Start with the lot’s evidence grade
The same merchandise description can support very different assumptions. A buyer who inspected every item has better evidence than a buyer looking at one wide photograph. Before estimating value, grade the information itself.
- Strong evidence: unit-level manifest, model numbers, clear condition notes, multiple current photographs, and an inspection opportunity.
- Moderate evidence: a credible manifest and useful photographs, but limited testing or incomplete condition notes.
- Weak evidence: category totals, stock photographs, unmanifested gaylords, customer-return labels without testing, or no inspection.
Evidence quality should change the assumption. Weak evidence calls for a lower base sellable rate and a wider downside case. It should never produce the same forecast as a tested, itemized lot.
Use condition buckets with explicit rules
| Bucket | Decision rule | Revenue treatment |
|---|---|---|
| Ready to sell | Complete, functional, legal to sell, and suitable for the intended channel after ordinary cleaning. | Use a conservative sold-comp value. |
| Repairable | A known, economical repair can produce a marketable item. | Use repaired value minus parts, labor, failure risk, and added hold time. |
| Parts or salvage | Not economical as a complete unit, but identifiable components have a realistic market. | Use net parts value after disassembly, listing, fees, and leftovers. |
| Disposal | Unsafe, prohibited, incomplete without value, contaminated, or uneconomical to process. | Use zero revenue and add disposal cost where applicable. |
Do not count a unit as repairable because it might be fixable. Identify the likely failure, parts availability, expected repair time, and the probability the repair succeeds. Unknown problems belong in a more conservative bucket.
Sample across the lot, not just the top layer
If inspection is allowed, choose units from different pallet positions, cartons, brands, and value bands. A convenient sample from the most visible boxes can be biased toward cleaner inventory. Record each sampled unit in the four buckets, along with missing accessories, cosmetic damage, power-on result, and any safety concern.
Suppose a 60-unit pallet allows a 15-unit inspection. You classify 8 ready to sell, 3 repairable, 2 parts-only, and 2 disposal. The raw ready-to-sell rate is 8 ÷ 15, or 53%. If half of the repairable units are expected to become sellable, their weighted contribution is 1.5 units.
That does not automatically justify a 63% forecast for the whole pallet. If the sample was limited or the uninspected boxes differ, apply an uncertainty haircut. A 10-point haircut would produce a 53% planning rate, or about 31 sellable-equivalent units out of 60 after rounding down.
Value each bucket separately
A single average price hides the difference between a complete item and a damaged item. For the 60-unit example, assume 27 units are ready to sell at a conservative $42 net sale price, 8 are repair candidates with a 50% success rate and $18 net contribution per attempted repair, and 10 produce $5 each in parts value. The rest have no revenue.
- Ready-to-sell contribution: 27 × $42 = $1,134
- Repair contribution: 8 × $18 = $144
- Parts contribution: 10 × $5 = $50
- Gross expected recovery: $1,328
The $18 repair contribution must already be net of parts, extra labor, selling fees, shipping, and failed repairs. If it is only the resale price, the model overstates value.
Account for correlated defects
Random sampling works poorly when defects are correlated. A pallet may come from one retailer return batch, one failed installation program, or one model with a common defect. Ten apparently clean units do not prove that the rest are independent. Watch for repeated return labels, identical damage, water exposure, crushed corners in one pallet zone, missing proprietary accessories, or a recalled model.
When defects appear correlated, build the downside case around the batch—not a single unit. For example, reduce the expected success rate for every unit of that model or treat a whole affected layer as disposal until better evidence exists.
Run base and downside cases
The base case should be conservative and plausible. The downside case asks what happens if the two weakest assumptions disappoint at the same time. For a mixed return pallet, reduce both the sellable rate and the average sold price. Also increase disposal or labor costs if those are uncertain.
- Base case: 53% sellable-equivalent rate and $42 net price.
- Downside case: 40% sellable-equivalent rate and $34 net price.
If the downside case creates an unacceptable loss, the bid needs to fall even when the base case looks profitable. The Liquidation Pallet Calculator lets you compare these assumptions with acquisition costs and target profit.
A pre-bid worksheet
- Record the total manifested or estimated unit count.
- Grade the evidence as strong, moderate, or weak.
- Classify an unbiased sample into ready, repairable, parts, and disposal buckets.
- Weight repairable inventory by a defensible success rate.
- Apply an uncertainty haircut for sample limits and correlated defects.
- Use sold comps for each meaningful value band, not stated retail.
- Subtract testing, repair, listing, returns, storage, and disposal costs.
- Run a downside case before converting recovery into a maximum bid.
The result is not certainty. It is an auditable estimate: another buyer can see why you chose the rate and which assumption would change the bid.