Shelf Pulls vs. Returns vs. Overstock

Shelf pulls are new, tagged merchandise that sat on a retail sales floor but was never sold and never owned by a customer. Customer returns were bought and brought back, so condition varies unit to unit. Overstock never reached a sales floor at all — it is factory-fresh excess pulled straight from a distribution center. Those three labels, not brand or retail value, are what decide whether a lot makes money for you.

Three pallets can carry identical retail value, brand mix, and footprint and still produce three completely different P&Ls. Buyers who gloss over condition blame the wrong thing when a lot underperforms: they decide they overpaid, when they paid a fair price for the wrong condition for their channel. Here is what each term means on a manifest in 2026, what each does to sell-through and labor, and how to match condition to how you sell.

What are shelf pulls, customer returns, and overstock?

All three are liquidation goods that leave the retail supply chain at different points. Shelf pulls leave from the sales floor, returns from a customer's house, overstock from the warehouse before it is ever shelved. That exit point determines condition, price, and how much work each unit needs before it sells.

Shelf pulls

Merchandise that sat on a retail sales floor but never sold. It comes off the rack at a season change, a planogram reset (the periodic remerchandising of a store's fixtures), a store closure, or a markdown cycle that ran out of road. Units are new, usually tagged, and the only wear is handling — a rack crease, a scuffed shoe box, a bent hangtag.

The defining trait: no customer ever owned it. Nothing was worn, washed, or returned.

Customer returns

Merchandise a customer bought and brought back. Because retailer return policies are generous, the condition spread inside a returns lot is wide. A large share of units are untouched — wrong size, changed mind, duplicate gift. A middle band was tried on or worn once. And there is always a tail: worn hard, missing components, damaged, or a box holding something other than what the label says.

The defining trait: condition varies unit to unit. Returns are graded in aggregate, never per unit.

Overstock

Merchandise that never reached a sales floor. It sat in a distribution center as excess — an over-forecast, a canceled wholesale order, a shipment that landed after its season window closed. Frequently still in original case packs, tags intact.

The defining trait: factory-fresh, but often narrow. Deep quantities of few SKUs, with a size curve skewed toward whatever did not sell through the original buy.

How much do shelf pulls, returns, and overstock cost?

As a share of original retail, overstock typically runs 18–30 percent, shelf pulls 12–22 percent, and customer returns 6–15 percent. Those are typical market ranges, not quotes on specific inventory — real pricing moves with brand mix, category, season, and lot size. The cheaper the condition, the more of your margin gets spent on labor and losses instead of goods.

Read the table left to right and returns look like the obvious win: cheapest entry, most units. That is true, but only if your labor is cheap or free, and only if your sales channel can absorb the damaged tail without damaging your reputation.

What does the same budget buy in each condition?

Put $2,400 against apparel in each of the three conditions and the returns lot buys roughly 700 units, shelf pulls roughly 500, and overstock roughly 400. The returns lot produces the most gross revenue per dollar spent on goods and the least revenue per hour worked. Here is the full comparison, using the ranges above as illustrative examples.

Three things fall out of this that are worth more than the specific numbers.

Returns beat overstock, but you pay for it in hours. The returns lot clears $1,282 more margin than the overstock lot — and costs you ten additional hours of sorting, or $128 an hour for the extra effort. If you are an owner-operator with time and no payroll, that is an excellent wage. If those hours would otherwise go to sourcing, listing, or running live sales, you just bought yourself a job instead of a margin.

Shelf pulls win on risk-adjusted return. They cleared $338 more than returns on nearly eight fewer hours of work, with zero disposal and no condition-complaint risk. For most single-location operators, shelf pulls are the quiet best buy — the reason they get overlooked is that the per-unit price looks high next to returns.

Average realized price is not the same across conditions. Shelf pulls sell for slightly more than returns of the same brand because the customer can see they were never owned — a premium that is small per unit and enormous in aggregate.

Why is processing labor the hidden cost of customer returns?

Because a returns unit takes roughly 45 seconds longer to handle than an overstock unit, and on a 700-unit lot that difference is about 8.75 hours of work. Nobody underestimates the cost of a pallet. Everybody underestimates the cost of touching it.

