In my first blog post I touched on a variety of Master Data Management (MDM) use cases outside of the traditional customer data or customer information integration scenarios. This month I would like to explore one of those use cases – reference data management -- a bit more deeply.
To reiterate, reference data is information such as country codes, state codes and gender identification, which is usually static and helps to categorize other data within code tables. Code tables are typically defined and managed on an application-by-application basis, meaning the values and codes for a specific reference data set differ within each application.
Reference data may be the best place to start on your MDM journey. Organizations sometimes struggle with getting started on MDM projects as they have difficulty agreeing on the definition of a customer or prospect across business units and systems. With reference data, it becomes more black and white. Reference data is often the easiest place to find common ground across an entire group and managing it correctly can reduce costs and risks and improve business efficiency.
One of the challenges with reference data is that loading information into the warehouse or an MDM hub typically requires reconciling reference data from multiple sources. Another interesting distinction is that the business drivers for managing reference data can vary by industry, for example:
- Banking: In Europe, different European Union countries can choose different versions of NACE industry classification code as their national standard. Banks with operations in multiple countries must manage the individual national NACE code sets and reconcile the differences across countries. Banks in all countries face the dual challenge of distributing reference data changes to downstream applications and assessing what applications are impacted by a reference data change.
- Healthcare: In the US, the government has mandated that healthcare payers and providers must switch to use ICD 10 codes for diagnoses and procedures by October 2013 (although the government has proposed a one year delay). There is not a one-to-one relationship between the ICD-9 and ICD-10 codes.
Strong governance, security, auditing and process control around the authoring, changing and usage of reference data is imperative. The business benefits organizations can derive from centralized management of reference data are:
- Reduced business risk – there is a reduced chance of error associated with manual, ad-hoc processes such as someone loading the wrong version of a reference data spreadsheet or file into a process.
- Improved efficiency for making changes to reference data and distributing to consuming applications.
- Lower support costs – there is a reduced burden on IT for managing reference data as well as a reduction in errors in integration jobs which leads to lower support costs.
Over the next few months I will explore the implications of the MDM use cases including product information management and threat and fraud analytics. Also, I invite you to leave comments and contact me with your thoughts. Some of our most interesting insights at IBM have come via our partnerships and conversations with organizations and individuals.