The Unsaleables Allowance: Paying for Damage That Didn't Happen

Nothing was damaged in March. Cinderhaven Provisions paid $7,500 for damage anyway, the way it does every month: a 2% spoils allowance deducted off-invoice from its natural-channel billing, $90,000 a year, whether the trucks arrive clean or not. Cinderhaven is a fictional company running on a synthetic dataset, built to illustrate an allowance structure that sits in distributor contracts across the natural channel, and that almost nobody reconciles against the damage that occurred.

Cinderhaven's verified unsaleables run 0.8% of the same billing, $36,000. The other $54,000 pays for damage that did not happen.

The allowance is priced like insurance and behaves like a fee

Unsaleables are products pulled from distribution for damage, expiration, or discontinuation. The industry's own benchmark series, run jointly by FMI and GMA through the 2000s, put the rate at 1.11% of sales in 2003, $2.57 billion in the warehouse-delivered grocery channel alone, down from a record 1.18% the year before. The mechanism most specialty brands pay through is older and simpler than the problem: a swell allowance, a flat 1-3% deducted from every invoice, agreed in advance, no claim filed, nothing shipped back, named for the swollen cans it once reimbursed.

The flat rate was institutionalized in the early 2000s as the adjustable rate policy: a pre-set percentage of sales paid in lieu of reimbursing actual damaged product through reclamation centers. The same white paper's data shows what happens to a rate nobody reconciles: the industry's unsaleables rate climbed from 0.75% of sales in 1994 to 1.14% in 2001. And the composition matters: per the 2004 benchmark, in rounded figures, damage accounts for 58% of unsaleables, expired product 22%, discontinued 13%, seasonal 6%. More than half the category is handling damage, which varies by channel, packaging, and distributor, none of which a flat percentage sees.

It began as reimbursement for swollen cans. It survives as a percentage of everything.

Cinderhaven pays 2%. Its damage runs 0.8%.

Cinderhaven's natural-channel billing through UNFI and KeHE runs $4.5M a year. Its contracts carry a 2% spoils and unsaleables allowance: $4.5M × 2% = $90,000, taken off-invoice, invoice by invoice. Its verified unsaleables, the spoils-coded deductions plus reclamation detail actually documented across the year, total 0.8%: $36,000. Most of the line is shelf-stable jars in corrugate, which ship with little damage; only the refrigerated fresh-salsa line runs higher. The gap is $54,000 a year. The allowance is two and a half times the damage it insures against, and it renews automatically with every order.

The same year, priced under the three models the industry has used:

| Payment model | What Cinderhaven pays | When the number gets checked | |---|---|---| | Flat swell allowance, 2% | $90,000 | Never; it is a contract term, not a measurement | | Adjustable rate policy | Pre-set %, in theory data-based | At renegotiation, only if someone brings data | | Pay on actual, via reclamation | $36,000 plus claim admin | Every claim, itemized |

The industry moved from the third row to the first in the name of efficiency, and the efficiency is real: no claims, no paperwork, no reclamation invoices. What was traded away is the connection between the payment and the event. A flat allowance rises with sales, not with damage.

The ratchet only turns one way

The natural channel adds a mechanism that makes the flat rate strictly worse than it looks. Confido's guidance for brands launching into KeHE puts it plainly: budget about 1% each for spoilage, returns, and merchandising allowances, and "if it ends up being more, KeHE will bill it". The overage arrives as the CS spoilage code already on the remittance. So when actual spoils exceed the allowance, the difference is billed. When actual spoils run under it, no credit arrives. The allowance is a floor for the distributor and a ceiling for no one.

The number never self-corrects, and there is research on why. An INSEAD working paper on the drivers of product expiration found that manufacturers compensate retailers for unsaleables without precise knowledge of what each party contributed, so the reimbursement mechanism favors whichever side holds the power. Nobody can tell from an expired jar whether slow rotation or over-ordering put it there, so each side attributes it to the other, and the flat rate settles the argument in advance. Meanwhile the allowance itself hides off-invoice, above the deduction ledger, in the tier of givebacks a gross-to-net analysis surfaces and a P&L review never does.

The rate settles an argument nobody can audit, and it settles it the same way on every invoice.

One spreadsheet is the whole renegotiation

Reconciling this takes one afternoon and two files: twelve months of spoils-coded deduction rows, and the reclamation or spoils detail the distributor will provide on request. Divide documented unsaleables by billing, per partner. That number against the contracted rate is the entire audit, and it converts the allowance from a fixed cost into a recoverable one, which is the same discipline that separates brands that recover deductions from brands that absorb them.

The precedent for data moving the rate exists on both sides. Publix cut its unsaleables return percentage by more than 16% in two years once it started refusing damage at the receiving dock: behavior changed when someone measured. And the industry's own 2005 joint recommendations, quoted in the same Supermarket News report, say allowance rates should be built on statistically sound data. A brand that arrives at renewal holding its actual rate is asking the contract to meet the standard the industry set for itself.

The allowance funds the meeting that shrinks it.

Your contracted rate against your actual rate

Pull the unsaleables or spoils clause from each distributor contract, and export the year's spoils-coded deduction rows to go with them. I will write back with your contracted rate against your actual rate, partner by partner, and the renegotiation number the gap supports. Two files, one afternoon. The 2% stops being furniture.