An Evolutionary Approach to Master Data Management
A little more than two years ago, Claudia Imhoff and Colin White wrote their first research paper on master data management (MDM). At that time, it was a fairly new concept, and the report was one of
the more in-depth studies of the new initiative. Since that time, hundreds of articles, presentations, blogs, white papers and case studies have been created and published. The discipline has become
more mature and the technology supporting it more sophisticated. Unfortunately, in the intervening time, more confusion surrounding master data implementations has also cropped up, leading to
conflicting and disjointed implementation advice.
The authors have taken a fresh look at where MDM is today with this second edition of the research report, pushing the ideas and ideals further than ever before. The goal is to establish a solid
baseline of understanding for MDM initiatives, the technological environment needed to fully implement a global store of master data, reasonable benefits that can be expected and how you can evolve
your MDM initiatives to move toward a fully integrated and functioning MDM environment.
The authors define MDM as:
A set of disciplines, technologies, applications, policies and procedures used to manage, harmonize and govern the master data associated with an organization’s main business
A fully functioning MDM environment can overcome many of the challenges faced by businesses today – poor data governance processes, redundant and inaccurate master data, inefficient business
processes, lost business opportunities, etc. Some of the benefits we found in our research include a reduction in risk, easier auditability of master data, lowered administration costs and improved
collaboration across the enterprise. In addition, the system of record, system of entry and system of reference for master data are all clearly defined and conflict-free.
A mature MDM environment includes three major technological components: master data applications for maintaining and managing master data in a global master data store, master data services for
use by master data applications and data management services that interoperate with the master data services. These, in turn, support three types of MDM processing: operational MDM, collaborative MDM
and analytical MDM. The report closes with practical advice on getting started with your MDM initiative. There are several starting points for an MDM project. Some organizations begin by gradually
upgrading operational applications to directly access a global master data store of the MDM environment. Other companies evolve a data quality management project into a full MDM initiative. Many
companies start MDM projects by extending a data warehousing and business intelligence (BI) environment to support master data requirements. Ultimately, the strategic goal is to move toward an
enterprise MDM environment whose purpose is an integrated MDM environment that supports all forms of MDM.
Read the entire study.