I have been participating in a series of events sponsored by DataFlux on strategies for long-term success for enterprise master data management projects. We are about halfway through the series, and so far I have noticed two common threads among the questions posed by the attendees. The first thread involves justifying the value of MDM knowing that there is significant upfront effort that might not lead to the commonly-noted benefits. The second is about herding the business managers together to have them discuss (and hopefully agree) about the impacts of replicated records and inconsistent semantics.
Filed under: Business Impacts, Data Quality, Master Data, Metadata
A few months back I shared a post about proper scoping of a master data management activity to focus specifically on a smaller subset of business activities that (a) are impacted by the absence of a unified view of data or (b) can be measurably improved through the facilitation of a unified view of data. Part of the actualization of that unified view involves selecting the right business activities and then getting a better understanding of the data needs of those business activities.
If you believe that there is business value in creating a unified view of a particular data concept and therefore create a master data asset for it, what would the first step be? I suggest that you would need to propose the creation of the master data domain to a data governance committee, who would evaluate the costs and benefits prior to initiating the sequence of tasks needed to materialize a master repository.
I will be the invited speaker at an exciting upcoming web seminar on Feb 10, 2011 at 2:00PM EST with Composite Software on “Data Quality’s New Best Practice – Data Virtualization.” Bob Eve, who is an Executive Vice President at Composite Software will join me as well. I hope you can make it! Click here to register.
I had two interesting briefings this week. One was from a company called Orchestra Networks providing a tool for model-driven master data management, and the other was a company called Collibra, which provides a model-driven tools for capturing metadata and semantics.