The price war is on. Your failure costs are winning.

Two headlines, same week, one conclusion. Xpeng, a company built to out-engineer the Chinese price war, issued a forecast the market read as capitulation: discounting is now outrunning the industry's cost roadmap. That same week, Ford walked the press through its Universal EV Production System at Louisville, a production machine assembled around a single obsession, unit cost. Neither is an EV story. Both say the easy savings are spent – platform consolidation, sourcing leverage, design-to-cost – and the industry has moved on to hunting margin in rooms it forgot had doors.

Here is what should worry you. When that pressure reaches the plant floor, I know what gets cut, because I have watched it more than once. Training goes first. Then supplier development. Then the hours reserved for PFMEA reviews and process trials – the entire prevention budget, cancelled by a controller who has never stood next to a press at two in the morning. Meanwhile the scrap bin keeps buying steel, resin and energy at full price and converting them into parts nobody ships. Most plants answer a price war by amputating the one function that still prints money.

Where the money actually hides

Price is set by your customer's purchasing department; material, by the mills and the index clauses; labour, by the agreement you signed in better years. Every traditional lever now points outward. The one large cost pool still inside your four walls is the cost of poor quality. Nobody owns it as a number. It is smeared across the P&L where nobody has to add it up: the sorting tent by the dispatch door, the rework cell of nine people (€320,000 a year fully loaded, before you count the parts), the premium freight chasing contained lots – I have seen €40,000 a month on that alone at one stamping plant – plus warranty reserves booked by finance and never read by operations, and the engineering weeks burned on customer escapes instead of new business.

The arithmetic nobody does at board level: failure costs scale with volume, not with your quality budget. Cut the budget 20% and the failure bill does not follow. Push throughput 15% and the bill comes with it. Procurement teams report the same squeeze from the other side of the table – cost, quality and resilience pulling against each other with no slack left. Shipbuilding is learning the lesson at national scale. You cannot buy your way out of accumulated quality debt.

A price war is a margin audit with a due date.

The tools are already on your shelf

I will not pretend this needs a digital twin. The playbook is two decades old and unglamorous. QRQC daily at the line, with the people who found the problem in the room. One A3 per quarter aimed at the top three failure-cost drivers – not the top three defect counts, the top three costs, because a cosmetic flaw on a high-runner and a functional flaw on a low-runner are not the same conversation. 8D discipline on every escape, run to closure rather than to the customer's patience. A visible quality wall, so the numbers survive the walk. None of it is new. All of it is rare.

At SNOP I built and ran quality for a greenfield plant of more than 900 people while a key customer demanded annual price-downs and material indices climbed the other way. We cut defect costs by 70%, carried customer satisfaction to 98%, and finished a quarter with zero critical customer escalations. That was not a trophy for the quality department. It was margin the plant rescued under commercial pressure and handed to sales as negotiating room, earned through cadence, discipline and the nerve to name the top three cost drivers in public. At WITTE Automotive the same architecture – QRQC, A3, a Q-Wall that made failure costs visible daily – produced substantial failure-cost reduction at a plant that had spent years paying for its own disorder. Through FOREAST I have since run the same arithmetic for ArcelorMittal and others. Steel and electronics differ in vocabulary, not in physics.

What the cuts do six months later

The sequence is mechanical. Prevention is cancelled in the fourth quarter – it is small, invisible, and defends itself poorly in a spreadsheet. Appraisal survives, because inspectors are countable and auditable. Then spring volume arrives, the pushed-throughput season every price war creates, and failure costs come with it: an extra shift of rework, freight upgrades on contained lots, the first escapes in eighteen months. The lag between cancelled prevention and arriving failure is roughly one volume ramp. That is precisely why plant leadership never connects the two.

It is also how defects reach customers at the worst possible moment. A price war compresses sourcing decisions; every award and every resourcing review is live. A single warranty claim inside that window does more commercial damage than the entire prevention budget would have cost across the year. The controller who cancelled the training keeps their job. The warranty reserve does the dying.

So when the disappointing forecast is someone else's and the cost programme is yours, the question for leadership is not whether quality is affordable. Price, material and labour have already been decided by people who do not work for you. Quality is the last margin you control, and the only one that grows when you pay attention to it rather than pay for it. A price war is a tax on plants that never measured their scrap. This season the taxman is collecting in public.