As master data management is widely adopted, effective MDM project management is also in wide demand. Professionals come to the MDM space from two distinct backgrounds, both of which can lead to gaps in their effectiveness. The generalizations that follow will be used for example only.

From one camp, there are people who arrive with a vast information management background. They understand data modeling, data integration, data profiling, data quality and related concepts necessary to achieve that elusive single version of the truth in an enterprise. Their blind spot can be the lack of an operational focus. Data warehouses, the epitome of information management to date, allow for many delays and timing errors due to their batch nature. MDM, which is operational, often needs to serve up data instantaneously to a variety of operational and analytical environments. This often occurs when data integration and quality work in concert through an MDM hub. The nature of the data also must be consistent. Data warehouses updated at 2 a.m. with no immediate users have a high degree of latitude when it comes to complete data consistency. MDM does not.

A further gap in the profile for this group is a lack of practice in actually forming data for the enterprise. More than 90 percent of data in the warehouse is not originated there, and the analytical data that exists in the post-operational environment is often derived from elements that originated elsewhere. MDM provides organizations vastly improved abilities to originate master data. Input portals served to customers, suppliers and various business departments, supported by a physical hub, often start the flow of data into the organization. Workflows can pick up the data and, along with data governance and robust process flows, ensure that the end result is a master data record ready for distribution. Such data conception and process flows require a different mindset for MDM.

The other camp of resources for MDM projects delivers MDM leaders who arrive with a background in the operational systems that either generate or need master data. This profile is larger than the analytical information profile for an MDM leader. While this group understands real-time operational issues, they are often unaware of the science and art to data profiling, quality and integration. Over-influence from this operational mindset can cause struggles with integration programs and ultimately, utilization of inferior data in MDM programs.

The Journey to Effective MDM Project Management

The first goal on the journey to effective MDM project management may be a self assessment along these lines and making sure that your gaps are filled by conscious education, team members and/or governance and stewardship input. One of the first steps in the project management journey should be to set up your data governance group. This group will perform several important functions to the MDM project.

Setting up data governance is not always the strong spot for technically focused managers, but it must be done.

The next goal is the initial requirements gathering, understanding the sources and targets of the data and determining where MDM capabilities such as workflow and hierarchy management are necessary. Gaining control of the data models and access to the data is essential to achieving this goal. The project plan can be drafted, but the eventual data profiling will have serious impact on the time frames.

The requirements process really cannot be shortchanged. The outcome needs to be data governance and stewardship roles, historical data requirements, conceptual hierarchies and workflows, a target architecture, definitions of the involved subject areas, and complete metadata on source and target systems. The elegant, final architecture is something that will be approached over time, so all of the above should be mapped onto the calendar and adjusted for any external dependencies. For example, in one MDM implementation, we must be ready with the master data for a new point-of-sale system by a certain date.

These requirements are contributed to and ultimately vetted by data governance. This phase is often best done by a small, experienced, agile team of the project leader and lead analysts, as opposed to the full cast of the development team. Speaking of the development team size, you cannot do the iteration plan without understanding the target team profile. Will the team be lean, mean and agile but conservative since the team size is small? Or will it be larger with a forceful agenda and a willingness to accept a lack of agility (and the potential of running faster than the company can absorb)?

Choose your poison, but choose excellence. Excellent MDM project leaders give you the most of all worlds: lean (small budget), agile and productive, able to work business and technology issues, and a strong sense of information management and operational focus.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access