By using website you agree to our use of cookies as described in our cookie policy. Learn More


Master Data Management Enthuses, Confuses Business Intelligence, Data Management Pros

No one seems to agree on just what Master Data Management is—and many DM pros still aren’t familiar with MDM as a technology vision, either.

If the tendentiousness which prevailed at the most recent TDWI conference is any indication, master data management (MDM) could soon amount to a huge—and apparently misunderstood—concern for many organizations.

There’s high-level confusion, for one thing. This is best reflected in a lack of consensus about just what MDM is, says Dave Wells, director of education at TDWI. “We ran six days of MDM content [at TDWI conferences] and heard about eight different definitions. I think this area will generate as much controversy and confusion as the Inmon [versus] Kimball noise [in the data warehousing space] a few years back.” DM vendors seem to have drastically different understandings of MDM, for one thing, while analysts and customers are similarly confused.

On top of this, many business intelligence (BI) and data management (DM) pros aren’t yet familiar with MDM as a technology vision, much less as a DM practice. “MDM is something that quite frankly I haven't really heard too much of. I'd seen the acronym thrown around a few times in [the] recent past in some of the BI magazines and websites I frequent, but it’s something that I felt was so far away and an idea that was so far from finished that I didn't pay it much attention,” comments Topher Thiessen, a manager of metrics and analytics at a subsidiary of a prominent lending and mortgage firm.

Ditto for Jerome Poudou, an R&D director with Merkurium, a Montreal, Canada-based BI ISV. He says that all that he knows about MDM he learned from a quick Google search. “My first impression is [that] it’s an SAP-like approach or a Biztalk-like approach [to application information integration]. SAP does the [same] job to insure a common definition of the objects—and more—[while] Biztalk integrates multiple systems and allows them to communicate easily.”

So what is MDM? It’s a good question. At a very high level, master data management describes a process of reconciliation and consistency: it’s about ensuring standard and consistent metadata definitions across all units or lines-of-business in a given organization. Perhaps most (business) units already have standard definitions for terms such as “supplier” or “product” or (even) “customer”, but—given the likelihood of M&A activity in large organizations, as well as the heterogeneity of data sources and applications—there’s bound to be renegade metadata somewhere in the wild.

That’s really the essence of MDM’s strategic value-add, proponents argue: MDM practices can underpin and enable a number of pipedream-like technology scenarios, including enterprise-wide reporting initiatives (in which companies create information hubs to consolidate data from operational systems and other disparate data sources) or—similarly—data synchronization efforts, in which operational data is synchronized between and among a master data hub and local systems, in this case, to support reporting, compliance, or other objectives.

Nevertheless, some DM pros suggest that MDM might be less a revolution than a wrinkle—i.e., it’s an innovation that could and should be incorporated into DM as it’s now practiced. “I think MDM tries to solve an operational-related issue and not a BI issue. In the BI world, the common definition of an object is here, so the MDM idea” doesn't provide anything new, argues Poudou. “From my experience, I think MDM is here already, but at a higher level. MDM says that enterprises have problems in their [metadata] definitions, but today we talk about cross-company integration, so cross-company MDM. This is an operational issue, not a BI one.”

At his own company, which develops and markets BI software for educational institutions, the problems addressed by MDM simply aren’t operative, Poudou says. “[T]he problem is not at the operational level but at the data warehouse usage level. Even if you have a good common vocabulary, the challenge is to use the right data at the right time to provide the right answer to the right question. A mistake in a question can result in a bad decision—and that’s the challenge we have today,” he concludes.

BI manager Thiesen concurs. “My initial thoughts on the matter are [that] … MDM will just fall into the DW platforms of today that are becoming increasingly flexible in how they store and access data,” he comments. “I'm sure we'll see each vendor adapt their own slightly different version of MDM, tout its capabilities over their competitors, and flaunt some major wins. At then end of the day it’s just part of the evolution of the DW into a more robust platform.”

This is the first in a series on MDM. In the coming weeks, we’ll speak with prominent industry watchers, vendors, and—of course—rand-and-file BI and DM pros, as well. Stay tuned.

About the Author

Stephen Swoyer is a technology writer with 20 years of experience. His writing has focused on business intelligence, data warehousing, and analytics for almost 15 years. Swoyer has an abiding interest in tech, but he’s particularly intrigued by the thorny people and process problems technology vendors never, ever want to talk about. You can contact him at [email protected].

TDWI Membership

Get immediate access to training discounts, video library, research, and more.

Find the right level of Membership for you.