Trumpf: How a Machine Builder Tried to Grow Through Service and Software

Why Trumpf is instructive on service and software

Hardly any advice to machine builders is repeated as often as this: do not just sell machines, sell services and software around the machine. Trumpf from Ditzingen near Stuttgart, a family-owned company making machine tools and lasers, is often cited as an example. The company founded its own platform for the connected factory early on and tried out new payment models.

This case study answers one question: according to publicly documented facts, what role did service and software actually play in Trumpf's growth, and what of it can be transferred to a small business? To answer it, I first set out the facts, then present a mental model from research on the shift from product to service, apply it to the documented steps, weigh two readings, and finally ask about the limits of transfer.

The documented facts: from workshop to laser maker

In 1923, Christian Trumpf and two partners took over a machine shop in Stuttgart that initially made flexible shafts. In 1950, the then 19-year-old Berthold Leibinger began an apprenticeship there as a mechanic, and in 1966 he became technical director. Under him, the laser became the core of the business: in the mid-1980s, Trumpf decided to build its own CO2 lasers, the first being the TLF 1000. In 2005, his daughter Nicola Leibinger-Kammüller became chair of the managing board. The company is still owned by the family and a family foundation.

Sales in recent years show how exposed even such a company is to the business cycle. In its anniversary year 2022/23, Trumpf reached about 5.4 billion euros, the highest sales in its history. Two years later it was 4.3 billion, a decline of 16 percent on the previous year, and reported operating profit fell to 59 million euros. In fiscal year 2025/26, sales of 4.34 billion euros were just above the previous year, operating profit rose to 480 million euros, and order intake rose by about 7 percent to 4.5 billion euros.

Employee figures differ depending on the statement: the preliminary company release gives 16,960 employees as of June 30, 2026, a report on the annual press conference 17,583, compared with 18,303 a year earlier. Both figures show a reduction. Incidentally, with 5.4 billion euros in sales in 2022/23, Trumpf was above the five-billion-euro threshold that Hermann Simon sets for hidden champions.

Service and software: what is publicly documented

Two steps are well documented. In 2015, Trumpf founded Axoom, its own software platform for connected manufacturing. It was meant to help manufacturers plan and control processes from order intake through material purchasing to shipping, including with machines from other makers. On July 1, 2019, the IT services company GFT took over Axoom's business operations and employees. According to press reports, the brand remained with Trumpf, and the two companies agreed on a development partnership.

In October 2020, Trumpf and the reinsurer Munich Re announced a model called pay-per-part. Customers would be able to use a laser cutting system without buying or leasing it and pay a pre-agreed price for each cut sheet-metal part. Munich Re was to finance the systems and bear the investment risk, and its subsidiary relayr was to analyze the data. The steel distributor Klöckner was named as a development partner. The German Federal Cartel Office cleared the partnership at the end of October 2020; the model was initially offered only in Europe.

What is not public is just as important: the statements reviewed give no sales figure for service, for software, or for the pay-per-part model. Whether these businesses contributed to growth, and to what extent, cannot be derived from the figures.

The mental model: two axes from product to service

In a qualitative field study, Oliva and Kallenberg (2003) examined eleven capital equipment manufacturers that were explicitly pursuing a service strategy. Their finding: the transition to service is a deliberate developmental process in which capabilities are built while two things shift. First, the nature of the relationship with the user of the product, from the single transaction to a lasting relationship. Second, the object of the offering, from the product itself to the customer's process in which the product is used.

From these two axes, a grid with four fields can be formed. It is my derivation from the finding, not the authors' own presentation. But it helps to classify a company's steps and to see how far each one moves away from the core business.

The further toward the bottom right an offering lies, the more capabilities it requires that a machine builder does not already have: software for other companies' processes, financing, risk-bearing. In my assessment, this explains why these fields are so often filled with partners. Gebauer, Fleisch, and Friedli (2005) also describe the service paradox from their work with more than 30 manufacturers: the companies invested in their service business, offered more, had higher costs, and did not achieve the expected returns.

