How MDM Fits with Big Data, Mobile & Cloud

More and more, organizations are asking how MDM fits in with other emerging trends – namely big data, mobile and cloud.  We’ll discuss how these trends intersect.

Big Data - For today's post, big data means the overall umbrella of social media data, unstructured documents, streaming data from instrumented devices, and more.

Master data is really just a subset of big data. I like to think of it as the subset of big data that organizations already have their arms around. This subset tends to be already in structured format, reasonably trustworthy, and shared and common across different lines of business or departments.

When clients start discussing big data projects, they usually want to start making better use of the all their data, beyond just the core master data.  As organizations plan these projects, it is crucial that they leverage their MDM hub as a foundation for big data.

  • The MDM hub can already tell you who your customers and prospects are, so use that knowledge to more efficiently sift through the rest of your big data to find more about those same customers and prospects.
  • The MDM hub is where you already keep the most complete view of your customers, products, accounts, and more.  As you uncover more information about those same entities, the MDM hub is the logical place to keep those new insights.
  • The MDM hub can let you have your cake and eat it too when it comes to trusting your data.  The MDM hub can keep a traditional 'golden record' of trusted information side-by-side with a less-trusted view of the same person or product based on what you find among your big data.  These two views can be combined to provide a more insightful complete view, but they can still be presented separately in cases where your business can't afford to base decisions on the less-trusted view.

That last point is key: when you combine the traditional 'golden record' with new information found among your big data, the superset of information can power even better business insights and business decisions that were not possible before. The MDM hub is the foundation.

Mobile - Mobile devices like smartphones and tablets are omnipresent touchpoints where end users have new and different ways to interact with master data.  They can create master data, put it into the hands of those who need it, and they can greatly extend the network of people who can help steward or govern the master data.

To me, it’s even more interesting that mobile devices are location-aware.  Every time you create or update a piece of master data on a mobile device, you add in new information about the geolocation of the transaction or user.  Those geolocation tags are brand new pieces of information that can and should be used with your master data. You can change the way you market to customers, stock your shelves, service your accounts, and more.

Cloud - In my experience, when somebody talks about MDM and 'Cloud,' they usually mean one of two things:

“Can MDM be deployed in a cloud infrastructure?”
“How does MDM leverage cloud-based applications as consumers or contributors of master data?”

Assuming you are comfortable with your master data living outside your firewall (which is a big assumption), the short answer to the first question is, 'Yes, MDM can be easily deployed on cloud infrastructure.”

The second question comes up more often for me. Most organizations have a standard for how they move data in batch between their on-premise applications. Most organizations also have a standard for how they move data in real time between their on-premise applications. However, most organizations do NOT have a standard approach or toolkit for moving data between on-premise and cloud-based applications.

In IBM's case, we have a data integration solution called Cast Iron that is purpose-built for bridging the connectivity gap to a whole host of cloud-based applications, so that the MDM hub can integrate just as easily (sometimes more easily) to cloud-based applications as it can to on-premise applications.


This post was originally published on August 14, 2012 on the Mastering Data Management Blog.