginning of the transformation project, it was clear that we needed a good data management framework to document processes and, in general, manage data properly. Although many accept the universal data management principles, the detailed approaches to managing data are not so obvious. We therefore decided to research several methodologies and frameworks on managing technology and data and adapt them to our case. While some of the approaches were very important to consider, some were of little help as they were too conceptual.

The primary reference methodology was the Data Management Body of Knowledge (DMBOK), developed by the Data Management Association (DAMA). DMBOK structures data management in ten logical functions connected to data governance. Each function consists of further activities. This circular representation of processes gives a good explanation about various functions, but we needed a more hierarchical and grouped view.

We also decided to create a continuous flow of activities in each process and delegate technical activities to the IT function. Before going further into details on how we modified DMBOK, we need to talk about COBIT 5.

COBIT 5 is a business framework developed by ISACA (Information Systems Audit and Control Association) for governance and management of IT. It perfectly bridges the gap between IT and business, by communicating business requirements starting at the top level with the board of directors down to the operational level with IT support and users. COBIT 5 consists of five process domains, where each one addresses different business and technical goals:

• Governing
• Planning and Organizing
• Building and Implementing
• Supporting
• Monitoring

We decided to create our representation of data management processes like these groupings in COBIT 5. Just like in COBIT 5, governance is the core process in DMBOK that is critical to the success of a data management initiative. Some activities it consists of are:

• Identifying strategic business requirements
• Developing data strategy
• Creating policies and procedures
• Managing data managers and stewards

The data governance process therefore became the roof of our data management framework.

We further grouped four processes vital for building and monitoring the data management infrastructure under one domain. While data architecture management and data development are two functions of DMBOK, data management monitoring and data lifecycle management are functions we developed, based on processes in COBIT 5.

The four essential building processes are:

• Data architecture management
• Data development
• Data management monitoring
• Data lifecycle management

The third domain consists of four processes specific to four data types:

• Master data management
• Document and content management
• Metadata management
• Business intelligence management

Each of these processes has a logically set flow of activities, starting with defining requirements to running day-to-day operations.

The described data management processes are designed using the same tool as other business processes, like finance, procurement, and IT. This creates an environment to integrate processes seamlessly using either of the following methods:

• Integrating using interfaces to related processes
• Assigning activities to each role of a process

One of the critical success factors in implementing a data management culture is the creation of data stewards in each business unit. Our plan was to assign specific activities to the stewards and they would further integrate with their business units. This approach simplifies complex details of the data management framework and opens door to some creativity in the way the stewards work.

Data Management Chart

As a last note, the main success factor in implementing the developed processes is communication.

Data managers and stewards need constant reminders of what proper data management is and the ways they can manage data effectively. One of such messages was to explain the processes in DMBOK. This is why we created the Data Management Chart that stewards can hang in their office. You can download it here.

Next Month’s Report

Our next article will cover metadata management. We’re going to share with you our practical experience on how to start managing metadata. You can read about that in the next DMFS Field Report – coming soon!

About the Author

Gizat Makhanov is an enterprise data manager on the Core Transformation Team.

His responsibilities include adopting and implementing data governance and data management processes, creating methodologies for data governance and data management using best practices and standards, and guiding portfolio companies to establishing effective data management culture.

He earned a Bachelor of Computer Information Systems from the University of the Fraser Valley in Abbotsford, BC, Canada. Gizat enjoys programming, mixed martial arts and studying urban street design.

Filed under: Architecture Frameworks, Best Practices, Data Governance, Data Quality, Enterprise Architecture, Master Data Management, MDM, Methodology, Politics, Strategy Tagged: Best Practices, data governance, data integration, data management, Data Quality, enterprise architecture, featured, Master Data Management, MDM, Reference Architecture