Returns management, and the returns that were never really returns

What returns management is

Returns management is the end-to-end process an e-commerce retailer runs when a customer sends something back: authorizing the return, getting the item moving, receiving and inspecting it, deciding what happens to it, and settling the refund or exchange. As drawn, it contains no step that asks why the unit came back. I once looked at a bundle that showed 157 units sold and 156 returned. The product was fine: a bundling app was splitting orders into individual line items for the warehouse.

That's not a brand opinion. The 2002 definition of returns management puts returns, reverse logistics, gatekeeping and avoidance inside one process, so deciding which returns should happen has been part of the job from the start.

Returns management is not reverse logistics

The two get used as synonyms, and the difference is the whole argument of this page.

Reverse logistics is the physical flow: transport back from the customer, receiving, inspection, restock or disposal. A movement problem. Returns management is the decision layer around that movement: whether the return should happen, on what terms, what the item becomes, and what the customer gets back. Reverse logistics is a subset of it.

Which is why you can run flawless reverse logistics on a return that should never have been created.

The returns process, end to end

Seven steps, and the names barely differ from one guide to the next. What differs is whether anyone tells you how each step fails.

1. Request: the customer asks, against a policy and a window. It fails at the reason field, where a dropdown flattens "did not fit" and "you sent the wrong item" into one word.
2. Authorization: the return is approved and an RMA is issued, which is the step where returns actually break. An approved return looks finished in every dashboard that shows it.
3. Label and transport: a label is issued and the item starts moving, or it does not. USPS return labels are scan based, so a merchant pays only for labels a customer actually uses.
4. Receipt: the parcel reaches a warehouse or 3PL and has to be matched back to its RMA.
5. Inspection and grading: condition decides what the unit is now worth. Grading that drifts between shifts quietly changes every decision downstream of it.
6. Disposition: somebody decides what the item becomes. It fails by default: nobody decides, and the unit sits.
7. Settlement: refund, exchange or credit, and the inventory reconciled.

What actually happens to a returned item

Disposition is a decision, and it's usually made after the parcel has moved, which is the most expensive moment to make it. Five paths, and recovered value drops hard from first to last.

• Restock: sellable condition, straight back into available inventory. The only path that recovers full value.
• Refurbish or repair: recoverable with labor, sold as open box or second quality.
• Liquidate: sold in bulk to a secondary market at a fraction of retail.
• Donate: written off, with whatever recovery the tax treatment allows.
• Dispose: the unit is worth less than the cost of handling it, and the freight has been paid twice.

Here's the part nobody mentions. For imported, duty-paid goods sold at retail and returned to the importer, drawback can recover up to 99% of the duties paid, provided the goods were imported within a year of the export or destruction and the claim matches on 8-digit HTSUS subheading and product identifier. Cross-border brands leave it there, because disposition is filed as a warehouse task rather than a financial one.

The returns that are not really returns

Split the pile. A return because the fit was wrong, the color wasn't what the screen showed, or the customer changed their mind is a cost of selling online, and the process above is the right tool for it. A return because the wrong SKU was picked in your fulfillment operation, the item arrived damaged, or the carrier posted a delivery exception nobody acted on is an operational break that already happened. The unit coming back is the receipt for a promise that broke days earlier, and processing it efficiently does nothing about the break.

One in five orders hits an operational break after checkout. At most brands nothing fires until the customer starts a return or sends an email, and then the break becomes a ticket. A helpdesk is a system for replying about problems, not for resolving them, so the reply goes out and the order stays broken. That's structural, not a staffing problem, and the work belongs to a different function: proactive e-commerce operations, which isn't a better helpdesk.

E-commerce is now 16.9% of US retail sales, and the National Retail Federation puts returns at 19.3% of online sales in 2025, inside $849.9 billion of retail returns, or 15.8% of annual sales. Reported rates, not benchmarks to hit.

The clock returns are already running against

Refunds and cancellations carry statutory deadlines, and they differ by market.

