AI’s integration into MDM shifts the focus of data teams towards more strategic issues, says Petr Žáček, Ataccama

Ivana Karhanová: Ataccama delivers an AI-powered data management platform, so you must know a lot about MDM. Why is the question “Is master data management still relevant?” still on the table?

Petr Žáček: I think it’s because MDM evolves in a very subtle way. Right now, all the focus is on new technologies, particularly AI, and we will definitely be talking about that. But for MDM, it’s more of a discipline that’s part of data management and data governance. It’s established, and companies understand that they need it. It’s not as much about new, emerging technologies, and it might seem like MDM is this old-fashioned approach to managing data. But it’s essential to recognize that master data management is all about managing the core data that underpins everything else. As long as we have master data, we will always need to manage it. The tools and approaches may evolve, but the discipline itself remains crucial. Sometimes companies try to replace MDM tools with new technologies, but in the end, MDM is still very much needed. The fundamental concepts and principles won’t change, so it’s up to companies like Ataccama to develop the right tools that guide people through this process and make it easier.

Ivana Karhanová: So, the tools might change, but the concept of MDM will always be needed?

Petr Žáček: Exactly. In the past, MDM was often viewed as a big IT project involving complex technology and implementations. But I think the key today is to deliver faster, to prototype solutions quickly, and to bring business value sooner. That way, companies can see the practical benefits of MDM early on, which makes it easier to onboard people and gain their buy-in. If you’re using a set of different tools instead of a dedicated MDM system, people might think that MDM is dead or outdated. But what they’re still doing is MDM—they’re just achieving it in a slightly different way. The principles remain the same. Master data management is all about organizing, improving, and understanding large sets of data, especially when you’re working with millions of records. So even if the tools evolve, MDM remains essential because it’s how we manage and make sense of this data.

Ivana Karhanová: So, simplifying it: MDM is about understanding the data?

Petr Žáček: Exactly. But it’s also about a lot of additional work, like data curation, improving data quality, matching records, and merging duplicates across data sets. These are technological challenges because they involve huge amounts of data that need to be processed quickly. You need to be able to identify and improve the data in real time, so whether it’s an MDM tool or another specialized unification engine, you’ll still need something to handle these tasks efficiently. The goal is to ensure that data is accurate, complete, and usable in real-time, and that’s where the technology comes in.

Ivana Karhanová: If you were to describe MDM in one short sentence, how would you define it?

Petr Žáček: That’s a good question. MDM is a discipline in data management and governance that connects data to the real world. Essentially, it’s the data that describes key business elements like customers, products, vendors, etc. For example, master data includes the names of your customers, the products you’re selling, or your suppliers’ contact details—information that is central to how a business operates. Without clean, well-managed master data, everything else you do with data—transactions, analysis, reporting—becomes much harder to understand. MDM is how you make sense of your business data and how it ties into your operations. It enables businesses to draw actionable insights from their data, which is essential for making informed decisions.

Ivana Karhanová: You’ve mentioned several ways to look at MDM. What exactly is it?

Petr Žáček: MDM is not just a set of features or a one-size-fits-all solution. In the past, organizations often thought of MDM as simply solving issues like duplicate customer records in their CRM or data warehouse. Some viewed it as a way to identify customers or other entities within the data and create a “single source of truth.” Others saw it as a system to enforce data quality rules, preventing mistakes and ensuring the consistency of information. MDM can be a combination of these approaches depending on the data domain and use case. For example, in the product domain, a system might be required to enforce policies about product codes or classifications. But when it comes to people, the challenge is slightly different. In that case, it’s about connecting data to real-world identities and ensuring that you have accurate, up-to-date records. MDM involves consolidation of data across multiple systems, and sometimes centralization as well. It depends on what you’re trying to achieve, and you may use a combination of tools and strategies to manage your data effectively.

Ivana Karhanová: Where was MDM five years ago?

Petr Žáček: Five years ago, MDM was still seen as something a bit more niche. It wasn’t as established as it is now. Back then, a lot of people still saw it as primarily an IT problem. There was a lot of experimentation with different technologies, like graph databases, and the industry hadn’t fully embraced the idea of MDM as a business problem. At that time, people were still figuring out how to implement MDM in a way that was more user-friendly and less reliant on technical expertise. It was still viewed by many as a highly complex area that required deep technical knowledge. Some industries like banking and insurance were ahead of the curve, having already adopted more mature data governance practices. However, many other companies were still in the early stages of understanding the true value of MDM and were in the process of figuring out how it could be leveraged to support their business needs.

