Master Data Management - An easy explanation

In simple terms, what is Master Data Management? It actually is a really simple concept but when one starts researching it on the web, it is not always that easy to get to the simple answer. So, here goes...

As we all know, companies’ software and database implementations have become very modularized. If it is properly modularized, then we in the data world can be lucky. But it is common now to have companies roll out data solutions in an uncontrolled way, with the same data captured in various places and stored in various data stores. How to fix this problem? MDM! Let’s work through an typical example to illustrate...

A customer is captured in the sales order system as living in Beach Road 10. The order gets fulfilled. A provisioning system uses this data to do some other task. Then, billing comes along and find out that the address is incorrect. It should be Beach Road 11. Billing corrects their data in their own application and data source, but the rest of the data sources are now out of sync. What to do? The answer is to implement a central store of correct data. The customer is initially registered in here with the incorrect address. Provisioning pulls this incorrect address, and so does Billing. Billing corrects it, and the correction is also written to the MDM side.

This all leads to step two of the MDM implementation... Writing the corrected data back to the systems that also store it. Intricate, but this is the best we can do if data duplication cannot be avoided in a company.

See how this all leads to Data Quality being addressed too? DQ can now be applied to the MDM solution, and fed back to the source systems. And finally, one can even pull in third party data sources to clean and enhance the MDM solution.

It is as easy as that J  Implementation is tricky, but rewarding.