Recently Joe McKendrick asked this question, "Is having SOA and Master Data Management at the same time a form of overkill?" . He argues that in fact they compliment each quite nicely:
SOA in and of itself holds little value to an organization unless it provides the capability to open up information to the enterprise. As is the case with SOA, successful MDM is a silo-breaker, invoking collaboration across the enterprise. MDM helps assure that the information populating SOA-based services is accurate, timely, and consistent.
While I agree that SOA needs to "open up information" to the enterprise to be successful, I'm not sure that I would say that MDM is the only mechanism to enable consistent, accurate, timely and accessible information-based services. There are multiple approaches to moving, managing and integrating data including ELT, Data Quality, Data Governance, Data Services and Data Federation (see a great post from Alex Kotopoulis), just to name a few. In fact, many of these mechanisms provide immediate "silo-breaking" value. And there is still disagreement on how these elements are used or not-used within MDM. I'll save that post for another day.
But if we go back to MDM as a broader definition for a moment - is it true that SOA and MDM are feasible to be put together? Does it mean we're exposing master data as "master data services" for re-use by a business process?
Regardless of your architectural recipe, SOA and MDM do have a place together. Provided that you undergo a bit of political wrestling to ensure its effectiveness (according to a post by David Linthicum) then will you start seeing the value out of SOA and MDM combined.