• United States: the FTC's Mail, Internet, or Telephone Order Merchandise Rule requires a seller that can't ship inside the window it advertised, or 30 days if it advertised none, to get the buyer's consent to the delay or refund. Where the refund isn't going back to a credit card account, 16 CFR 435.1 defines a prompt refund as one sent within 7 working days.
• European Union: a consumer has 14 days to withdraw from a distance purchase without giving a reason, and the trader must refund within 14 days of being told, though it may hold the money until the goods arrive.
• United Kingdom: the Consumer Contracts Regulations 2013 set the cancellation period at 14 days from the day the goods come into the consumer's possession, or the last delivery of a multi-item order.

The law treats a refund as a deadline rather than a courtesy, and the clock starts from an event the brand often doesn't notice.

Where a return breaks after you approve it

Every process diagram on this subject runs forward and finishes. These three states are where the "where is my refund" contact gets manufactured.

The label that was never used. The return is approved and open in the portal, and the parcel never moved. Nothing is billed for an unscanned label, so nothing prompts anyone to chase it, and the customer waits on a refund.

The parcel that arrived and was never matched. The unit is on the dock and unmatched to its RMA, so the warehouse reads it as unexpected stock and the returns system still reads it as in transit. Nobody notices until the customer asks.

The refund that was approved and never issued. Disposition is done, the record says resolved, and the money never left. This is the one with a legal deadline attached to it.

Each system is telling the truth about its own leg. The return is broken between them.

Running returns as a decision

A real decision layer runs in three moves. Detect the break from signals your systems already emit, a late scan, a short pick, a damaged delivery, not from the return request. Decide against the promise made at checkout, which sometimes means refunding or replacing before the item moves, and sometimes means the item shouldn't move at all. Act, and tell the customer before they ask. That's what post-purchase operations is for.

The edges matter. Keeyu isn't a returns portal, not a returns platform, not a 3PL, not a WMS, not a carrier and not a helpdesk. We don't print labels, receive parcels or grade returned stock. We work on the break that caused the return, and on the break that happens inside one.

If your returns process is efficient at handling consequences and blind to causes, the queue won't shrink, because a good share of what sits in it was created upstream and arrived wearing a returns label. Every order is a promise, and Keeyu keeps the promise: we watch every order against what it was promised, catch the ones that break, and act before the customer opens a return. See what the Keeyu platform does after checkout.

• Rogers, Lambert, Croxton and Garcia-Dastugue. The Returns Management Process, International Journal of Logistics Management 13(2), 2002. The definition that places returns, reverse logistics, gatekeeping and avoidance inside one process.
• National Retail Federation and Happy Returns. 2025 Retail Returns Landscape. An estimated 19.3% of online sales returned in 2025.
• National Retail Federation. Returns projected at $849.9 billion in 2025. $849.9 billion of total retail returns, 15.8% of annual sales.
• United States Postal Service. Business Return Services. Return labels are scan based, so a merchant is charged for labels that are actually used rather than for labels issued.
• Legal Information Institute. 19 CFR 190.45, returned retail merchandise. Drawback of up to 99% of duties paid, the one-year condition, and the 8-digit HTSUS subheading and product identifier match.
• US Customs and Border Protection. Drawback overview. Plain-language framing of drawback, cited alongside the regulation rather than as sole support for the figure.
• Federal Trade Commission. Mail, Internet, or Telephone Order Merchandise Rule. The advertised shipping window, the 30-day default, and the seller's obligation to obtain consent to a delay or refund.
• Legal Information Institute. 16 CFR 435.1. Prompt refund defined: 7 working days where the refund is not made to a credit card account, one billing cycle where the seller is the creditor.
• European Commission, Your Europe. Returns and the right of withdrawal. The 14-day withdrawal period, the 14-day refund deadline, and the trader's right to hold the refund until the goods arrive.
• UK legislation. Consumer Contracts Regulations 2013, regulation 30. The 14-day cancellation period, running from possession of the goods or the last delivery of a multi-item order.
• US Census Bureau. Quarterly Retail E-Commerce Sales. The scale of US e-commerce that the returns leg sits behind, at 16.9% of total retail sales in Q1 2026.