Loss to the Floor

The andon cord went a lot in his zone. That was the first thing you noticed, and if you did not know what you were looking at, it looked like a problem.

Pulls were high. Stops were almost nonexistent.

Anyone who has run an assembly floor just sat up, because that combination is the whole thing. High pulls mean the team is catching problems — they see it, they say it, they are not hiding anything to protect a number. Few stops mean they are solving it inside the takt time, before the line ever goes down. Pulls without stops is not a quiet team. It is a team that is awake and competent at the same time, which is the rarest condition in manufacturing and cannot be ordered, bought, or installed.

His standard work sheets were perfect. Training and certification records complete, every one. Almost no defects out of his zone.

His name is Ramon Hernandez and he was the best team leader I ever had.

I ran the Trim Shop at Lansing Grand River for General Motors. Ramon took care of his team. His team took care of the product. That sentence is not sentimental; it is a causal chain, and it ran in that order, and it never ran in the other order.

He had a training instinct I have never seen matched. Put a new team member in front of him and inside the first hour Ramon had read the hands, the posture, the hesitation pattern — and calibrated his coaching to that specific person. Not to a curriculum. To that person. And he carried an encyclopedic mental model of every bottleneck the trim line had ever produced, which meant that during a launch, when GM was bleeding time against a hard date, Ramon found the constraint before the engineers finished their first meeting. He solved problems that did not have names yet, because the vehicle had not existed long enough to generate a problem history.

None of that came from a syllabus.

We paid for it

Here is the part that matters, and it is the part everyone gets wrong about Grand River.

We made him. Deliberately. On purpose. With money.

A billion-dollar plant. Body shop, paint shop, a novel general assembly. Tom LaSorda’s plant, the first one he built in a unionized shop. I was one of the first shift leaders hired, and production was a full year out, and a few of us looked at each other and said: there is no plant here. What are we going to do for a year?

Invest in people. That was the answer. That was the design.

So we built a little assembly line out of wood. Little wooden cars, with andon pull cords, standardized work, team leaders, quality stations — the whole architecture in miniature. Ken Knight’s idea, one of LaSorda’s apostles. We put out a brochure to Local 652 saying come work at the new plant, and people came in droves, and we could afford to be selective because we had the volume. You cannot interview a floor worker off a résumé. So we watched them build wooden cars.

That plant went on to JD Power Gold.

Now find that year in General Motors’ accounts.

You cannot. The building capitalized over thirty-nine years. The tooling over seven. The presses, the conveyors, the paint booths — every one of them on the balance sheet, depreciating politely, visible to any analyst who cared to look.

The year we spent making Ramon Hernandez was expensed in the period incurred. Gone by the next 10-K.

A billion dollars of that plant is still on the books. The single best thing in it never was.

Loss to the floor

Then Pilot came for him.

We needed a Pilot team member to work the next car, and Ramon was the obvious choice, and I let him go. He went on to become one of the best pilot team leaders GM had — he brought in two vehicles after the first CTS, with results.

Loss to the floor. Ascension for him.

I felt it every shift. My numbers got worse. Measurably, immediately, for months. And GM’s asset got better by exactly the same event.

There is no system anywhere on earth that connected those two facts.

My P&L recorded the loss. Nothing recorded the deposit. If you had audited my shop that quarter you would have found a supervisor whose performance declined and no explanation, because the explanation was that I had transferred a capital asset out of my cost center and there was no line to transfer it on.

That’s the argument. Not that companies are cheap. That the instrument only has one column, and it says loss.

Producing from inventory

An oil company that pumps a reservoir and books the cash as earnings is not more profitable than its neighbors. It is producing from inventory.

Every accounting regime knows this. Under ASC 932 an extractive firm cannot simply report the revenue — it must charge depletion against a recorded reserve and publish a standardized measure of what remains in the ground. That regime exists for one reason: a company that is earning and a company that is selling what it has left look identical on an income statement. Right up until they don’t.

Formation is the only reserve in the modern economy that can be pumped to zero with no depletion charge, no reserve disclosure, and no line item anywhere that would reveal it happening.

Sixty people report directly to Jensen Huang. This gets offered as proof that AI has changed the architecture of leadership. Ask one question about those sixty: where were they formed? Not one of them at Nvidia. They arrived formed — decades of deposits made somewhere else, at somebody else’s expense, by people whose names are not in the article. Nvidia did not build a structure that produces sixty executives. It built one that consumes sixty.

Grand River made Ramon. Somebody else got him. That is not a scandal — that is what should happen, that is what ascension is. The scandal is that no ledger anywhere records th