Master Data Management - An Imperative!
A typical enterprise stores transactional data which is processed with reference to master data. A purchase order is made to a supplier, a sales order is created for a customer and products are manufactured and sold. Master data is undoubtedly the key business entity which defines an enterprise. In fact, master data sets the context for transactional data in an IT system.
Master data is used in multiple systems that are interlinked in an enterprise, which has a legacy of homegrown systems and packaged applications. A typical problem in this scenario is that master data is not common to these systems. On the contrary, the same set of customers, suppliers, products and employees is identified differently in these disparate systems which thwarts the fusion of data from all these systems. In a nutshell, MDM is a journey from unmanaged master data to managed master data. The offshoot of not having managed master data is a lack of integration among systems in the true sense. When data from multiple systems can be brought together and analyzed to get a holistic picture of performance, master data plays a critical role in enabling that integration. Not having managed master data dilutes the benefits of an integrated enterprise system and can lead to diminished performance. On the other hand, an MDM program can improve customer service, reduce inefficiencies and help spot opportunities for raising employee productivity.
There are different strategies to manage master data. It must be borne in mind that it is a process and not a destination. Having an infrastructure in line with the chosen strategy contributes to that process but importantly, a master data governance model is fundamental to set the rules for creation, consumption, modification and deletion of such data. Technology can support but not solve the problem in its entirety. While on this topic, it is worth noting that improving data quality not only supports a MDM program but is also sustained by it - a symbiotic relationship at best!
In my next post in this series, I will write about MDM in general with a greater emphasis on typical scenarios which expose the realities and consequences of unmanaged master data.
Comments
A good insight. In my view, a key to creating an enterprise with solid MDM is having an enterprise wide data model and clear identification of data ownership.
Doing these solves a good deal of MDM woes.
Posted by: Somnath | March 4, 2009 1:21 PM
A very important topic, as ERP systems have grown in usage. Modules with master data have been common. MDM as an approach is gaining popularity to separate modules that need the data from the modules that creates them in a centralized manner.
Posted by: Venu | March 9, 2009 7:09 PM
It is true that CRM has given an offshoot to MDM, where the inefficiencies have been addressed. But, with MDM itself maturing and various vendors now looking at multi-entity hubs and hub of hubs implementations, the time now has come where we need to look on how MDM will evolve over a period of time to sustain. In fact, my blogs on MDM staying put and MDM being able to sustain speak more about this.
The social imperative the 2.0 version is something that will be a game changer over a period of time to harness the potential of the same.
Posted by: venkataraman.r | September 24, 2009 1:50 PM