The MDM Institute held its annual MDM & Data Governance Summit in New York City last month. Judging from that event, MDM is well past its infancy and getting into late adolescence. It's grown up but has more growing to do; it's somewhat mature but has a ways to go; it doesn't need handholding but still requires guidance.
One indication of MDM's maturity is that discussions about "type of MDM" were sparingly few--an indication that the MDM community is past the initial confusion. (That's not to say that MDM style doesn't matter, it does. It remains one of the most, if not the most, important architectural decisions you'll make in the early stages of your MDM program.)
More Insights
Webcasts
- Why is Information Governance So Important for Modern Analytics?
- The Critical Importance of High Performance Data Integration for Big Data Analytics
White Papers
- DevOps for a New Millennium: A Lifecycle Perspective Supporting Business Growth in an Altered Economy
- Agile for Dummies
Reports
More >>A presentation at the conference on unstructured MDM shows that we're comfortable with the core concept of MDM, and ready to extend it to areas like unstructured master data. However, my advice is to let it rest until you've mastered the management of structured master data.
It certainly makes sense to keep the two initiatives aligned since unstructured MDM uses structured master data as metadata. In other words, structured and unstructured master data are bridged by metadata, so you'll need to have a good handle on your metadata management practices as a pre-requisite to unstructured MDM.
How about using your CRM product as your customer (master) hub? Don't do it, was one presenter's response: CRM is about business functions and transactions, whereas MDM is about hierarchies and uniqueness, and never the twain shall meet.
[ Want more on enterprise data management? See 10 Tenets Of Enterprise Data Management.]
How important is data security, and how do you secure your data? Different industries have different terms for sensitive data--personal health information (PHI), personal or protected private information (PPI), etc. Chances are you'll need to maintain some confidentiality around sensitive information. Role-based security and relying on legal and contractual obligations were suggested, but the responses lacked specifics or innovation.
Incidentally, vendors are enticing us with software-as-a-service MDM solutions, but it doesn't look like customer data is moving to the cloud any time soon. Security is the reason, of course. Making the case for placing your crown jewels—and corporate skeletons—in someone else's custody is tantamount to putting your job on the line, and vendors aren't providing much help here. As one attendee aptly said: Don't just sell us cloud solutions, help us get through the regulatory and compliance morass that comes in the way.
Data Governance
One speaker tried to distinguish his services company by touting its approach to "inside-the-system" data governance: "We do that, like all others, of course, but we really like to focus on outside-the-system governance". I offer them the "Consulting speak of the Day" award: Data governance is about information management processes, which may or may not use systems like MDM (or any systems, for that matter). There is no "inside-outside" to data governance.








