pressmaster - Fotolia

Data governance and your master data management strategy

Strong data governance and master data management strategies typically go hand in hand. Read on to see how key factors of data governance can support your master data management.

The value of data to an organization may be hard to pin a dollar value on, but it has become increasingly clear that data is an important business asset. Given this, it makes sense to manage data as effectively as possible, which is what data governance and master data management do.

Traditionally, data governance had been considered the remit of the IT department, but more recently it has become clear that business leadership needs to take a lead on the management of data with support from IT resources. This is partly because data in most organizations is a mess, with multiple competing and overlapping versions of data about customers, products, suppliers and assets strewn across multiple applications.

One global company I worked with years ago decided to radically reduce the number of brands the company ran, focusing only on the most profitable ones. The CEO was disappointed to discover the systems in the company were unable to determine the true profitability of any individual brand, due to inconsistent allocations of costs, among other reasons. This was not a rare exception: A KPMG survey in 2018 found that just one-third of CEOs trust the accuracy of their data.

This sorry state of affairs is partly due to the amount of overlap in corporate data. An Information Difference study some years ago found that the average organization has six competing sets of master data for customers and nine different sets for products. Five years later a repeat of the survey showed the situation had not improved.

What is master data management?

Master data such as customer, product, asset and location data are embedded in a number of applications in a typical enterprise, only one of which is its ERP system. The field of master data management (MDM) was developed in order to try to get a handle on this.

The idea behind a master data management strategy was to either copy data out from other systems into a trusted master data store that could then be used by other systems or to map where it was stored and document the differences -- e.g., a product might be classified in a simple hierarchy by one department or business unit but a more detailed way by another.

For example, a marketing department might care about the brand of a product, its packaging and whether it is on special offer, but a logistics department cares about the number of products in a palette, its dimensions and weight and where to deliver it. These different needs drive different categorizations, which in turn become embedded in different computer systems.

Why is MDM important?

A successful master data management strategy most likely has a business-driven imperative and operates within an established data governance program.

Some MDM projects run as large as $100 million in scale, but these projects often have an excellent ROI despite their considerable cost. On one project I worked on, a single pricing error on one product that had remained hidden until the master data management strategy came along paid for the entire cost of the regional MDM project rollout.

Can MDM help data governance?

A data governance strategy needs to work hand in hand with a master data management strategy. Many early MDM projects failed because they failed to engage with business leaders, and the IT department simply doesn't have the authority to tell the business to change its current processes in most organizations. Data governance and master data management working together can accomplish much in the way of data accuracy and data trust across business divisions.

A successful data governance program will have senior businesspeople working alongside the IT department and will assign data stewards across various departments. A good program will have dispute resolution and scalation mechanisms to address the inevitable disagreements about whose customer hierarchy is better and who has to change to the new standard.

Analysis of an Information Difference data governance benchmark database showed that successful data governance programs, statistically speaking, had a high correlation with a number of common traits. Among those were:

  • a clear and documented process for resolving disputes;
  • documented business processes;
  • regular data quality audits; and
  • data models for key business data domains.

The key message is that data governance and master data management need to be aligned. A master data management strategy in the absence of a solid data governance program is likely to encounter difficulties, because the heart of such a project is to agree on a common golden copy of key data across an enterprise.

This synchronicity can only happen with businesspeople buying into the program and being prepared to change their current processes. The technology matters, but it is less important than the degree of business buy-in. The latter can be facilitated by a well-reasoned, quantified business case showing ROI projections.

A well-motivated, business-led project with a decent MDM tool has more chance of succeeding than one with a state-of-the-art MDM product with every feature under the sun where business leaders are resistant to change. By contrast, successful MDM projects that operate with the support of a solid data governance organization can achieve considerable ROI.

Dig Deeper on Data governance

Business Analytics
Content Management