Don’t Forget Reference Data When You’re Managing Master Data

I was asked the other day how to embark on an initiative to manage master data and I recounted a story of how I introduced the concept of Master Data Management (MDM) to an Australian headquartered global organisation to demonstrate where to start.

This company had grown, and continues to grow, by acquisition as well as organic growth. They have many different businesses operating in one particular vertical and they’ve permitted a great deal of autonomy on how its various business units operated.

This autonomy of operation, rapid growth and acquisition had led to siloed systems and processes which in turn leads to problems with producing data that was consistent and controlled across the organisation. For example, having a definitive list of the business units, a list of employees, or an organisation-wide list of capital projects that the company was working on.

Even the home page of their corporate website had different descriptions of the businesses that the group operated, variously describing four, five or six businesses in the group. This situation is very common in large organisation and is often a driver for organisations to start an initiative to manage the master data.

When approaching this type of problem I split the types of data that need to be managed into two groups: Reference Data and Master Data.

Reference Data

Reference Data is simple lists of data used to control other data, for example, lists of countries, states, currencies or postcodes. Reference data most often only has one relatively simple source; such as an ISO standard and is relatively easy to define and control. This a great place to start, as good management of Master Data often requires good management of the reference data.

Master Data

Master Data on the other hand refers to entities that a company defines and manages such as a customer, a product or a service. Master Data is often harder to define and manage with multiple owners at different points in the lifecycle, different departments that use the data differently, multiple systems that affect different aspects and often complex hierarchies to ‘roll up’ or aggregate the data.

Following are two examples of where attributes of Master Data are made possible and more manageable with the use of Reference Data:

• The Master list of customers of an organisation needs an attribute of an address
  • Maintaining a list of postcodes or cities as Reference Data makes this much easier to manage than simply allowing free text

• The sales figures for a product need to be recorded against a currency
  • Using an industry standard list of currencies makes exchanging and converting this to a common currency much easier

At this global organisation starting with the reference data meant that staff could learn to manage the simpler reference data first, processes could be worked out, the wider community could understand the advantages and very little technology was required. We started with a few database tables and interfaces for systems to make use of these simple lists of reference data. This provided a platform over time to address dealing with the more complex concepts and additional effort needed to tackle the master data items.

When tackling the challenges of Master Data Management in your organisation, first consider managing the Reference Data as a small investment with good returns on which to base the next steps in managing the Master Data.

About the Author: Peter Hopwood is a Principal Consultant at Altis Consulting and has extensive hands-on experience managing, architecting and delivering Information Management, Data Warehouse and Business Intelligence solutions. Peter also manages a number of the training courses delivered by Altis consulting.
Peter has spoken recently the IBM Business Analytics Forum, The Australian Computer Society and the Australasian Association of Institutional Research (AAIR). Peter has been published in DM Review, and he has been interviewed in MIS Magazine.

No related posts.