Select Page
9 August 2022

Master Data Management

The continuing importance of master data management

Simply put, master data management helps you manage the company’s master data. Find out everything you need to know about master data management, from its qualities to its limitations.

Reference data

Reference data, or master data, change little over time. Reference data is the structuring information for the company: the list of product data, the list of customers, the data on employees, etc. You will rarely have to modify it. The reference data may be distributed in various places depending on your company’s organization. Customer data, for example, can be found in the company’s ERP, CRM, and accounting management solution, or in an Excel file that Fred from Sales made by hand but which is not accessible to other teams.

Master data management defined

Master data management (MDM) is the set of techniques and processes that ensure the quality of master data. It is about applying data governance to a group of data essential to the company. The reference data must be:

  • Qualitative
  • Error-free
  • Familiar to all employees in the company

With an MDM solution, you can describe the master data, define its value, and reference it in a single source. If someone is looking for master data, they’ll know where to find it. MDM will answer the question: Is the customer list in the ERP, CRM, or Fred’s Excel file?

Master data management also includes other services: cleaning up data, applying updates, adding data descriptions, etc.

Benefits & limitations of master data management

Master data management: The advantages

Master Data Management brings many benefits to your company, including a standard view of master data. It helps you significantly reduce the risk of errors and duplication and ensures better control over the life cycle of your master data. Controlling your company’s data means meeting today’s data knowledge challenges.

You can more easily implement data-related regulations, such as GDPR. Beneficial for large companies that use a lot of IT applications, Master Data Management makes it possible to simplify the IT architecture, thus reducing operating costs. Centralization is essential for setting up a data repository accessible to all the company’s employees.

Limitations of master data management

The main difficulty is change management. It is impossible for business experts to keep their old habits, especially Fred, who will have to abandon his personal Excel file. MDM is about defining the steps to find and use reference data. Consensus can be difficult to achieve when sensitive information is involved. Adopt an agile and pragmatic policy to avoid seeing your Master Data Management project fall by the wayside. Start small and work your way up. The company’s employees will gradually adopt the new processes.

What style of master data management is best for your company?

Master Data Management is divided into four main styles: Registry, Consolidated, Coexistent, and Transactional. Each style offers distinct data control. You can manage your data from a central information system or manually synchronize the different sources.

 #1: Registry

Of the four main styles of Master Data Management, Registry is the simplest to implement and use. It uses stubs, which are records that indicate the source and storage location of the data.

It is inexpensive and does not transmit changes made to data to other computer applications (basically, no synchronization). The risk is that inconsistencies between different types of data may appear. The Registry transfers the data through the Master Data Management tools, but the data never returns to the source for adjustment. It ensures that master data is put into secondary repositories.

#2: Consolidated

The Consolidated style follows the principles of the Registry, but instead of sending master data to secondary repositories, it sends it to the enterprise’s central data library, such as the Data Catalog. Consolidated should be adopted if your IT systems experience high latency: data transfers can be scheduled to occur in clusters.
If reference data is sent to the central data library, it does not automatically synchronize with remote sources. So you have to deal with integrating the data by hand.

#3: Coexistent

An improved version of Consolidated, the Coexistent style synchronizes the reference data with the data source. The reference stubs are identical in the different repositories of the company (ERP, CRM, central data library…). As the data transfer system is more advanced, the synchronization is slower. This is the best master data management for small and medium-sized companies, which can update the master data and synchronize it to the remote data several times over a given interval.

#4: Transactional

If you want to adopt an absolute master data management strategy, turn to the Transactional style. Although more expensive than the others, it has all the advantages, especially regarding data quality. Data is transferred from the source to the central repository. Once ready, it returns to the source (without leaving the central repository). Then, the data is processed from A to Z: cleaning, standardization…

With the Transactional style, you can:

  • Implement your data governance strategy;
  • Guarantee the quality of reference data;
  • Ensure data sharing between different departments of the company;
  • Reduce latency during data transfers.

The limitation of the Transactional style is that you need specific software to code the master data transfer paths.


To conclude, you can start with a rudimentary master data management and then move towards more efficiency and complexity. Every organization needs to manage its master data properly. MDM can be an effective solution that helps you meet your company’s specific requirements and organize your data to increase efficiency and save time.

To learn how DataGalaxy can help data professionals improve efficiency and drive data management excellence using DataGalaxy’s Data Knowledge Catalog, request a demo or sign up for your 15-day free trial!

Structuring a data-driven organization

Other articles