Master data management: a best practice guide

10 August 2015 Karan Vidal

Master data management (MDM) is the process of identifying and managing the most important information that a company must hold a single copy of.

The company’s master data is information about customers, partners, suppliers, products and other critical records stored on duplicate systems across the organisation. This data is vital to support different business processes.

Compliance requirements and efficiency drives fuel the need to have accurate MDM. Handling master data can be seen as the one of the responsibilities of working with big data.

Why MDM is chief

Sorting out master data should be a top priority as it gives a true picture of the company’s records. Without MDM, there’s a risk that companies can have inconsistent and duplicate data. HR’s master data may be different from Finance’s for the same customer, leading to confusion and a waste of time.

MDM can truly only be the boss of all data if it achieves these objectives:

  • A distinct and accurate view of the data.
  • Show all interactions and relationships between all elements of the data.

MDM in action

Ever wondered how to get started on integrating only the master data from hundreds of systems into something that’s useful? This is exactly what SAP did for Merck, one of the largest pharmaceutical and chemical companies in the world. It has 39,000 employees based in 66 countries, with an annual revenue of €66bn. Needless to say that organising data for such a huge organisation would present a challenge.

Merck’s ‘M3’ initiative set out to coordinate master data to enable uniform processes across the company. It used SAP’s master data software and methodology to achieve its goals of MDM. Merck also enlisted the services of Camelot Management Consultants to provide the change management aspect of the MDM.

A road map was developed to match the master data to the relevant business processes. The MDM implementation was a four-pronged approach:

  • Technical.
  • Workflows.
  • Data quality.
  • Roll out.

After implementation, controls needed to be in place to maintain the integrity of the data. After all, it would be a waste of money and time if the master data was perfectly classified and anyone could make changes to it willy-nilly. Reports could be produced that showed who made changes to the data. Given the strict regulations around the food and drug industry, this feature was a must-have.

you can download the free PDF if you want an in-depth look at how Merck implemented its MDM.

The MDM Wheel

The following is the basic cycle of master data governance:

  • Definition of the scope of data quality.
  • Measurement of data quality.
  • Assessment of the business impact.
  • Analysis of the root cause.
  • Plan to tackle data quality.
  • Carry out data quality plan.

MDM best practices

Given that successful MDM involves data from the whole organisation, it’s vital to have a systematic and structured approach to implementation.

Resistance might result from departments who are happy with their record-keeping and will oppose attempts to ‘touch my data’. As well as the MDM wheel, ensure these principles are in place to give implementation the best chance of success.

1.I have a vision

Having the foresight to understand the importance of MDM isn’t enough. You’ll need to communicate this to all the captains of the departmental ships.

The suggestion that you want to take away the control of one of their most precious assets might not go down very well. To make change management easier, get the highest person in the organisation involved in the implementation.

2.State your case

The business case details what needs to be done and why. It also needs to include measures of success and return on investment. These success criteria will be your life raft when soliciting support from departments.

3.Let’s get down to business

Data management deals with systems, so it’s easy to ‘palm-off’ MDM on IT. This thinking will ensure that your MDM implementation is doomed. The business has to be in the driving seat as they will be the ones using the data. This approach will lead to less teething problems down the line.

4.Power to the project team

‘Messing’ with the way people have always done things is never an easy task. Solid project and change management practices must be in place. If there’s resistance to MDM implementation, slap-dash project management can be the excuse that’s required to derail progress.

Make sure that your project approach and documentation is water-tight so there’ll be no sinking of the MDM ship.

5.We’re better together

You’ll be working across all business units to achieve your goal. You need to get everyone to work together for MDM success. Change management is crucial for agreement on a useful solution.

Depending on resources, it may be a good idea to get external consultants to manage this part of the process. As in Merck’s example they used a consultancy firm to handle the change aspect of MDM.

6.Take baby steps


Rome wasn’t built in a day and neither will the MDM of a big organisation. Aim for small and consistent steps to lessen the chance of things going awry. Focus on getting the processes right for one sub-set of master data.

Complete that process and move on. It’s far easier to fix one data sub-set than all of the data. Piecemeal implementation gives a better overview of what will happen after the roll-out is scaled.

7.There’s no end in sight

The sigh of relief that accompanies the end of a project should be absent in when dealing with MDM. The implementation phase may be over but (in the words of Karen Carpenter) “We’ve only just begun”. The purpose is to keep the data clean and integrate it into the way of working. To achieve this, the following must be in place:

  • Business rules.
  • Data control.
  • Compliance structure.
  • Data stewardship.

8.Lean on your vendor

There are many considerations when choosing a vendor for your MDM implementation. Who you choose will depend on your business needs. The following will help you to get the most out of the software:

  • Understand how your vendors’ model integrates with your software.
  • Keep them involved in every stage of the implementation.
  • Avoid making too many customisations as this will prevent issues with upgrades.
  • Stay in the loop with vendor’s patches to benefit from upgrades and fixes.
  • Test everything, every time. Don’t just leave this in the hands of your vendors as you know your systems better than anyone else.

MDM is a must-have for big organisations. Although potentially painful to implement, the benefits of compliance adherence, better decision-making and efficiency may not leave organisations with a choice.