Ask ten liquidation operators how their last truckload performed and most will tell you what it sold for. Fewer can tell you what it recovered. The difference between those two answers is the difference between a business that knows which suppliers to buy from and one that finds out eighteen months late.

What recovery rate is

Recovery rate is what a load returned, expressed against what the goods were worth at retail.

At its simplest: total sales from the load, divided by the load’s gross retail value, times one hundred. If a pallet carried eight thousand dollars of extended retail on the manifest and produced twelve hundred dollars in hammer prices, recovery on that pallet was fifteen percent.

It is a deliberately blunt instrument. It does not tell you profit, it does not account for labour, and it does not care what you paid. What it does is give you one comparable number across loads of wildly different sizes and categories, which is exactly what you need when deciding whether to bid on a supplier’s next truck.

Why the published benchmarks contradict each other

If you go looking for a benchmark you will find a mess. Some sources cite recovery in the region of fifty to seventy percent. Operator-facing sources routinely cite fifteen to thirty-five percent for mixed customer returns, and category medians well into single digits for things like apparel and beauty.

Both can be true, because they are measuring different things from different sides of the transaction.

A marketplace reporting recovery for a retailer client is measuring what the retailer got for goods it already owned, often against a cost basis rather than full retail, and often on cleaner merchandise. An operator measuring recovery on a pallet they bought is measuring resale against the manifest’s retail figure, on goods that have already been picked over.

So before you compare your number to anyone’s benchmark, check the denominator. Recovery against gross retail value, recovery against your landed cost, and recovery against a wholesale reference are three different metrics wearing the same name. Pick one, define it in writing, and use it consistently. An internally consistent number you trust beats an industry-standard number you cannot reproduce.

Measure it per pallet, not per load

The most common mistake is calculating recovery at the truckload level. It feels natural, because that is the unit you bought. It is also where the useful signal disappears.

A truckload is a portfolio. Inside it there will be pallets that recovered forty percent and pallets that recovered four. Averaged together you get a number that describes nothing you can act on. Measure per pallet and patterns start appearing: a particular category consistently underperforms, one supplier’s pallets are consistently light, general merchandise recovers predictably while a specific department never does.

That is the level at which sourcing decisions actually get made.

Count what did not sell

Here is where most recovery calculations quietly flatter themselves.

If you divide sales by the retail value of only the units that sold, you are not measuring recovery. You are measuring the performance of your winners. The units still sitting on a rack, the ones that were never listed, and the ones written off as damaged are all part of what that load returned. Excluding them makes a bad load look acceptable.

Three buckets need to be in the calculation:

• Sold. Straightforward, and the only part most operations track.
• Unsold but listed. Stock that went to auction and did not clear, or has not yet. This is inventory carrying cost accruing against the load.
• Never listed. Units that came off the pallet, got a bin, and were never published. This is the bucket that hides. It does not appear on any sales report, because it never entered the sales process.

An honest recovery number takes the full retail value of the load as the denominator, including everything that never moved.

The time dimension

Recovery is not a fixed property of a load. It decays.

A lot priced correctly on day one commonly needs a meaningful price reduction by week six, and the cost of holding it is running the entire time. Pallet storage is not free, and floor space that is holding aging stock is not available for the next truck. There is a real scenario where a warehouse cannot receive because it cannot ship, and the load blocking the dock is one somebody bought eight months ago.

So track days to sale alongside recovery. A load that recovers twenty-five percent in three weeks and a load that recovers twenty-eight percent over five months are not comparable, and the faster one is usually the better buy.

The other numbers worth having

Recovery rate is the headline, but it works better with company.

Sell-through rate, the percentage of lots that clear per auction cycle, tells you whether your lotting and pricing are right independent of what the goods were worth.

Landed cost per unit, the bid plus buyer’s premium plus freight divided by units, tells you the floor under which selling is pointless.

Cost per pallet processed, your labour and overhead divided by pallets through the door, is the number that decides whether a marginal load is worth touching at all. Every additional touch has a cost, and on low-value goods the processing can exceed the recovery.

Put those next to recovery and you can answer the real question, which is not “did this load make money” but “should we buy the next one”.

Making it something you actually see

The reason most operations do not track recovery properly is not that the maths is hard. It is that the inputs live in different places. Manifest retail is in a spreadsheet or an email attachment. Sale prices are on the auction platform. Unsold stock is a walk around the warehouse. Assembling those three by hand is a month-end job, and month-end jobs get skipped in busy quarters.

It becomes routine when the manifest is imported at intake so retail value is captured on the unit, when every unit carries the load and supplier it arrived on, and when invoices come back from the sales channel against those same units. At that point recovery is a report rather than a project, and unsold and unlisted stock are already counted because the system knows they exist.

The operators who scale past the point where intuition works are the ones who made that number automatic. Not because the number is magic, but because they stopped buying second truckloads from suppliers whose first one did not perform.

See how recovery and revenue reporting works in LiquidationRoute.