A recent TDWI survey found about half of the organizations questioned were still in the early lifecycle stages of master data management programs. Given that, you're probably thinking it's a wee bit premature to be talking about next-generation MDM.
But no. And, actually, talking about what happens next with MDM may help you avoid mistakes now, depending on where you are in the process.
Phil Russom, research director for data management at TDWI, is coming out with a 40-page best practices report, "Next Generation Master Data Management," next month. But he's already sharing parts of the report, including the 10 top priorities for next generation MDM.
I think as you read his list, you'll see he's actually talking about a lot of issues that may have been on your agenda all along. And if they're not on your agenda, they probably should be.
For instance, item number two on his list is "multi-department, multi-application MDM." Basically, this means if you played it safe and deployed MDM for a single application - be it ERP, CRM or BI - you're probably going to want to expand it beyond this one application. Likewise, if you only rolled out MDM to one department, you've probably already figured out for it to really work, it needs to be an enterprise-wide deployment.
Other examples of "next generation" capabilities you'd probably figure out anyway:
- Multi-data-domain MDM. Most organizations started with customer data, but soon learn they'll want to add other types of master data, such as products or financial master data. This could actually become a huge technology problem, if you didn't think ahead, because MDM solutions tend to have evolutionary roots in one or the other. That said, more vendors are moving toward multi-domain MDM.
- Bidirectional MDM. It's possible to pull out your reference data and put it into a master data hub that's only feeding one system and never update it - but that's not a good idea. Bidirectional MDM acts more as a central hub, constantly resolving data differences and making the data available to multiple applications.
But there are also some next-generation priorities that aren't so widely discussed, including:
- Real-time MDM
- Consolidating multiple MDM solutions (apparently, some of you made silos out of MDM, which is a bad idea for many reasons, not the least of which is it defeats the purpose by creating multiple versions of your master data!)
- Richer modeling
- Expanding MDM into the customer channel - which may include the Web or - dare we say it - social media!
- Workflow and process management integration - particularly, a way to automate the creation or change of data
- MDM solutions built on vendor tools and platforms - in other words, moving away from hand-coded MDM
That's what TDWI sees on the horizon for MDM. Most of these things are possible or within shooting distance already.
If you really want to look in the crystal ball to see what the distant future holds for MDM, you'll need to read Jim Harris' recent piece, "The Semantic Future of MDM." I have to say, it's a nice vision - where the actual person of record controls his or her own master data via a personal data locker and allows you to access it - or not - depending on need.
No more "digital clones" or wrong information. No more updating your file on a customer.
It's a bit over-the-rainbow, to be sure, but along the way, Harris points to some of the flaws inherent in MDM as it exists today, making it a relevant read.