Free Returns Are Not Free: The Cost Per Order Nobody Budgets
A free-returns policy costs you on every order, not just the returned ones, because the return rate, the outbound shipping, the processing, and the fraud all come out of the same margin pool. NRF and Happy Returns estimate 19.3 percent of online sales will be returned in 2025, worth $849.9 billion, and about 9 percent of those returns are fraudulent. Price the policy per order, not per return, or it will eat your margin quietly.
Most stores treat free returns as a conversion lever and never as a line item. The policy goes on the product page, conversion ticks up, and nobody runs the math until the quarter closes and the margin is gone. That is the wrong order. Returns are a cost of selling, and they land on every order you ship, not only on the ones that come back in a bag.
NRF and Happy Returns estimate that 19.3 percent of online sales will be returned in 2025, about $849.9 billion of merchandise, and that roughly 9 percent of all returns are fraudulent. Loop Returns puts the adjusted apparel return rate at 23.2 percent across 22 million returns from more than 4,000 Shopify merchants. If you sell apparel, close to one order in four comes back. The question is not whether to offer returns. It is what each return actually costs you, and who pays for it.
What one return actually costs
Start with the return itself. Outbound shipping you already paid, return shipping you now pay, warehouse receiving, inspection, steaming or repackaging, restocking, and the refund. A $60 order that comes back can easily carry $15 to $20 of handling before you count the product at all. Add the refund, and the order is a loss even when the item goes straight back on the shelf in sellable condition.
Then count the orders that never return anything. The policy still costs you on those, because the return rate is priced into every order whether or not that customer uses it. If 20 percent of orders return and each return costs $18 to process, the policy costs $3.60 on every order you ship. That is the number that belongs in your contribution math, not the headline return rate.
The fraud slice nobody models
About 9 percent of returns are fraudulent, per NRF, and that slice behaves differently from the rest. Wardrobing, empty-box returns, and serial returners do not show up as a return rate problem. They show up as inventory shrinkage and customer-service hours. A returns portal that auto-approves everything hands those customers a faster machine for taking product.
The fix is not to interrogate every customer. It is to tier the policy. First-time returns get the fast path. Repeat returners, high-value items, and worn-condition claims get a human look. Most stores find the fraud concentrated in a small slice of accounts, and a tiered flow catches most of it without slowing down the honest majority.
Pricing the policy instead of fearing it
Run the numbers per order, not per return. Take last quarter's returns, multiply by your true cost per return, and divide by total orders. That is your returns cost per order. Then test the policy against it: a paid-returns option, a store-credit incentive instead of a refund, or an exchange-first flow that keeps the revenue. Store credit costs you far less than a refund, because the margin on the replacement sale is real money back.
Watch what changes when you test. A $5 return fee usually cuts the return rate by a few points and barely touches conversion, because the customers who return constantly are not your best customers anyway. The ones who matter are the first-time buyers deciding whether to trust you, and for them the policy text on the product page matters more than the fee buried in the returns page.
Related reading
For the background behind “Free Returns Are Not Free: The Cost Per Order Nobody Budgets”, these earlier guides help: How Edge Personalization Works Without Slowing the Storefront Down, Edge Observability: Seeing Storefront Errors Where They Actually Happen, and Image and Font Delivery at the Edge: The Quiet Speed Win.
Frequently asked questions
What percentage of online sales get returned?
NRF and Happy Returns estimate 19.3 percent of online sales will be returned in 2025, about $849.9 billion in merchandise. Apparel runs higher: Loop Returns puts the adjusted apparel rate at 23.2 percent across 22 million returns from more than 4,000 Shopify merchants.
How much does one return actually cost a store?
Outbound shipping, return shipping, receiving, inspection, and restocking typically run $15 to $20 per return before the refund, and the refund itself makes the order a loss even when the item resells. The honest number is returns cost per order: total return cost divided by total orders, which usually lands around $3 to $4.
Should small stores charge for returns?
A small fee usually cuts the return rate a few points and barely touches conversion, because heavy returners are rarely your best customers. Offer store credit as the free option and charge for refunds to cash, so the policy still reads generous on the product page while the expensive path costs something.
How do you stop return fraud without annoying real customers?
Tier the flow. First-time returns and low-value items get instant approval. Repeat returners, high-value items, and worn-condition claims get a human review. NRF puts fraud at about 9 percent of returns, and it concentrates in a small slice of accounts, so tiering catches most of it without slowing the honest majority.