Management Data Master (MDM) – Back to the Future!
Let me be upfront, I am a self proclaimed “dinosaur” from the good old days of computing (YES, I was a COBOL programmer). But over the years having watched technology grow and change and morph, got me wondering if we were really going back in time to that golden era of IT. Let us consider the following scenario:
1980’s (& before for those older!): COBOL was the rage of the IT professionals. Strict coding guidelines, very strict data definitions, batch input of transactions with check digits and batch totals to ensure accuracy and consistency of data. Of course life was simpler then – single systems, simpler complexity of the business landscape. But then this was very strict meta data management and very controlled. We all understood the rule of GIGO – garbage in, garbage out!
1990’s: The age of Database management Systems – first networked and then relational. It was all about creating that Master DB, the holy grail of all us IT professionals. We could create data entity structures on screen, link them to other entities, create relationships and modify them on the fly. We could also create multiple instances of the same schema. But then while this was great from the point of creating relationships between data elements and maintained a fairly high degree of common data definitions, the spread of db instances across multiple systems caused the actual data to go out of sync, not to mention the fact that different definitions, representations and versions could exist for the same data entity across multiple instances.
2000’s: Enter the age of Microsoft and the most popular “Management Tool” – Microsoft Excel. Data was now maintained by their owners, good idea, but locally and most often in MS Excel – bad idea!! So the silo’s of data that relational databases created just multiplied and this time “under the desk”.
2010’s: With the data explosion of the 2000’s, much like the population explosion of a few decades ago, came the realisation that we have data everywhere but no idea of where it is let alone whether the quality is any good. This has created huge challenges as enterprises and departments consolidate and everyone tries to move to being a customer-centric organisation. In came vendors with MDM technology and best practices – controlled meta data centrally managed and shared; data managed not just for accuracy/ consistency in its’ own domain but in all related domains too. In came also the concept of “Conditional master data”. And most importantly the mantra of all consultants today is GIGO – garbage in, garbage out! Fix your data or all else is useless information.
So tell me – is it “Back to the Future” or am I missing something!