Six Steps to MDM Success
Steps you should take before embarking on a master data management initiative.
It's been nearly a decade since master data management came on the scene, and yet, according to recent assessments of MDM, it's still an immature discipline, both in terms of the tools and how it's used by companies. How is that possible?
Recently, David Littlewood, a senior consultant with the Australia-based Altis Consulting and 15-year information management veteran, assessed the maturity of MDM tools as compared to the vendors' own promises about what MDM would deliver.
"Now that Master Data Management (MDM) technology has been around for nine or ten years, what is the state of play in the MDM tool market? Are they relatively mature, similar to BI or Portal technology, for example? Or are they still somewhat immature?" writes Littlewood. "My contention would be the latter, unfortunately. In Gartner terms they are somewhere between the Peak of Inflated Expectations and the Trough of Disillusionment (my view not Gartner's, read about the Gartner Hype Cycle)."
For instance, when I started covering MDM about five years ago, IBM was talking about offering multi-domain MDM. But if you look at most deployments, they're still divided between customer and product master data, in large part because the solutions themselves are divided that way. In fact, in Gartner's 2010 Magic Quadrant assessment, Gartner refused to offer a magic quadrant for multi-domain MDM because it saw the offerings as still too immature.
Although Gartner expects multi-domain MDM to make some headway this year, Littlewood's skeptical. Despite consolidations in the space, niche solutions are still popping up like dandelions in spring, and the mega-vendors still haven't manage to show their mega-ness in MDM the same way they have in the BI market, he writes.
Maybe we shouldn't be surprised, though. As Littlewood notes, MDM was born out of ERP's failure. I assume the logic was since ERP wasn't bringing all the enterprise components together, maybe companies could at least pull out the master data and solve that piece?
This "looking back" approach also highlights another self-defeating aspect of today's MDM market: Many of the big vendors offer more than one MDM product, and not necessarily from the different niches of product or customer.
While Littlewood points out the immature nature of MDM tooling, Forrester contends that customers are actually become more mature at implementing MDM-but I have to say, I think their standard is too low.
"Master Data Management: Customer Maturity Takes a Great Leap Forward" is the name of the Forrester Report, but reading the Information Management article, it seems this conclusion is partly based on the fact that customers are asking fewer questions about the tools and more about best practices. It's impressive that Forrester reports a 90 percent increase in MDM-related inquiries since 2008, and it's nice that companies now know what MDM means. Obviously, there is some customer maturing going on-but it still seems a far cry from maturity.
As the article points out:
"Implementation is still somewhat in the early stages, as about one-third of requests on best practices related to getting started, and only 15 percent centered on improving existing programs. But as interest and planning mature, (Forrester MDM analyst Rob) Karel anticipates implementation, strategy and fine-tuning of data oversight to develop along the same lines over the next few years."
MDM may be on the rise, but it's still not what you would call common practice. Today, Kalido released a report, "The State of Data Governance Maturity 2011," which queried 200 companies and found that 49 percent still do not have any form of centralized master data management.
Another reason I think MDM is still immature as a discipline is this recent piece by William McKnight, president of McKnight Consulting. He writes that many MDM solutions are focused still on "simple" issues, and that he's "finally starting to see some pioneer MDM implementations explore the full range of benefits."
His post at the Bloor Group's Virtual Circle explores what these full range of benefits are and the breadth of business benefits they can bring. It's striking, though, that we're just now getting to the point where people are talking about secondary benefits that, in several instances, go beyond managing data to actually providing benefits to grow business. For instance, he talks about the value of having a leverage for syndicated data. Many organizations already understand the value of syndicated data, but how many have thought about actually offering their own syndicated data?
Organizations often struggle with syndicated data because it is expensive. It is also brought into a single-use system that lacks the ability for distribution to all the places that could use it and, consequently, the multiple departments that could contribute to paying for it. Master data management, with hooks into many enterprise systems, solves that issue. Many MDM programs have been launched to create an elegant place for distribution of syndicated data.
It'd be nice to hear more examples of how MDM can add value. As Abhi Beniwal, Daymon Worldwide senior IT director and one of Information Management's 25 Top Information Managers of 2011, recently said of Daymon's own MDM project, MDM should help "bring data to life" and solve tangible business problems. Even if it is still immature, it's had enough time to at least demonstrate that.