Ivana Karhanová: What are your most important findings when talking to clients?

Petr Žáček: It depends on the client. Some companies are new to MDM and are just beginning their digital transformation journey. Others have been dealing with large amounts of data for years but are now realizing that they need a more structured approach to manage it. Recently, there’s been a lot of excitement around AI and how it can support MDM. Clients are excited about the potential for AI to automate a lot of the work involved in data management. However, there’s also been some disappointment when they realize that AI, while incredibly powerful, doesn’t necessarily do everything for you. It can automate tasks, but the critical work of ensuring data quality and governance still requires human oversight. The challenge for businesses is understanding that MDM is more than just a technology—it’s a process that requires a strategy. AI is a valuable tool, but it won’t replace the need for well-managed master data or the discipline behind it.

Ivana Karhanová: Could AI disrupt MDM?

Petr Žáček: AI won’t completely disrupt MDM in the next few years, but it will certainly improve and accelerate existing processes. AI can help automate some of the more repetitive tasks involved in managing master data, like cleaning up data, enriching records, or categorizing information. For example, AI can help you match records across systems more quickly or suggest improvements to data quality. However, AI isn’t a magic bullet—it can assist with tasks, but it still requires human oversight. The understanding of the data, how it impacts the business, and ensuring that it’s accurate and consistent are areas where human involvement is still crucial. So, AI will definitely help speed up processes and improve efficiency, but it won’t eliminate the need for good MDM practices.

Ivana Karhanová: Does this mean AI will require more senior people to oversee data management?

Petr Žáček: Not necessarily more senior people, but the focus of their work will shift. In the past, people in data teams often focused on technical tasks, like defining data models or writing rules for data quality. With AI helping automate some of those tasks, data teams will be able to focus more on strategic decisions and high-level business objectives. This will allow business users who are not necessarily technical experts to be more involved in data management and understand how the data is affecting their business. The role of data professionals will evolve from being focused on day-to-day technical tasks to being more focused on governance, strategy, and the business impact of data.

Ivana Karhanová: Can AI solve problems like poor data quality?

Petr Žáček: Yes, AI can help improve data quality, particularly when it comes to enrichment. For example, if you have incomplete or missing data, AI can pull in data from third-party sources or public databases to fill in those gaps. It can also help by breaking down data silos, bringing together data from various systems, and ensuring that it’s consistent. However, AI can’t solve all data quality problems. If you have poor data to begin with—like outdated or incorrect information—AI can help clean it up, but it can’t fix fundamental data issues. AI is great for automating processes, but the underlying data still needs to be accurate and properly governed.

Ivana Karhanová: How will cloud or multi-cloud strategies impact MDM?

Petr Žáček: The impact of cloud and multi-cloud strategies on MDM is still unfolding. Many companies are moving their data to the cloud, and some are even using multiple cloud providers. This introduces complexities for MDM, especially when it comes to ensuring data governance and compliance. For multinational companies, there are often country-specific regulations that complicate the process. For example, if you have data in multiple geographies, such as the US, Europe, and China, you might face challenges around data sovereignty and where the data can be stored or processed. It’s possible to do MDM in a multi-cloud environment, but it requires careful planning and often some customization. As more organizations move to the cloud and get used to these new infrastructures, they will see the benefits of cloud for MDM, such as scalability and cost-effectiveness. But it’s important to make sure the right data governance measures are in place to protect sensitive information.

Ivana Karhanová: How can we measure the success of an MDM strategy?

Petr Žáček: Measuring the success of an MDM strategy can be difficult because the impact is often seen downstream. For example, if a company improves its address data, it might save money on shipping costs by reducing duplicate catalogs being sent to the same address. However, it’s essential to start with clear goals before implementing MDM. Whether it’s improving customer satisfaction or meeting regulatory requirements, you need to have a specific use case that you’re trying to solve. If you don’t have a clear objective or pain point, the project could fail. But once MDM is in place and the data is clean and accurate, companies will see improvements in operational efficiency, decision-making, and even customer experience.

Ivana Karhanová: Thank you for sharing your insights today.

Petr Žáček: Thank you for having me.