Leveraging Information and Intelligence

In this Infosys article entitled "A Retail perspective on MDM," the importance of MDM in the retail vertical is highlighted.

"Retailers are increasingly paying closer attention to the customer data to analyze consumer behavior. Consumer behavior is not the only reason to start maintaining enterprise data in a better way, master data management is an approach of holistically identifying enterprise wide key entities and maintaining them in centralized manner."

1. Items

2. Customers

3. Suppliers

Items, is the most complex notion within retail. Indeed, with the number of channels and supplying relationships, "the same item can be called with many different names depending on which channel it is being sold or which supplier it is being sourced from."

Customers are another issue. Retailers often have a very high amount of customer data, and most have no idea how to mine that information for the BI goodies they need to make their business better.

Suppliers, create the challenge of numbers and dynamic nature. Typically, there are "as many as 40,000 active suppliers," and the number and type of suppliers is continuously shifting.

What I found most interesting about this vertical use case is that similar problems are occurring within the other verticals as well, including health, finance, and manufacturing. What's core here is that there should be a certain amount of abstraction that occurs within each entity, from the least descriptive to the most. I would consider the use of ontologies here, but traditional MDM best practices would work as well.

The most important MDM concept is that you create common entities, such as items, customers, and suppliers, and learn to create general descriptions around those entities, and then bind core data to those entities. While this is easier said than done, most organizations will find that this will provide common guidance for defining data within new and existing systems.