If you don’t have to implement master data management (MDM), then don’t. That’s the surprising advice given by Forrester MDM and data expert, Michele Goetz.
I’ll be honest, I can’t recall anyone having said that previously. In fact, the general assumption, from vendors to analysts to authors, has been that if you have master data, you need MDM.
If you’re unfamiliar, MDM is a discipline and a technology that sets in a separate layer from your data storage and applications. As a discipline, MDM requires you to establish rules about things like which data to overwrite and which to accept as the “golden copy.” That’s the role of MDM: to establish a trusted version of your master data, to which other systems can defer.
As a technology, it’s either a hub or a registry that manages all of the MDM rules and data.
It seems so obvious what the benefits would be, and that’s lead to many adopting it automatically as a best practice. And that’s the problem, according to Goetz: It’s not a best practice. MDM is expensive, an ROI can be hard to pinpoint, and MDM can take months or years to implement. On her Forrester blog, Goetz writes:
“What concerns me is that MDM is still misunderstood. Even as companies take the recommendations to implement it, they don't really know what they have. They want to load data into the hub, standardize the view, then push the data. Huh, sounds like ETL. Thus, how do you justify the value?”
That’s why it’s “not a check box in your data strategy,” she states.
She’s not the only one who thinks organizations are making serious mistakes with MDM. Gartner analyst Andrew White has written several blog posts on common misconceptions about MDM.
Judging from the number of question marks he uses (six), he must be particularly aggrieved about organizations using MDM hubs to “master” data that is, in fact, not master data at all.
Part of the problem is that there actually is no good way to govern some of this data, e.g., unstructured data, White explains in another Gartner blog post. But that doesn’t mean because you have an MDM “hammer,” everything is master data.
In the meantime, if you are considering MDM, check out Goetz’s post, which also appeared recently on Information Management. She describes specific situations in which MDM is appropriate. Goetz also includes a list of what you should know about an MDM tool before you commit to it.