Six Steps to MDM Success
Steps you should take before embarking on a master data management initiative.
Say what you will about Gartner - and many of you do - those people don't shy away without putting up a fight, even when you aren't sure there's actually a fight to be had.
Take, for example, this recent post by Andrew White, a research VP and agenda manager for MDM and analytics at Gartner.
"Blogs would be boring if they didn't incite a little riot now and then," he wrote Wednesday. "So here's my effort."
Everybody chant with me: "Fight! Fight! Fight!"
The survey found - among other things - that nearly 70 percent of its respondents said MDM and data governance were mutually supportive. That basically confirms what experts have said all along, which is you can't just clean up data and call it master data management. MDM is a long-term commitment that requires governance.
So what's got White up in arms?
Let me give you a head's up: You may think he's picking at nits - but you should expect that when you deal with data people, because it's what makes them good at what they do. And, in their defense, nits seem to matter a lot more in data - particularly master data -because a small problem ignored once can have long-term repercussions.
But back to the particular nits he's picking on. White takes the survey and article to task on the following points:
1. The title, "Data Governance and MDM More Successful Together." It's just wrong to even separate these two disciplines like that, he contends. "MDM includes governance - it is about governing (for better business value) master data. So you cannot actually do' MDM without a governance component. arguably data governance' is very broad and not very specific, and MDM is data governance applied to master data,'" he writes.
2. The first sentence, which calls MDM a technology. You notice in the paragraph above I said MDM is a discipline? That's so the analysts won't come after me next. White's not the only one who sees red over this issue - many data people will hunt you down and beat you for saying MDM is a technology.
Remember that English teacher who became very angry if you used "whom" wrong? Well, that's how analyst are about calling MDM a technology. So let's be very clear: It's like saying a budget is an Excel spreadsheet. You may put your budget into an Excel spreadsheet, but buying Excel won't "do" budgeting for you. In the same way, MDM can be supported by technology, but if it's going to work, then the technology can only be a tool, not the entire program.
3. He suspects an underlying problem with this and any survey on MDM is that people are confusing MDM with "yet another data integration effort to tackle the same problem!" he writes. Translation: They're going to fail, because they're doing the same old thing. And as an extension of that, White takes issue with calling MDM a "project."
"Well, MDM is not a project! MDM is a program, that will spawn off and launch all manner of projects, that span and include DQ work, DI work, deployment of a hub, data modeling, and the list goes on," he writes. "None of these individually are' MDM. But when organized in a specific way, oriented around a discipline to change the way in which the business users crate, use and abuse their own data, and then MDM persists."
Yes, it's true: Analyst fights are like watching the guys from "The Big Bang Theory" duke it out: It's all 25-cent words, no blood and not nearly as satisfying as a brawl between the Bruins and Canadiens.
But don't think this is about semantics. Picking the right words is just part of trying to get his real point across in a way that just isn't happening - as the survey shows.
White's main point is that people don't understand master data management in a fundamental way. If they did, they'd understand that governance isn't just something that makes it better; it's critical to the process. Heck, it is the process.
You can start to appreciate his point when you consider that two-thirds of the respondents were from really big companies, with annual revenue of over $1 billion. MDM isn't cheap; there's no doubt they've spent a lot of money on it. And yet, 19 percent called their MDM program unsuccessful while 54 percent described it as "at least moderately successful." Are you starting to see how this could become a major problem? Given the costs and effort required for MDM, 54 percent should be the starting point for "at least freakin' ecstatic."
Then look at the results for data governance: Only 10 percent described their data governance as very successful. That's a really low rate. Roughly 50 percent described these programs as "moderately successful."
So, yes, this is more than a question of linguistics. Maybe part of the problem was the wording of the survey, but given the responses, it seems more likely it's just an indication of the larger, real problem, which is how little MDM is understood by those who apparently think they're doing it.