A returns unit needs inspection, a condition call, sometimes a wash or a steam, sometimes a photo, and a price-tier decision. An overstock unit needs a cut from the case pack and a price. Run the arithmetic before you buy: units in the lot, seconds per unit, divided by 3,600, times your real hourly cost. If you are the labor, use the number you would charge a client, not zero.

Which condition should you buy for your sales channel?

Bin stores should buy customer returns, boutiques and live sellers should buy shelf pulls, and marketplace and export sellers should buy overstock. The rule behind it: the more directly your customer inspects the item before paying, the more you should pay for predictable condition.

A bin store is a retail format that prices an entire table of mixed merchandise at one flat price per day, dropping through the week — which is why it tolerates condition variance better than any other channel. If that is your model, the volume logic flips the whole analysis; read where bin stores get inventory alongside this.

Where does each condition actually come from?

Shelf pulls and overstock come out of retailer and brand channels directly; customer returns come out of returns-processing centers that consolidate what shoppers send back. The number of hands between that origin and you is the single best predictor of what shows up on the dock.

Marketplace returns are the most heavily traded of the three and therefore the most picked over — a lot advertised as returns may have passed through two or three intermediaries, each pulling the best SKUs. If that is the lane you are shopping, our guide on where to buy Amazon return pallets covers how to tell a direct-source seller from a reseller of resold goods. Shelf pulls and overstock change hands less often, because fewer operations are set up to move them in volume.

WCD Wholesale, a B2B marketplace for manifested apparel and footwear liquidation, sells manifested lots by stated condition rather than by mystery grade — a manifest being the line-by-line document listing SKU, description, quantity, and retail price for every unit in the lot.

How do you verify what you are actually buying?

Get five things in writing before you wire money: the source tier, a unit-level manifest, a contractual defect tolerance, confirmation the lot was not picked, and the size curve. Any seller who will not put those in writing is selling you optionality at your expense.

• What is the source tier? Direct-from-retailer, or has the lot been through one or more middlemen? Every hop increases the odds it was cherry-picked.
• Is there a unit-level manifest? A real manifest lists SKU, description, quantity, and retail price per line. A category summary reading "500 assorted apparel" is not a manifest. Our walkthrough on reading a liquidation manifest covers what a legitimate one looks like line by line.
• Is the defect rate contractual? "Approximately 10% damage" in an email is a hope. A written tolerance you can claim against is a term.
• Has it been picked? Ask directly whether high-value SKUs were pulled before the lot was assembled. Sellers who do not pick will say so plainly.
• What is the size curve? Overstock in particular can arrive as 60% XS and XXL. Ask for the size distribution, not just the unit count.

What are the most common mislabeling traps?

The three most common are returns sold as shelf pulls, picked-over remainder sold as overstock, and grade letters used in place of an actual condition description. All three work because the buyer cannot verify the claim until the freight is already paid for.

"Shelf pulls" that are returns. The most common substitution in the market, because shelf pulls command a higher price and the difference is not visible until you are elbow-deep. Tell: a defect rate above roughly 10 percent, or wear rather than handling marks.

"Overstock" that is a picked-over remainder. Genuine overstock has a coherent size curve, because it is what was left of a real buy. Remainder posing as overstock has a broken one.

Grade letters with no definition behind them. A-grade, B-grade, and C-grade mean nothing standardized across the liquidation industry — one seller's B is another's A. Insist on a plain-English condition description and a defect tolerance instead of a letter.

The broader mechanics of how lots are assembled, priced, and freighted are covered in the liquidation pallets buyer's guide.

Is one condition better than the others in 2026?

No — cheapest per unit is not cheapest per sold unit, and the right condition is the one that matches your channel. Returns give you volume and margin if you have hours to spend and a channel that forgives variance. Shelf pulls give you the best margin per hour worked and the lowest complaint rate. Overstock gives you predictability and listing efficiency at the cost of assortment.

Pick the condition that matches how you sell, then negotiate on price. Buyers who work in that order stop having bad pallets — they have pallets priced correctly for what they were. Every lot in our current inventory at WCD Wholesale ships with a unit-level manifest, a stated condition, and freight included, so you can run this math before you commit rather than after.