01The word means whatever the yard says it means
I have graded pallets for nineteen years and I still get calls that start with 'this isn't Grade A.' The caller is usually right, by their definition. They are also usually comparing our pallets against a competitor's, and that competitor wrote a different definition. Both definitions are legitimate, because there is no authority that publishes one.
New pallets have standards. Lumber has grading rules enforced by accredited agencies. Recycled pallets have nothing of the kind. What exists instead is a trade vocabulary — Grade A, Grade B, #1, #2, premium, combo — that everyone uses and nobody governs.
02Why the definitions drift
Grading is not a machine process. It is a person picking up a pallet, turning it, running a hand along the stringer, and making a call in about four seconds. That person's employer pays them by the hour and sells the result by the unit.
Here is the uncomfortable arithmetic. Take a thousand cores. Grade them tightly and you might get 180 Grade A, 520 Grade B, 240 Grade C and 60 for dismantling. Loosen the Grade A definition by allowing one companion stringer and suddenly you have 310 Grade A. That is 130 pallets that just moved up a price tier worth roughly two dollars each, for zero additional labour.
Nobody sat in a meeting and decided to do this. It happens gradually, one borderline pallet at a time, over a year, and then a customer calls.
- Time a grader spends per pallet
- ~4 secTime a grader spends per pallet
- Grade A yield from one loosened criterion
- +72%Grade A yield from one loosened criterion
- Value of that drift per thousand cores
- $260Value of that drift per thousand cores
03The four criteria that actually matter
Strip away the vocabulary and there are only four things a grader is really assessing. Everything else follows from these.
| Criterion | What it tells you | Why it moves the grade |
|---|---|---|
| Companion stringers | The pallet has already had a structural failure | Adds 3–5 lb, changes notch geometry, visible |
| Replaced deck boards | How much abuse the pallet has absorbed overall | Correlates with remaining service life |
| Plugs and mend plates | A notch or stringer has failed once already | The notch is where the next failure starts |
| Squareness | Whether it will stack, rack and convey cleanly | The criterion buyers check least and suffer most from |
Squareness is the one I would push hardest on. Measure both diagonals with a tape. If they differ by more than an inch, that pallet will not sit properly on a rack beam and will jam an automated line, regardless of how good the boards look. It takes five seconds to check and almost nobody does it.
04Language that actually gets enforced
The fix is not to find an honest supplier. It is to write a specification with numbers in it, so that 'Grade A' becomes a shorthand for something measurable rather than a marketing tier.
- Grade: A, as defined below. Supplier grade names do not govern.
- Companion stringers: zero permitted.
- Replaced top deck boards: two maximum, none in a lead position.
- Plugs: one maximum. Mend plates: zero.
- Diagonal variance: 0.5 in maximum.
- Protruding fasteners: zero on either face.
- Inspection: buyer may sample 10 units per 500 and reject the load at over 2% failure.
That last clause is the whole thing. Without a sampling protocol and a rejection threshold, the preceding six clauses are aspirations. With it, a supplier knows exactly what happens if the load drifts — and drift stops happening.
05What we did about it
We publish our criteria on the pallet grades page and on the recycled wood pallets page. That was not a marketing decision; it was a defensive one. Publishing a number means we have to hit it, and it means an argument at a receiving dock resolves against a document rather than against whoever is more insistent.
We also grade twice — once at intake, once after repair, by a different person. A grader who decided a pallet was repairable is invested in that decision. A second pair of eyes at the exit costs almost nothing and it is why our reject rate on repeat accounts sits under 0.4%.
“The customers who inspect are the easiest customers we have. They tell us within a day when something drifts, and we fix it before it becomes a truckload.”
