The Importance of Identity Resolution to MDM
By Dan Power, President and Founder, Hub Solution Designs
In a recent post on the Hub Solution Designs blog, I wrote about the importance of integration to Master Data Management (MDM). Today, I’d like to dig into the importance of identity resolution to MDM.
In a previous life, I worked for Dun & Bradstreet, and I had a chance to learn quite a bit about D&B’s technology for identity resolution, which it calls Entity Matching.
But the general question of identity resolution, or matching, is critical to a successful MDM initiative. There’s a good article by Roger Wolter and Kirk Haselden of Microsoft which says “the matching accuracy of MDM tools is one of the most important purchase criteria.”
I agree with that. If your goal is to build an MDM solution, with a central hub, integrated data quality, robust middleware, and useful external content, all driven by a data governance organization composed of experts from the business and from IT, then you’re going to need highly accurate matching tools.
The consequences of false matches (showing one customer record as a duplicate of another when that isn’t the case) or missed matches (not grouping two customer records together when they should be) will strongly affect the level of trust your MDM solution receives from the organization as a whole.
And depending on your industry, too many false positive or false negative matches could have serious business impacts, and in a medical or law enforcement MDM application, could even have life and death consequences.
The crucial question is “are the built-in matching tools provided by your MDM hub vendor good enough, or are you better off integrating a purpose-built identity resolution engine with your MDM hub?”
There’s one good way to answer that question - do a “bake off” or side-by-side comparison of the MDM hub with the dedicated matching tool, and see if the difference in results warrants integrating a matching tool with your MDM solution.
I know the first time that the business users of your new MDM solution find records grouped together that shouldn’t be, or find a large number of duplicates that to the human eye are obviously the same, their level of trust in your MDM hub will drop, and it may be very difficult to get it back again.
Dan Power is president of Hub Solution Designs, Inc., a consulting firm specializing in master data management and data governance. He has 20 years of experience in management consulting, enterprise applications and strategic alliances at companies such as D&B, Deloitte & Touche, Computer Sciences Corporation, eCredit and Parson Consulting. He speaks frequently at technology conferences, has written several articles for DM Review magazine, and regularly advises clients on developing and implementing high impact MDM strategies.
Subscribe to this blog