Here is the uncomfortable truth most retail operators refuse to sit with: you already have a returns strategy. You just built it by accident, one packing slip at a time.
Returns get filed under reverse logistics. There is a cost center, a 3PL contract, a restocking workflow, and a VP who owns the number. The whole apparatus is designed to move a box backward as cheaply as possible. It is efficient. It is also solving the wrong problem.
Because the return did not start in the warehouse. It started at the price tag, the size chart, and the product description. By the time a box comes back, the merchandising decision that caused it was made weeks earlier. Reverse logistics just pays the invoice.
Return rate is the least controllable number you track
A recent Richpanel analysis said the part most teams talk around. One in five online orders comes back, and you cannot engineer that to zero. The brands that protect their margin track refund rate and cost-per-return, not the headline return rate.
Sit with that. The number your logistics VP is graded on is the number they can least influence. There is no single good number, because return rate is mostly set by category. The NRF’s 2025 Retail Returns Landscape puts the overall ecommerce rate at about 19 to 20% of online orders. Apparel alone runs far higher. Apparel runs 20 to 40%, footwear 17 to 30%, electronics 8 to 15%, and beauty 4 to 12%.
If your rate is structurally set by what you choose to sell, then chasing the portfolio average is theater. The signal you actually want is buried underneath it: which specific SKUs are returning above their own baseline, and why.
That “why” is almost never a logistics answer. Sizing, fit, and color are the primary reasons for 45% of all retail returns. 16% of returns are due to damage, while 14% are due to inaccurate item descriptions. Fit. Color. Copy that oversold. Those are merchandising and pricing decisions. Every one of them.
Every return is a repricing event you did not authorize
Here is the reframe. A return is the customer telling you the item was mispriced for the value delivered, mis-sized against the chart you published, or misrepresented in the photos you shot. You set a price assuming the sale would stick. When it does not, you eat the full round trip and resell the unit at a markdown, if you resell it at all.
That is a price cut. You just did not put it in the merchandising plan.
The math is brutal at scale. Returns processing costs $5 to $15 per item in 2026, but the all-in cost of a return reaches $17 to $29 once return freight, depreciation, and support labor are counted. Now multiply by the recovery gap. Once you add the difference between original selling price and what the item ultimately resells for, the total cost of a return diverges sharply by category. Bulk, fragility, and resale depreciation are what drive the spread.
So the effective price of a returned unit is negative. You paid to ship it out, paid to ship it back, paid labor to inspect it, and then discounted it to move it. If a merchant proposed that pricing structure in a planning meeting, they would be fired. Do it silently through returns and it is just “the cost of ecommerce.”
Route the signal to merchandising, not the dock
The fix is not a better 3PL. It is treating the return reason as first-class merchandising telemetry.
Richpanel’s own prescription points the right direction. Find the specific products driving an above-baseline rate and fix the real cause (sizing, photos, packaging) instead of chasing a portfolio average. That is the whole game. Stop averaging. Start diagnosing at the SKU.
From the forward-deployed seat, this is where production AI actually earns its keep in retail. Not another chatbot on the returns portal. An agentic loop that reads return-reason text at the unit level, clusters it against price band, fit data, and PDP copy, and hands merchandising a ranked list: this SKU is returning for “too small,” reprice the size curve or fix the chart; this one is returning for “not as pictured,” reshoot; this one is a genuine price-value gap, adjust or discontinue.
The cost of doing nothing here is not a one-time hit. It is a subscription. You keep repricing the same defect thousands of times a season, financing a merchandising error through a logistics budget, and reporting it as operational efficiency because the box moved cheaply. The CODN is the entire recovery gap, compounding, quietly, quarter after quarter.
Free returns will not save you, and neither will fees. Both are policy levers on a merchandising problem.
Stop asking your logistics team to win a race that starts at the price tag. Route the return reason to the people who set price, fit, and copy. Because you are already making a pricing decision every time a box comes back. The only question is whether you make it on purpose next season, or keep making it by accident.