Back in 2011 Gartner, the analyst firm, predicted that these three things would shape the Master Data Management (MDM) market:
- Multi-Domain MDM
- MDM in the Cloud
- MDM and Social Networks
The third point was in 2012, after the raise of big data, rephrased to MDM and Big Data as reported in the post called The Big MDM Trend.
In my experience all these three themes are still valid with slowly but steadily uptake.
But, have any new trends showed up in the past years?
In a 2015 post called “Master Data Management Merger Tardis and The Future of MDM” Ramon Chen of Reltio puts forward some new possibilities to be discussed, among those Machine Learning & Cognitive computing. I agree with Ramon on this theme, though these have been topics around in general for decades without really breaking through. But we need more of this in MDM for sure.
My own favourite MDM trend is a shift from focussing on internally captured master data to collaboration with external business partners as explained in the post MDM 3.0 Musings.
In that quest, I am looking forward to my next speaking session, which will be in Helsinki, Finland on the 8th December. There is an interview on that with yours truly available on the Talentum Master Data Management 2015 site.
Nice blog again. The future of MDM is mainly impacted by the following areas:
1. The age of relationships: leveraging and visualizing the power of relationships
2. The age of bimodel governance in a world where data is decomcratic exercise
3. Master data fueled apps for specific use cases, industries and roles
4. Hybrid IT: Running business critical data on hybrid environments across cloud, on prem and hosted.
Happy to chat more.
I agree that the most interesting area recently has been the collaboration with external business partners, particularly companies selling data. There are a lot more of them now! And accuracy, as we all know from our internal data sources can be a challenge. While different sources can have different specialties and quality associated with their own sources, they will have different probabilities that they will be correct overall. Where the tricky problem exists is trying to apply this data to a particular individual. As my statistics professor used to say: On the whole, you can calculate the probability of the outcome. But that does not help you know if your lottery ticket is the winner!
Thanks a lot Ben and Gino for elaborating on this everlasting subject.
To Ben’s four points:
1: Relationships are indeed often forgotten in our quest to get the entities right. In my previous role at an end MDM user site I fought a yet to be won fight (mainly with IT) for taking master data relationships seriously.
2: Bimodal data governance is a hot topic and something close to me whether you call it “top down and bottom up” or “centrally governed and self-governed” or bimodal or something else. There is a balance to strike. I will blog about that soon.
3: Fueled apps is a good term. I like MDM aware applications maybe better. Not at least when looking at multi-brand landscapes 🙂
4: Your right. It is better to talk about hybrid instead of cloud vs on-premise.
To Gino’s point about more data sellers out there, I will add available open sources and public sector sources. In my home country, Denmark, the government has realised that offering (almost) free big reference data is a plus for the society as a whole.