All of your data in order: The what and why of master data management

This article was originally published in Dutch by ITDaily. It has been translated into English below.

Master data management (MDM), often—unfairly—has the name of being a monolithic and challenging cost. The truth today is different: MDM is more accessible, pays for itself, and serves as a foundation for a data strategy in which artificial intelligence (AI) can play an important role.

AI is the hype of the moment, but there no AI without quality and accurate data. Master data management is the name for a strategy that provides consistency, accuracy and accountability for all corporate data, across all systems. Such a data strategy, with a single source of truth for all critical data, is essential for any organization with AI ambitions. Even without AI, MDM has numerous tangible benefits, especially in the case of acquisitions or divestitures:  

• Consistent and accurate data
  MDM ensures that data is not in silos, but all departments of the organization have access to the same up-to-date data, without inconsistencies and errors.
  
  
• Better decisions
  With reliable and unified data, companies can make informed decisions.
  
  
• Compliance and quality
  MDM provides a foundation for clear data governance so that compliance with rules and standards follows smoothly.
  
  
• Operational efficiency
  Getting rid of duplicate and conflicting data also simplifies data source management.
  
  
• More insight
  Thanks to master data, you get a total, complete and accurate picture of customers. You can then use that for more personalized service.
  
  
• More turnover
  Better insights and fewer errors lead to an increase in turnover.
  
  
• Data as an asset
  In an MDM strategy, all data have a responsible owner, and are linked to the business.
  
  
• Growth and transformation
  MDM is crucial in restructuring or acquisitions, where a multitude of data sources come together.

Perception problem

There is reason enough to embrace MDM, though there is still a perception problem today. "MDM had a reputation for being monolithic and costly," says Sumit Jha, Associate Director & Data Management Lead at Cognizant.

"Organizations see Master Data Management as a complex project that involves a lot of changes, and only very stubbornly delivers concrete value."

Jha acknowledges that the situation used to be that way, but points out an important evolution. "MDM projects only delivered value after months, or even years. With the advent of SaaS, however, everything has changed. MDM is part of a bigger picture and the monolithic aspect is gone. With SaaS, you can start an MDM project small and reap the benefits in just a few weeks."

Many systems, one truth

"If you have a 500-page book, with all the information in it, MDM is the index to that book," says Jha. "If you need something, you know where to go." He clarifies the role of master data management in a modern enterprise. "MDM is what you need when you have more than one system, such as a customer database, an enterprise resource planning (ERP) system and a client relationship management (CRM) system. This creates different records that refer to the same entity. In a first step, you unify those so that only one record exists for all systems."

A simple example clarifies: ITdaily can be a customer record, a partner for a marketing department, and a vendor in a CRM system, all from the same company. By fusing those records together, so that instead of ITdaily Marketing, ITdaily Customer and ITdaily Supplies, only ITdaily exists, and the three domains are linked together, you get a single source of truth. All the data on the ITdaily record is now in one place, with correct spellings, addresses, phone numbers... "But that's not MDM yet," cautions Jha.

He continues, "True master data management begins when you roll out the right measures to ensure that the record never splits and the data stays correct across all systems." In other words, if a marketer changes the address of ITdaily's headquarters, it should also be visible in ITdaily's record in the CRM system.

The (small) role of technology

"The migration to SAP S/4 HANA has accelerated the adoption of MDM," Jha believes. "Keeping data clean is important in S/4. Moreover, migrations run smoother when data is cleaned up. The manufacturing industry relies heavily on SAP, and that approach helps."

Still, technology is only a small piece of an MDM strategy. "A minority of the work has to do with technical complexity," Jha estimates. "The rest is about people. MDM is not an app you roll out; it's about people and processes. You need active participation from the business. There has to be enthusiasm." He estimates that a mature MDM strategy brings a two percent increase in revenue, thanks to better insights and greater efficiency.

MDM is exceptionally relevant in acquisitions and mergers.

"MDM is exceptionally relevant in acquisitions and mergers," Jha emphasizes. "When you bring together multiple ERP systems, CRMs and other databases, master data is a must. This is true even in a company split. You need to know what data you have, where it sits and who is responsible."

Measure first

Getting started with MDM is easier today than it used to be. "You have to do your homework, though," Jha says. "Where do you want to roll out MDM first? And why? Only when you have answers to those questions is it time to work out a strategy. Moreover, measure the problems before you start the rollout, otherwise you will have a hard time explaining the value to management."

If you have one record, ITdaily, and another with a spelling error, like ITdialy, which contains similar info, it's certainly helpful when you combine them under the right name, but that action alone is hard to translate into value. Jha: "Examine how much time is lost, how many customers are not approached correctly, or how much unnecessary work is done. You can improve those parameters with MDM."

Concrete benefits

Jha gives another example. "Manufacturing companies need to correctly identify their products and struggle with that. However, correct product data has a big impact. When a customer calls the help desk with a problem, the helpdesk employee who has access to all relevant information about a product and a customer via MDM may be able to solve that problem over the phone. Without MDM and access to the data, the alternative is to send a technician. It is clear, though, which method saves the most time and money."

A good data strategy without silos and duplicates opens the door to further applications. Jha continues, "Again in the manufacturing industry, products need to be classified correctly. It is important whether they are green or not, for example, or fall under certain regulations. This is usually a manual and time-consuming process, but those who have their affairs in order with MDM can unleash generative AI on the data and save a lot of time and effort."

An end goal

Master Data Management is not an end in itself, but a means. A large organization without MDM is like a huge room with stacks of books in it. With MDM, that room becomes a library. Moreover, MDM is never finished; it is a never-ending process. If you give up managing the library altogether, in no time you'll have another messy pile of books in which you can't find anything.

Sumit Jha continues, "If you stop when you have your data in order, it takes a year before 20% of your data is bad again," Jha estimates. "After two years, it's over 60%." The quality of your data deteriorates quickly. MDM you have to maintain."

That too should not deter you. Actually, MDM is simply a tactic for creating order out of chaos and maintaining that order structurally. So it should come as no surprise that companies active in life sciences have very high maturity levels. "Everything is taken care of by MDM," Jha knows. "Right down to employee salaries. When you are subject to a lot of regulations, you have to be able to track everything. You can imagine the value of master data in such a case."

"There are many companies that are just starting MDM today," Jha reiterates. "Start small and expand. Pause if you don't see added value, and expand the project if the value is there. And above all, don't go too fast," he concludes.