Work out what returns actually cost, per return and across a month. Enter your order economics and your return rate, and this returns the true cost of a single return, the monthly total, and what the return rate does to the margin on everything you sell.
The refund is the part everybody counts and the smallest part of the bill. You also paid to ship the order out, and that money is gone whatever happens next. If you offer free returns you pay for the label back. Somebody has to open the parcel, check the item, repackage it and put it back on a shelf, and that time costs money whether or not anybody writes it down. And the item that comes back is rarely worth what it was worth going out.
Apparel is the extreme case, where a third of orders coming back is normal and bracketing, ordering three sizes intending to keep one, is a deliberate customer behaviour rather than a failure. But the arithmetic is the same in every category: a return rate of ten per cent on a thin margin can wipe out more profit than a ten per cent discount would, and most sellers have never put a number on it.
The output worth looking at is the last line: the margin you actually run at once returns are paid for, against the margin you think you run at. The gap between those two is the number that decides whether a change to sizing guidance, photography or packaging pays for itself.
Opening, inspecting, repackaging and restocking. If you have never measured it, ten minutes of somebody's time is a fair starting guess.
A hundred means it goes straight back on the shelf at full value. Sixty or seventy is realistic for apparel. Zero for anything you cannot resell.
What a return actually costs
| Line | Why it counts |
|---|---|
| Outbound shipping | Already spent. It does not come back with the parcel. |
| Return label | Only if you offer free returns, which most sellers do because the alternative costs conversions. |
| Handling | Somebody opens it, checks it, repackages it and puts it away. It is labour whether or not it appears in the accounts. |
| Value lost on the goods | Opened packaging, worn once, last season. Very little comes back worth exactly what it went out worth. |
| Unrefunded fees | Payment processors often keep some part of the fee on a refunded transaction. |
What actually moves the rate
Show the product as it is. The majority of non faulty returns are the item not matching what the buyer expected, and expectation is set entirely by the photographs and the description. A photograph on a model with their measurements printed under it does more for returns than any policy change.
Show the right photograph for the variant they chose. A gallery that does not change when the shopper picks a colour is a shopper buying from a picture of a different colour. It is the easiest expectation mismatch to fix and one of the most common.
Measure by product, not by store. A store level return rate is an average of a long tail of fine products and a short list of terrible ones. The fix is almost always specific: one supplier, one size run, one photograph.
Do not fix it by making returns hard. It works, in the sense that the rate falls, and it costs more in conversion and repeat purchase than it saves. Fix the expectation instead.