The home price war should not cross the border

Since the middle of 2023 the market for industrial robots in China has behaved like the market for any crowded manufactured good. Dozens of domestic makers, most of them with new capacity and investors expecting growth, have cut prices for six-axis arms and SCARA robots quarter after quarter. Our own average realised price at home fell 22 per cent between 2023 and 2025. Some of that fall was earned: cheaper servo motors, better yields at our Changzhou plant, a simpler controller. Much of it was simply conceded.

When prices fall that fast at home, export markets begin to look like relief. Abroad, the incumbent Japanese and European brands still charge prices that look generous from Nanjing. The arithmetic proposed inside every Chinese robot company I know is the same: take the domestic cost advantage, price 15 or 20 per cent below the incumbents and buy share. In late 2023 our own European sales team proposed exactly that for the 2024 price list, and it was a reasonable proposal from people under real pressure. I argued against it, the board backed me, and we paid for the choice in 2024. I still think it was right, and I think the reasoning holds for most Chinese industrial makers that sell through partners abroad.

What an integrator hears when the price falls

In Europe, Turkey and Mexico we do not sell most robots to the factory that will run them. We sell to system integrators: engineering firms of twenty to two hundred people who design a welding, palletising or machine-tending cell, buy the robot, write the programs, install the cell and answer the telephone when it stops. The integrator quotes the factory a fixed price for the whole cell, often for delivery many months later, and carries the risk of everything inside it.

An integrator reads a price cut differently from a buyer of finished goods. First, it devalues work in progress. An integrator that quoted a cell in September with our robot at one price discovers in January that a rival can quote the same cell for less. Second, it raises a question about the next decade. A robot runs for ten years or more, and an integrator that recommends a brand is betting that the maker will still supply reducers, controllers and software updates in 2035. A maker that cuts its price by a fifth looks like a maker in difficulty. Third, it signals what the brand thinks of itself. For a Chinese brand abroad, price is the first quality statement a buyer reads, and a deep discount confirms the very doubt we are trying to answer.

None of this is sentiment. When we surveyed 212 European integrators in 2024, the reasons they gave for not quoting a Chinese robot were, in order: uncertainty about spare parts, uncertainty about the maker's staying power, documentation and software support, and only then price. Nobody told us our robots were too expensive. Several told us that very cheap robots made them nervous.

A corridor, not a list

Holding price abroad is not the same as refusing to compete. Our answer was a price corridor. Each region has a published list price, and every project price must fall inside a band beneath it. Discounts up to 8 per cent are the sales team's to give. Anything deeper needs my signature and the chief financial officer's, with a written reason and the project's expected service revenue. The band is narrow on purpose. It means an integrator in Bursa and one near Stuttgart can trust that a rival quoting our robot is not being quietly subsidised by us.

The corridor also tells us where to spend instead. Money we did not give away in price went into what integrators said they lacked: parts depots in Ludwigsburg, Bursa and Monterrey, application engineers who answer in the customer's language, and documentation rewritten by people who program robots for a living. In 2024 we spent about RMB 38 million on those things in Europe, a little less than the proposed cut would have cost in margin on that year's European sales.

There are exceptions, and we write them down. A first order from a large manufacturer that will standardise on one brand across several plants can justify a deeper price, because the installed base pays back in service and in further orders. We have approved four such exceptions since 2023. Each is reviewed a year later against what we expected, and two of the four have not yet earned their discount. I would rather know that than pretend every exception was strategic.

The cost that never reaches the plan

The strongest reason not to export a domestic price war rarely appears in a marketing plan. Trading partners have tools for responding when imports arrive at prices their own producers consider unfair. The European Union, among others, can investigate imports for dumping or subsidy when its industry complains of injury, and can impose duties that last for years. I am not a trade lawyer, and I will not predict any particular case. But I know what such an investigation examines: export prices compared with prices in the home market and with costs, and whether producers in the importing market are losing ground to imports sold at those prices.

A Chinese maker that cuts export prices to match a collapsing home market is producing exactly the evidence such a complaint needs, and producing it on behalf of every other maker in its sector. The cost of a duty does not fall only on the company that cut. It falls on every integrator that has standardised on a Chinese brand and on every factory whose cell was quoted at yesterday's landed cost. That is a risk a board should price, and very few export plans include it.

The United States shows what happens when a door closes for other reasons. Since 2018 it has applied additional tariffs to a wide range of Chinese-made industrial equipment, robots among them. We decided in 2021 not to build a sales organisation there, and our US revenue remains below 2 per cent of the total. I would rather build a durable business in the markets that remain open than help close them one by one.

The corridor was not free. In 2024 we lost two large tenders, one at an appliance plant in Poland and one at an automotive supplier in the Czech Republic, worth RMB 64 million together. European revenue grew 7 per cent that year against a plan of 25. For three quarters our European colleagues watched other makers' robots go into cells they had expected to supply, and two of them left.

In 2025 European revenue grew 31 per cent to RMB 520 million, at a gross margin of 38 per cent, and the number of certified integrators in Europe rose from 54 to 81. I cannot prove that the corridor caused all of that; the European market itself recovered in 2025. What I can say is that our integrators stayed, and that several who joined told us they had watched what we did with price before deciding.

My advice to other Chinese industrial makers is narrow. Separate the price you need to survive at home from the price at which you want to be judged abroad. Write the exceptions down and review them. And spend the difference on the things your partners would otherwise have to promise their customers on your behalf.