April 28, 2020
The synergy of MDM and DG solutions
In this article, we will explain how different business key data management concepts can be used together, and why their combination has a greater effect than its separate elements.

Firstly, it is important to understand that the concepts of Master data management (MDM) and Data Governance (DG) imply differences in the functions performed:

  • The MDM concept describes the Executive function. In other words, by using MDM solutions, you can organize the practical implementation of data management: quality assurance, standardization, validation, etc.
  • The DG concept describes the Supervisory function, as in you can use DG solutions to create data management requirements and monitor their implementation: form business terms, link business concepts to the technical implementation of the IT landscape, define requirements for maintaining MDM, and so on.

In other words, if MDM is primarily a set of tools for managing master data, then DG is a set of tools for working on data management policies.

However, both MDM and DG solutions are designed to manage core data, helping businesses achieve a global goal: a Mature, optimized management system.
What does MDM include?
Using an exclusively MDM-class solution, you can clean, enrich, validate, and consolidate data to produce reference records that will be used in multiple business information systems at once. All this is achieved using standard MDM tools that are configured at the solution implementation stage.

This is not difficult at the initial levels of data management maturity when the primary tasks are obtaining reference data and consolidate the practice of working with data in a new scenario. MDM performs well in individual projects and medium-sized organizations (independent business or part of a holding company).

However, if a business uses multiple MDM systems, or if data management is affected by external regulators whose requirements must be met, configuring MDM tools becomes noticeably more complex. There may be errors or additional discrepancies in understanding the correctness of the reference data. Because of this, for example, duplicate search rules or data quality rules may differ, which will affect both the data itself and different reports.

In addition, MDM systems in different enterprises of the same holding may initially have different data models, which will be difficult to compare or bring to a single standard. And differences in the understanding of terminology or rules further aggravate the situation.
What does DG include?
With the problems described above, it is not possible to solve strategic data management tasks, so there is a need for data governance. Implementation of the DG system allows you to:

  • Create a single top-level data model;
  • Evaluate data as assets and manage these assets;
  • Define a common understanding of business terms;
  • Manage all available data sources;
  • Create data management policies/regulations;
  • Organize data management processes;
  • Report a unified reporting system;
  • Solve other tasks related to data management (additional tools can be implemented in separate DG systems).

DG class solutions are suitable for large businesses, for example, for widespread implementation in holding companies. The DG system allows you to move to the next levels of data management maturity, when all the necessary tasks and projects will be implemented on time, efficiently, predictably, and within the budget, bringing a much greater positive economic effect.
Combining MDM and DG
Now let's look at the common ground between MDM and DG.

Point 1. The top-level data model in the DG system is created based on the business IT landscape when all available data sources are combined. Sources are:

  • Databases.
  • Information systems.
  • Datastreams.
  • Reports.
  • Data models of MDM systems.

The DG system must parse the structure from the database, other systems, streams, or reports to create suitable data models and link them together. As a rule, ready-made models of MDM systems do not need to be parsed, which significantly simplifies both the process of the initial loading and updating when changes are made to the data source.

Point 2. The data model of an MDM system is much easier to evaluate from the asset's point of view than with other types of data sources. In addition, the data model already combines several sources that are unified and contain reference data, i.e. it is an agglomerate.

Point 3. Business terms, data management policies, and regulatory requirements are easier to translate to performers through MDM systems since MDM contains all the necessary tools for data management.

Thus, even if the DG class solutions are independent and can solve all the necessary tasks, but the integrated use of DG and MDM solutions gives an amplifying effect that affects both the integration process of new solutions and the data management itself.
Author
Daniko I.