The Call
November 14, 2023. 6:52 a.m. My phone buzzed before I'd finished my first coffee — our line supervisor, calling from the plant floor. "Check your email. The customer just flagged the QC report on the connector lot."
I opened it. 12,000 CNC-machined terminals — brass contact pins, the same style of terminal that finishes a beru BE6/1 0300.111.00 spark plug connector, and that also goes into the connector end of our beru coil pack subassemblies. The customer's incoming inspection had found something they didn't love: the plating had a mottled, slightly gray look in one corner of the tray. Not an obvious defect. Just off.
Two months earlier we had moved plating for this part family to a new supplier. The quote was about $0.03 cheaper per terminal. On 12,000 pieces, that's $360 in savings. I signed off on it after a first-article run because the dimensions were dead-on and the sample I held in my hand looked fine.
Honestly, I didn't think much about it. That's the part of the story I still replay in my head.
What We Actually Shipped
The switch looked like a clean win. Same base brass, same machine shop, same packing spec. The new plater's line was calibrated to a slightly different bath chemistry than our previous vendor, but their tech said it was "industry standard" and our first-article sheet came back with a green checkmark on every parameter.
What we didn't do — and I own this — was run a full salt spray qualification on the new plater's first production lot. We eyeballed a tray, measured thickness at three points, and shipped. Total saved on the qual work: maybe two working days and about $400 in lab time.
Fast forward to November. The customer's complaint came in as a cosmetic issue: surface discoloration, nothing more. But their quality team was sharper than ours. They asked one question on the call that made my stomach drop: "What's the contact resistance on the affected terminals?"
We hadn't tested it. Their team had. The reading came back at 12–18 milliohms on the discolored region. Our spec caps it at 5.
The Turn
Here's where the story stops being about plating and starts being about consequences. Roughly ten days after the batch shipped, the customer's downstream network — the repair shops that install our beru coil pack assemblies in the aftermarket — started logging intermittent misfires on vehicles that had just had a coil pack replacement. Their field techs did what any good tech does: they checked the obvious things first.
One shop kept chasing water pump failing signs because they assumed the misfire was heat-related. They replaced the coolant. They replaced the pump. The leak wasn't the problem.
Another shop swappped a sealed fuel filter assembly on a completely unrelated diagnostic path — the vehicle had a lean-bank code, so the tech went after fuel delivery. Wrong tree.
And at least two technicians spent a good chunk of a Friday arguing about where the O2 sensor sat relative to the exhaust manifold, because if it was reading slow, that would explain the misfire. It wasn't reading slow. The O2 sensor was fine.
Every one of those checks was reasonable in isolation. None of them pointed at the actual root cause: a 4-micron plating shortfall on a $0.60 connector terminal.
The customer ran their own salt spray test on our lot the same week. White corrosion at 20 hours. Red corrosion before 24. Our master spec had always been 96 hours to red rust. We never got near it.
The Damage
The lot was rejected outright. Twelve thousand pieces. We quoted the rework internally: strip, re-plate, re-inspect, re-pack, emergency air freight to hit the customer's replenishment window. That came to $3.20 per terminal, or $38,400. Not counting the freight, not counting the two days their assembly line sat idle waiting on a replacement lot, and not counting the fuel pump and fuel filter work those repair shops did for free to chase the wrong diagnosis.
Compared to the $360 we thought we'd saved. I don't need to do that math for you.
What I Changed
I keep this one on my whiteboard now. Not for drama — for the checklist.
Reverse validation, in my own career: everyone told me to always run a full salt spray on a new plater's first lot. I only believed it after skipping that step once and eating a rework bill big enough to fund a year of lab testing.
Here's what I now insist on, every time:
- First-article isn't first-lot. A sample run through a plating bath tells you the bath can hit spec. It does not tell you the plater can hit spec on 12,000 pieces over three shifts. We now require a 500-piece qualification lot before we release full production.
- Salt spray every new vendor, no exceptions. 96 hours to red rust per the customer's spec. We pay for the lab work up front. It's cheaper than freight on a rejected pallet.
- Contact resistance is a design parameter, not a nice-to-have. I don't care how good the plating looks. If we can't read it, we don't ship it.
- Pass the full spec down the supply chain. The new plater's tech said their process was "industry standard." I never checked what their internal standard actually said. That's on me.
The harder thing to fix was the downstream mess — the repair techs chasing water pump failing symptoms on a coil pack that was fine, the free fuel filter swaps, the O2 sensor goose chase. If you ask me, that's where cheap plating really shows its cost. Not on the invoice. On a repair shop's diagnostic bay, hours after the part left our dock.
I'm still not 100% sure why the discoloration showed up on only about 8% of the batch and not the rest. My best guess is bath agitation variation on the third shift — the affected tray was packed from a later run. Don't hold me to that. What I am sure of is that a full qualification lot would have caught it before it became 12,000 terminals and a $38,400 rework order.
That batch taught me something I should have known from day one: the cheapest supplier on paper is expensive the moment it fails a spec you didn't verify. And the customers who trust our beru coil pack assemblies — the ones running diagnostics in shops where nobody's watching — pay the price in their own time when we get it wrong.