Intelligent Enterprise

A comment on the Gartner Magic Quadrant for Customer Data Hubs from our own Brain Food Blogger served, inadvertently, to highlight the confusion that prevails in the CDI/PIM/MDM space. That's particularly unfortunate, because one simple mathematical equation could set the confusion to rest.

First, here's the equation, in terms of set theory: {CDI, PIM} Subset {MDM}.

In other words, customer data integration and product information management are both subsets of master data management, one focused on customers and the other on products. Too bad, they don't have conforming acronyms (who would think that CDI and PIM are so closely related?)

Now, the (opening) comment from Brain Food Blogger: "One indication that the customer data integration (CDI) space isn't very mature is that there are only 11 vendors in Gartner's just-released Magic Quadrant for the technology."

Yes, indications are that the solution space has indeed yet to mature, but the number 11 is only half-true at best. The confusion arises, I think, from the Gartner report quoted.

Gartner has a Magic Quadrant for Customer Data Integration (CDI) and another for Product Information Management (PIM). Now, both CDI and PIM are essentially subsets of MDM yet, as far as I am aware, Gartner does not have a Magic Quadrant for MDM itself. This, in effect, partitions each of the MDM quadrants into two, one for CDI and one for PIM. Combined, there are 18 vendors in the MDM space. Of these, some focus on CDI (e.g. Initiate), some on PIM (e.g. Tibco), some on both CDI and PIM but separately (e.g. IBM), and others on relatively pure MDM ("bring on your master entity; we can handle it," e.g. Siperian).

If your head is beginning to spin, hang on... this gets worse. Over time, most vendors have realized that at some level, it's all MDM – and seem to be moving towards a "specialized generality" (or generalized specialty, if you like) by extending their entity-focused solution into a more generalized MDM solution.

How can we reduce the confusion? Here are some thoughts.

First, I suggest we settle on a conforming set of acronyms, such as CDM, PDM and MDM, or cMDM, pMDM and MDM (I know – "as if we didn't have enough acronyms going around already"). Remember, master data management is all about standardization and nomenclature.

Next, I suggest that Gartner should publish the Magic Quadrant for MDM as well. Would this include the cMDM and pMDM vendors? Well, any vendor that claims to have generic MDM capability should be included and evaluated (separately from the unique entity-specific evaluation). In addition, Gartner should publish the Magic Quadrants for all three simultaneously, so that we can refer all three together. This should help reduce confusion considerably.

Third, I suggest that vendors should take a close look at their offerings and avoid any subterfuge. It's all very well to claim generic MDM capabilities, but can we really use a highly specialized product data model for all kinds of master data? Probably not. If the vendor does not have an entity-specific data model, can they claim to have a specialized solution for that entity? Again, probably not. We need to be clear on what's out there.

Last, for customers of MDM-type solutions, use the above equation to determine the best-fit vendor. If your main requirement is for customer data, you are better served by going with a cMDM vendor, and if it's product data, a pMDM vendor. For more generic master data management, go to the more generic vendors.

Your opinion and views?