Our grading floor is deliberately boring

Visitors find the grading floor the least interesting part of the tour, which is the point. Every design decision here is aimed at removing discretion from the moment when discretion is most expensive.

Why grading drifts

By the time a unit reaches grading it has been drained, stripped, washed, triple-rinsed, camera-inspected, cage-straightened, re-gasketed and pressure-tested. That is roughly an hour of yard time and real consumable cost.

A person looking at that unit at four in the afternoon, under pressure to clear the floor, with a borderline odour call to make, is being asked to throw away an hour of their own colleagues' work. Every incentive in that moment points toward passing it.

Grade drift is not dishonesty. It is a tired person being asked to discard work they can see, on evidence they can barely smell.

The six bays

Each grade is a physical location with a painted line and a sign. A unit is not graded on a clipboard; it is pushed into a bay. That sounds trivial and it is the most effective control we have, because moving a unit between bays is visible to everyone on the floor.

The fixed sequence

Six questions, asked in this order, every unit, no exceptions. The order matters because the disqualifying questions come first, so a unit that fails early never reaches a judgement call.

1. Is there a prior-contents record? No record, bay 6.
2. Is there any odour in the headspace after the 48-hour hold? Odour, drop at least one grade and never into bay 1.
3. Did it pass the 3 psi leakproofness test? No pass, maximum bay 4.
4. Any crazing, haze or stress cracking at the bottom radius? Crazing, maximum bay 4. Cracking, bay 6 for the bottle.
5. Cage welds intact, uprights bowed rather than kinked? Fails, bay 6 for the cage.
6. Does the paperwork that will ship with it actually exist? If not, it cannot be sold as the grade that requires it.

The rule that does the work

A unit moves down the list and never up. Once it is in bay 4 it cannot be reconsidered for bay 3 because somebody re-smells it and decides it is fine now.

This closes the loop that produces drift. Reconsideration always runs in one direction — nobody ever re-examines a passing unit hoping to fail it — so allowing reconsideration at all is equivalent to allowing upgrades.

The grader does not see the cost

We deliberately do not show the grading station what a unit cost to process. No wash-line time, no consumable cost, no indication of whether it came from an expensive inbound lane.

It is the only station in the yard where we withhold operational information from the person doing the work, and it is withheld precisely because it would be relevant to a decision we do not want them making.

How we know it is working

Two measurements, both cheap, both run monthly.

• Blind re-grade. Twenty units are pulled from outbound stock, their grade labels covered, and re-graded by a different person. Agreement has run between 94% and 98% since we introduced the bays. Before them it was in the low eighties.
• Downgrade rate by time of day. If the afternoon downgrade rate is materially lower than the morning rate, the process is drifting with fatigue. We watch the gap rather than the absolute number, and it has stayed within two percentage points.

Both measurements exist because the first version of this system had no way of telling whether it was working, which made it a belief rather than a control.

What a buyer should take from this

If you are evaluating a reconditioner, ask to see the grading floor and ask two questions: can a unit be regraded upward, and does the grader know what the unit cost to process?

A yard where the answer to either is yes is not doing anything wrong. It is simply exposed to a drift mechanism it has not closed, and over a few thousand units that shows up in what arrives at your dock. Our grade definitions are on grades explained.