MDM: With Great Promise, Comes Great Work


There's no doubt about it-CIOs would like to get information under control. The conflicting numbers resolved, the misspelled names corrected, the duplicate records gone-it's like the Promised Land of IT, flowing with the milk and honey of pure data.


Just as a crazy "for instance," wouldn't it be great if next year's calculation of customers didn't include footnotes such as "+/- half a million, depending on whether numbers are from marketing, sales or customer service" and "assuming the 10,000 listings of John Q. Public are NOT the same person?"


The results of an Accenture survey on Information Management suggest that, yes, CIOs think that would be great. In fact, an impressive percentage of you dare to dream: Seventy-five percent of organizations want a "comprehensive, enterprise-wide information management" in three years time. Good luck with that. And 98 percent of you want, at a minimum, an organized MDM (master data management) and sharing program in place for key data.


Ah, dreamy MDM. From cutting costs to easing mergers and acquisitions, and even boosting SOA's success-is there nothing it can't do?


From where I stand, MDM looks a lot like ERP in the early part of the decade. There are a lot of promises about MDM's potential - and they may be true - but it's also very expensive and, by all accounts, much harder than the vendors would have you believe.Those who know even a fraction of IT history will see where this is headed. But it doesn't have to end in disillusionment and backlash-at least, not for you.


Recently, I've found two points all companies should know about MDM. You'll notice I'm not saying there are only two - I'm just saying I've recently found two that I think are particularly critical.


First, as Contraste Europe consultant Tod McKenna recently points out on his "Tod Means Fox" blog, MDM is a not a technology, which is to say, you can't just buy it:


"Remember that MDM is a capability and not a technology. You cannot buy MDM, but you can build a MDM strategy. This strategy will likely cross several technologies and platforms. ... Vendors will continue to push their MDM solutions, but be careful not to trap yourself into thinking that the job is done once you've installed. Vendors can wrap most technologies necessary for MDM into a single package, but they cannot provide you with a strategy or the personnel to make it work for your organization."


I love his word choice there-capability, as in something that's possible, but not certain. As in, you may be capable of becoming a plumber, but you can't just buy the plumbing tools and open a plumbing business. You also have to learn how to do it, through an apprenticeship, and pass an exam. Too bad there are no journeyman exams for MDM.


Second, MDM can solve data quality and integration problems-but done incorrectly, it can also create silos. And it's tricky to get right, as a recent Information Management article, "The Wide-Ranging Effects of MDM on BI Systems," shows.


The article is written by David Templeton, a self-employed consultant. To be honest, it's dry and difficult to read for non-techies, but I think it's an essential read for IT leaders contemplating MDM.


Templeton explains that MDM changes the collection and integration points of other systems-particularly the business intelligence layer. How you handle this can make or break MDM:


"When MDM comes to an organization for the first time, a lot of assumptions about what it is and its effects. In fact, the effects of MDM on an existing architecture can be broad and deep. By knowing the high-level impacts of MDM on an existing business transaction/BI architecture, you can preempt misunderstandings that, down the road, affect your MDM implementation and the business value and cost of improving the quality of your master data."


As I read it, Templeton is saying MDM can help you with data integration, but you've got to make sure your system and application integration is spot on.