Interview with Stephan Müller:
The factory you can’t see

Stephan, you have been working intensively with master and product data for around 15 years. Why does the topic never let go of you?

That has to do with my background. At ETH, I was fascinated by how a manufacturing company works. That is where I came across supply chain management, the question of how to organise the real value chain in a well-considered way. At Bain, I worked on strategic questions of this physical supply chain. At the web agency Smartmedia/Unic, I then explored how web shops and content management systems work, where their data comes from and which challenges come with it. As early as 2002, we did pioneering work in the field of Product Information Management (PIM).

Imagine a CFO is listening to us. How would you explain to them what this is about?

I don’t start with the tools, but with the context. To bring its goods and services to the customer, an organisation today has to master its Digital Information Supply Chain. I deliberately modelled the terms on the real supply chain. A CFO knows what a factory is. So I ask: What does your data factory look like, and how are you developing it further? The software systems are the machine tools in it. They, too, have a life cycle, and it is valuable to know which tool is in which condition.

Have you ever seen anyone put a concrete figure on the economic damage caused by poor master data?

No. Customers have always dodged that question. I describe it with the image of an iceberg. The demands placed on data are rising, in terms of data variety as well as regulation, and costs are exploding along with them. Because the Digital Information Supply Chain runs across all departments, these costs are scattered throughout the organisation. Usually only the tip is visible.

How do decision-makers react when you show them the iceberg?

With “Yes, that’s exactly how it is for us, too.” That applies to the iceberg just as much as to the other challenges I describe in the Digital Information Supply Chain, DISC for short. Nobody has ever contradicted me. It’s like a story you didn’t have in mind yourself. As soon as someone tells it, you say: “Exactly, that’s my reality too.”

Why does so little action follow from that?

If I tell the story to the C-level in general terms, without tailoring it to the organisation, it fizzles out. It remains a brief aha moment. At C-level, people constantly hear all kinds of stories; there is a veritable overload of stimulating stories.

A preliminary study is needed. The way there leads through someone at the operational level in the organisation who says: “I have this problem, and I’ll organise the budget for the preliminary study.” In the preliminary study, we enrich the story with the organisation’s context. We show how mature the systems are and make connections and challenges visible. That makes it concrete for this context, and a decision can be secured.

“It is an illusion to believe that change in organisations works primarily through the C-level. The key is the people within the organisation who do the groundwork.”

And ROI calculations?

They were only rarely part of the project proposal. It is well established in organisations that such calculations feign a false precision. Decisions are made out of a qualitative conviction that it is necessary.

What does AI change in the handling of master data, and what stays the same?

Mastery of data is becoming even more important. A good data architecture is becoming even more decisive, because it allows you to unlock enormous potential with AI. What AI will not do: turn a structural data problem into a data paradise.

“If you have a data desert or a data purgatory, AI won’t come along as an angel and build a paradise out of it.”

So more data quality?

I am also concerned with the quality of the data structure. It is just as important as the quality of the individual data. What does your data architecture look like from end to end? Do you understand how everything fits together? In a real factory, someone designs the entire factory layout: how the goods flow and where the production departments are located. Data needs the same care. Instead of each department optimising for itself, a Gesamtkunstwerk, a total work of art, emerges.

Where does AI help in concrete terms?

It makes an idea realistic that I have carried within me for a long time: a control room for the data factory. In the process industry, you cannot climb into the pipes to see how things are running. Sensors continuously capture data. If something moves outside the tolerance range, a message appears in the control room, and someone intervenes precisely at the right place.

Who sits in this control room?

It is the tool of a Data Governance Board. My military background plays into this, too: in a battalion staff, you run a command room where the relevant information is kept up to date on a command wall. Until now, the investment in such a control room was illusory. With AI, a realistic option opens up for it for the first time. And AI will take over subtasks. Just as production robots and automated transport systems increase automation in factories today, agents will do so in the data factory, working together with people.

What is the mistake clients make most often?

The projects are staffed too thinly. In a real factory, people are on site with their whole being. In the data factory, staff have to work on too many topics at the same time. When the investment is approved, the right people must also be made available with enough working time.

And if you could give one single piece of advice to an executive board that commissions an MDM programme tomorrow?

Trust in the expertise of people who live in this field and feel at home in it. For them, it is a passion. It is like a candle: let your people catch fire, so that the candle in your organisation burn

Stephan Mueller CEO foryouandyourcustomers

About the person
Stephan Müller studied electrical engineering as well as operations and production science at ETH Zurich. After stints at Bain and Smartmedia/Unic, he founded the foryouandyourcustomers group in 2010 together with Jonathan Möller and Josef Stadler, and he is its CEO today. He developed the Digital Information Supply Chain (DISC) and works as a consultant on customer projects. The conversation took place in April 2026.

Interview conducted by Andreas Drexhage with Stephan Müller on 1 April.