In my previous post, I suggested that if we were to better articulate how master data management (MDM) is typically used, we could develop the components of solution templates that can speed the integration process. In this post, we’ll start to look at some common ways that the capabilities that comprise an MDM system are used.

First, a quick review: fundamentally, MDM provides a means for enabling enterprisewide visibility into a shared representation of unified information about uniquely identified entities. A common example enables customer data visibility to information culled and referenced from across a number of source systems.

That being said, once the information about the specific managed data domains (such as customer or product) is created, how is MDM used? Here is a sampling of common usage scenarios:

• Identity search and verification of membership – A candidate entity’s identifying attributes are used to search an existing registry or index of known entities to determine if that candidate has been registered within the system. An example is checking to see if a person creating a new customer account on your web site has already registered as a customer.
• Identity search and access – A candidate entity’s identifying attributes are used to search an existing registry or index of known entities and if the candidate has been registered, return the reference to the sources for the entity. An example is a call center representative searching for all information about a caller’s previous interactions with the company.
• Identity registration – If a candidate entity is not found within the system, create a new reference profile and register the candidate’s identifying attributes. An example is creating a new customer account for an individual who has never been registered in the past.
• Identity merge – Merge two unique identity records into a single identity record. An example might be when two records with different customer names that share the same telephone number, email address and birth date are determined to represent the same individual.
• Identity split – Split one identity record into two when it is determined that a single record inaccurately represents multiple individuals. An example is determining that a single record represents a pair of twins because their names are similar and they share a birth date.
• Data set cross-reference – Find all records from a query data set that have a corresponding match in the master data set. An example is a batch check of a collection of applications for customer eligibility for a new promotion to determine if any have already been given the promotion.
• Cross-reference update – Update the master data profiles for all records matched from a query data set. An example is a health insurance company updating information for a collection of member records within a single health insurance group.
• Duplicate elimination – Find all duplicate records in a data set and schedule selected duplicate records for elimination. An example involves identifying when customers have already created multiple online accounts so that they can merged together into a single unified customer account.

Typical business use cases leverage these capabilities. For example, a targeted marketing campaign often begins with identifying a market segment with an expectation of a level of data quality that is rooted in customer identifiability and uniqueness.

An offer that is presented to the same customer multiple times may skew the analytical results – either the customer responds multiple times (and is counted multiple times as well) or ignores the offer multiple times. Neither option is desirable from an effectiveness perspective. However, adoption means more than recognizing the value proposition. The MDM program manager must make it easy for new customers to use the master data system, and that is the subject of my next post.