Returns Processing That Doesn't Bury Your Warehouse: Disposition Rules, Restock Speed, and What to Measure
A return costs you twice if it sits on a bench for a week. How to write disposition rules a new hire can follow, get sellable units back to the pick face in days, and measure the part that actually moves money.
A return costs you twice. Once when the sale reverses, and again for every day the unit sits on a bench instead of back on the pick face where it could sell to somebody else. The second cost is the one nobody budgets for, and it is almost entirely within your control.
Most warehouses treat returns as an interrupt — a pile that gets worked when the outbound wave is clear. That ordering is backwards. Returns are inbound receiving with worse paperwork, and they respond to the same discipline.
The clock starts when the parcel arrives
Measure the interval from carrier delivery to sellable-and-slotted, not to "processed." A unit marked processed but sitting in a staging cage is not inventory; it is a promise.
Two-day cycle times are achievable for straightforward goods. Two-week cycle times are common and expensive: a $40 item returned in March and re-slotted in April has missed a full month of demand, and in seasonal categories it may have missed the season entirely.
The fix is scheduling, not headcount. Give returns a fixed daily window and a named owner. A pile that is somebody's job at 7 a.m. gets worked; a pile that is everybody's job at the end of the shift does not.
Give the return an identity before you open it
The most common failure is a carton that arrives with no reference — no RMA, no packing slip, sometimes not even a note. That unit becomes a research project, and research projects go to the bottom of the pile.
Three things close the gap:
- Require an RMA reference on the outside of the parcel, issued by the storefront, and match on receipt.
- Capture the carrier tracking number at the door so an unidentified carton can still be traced back to an order.
- Photograph at the bench — a five-second photo settles most damage disputes and every customer-claims argument that follows.
Once identified, the return should be scanned into a returns location the same way any inbound receipt is, so the unit exists in the system before anyone decides what to do with it.
Write disposition rules a new hire can follow
Disposition is where returns quietly bleed money, because it is usually a judgment call made by whoever is standing there. Judgment varies; rules do not. Write them down, post them at the bench, and make them binary wherever possible.
| Disposition | Trigger | Where it goes |
|---|---|---|
| Restock as new | Sealed, undamaged, within return window | Back to primary pick face |
| Restock after rework | Opened but complete; needs repack or new outer carton | Rework queue, then pick face |
| Quarantine | Suspected defect, missing components, hygiene category | Hold location, escalate to brand |
| Scrap | Damaged beyond rework, or scrap-cost exceeds unit value | Waste stream, write-off posted same day |
The judgment call worth removing first is "does this need rework?" Define it physically — outer carton crushed, seal broken, insert missing — rather than as a quality impression. A rule a temp can apply on their second day is worth more than a policy that requires five years of context.
Packaging decides how much of this you ever have to do
A meaningful share of returns are damage returns, and damage returns are usually a packing-station decision made weeks earlier. Undersized cartons, thin void fill, and the wrong box for the weight class all show up later as rework labor.
If your damage-return rate is climbing, the investigation belongs at the pack bench — the same place your carton selection and dimensional weight decisions get made — before it belongs in the returns aisle.
The inventory lie returns create
Every return in limbo is a small inaccuracy. The storefront may have already credited the customer and incremented available stock, while the physical unit sits unopened in a cage. Sell that unit and you have an oversell, a second refund, and a customer who is now unhappy twice.
Two guardrails prevent most of this: do not increment sellable stock until the unit is physically restocked, and keep returns holding areas inside your cycle count rotation rather than treating them as outside the counted world.
What to measure
- Return-to-restock cycle time, in days, median rather than average — averages hide the stuck tail.
- Recovery rate: share of returned units that go back to sellable stock. Below 60% in a general-merchandise catalog usually points at disposition rules, not at customers.
- Cost per return processed, all-in with labor and materials. It is the number that tells you whether an automated returns portal pays for itself.
- Damage-return rate by SKU, which reads as a packaging report card.
Returns will never be the profitable part of ecommerce fulfillment. They can be the fast part, and speed is what converts a reversed sale into a recovered one.
If disposition rules are the thing living in one supervisor's head today, AIDWMS can hold them as routing logic at the returns bench, so the same carton gets the same decision on a Tuesday in November as it does on a quiet Thursday in March.
