Master data management is an established industry, but there’s still a lot of movement and innovation, which continues to surprise me. Just last month, for instance, Informatica arranged a stock buyout of the Germany-based Heiler Software MDM.
Of course, a lot of this change is happening because vendors are moving toward multi-domain MDM. While vendors don’t like to talk about it, the fact is MDM tends to come from one of two heritages: customer data integration or product information management.
Generally speaking, the existing software tends to favor one of those traditions, but the end goal for vendors is multi-domain MDM, which you could use to master whatever you want.
Then, too, a lot of the innovations focus on social media and the cloud; although, as I shared last week, there’s also interest in marrying MDM and Big Data solutions, especially Hadoop.
So I guess we shouldn’t be surprised to learn there’s now an entirely new MDM solution on the market. Pitney Bowes Software is offering its own MDM hub with its Spectrum data management suite.
But here’s something that I did find surprising: It’s built on a NoSQL database, according to Andy Hayler, CEO of the Information Difference.
What’s that mean? Well, other vendors use a relational database as the platform for MDM, Hayler explains, but Pitney Bowes’ MDM solution uses a graph database.
Graph databases are great for picturing and exploring relationships between data, even in Big Data, which means you can do things like look at the links between customers or check behavior patterns.
That could be particularly useful as organizations link MDM to unstructured data, particularly social media data.
Still, you should get your structured data mastered before you start adding unstructured information, according to Arvind Singh, co-founder and CEO of the enterprise data solutions provider, Utopia.
Singh shared what he sees as the seven best practices for master data management with Information week recently. You probably won’t be surprised to learn that his first piece of advice was to state the business problem you’re trying to solve.
Actually, the first three best practices all focus on being very clear about the business goals and ROI of what you want to accomplish. Given how master data management projects tend to be IT projects rather than business projects, I think three best practices devoted to avoiding that sounds about right.
For more on MDM best practices, you might also want to check out the TDWI’s whitepaper on Next Generation Master Data Management, which includes a list of best practices for multi-domain MDM.