A supplier's capacity should follow the volume its customers will pay for, not the volume they forecast.
I am executive vice president and chief strategy officer of Tallisford Thermal Systems, a fictional NYSE-listed supplier to vehicle makers, based in Southfield, Michigan. We earned US$3.72 billion of revenue in 2025. About two-thirds came from light vehicles: heating and air-conditioning modules that every powertrain needs, and the cooling plates and chillers that keep electric and hybrid batteries at working temperature. The remaining third came from cooling packages for heavy trucks, construction equipment and farm machinery, made in 15 plants across the United States, Mexico and Canada by 14,600 colleagues.
My job is to decide, with our chief executive and board, where Tallisford's capital and attention should go and where they should stop going. Since 2021 that has meant buying a heavy-duty cooling business in Wisconsin, selling the engine-cooling business I had previously run, building one battery plant where our awards implied three, and keeping our Mexican module plants open through the 2025 tariff changes. I began as an application engineer on a Dayton module line in 1997, and I have since been a strategy consultant, a private-equity operating principal and a general manager. Each role taught me to distrust a different kind of optimism.
The lesson I learnt first, and still apply most often, is that a supplier's volume belongs to someone else. A vehicle maker awards a programme with an estimate attached, and the supplier buys the tooling, the floor space and the people. When the estimate proves wrong, the customer adjusts its weekly schedule and the supplier keeps the plant. Strategy in our part of the industry is mostly the discipline of deciding how much of that risk to accept, at what price, and on what paper.
Dependence on one customer cut by a third
Our largest customer took 34 per cent of revenue in 2021 and 23 per cent in 2025. The US$640 million purchase of Quenmore Heat Transfer in 2022 and the sale of Engine Cooling in 2024 did most of that work; new heavy-duty and off-highway awards did the rest.
Battery capacity staged instead of promised
We built one battery-thermal plant, in Tennessee, rather than the three our 2022 awards implied. When programme volumes ran at 38 per cent of quoted peak, we had avoided about US$230 million of idle capital and recovered US$41 million in shortfall payments.
A legacy business sold while buyers still wanted it
Engine Cooling went to Corrander Capital Partners for US$790 million in March 2024, and net debt fell from 2.6 to 1.5 times adjusted EBITDA within the year. Hybrid demand later made the buyer's assumptions look better than ours, and I say so on this site.
Tariff cost reduced without moving a plant
During the 2025 tariff changes we raised the share of Mexican module value documented as qualifying under the USMCA's rules of origin from 58 to 91 per cent, and agreed tariff cost-sharing with three of our four largest customers.
Current focus
I am preparing the board's 2027 decision on converting part of our Tennessee battery-thermal plant to hybrid and power-electronics cooling, now that two customers have moved programmes from fully electric to plug-in hybrid drivetrains. Alongside it I am leading a review of our heavy-duty business through the bottom of the truck cycle, to decide which Quenmore product lines deserve more capital and which should be licensed or sold.
Conversations I welcome
I am not looking to leave Tallisford. With my board's consent I would consider one independent directorship at a US-listed industrial company outside the vehicle supply chain, joining its finance, compensation or nominating committee and, after a year, chairing a finance or strategy committee if it has one. I can give about 28 days a year. I also accept a few invitations to speak to boards and investors about capacity risk and portfolio decisions.
