I have worked on projects that today would be referred to as master data management (as well as customer data integration) for almost twenty years, so I have had some confidence in, over the years, referring to myself as an “MDM expert.” Yet the focus of most of the projects that I had worked on involved record linkage for the sake of data cleansing and resolution, as a way to identify records in different data sets that referred to the same individuals and then merging sets of records together into a single record, effectively combining the data into a single result. More to the point: the essence of the activity was integrating records together.
That being said, I recall one of my favorite SAS DataFlux IDEAS conference moments back in 2006. It was in Las Vegas, when Cathy Willot from Amgen described the process her company had gone through to employ MDM techniques for regulatory compliance. As a pharmaceutical company, Amgen is obliged to comply with numerous rules and regulations, and Cathy’s talk described how MDM and CDI techniques were used to increase productivity and reduce process complexity. She also noted something I had not really ever grasped before: the use of these data integration techniques for compliance with rules directing segregation of data.
As I recall, the specific challenge involved understanding certain (compliance-oriented) business processes where data records specifically could not be linked together. Similar to the concept of a “Chinese wall” in the financial industry (that separates individuals or groups and restricts the flow of information), her company needed to know where records in two different data sets and applications referred to the same individuals and therefore needed to remain segregated.
This was the first time I truly grokked the idea that MDM techniques could be used for more than just data integration and consolidation, and could be employed as tools to support different types of business intent. This has influenced our approach to considering master data management programs and in our consulting engagements, specifically in helping our customers differentiate between the adoption of MDM based on what the analyst/press suggests (that is the dreaded “single source of truth”) and sentient use of available technologies to help solve specific business challenges.