A manifest is the seller’s line-by-line claim about what is on a pallet or in a truckload. It is not a guarantee, and it is not an appraisal. It is a starting point for your own arithmetic, and the operators who make money on returned goods are the ones who treat it that way.

Here is how to read one properly, and what to check before the load is paid for.

The columns that matter

Most manifests, whatever format they arrive in, carry some version of the same fields.

Quantity. How many units of that line are claimed. Simple, and the first thing to verify against the physical count when the truck arrives.

Unit retail. The claimed retail price of a single unit. This is where optimism creeps in. A unit retail pulled from a launch-day list price is not the same as what that item sells for today, and nobody is obliged to update it.

Extended retail. Unit retail multiplied by quantity. Sum the column and you have the gross retail value, or GRV, of the load. This is the number sellers lead with, because it is the biggest number on the page.

UPC. The barcode. The single most useful field on the sheet, because it is the one thing you can independently price. If a manifest has no UPC column, you are buying a description rather than a product.

Condition. Sometimes a letter grade, sometimes descriptive text, sometimes absent. More on this below.

Category or department. Useful for a fast sanity check on whether the load matches what you actually sell.

Gross retail value is not a price

The most common way to lose money on a load is to treat GRV as if it were the value of the goods. It is not. It is the value the goods had when they were new, on a shelf, in packaging, with a warranty.

What you are buying is a mixed pile of customer returns. The useful number is the cost percentage: what you paid divided by the extended retail. A load with ten thousand dollars of extended retail bought for fifteen hundred is a fifteen percent cost percentage. Whether that is a good buy depends entirely on the category, the condition mix, and what you can recover.

Do the second calculation too. Landed cost per unit is the bid plus the buyer’s premium plus freight, divided by the number of units. Freight on liquidation loads is not a rounding error, and it is frequently a meaningful share of what the load actually costs you. If your landed cost per unit is above what the average unit will fetch, no amount of extended retail on the manifest will save the deal.

Condition columns, and what they leave out

Condition grading in this industry is not standardised, which means the same word means different things from different sellers.

You will see two systems. Letter grades run A, B, and C, roughly mapping to like new, open box, and damaged or salvage. Descriptive grades are more common on manifests and more useful: new, like new or open box, used or good, untested, and salvage or as-is.

Pay particular attention to untested. It means the seller processed the returns without checking whether they work. Assume a mix of working and defective units, and price accordingly. Buyers who have been in the business a while apply a substantial discount to untested loads, and they are right to.

Scratch and dent is appliance-specific and usually means cosmetic damage with function assumed intact unless the manifest says otherwise. Assumed is doing real work in that sentence.

If the condition column is missing entirely, that is not a neutral omission. It is a decision someone made.

Six red flags

Round-number retails. A manifest where a suspicious number of unit retails land on clean figures has usually been estimated rather than pulled from a system.

No condition column. As above. Sellers who grade honestly generally say so.

One line carrying most of the value. If a single high-retail item accounts for a large share of the extended retail, the load’s economics depend entirely on that one unit being present and working. Frequently it is neither.

The same manifest on multiple listings. If you recognise the file from a different pallet, you are looking at a template, not an inventory.

Manifested versus unmanifested confusion. A partially manifested load gives you category counts rather than item lines. That is a different product from a manifested load and should be priced as one.

Count variance on arrival. The manifest says four hundred and twelve units. Somebody needs to confirm that four hundred and twelve units actually came off the truck.

That last one is the only red flag you cannot check before you bid, and it is the one that costs the most.

Reconciling the manifest against what arrived

Here is the part most operations skip, because it is tedious and the truck is blocking the dock.

The manifest is a claim. Your scan-in is the fact. Reconciliation is the process of putting the two side by side and finding the gap while you still have standing to raise it.

In practice that means importing the manifest before the load is unloaded, so every claimed line exists in your system as an expected unit. Then, as your crew scans units in and grades them, each scan matches against a line. When the load is checked in, three things are visible:

• Lines that were claimed but never scanned. These are your shortfalls.
• Units that were scanned with no matching line. These are unmanifested extras, which are sometimes a bonus and sometimes a sign the manifest belonged to a different pallet.
• Lines where the condition you recorded is materially worse than the condition claimed.

None of those are recoverable if you find out three weeks later while wondering why recovery on that load looked bad. They are recoverable, or at least arguable, on the day.

Storage time is pure cost, so fast receiving matters for its own sake. But receiving fast and receiving accurately are not in conflict when the manifest is already in the system and your crew is scanning rather than writing on a clipboard.

What good looks like

An operator who reads manifests well does four things consistently. They price off cost percentage and landed cost per unit rather than off gross retail value. They discount hard for untested and for missing condition data. They check the structural red flags before bidding rather than after. And they reconcile every load against its manifest at the dock, so a bad supplier is identified after one load instead of after five.

None of that requires a bigger warehouse or more staff. It requires the manifest and the scan data to live in the same place, so the comparison is automatic rather than a job somebody has to remember to do.

See how manifest import and reconciliation works in LiquidationRoute.