“That which we call a rose, by any other name would smell as sweet…”
We take it for granted. When we call a friend in France, we automatically add the internationally recognised dialling code of 33. An entire orchestra is tuned to an A at 440 Hertz frequency. Even nature has its golden ratio for reference when designing the arrangement of petals in a flower.
There are many bits of reference information that are internationally recognised and in common usage – grocery bar codes, medical conditions, industry classifications and postcodes to name but a few. Some reference data can also be specific to just one sector, where it can support a unique business process or describe the configuration of an individual application.
Reference data is generally not very volatile in nature but, despite this, any changes can have significant impacts on IT systems and business operations. Consider, for example, the widespread implications of a country or tax rate code changing. Yet despite the repercussions of such a change, many organisations have only limited business controls or governance on reference data’s definition, usage and quality.
Because it rarely changes, issues with reference data governance often become apparent at times of transition, such as when integrating or consolidating systems. Problems also occur during the production of analytical reports, when reference data from multiple sources is in differing formats. Issues can also occur when there’s a need to identify a person who has been using information for a purpose for which it was not designed. The wider point is that transition projects can be hindered by non-coherent descriptions of how data is being defined or who is using that data. This confusion can also add extra costs to each project because once the conflicting reference data has been located, a potentially an onerous task in itself, the quality of the information must then be reviewed, often manually.
Reference data management solutions provide support to the implementation of governance best practices. They provide places to document the data’s purpose, control its definition and distribution, and describe its relationship with other types of reference data.
Consider a company’s commercial organisation. It might have regional offices along with assets and employees in numerous locations and with a variety of roles and functions. A reference data management solution can define the types of business processes and reference the job roles required to perform each one. These roles are then able to have locations and employees assigned to them.
For a second example, let’s take a healthcare institution that needs to provide regulated product labelling. The reference data describing medication types and illnesses are already internationally recognised. A reference data management solution would be able to load existing sources of trusted information and format/transcode it in a way that is suitable for all healthcare systems to interpret. In this way, data references are actively governed but the existing systems do not need to be changed to accommodate new standards.
Finally, we’ll look at a bank that is implementing a governance strategy for financial services reference data. Since the data comes in multiple forms across many business units, it will have disparate systems, processes and people (data stewards) managing it. Reference data is required to support both the definition of master data (party, account, product, user, etc) and transactional data (trade order, service event, contact, etc). Regulatory compliance initiatives such as BCBS2391 require the support of reference data to successfully aggregate risk data and support reporting requirements.
The importance of managing reference data at an enterprise level is becoming increasingly important as 3rd platform initiatives tend to push information back into silos. Integrating those silos to support new business functions requires a deeper understanding of how the underlying reference is defined.
Perhaps if reference data management systems had been available to Mr. Shakespeare at that time, he might have changed his mind about what he understood a rose to be… or not to be.
 Principles for effective risk data aggregation and risk reporting. https://en.wikipedia.org/wiki/BCBS_239