Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Metric

Return Rate

Share of orders or units sold that customers sent back.
BusinessmetricUpdated

Formula

Orders (or units) returned ÷ Orders (or units) sold × 100

Unit: % · lower is better

Definition

Return rate measures sales that were reversed. For a chosen set of sales, count how many were sent back and divide by how many were sold. Two counting bases are in common use and they answer different questions:

  • Order return rate counts an order as returned if any item in it came back. It tells you how often a customer was dissatisfied with something.
  • Unit return rate counts items. It tells you how much stock and revenue flowed back, and it is the version to use when baskets contain several items, especially when shoppers buy more than one size to try.

Orders (or units) returned ÷ Orders (or units) sold × 100(%)

  • Orders (or units) sold are the sales in a fixed period, counted on the same basis as the numerator.
  • Orders (or units) returned are the returns that belong to those sales, wherever they arrived in time. Matching each return to its original sale is what makes the rate attributable to a channel, campaign or product.
  • The return window creates a lag. Returns arrive days or weeks after the sale, so a rate calculated on last week's sales is incomplete. Rates should be reported for sales cohorts whose return window has closed, or labelled as provisional.

A value-based version, refunded revenue over sold revenue, is often more useful for margin work, because a returned dress and a returned pair of socks are not the same loss.

How to interpret it

Lower is better, but a single store-wide figure hides most of the information. The readings that change decisions are:

  • By category. Clothing that depends on fit, items bought as gifts and products whose colour is hard to judge on screen come back more often than consumables. The fashion industry guide treats returns as a structural cost of selling apparel online for exactly this reason.
  • By acquisition channel and creative. The same product can be returned at very different rates depending on what the ad promised and who it reached. This is the reading platforms cannot do for you.
  • By promotion. Deep discounts and free-returns offers can lift orders and returns together, so the net effect has to be measured, not assumed.
  • By customer. A small group of serial returners can account for a large share of returns; that is a policy question rather than a marketing one, but it distorts channel comparisons if left in.
Illustrative example — not a real result

A baby clothing shop nets out its returns

Thistledown Baby Wear compares two sources over a quarter whose return window has closed. A social video campaign spent £6,000 and reported £30,000 of revenue, a ROAS of 5.0; search spent £6,000 for £24,000, a ROAS of 4.0. Matching returns to orders, 38% of social revenue came back against 9% from search, because the videos showed newborn outfits on babies who were clearly older and the size chart did not match. Net of refunds, social kept £18,600 (a ROAS of 3.1) and search kept £21,840 (3.6). With gross margin at 50% and a return handling cost on top, social was the weaker channel. The shop reshot the videos with a size caption before spending more.

What moves it

  • Sizing and fit information. Accurate size charts, fit notes and model measurements reduce returns in apparel more than any other single fix.
  • Imagery and description. Products that look different on arrival than on screen come back. Colour accuracy, scale and close-ups set realistic expectations.
  • What the ads promise. Creative that overstates a product, or shows a version the customer will not receive, raises returns from that ad's buyers specifically.
  • Checkout prompts. Size or variant reminders at the last step catch mistakes before they ship, which the checkout optimisation guide places among the late-funnel fixes.
  • Returns policy. Free and long return windows reduce purchase risk and often lift conversion, while also lifting returns. Whether the trade is worth it is a margin calculation, as the free shipping threshold experiment shows for the related question of delivery charges.

What good looks like

There is no reliable benchmark, and the site does not cite one. Acceptable return rates differ enormously by category, price point and returns policy, and published averages rarely say which base they used. The useful tests are internal. Is the rate stable or improving for each category against its own history? Does any channel, campaign or creative return far more than the same products sold elsewhere? And after returns, handling and lost resale value, does each channel still clear its margin threshold? That last test matters most, because a return rate the business can afford in one category can wipe out the profit in another.

Returns sit between revenue and Contribution Margin: every return removes revenue and adds handling cost, so contribution per order falls faster than revenue does. The contribution margin and break-even guide shows where returns enter that calculation. Gross Margin sets how much a returned sale costs the business, and ROAS is the metric returns distort most, because platforms report revenue at the moment of purchase. AOV can mislead in the same way when shoppers buy several sizes and return all but one. Product listing ads add a further wrinkle, since feeds and catalogue ads reach shoppers who compare quickly and return freely, which the shopping and catalogue ads guide covers. Affiliate programmes face the same question when commissions are paid on orders later refunded, a point the affiliate measurement guide addresses through validation periods.

Common mistakes

  • Reading platform ROAS as kept revenue, when the platform never learns which orders came back.
  • Calculating the rate before the return window closes, which makes recent periods look better than they are.
  • Mixing order and unit bases between reports or between categories.
  • Reporting one store-wide rate, which hides the channel or creative that causes most returns.
  • Ignoring handling and resale loss, so a return is treated as cost-free once the refund is netted.
  • Setting ROAS targets on gross conversion value without raising them to allow for the share of revenue each category will send back.

More terms to look up

Where this term is used

Related