The date is met. The standard isn't.

The only genuinely surprising thing about the foldable "Crackgate" affair is that somebody said it aloud. Insiders at the assembler reportedly conceded that quality standards were quietly lowered to protect the ship date. That is not a consumer-electronics scandal. It is the oldest bargain in manufacturing, and the sentence is usually spoken in a room with four people, one door and no minutes. I have heard versions of it in Polish, German and English across two decades of automotive and aerospace launches. The language changes. The arithmetic does not: when the launch date is immovable, the acceptance criteria become the release valve.

How a standard actually gets lowered

Nobody writes a memo headed "we are relaxing the standard". The act is always smaller than that, and it always looks technical.

• A drawing revision moves a tolerance from 0.08 mm to 0.12 mm — still within process capability — and nobody red-lines it, because the change lands mid-ramp, when engineering change review is running on four hours of sleep.
• An AQL slides from 0.65 to 1.5 for the ramp period. The ramp period has no defined end.
• A sample plan gets halved, 200 pieces down to 80, to protect takt. Nobody in the history of manufacturing has cut a sample plan because the parts got better.
• A durability test now stops at spec pass instead of running to failure. The green tick still arrives. You simply stop learning where the margin is — the one number you will want in week three.

The tell is not any single change. The tell is that the paper trail thins on purpose. Verbal waives at 21:40 on the night shift. Deviations approved in a corridor between meetings. If your audit finds launch documentation thinner than serial documentation, you are not looking at an efficiency gain. You are looking at a date being protected.

Move the date and the plan forgives you. Move the tolerance and the customer does not.

Why the date wins every planning meeting

The date wins because the incentives are asymmetric, and pretending otherwise is how good engineers end up signing bad paper. A slip is certain, visible and booked to one budget line in one quarter. A field failure is probabilistic, delayed, and lands in somebody else's quarter — often on another site, in another performance review. Governance exists to make that cheap bet expensive.

I lived the clean version of this. At SNOP I built the QA/QC function for a 900-plus-employee greenfield plant against a fixed start-of-production date — a date wired into customer commitments and hiring plans, about as movable as bedrock. The plant went on to a 70% cut in defect costs and a quarter with zero critical customer escalations, and none of it came from courage. It came from gates: entry and exit criteria written down before the pressure arrived, a named signature on every gate, and a standing rule that no deviation was valid without an expiry date. A waiver without an expiry date is not a waiver. It is the new standard wearing a disguise.

The catching mechanism matters as much as the gate. At Witte Automotive we ran a Q-Wall — every defect, every deviation, physically posted and confronted daily — with QRQC discipline underneath it. That is what picks up a quietly widened tolerance within a shift instead of within a quarter: the data gets argued over every morning, in front of the parts, by the people who made them. A monthly report can hide a loosened AQL. A wall at 07:00 cannot.

The bill arrives with the customer's name on it

Do the arithmetic both ways, in the open, at the gate. A two-week slip on an assembly line costs, very roughly, €200,000 to €500,000 in idle capacity and deferred revenue. Painful, visible, survivable. A field campaign costs reverse logistics, warranty, a service programme, and then the multiplier nobody books: your competitor's marketing department working for free. Bending-hinge test footage from a rival's lab does more brand damage in a week than a slip would have done in a month.

The old 1–10–100 rule holds in every industry I have worked in. A defect caught on your own line costs one unit. At the customer's dock, ten. In the customer's hands, on camera, one hundred — plus the premium on every future negotiation, because your acceptance criteria now start from a position of distrust. Automotive and aerospace run the same mechanism at higher stakes: a cracked hinge is a warranty item, a cracked caliper is a headline, a cracked structural rib is an airworthiness question. The physics of the bargain is identical. Only the invoice changes.

Here is the part nobody puts in the launch review: a standard lowered for a launch never comes back at full strength. Re-tightening an AQL means admitting it was loosened, so the "temporary" ramp tolerance quietly becomes the serial tolerance, and the next programme inherits it as baseline. The date is remembered for a quarter. The defect is remembered for a product generation. Hold the gate and sign your name to the criteria, or move the date out loud. Those are the only two honest moves on the table, and a quality organisation earns its keep by making sure there is no third.