The further a service is from your own product, the more likely it is to need a partner who can do what you cannot.

What the model explains about Trumpf

If you place the documented steps in the grid, a clear picture emerges. Axoom was in the field of the customer's process: the platform was meant not only to support Trumpf machines but to control entire processes in manufacturing businesses. Pay-per-part was in the bottom-right field, because the customer no longer buys a machine there but a result per part.

In both steps, Trumpf did not carry alone the capabilities that are far from machine building. The software went to an IT services company after four years, and financing and investment risk of the payment model were taken on by an insurer. Whether this was a retreat or a deliberate division of labor has not been publicly explained. But the sequence fits what the model leads you to expect: the further an offering is from your own product, the more outside capabilities it requires.

The strongest documented growth area, by contrast, is not in service at all. It is the EUV laser that Trumpf supplies for systems from the Dutch manufacturer ASML, which are used to expose semiconductors. It comes from the company's technical core, the laser, and in 2025/26 contributed 791 million euros, almost a fifth of sales.

Two readings: growth through service or through technology?

The first reading is the common one. It carries weight, because Trumpf did invest in software early, and partnerships are no sign of failure. The second reading relies more closely on what is documented. It contradicts the obvious assumption that a machine builder grew mainly through service and software.

I consider the second reading better supported, without rejecting the first. According to the public figures, the big spurts came from technology, most recently from the EUV laser. In my assessment, service and software have a different task: they deepen the relationship with the customer and make the machine more useful. Whether they themselves carry a significant share of profit cannot be judged from the outside.

What small businesses can transfer

A business with twenty employees will not found a platform or set up an insurance model. What can be transferred, however, is the sequence the grid suggests.

1. Start with your own product

   Begin in the top-left and bottom-left fields: deliver spare parts quickly, offer maintenance as a contract. How such contracts are costed is described in the guide building recurring revenue.

2. Only then move to the customer's process

   Move into the right-hand column only once you really know the customer's workflow. An offering for other companies' processes requires knowledge you gather over years with the customer.

3. Do not build missing capabilities alone

   Software, financing, and risk-bearing are businesses in their own right. Check whether a partner can do them better, as Trumpf did with software and financing.

4. Cost every service

   The service paradox arises when services grow without costing. For every service, define what it costs and what it should deliver, and review this annually.

5. Maintain the technical core

   The case shows that the core must keep carrying the business. A service meant to cover up a weak offering does not last long.

Why reliable extras in particular keep customers is described in the essay service as a differentiator. And anyone who grows through one large buyer, as Trumpf does in the EUV business with a single partner, should keep an eye on the concentration risk described in the analysis customer concentration risk.

Limits of transfer

Three limitations belong to this case. First, the decisive figures are missing: without published service and software sales, every statement about their contribution remains an assessment. Second, the findings of Oliva and Kallenberg (2003) are based on eleven manufacturers and a qualitative study, and the grid is my derivation from them. Third, Trumpf is a group with its own research and international sales. What it solves with partners, a small business often cannot even begin.

How a family business is run across generations, and why family and business belong apart, is a separate question. The article running a family business addresses it.

Conclusion: what the Trumpf case teaches

To answer the guiding question: according to publicly documented facts, service and software played a visible role in Trumpf's growth, but not a documented load-bearing one. The own platform Axoom went to an IT services company after four years, and the pay-per-part model was created with an insurer that bears financing and risk. The documented growth spurts came from technology, most recently from the EUV laser with 791 million euros in sales. The model of Oliva and Kallenberg (2003) makes it plausible why offerings that move away from your own product need partners.

For small businesses, this is not a rejection of service but a sequence: first, service on your own product, properly costed; then, carefully, toward the customer's process; and for outside capabilities, a partner. The case also shows that good service complements the core and does not replace it.