What is a good return rate for online stores?

Compare yourself to yourself, because a store average across a mixed catalogue is not comparable to anything.

The biggest driver is whether the product has to fit. Anything sized, worn or matched to a body will return at a much higher rate than anything consumed or used as delivered, and no amount of good practice closes that gap. A store selling clothing and a store selling coffee are in different businesses as far as returns are concerned.

The second driver is price. Higher priced items get more scrutiny after arrival and come back more often. The third is whether the customer can see the real thing before buying, which is why colour, texture and finish drive returns in furniture, paint and cosmetics.

The fourth is your own returns policy. A generous policy raises the return rate and often raises conversion and repeat purchase at the same time. That is a trade, not a failure, and judging it by return rate alone gets the decision wrong.

So the useful question is not what is normal. It is whether your rate is moving, and which products are moving it.

By units returned as a share of units sold. This is the operational number. It tells your warehouse how much work is coming and it is the one to track per product.

By value returned as a share of revenue. This is the financial number, and it can look very different, because the expensive items usually return at a different rate from the cheap ones. A store with a modest unit rate and a high value rate has a problem in its premium lines.

By reason. This is the only version that tells you what to fix. Wrong size, not as described, colour different, damaged in transit, arrived late, changed mind, ordered multiples to choose from. Each reason points at a different part of the business: sizing guidance, photography, packaging, carrier, or a marketing message that overpromised.

Also split by channel and by campaign. Traffic from broad social prospecting often returns more than traffic from branded search, because the buying decision was faster and less informed. If one campaign has a much higher return rate, its true return on spend is worse than the report shows, and that changes where the budget goes.

Use a consistent time window, since returns arrive weeks after the order.

Stores sometimes celebrate a rate near zero. It is worth checking why before treating it as a win.

It can mean the policy is hard to use. A returns process that requires an email, a wait for approval, a printed label and a trip to a depot will suppress returns and produce customers who never buy again and tell people why. The refund you avoided cost you the customer.

It can mean the return window is too short for the product. Someone buying a seasonal item in advance may not open it within a fortnight.

It can mean returns are being handled as something else, such as refunds without a return, or replacements logged as new orders, which makes the reporting look clean while the cost sits somewhere else.

Or it can genuinely mean the product pages set expectations well, which is the outcome you want.

The way to tell them apart is to look at repeat purchase rate and support contacts alongside the return rate. A low return rate with high repeat purchase and few complaints is a healthy store. A low return rate with poor repeat purchase and a stack of support tickets is a policy problem wearing a disguise.

Work out what one return actually costs you before deciding how much effort to spend on the rate, because the figure surprises most stores.

There is the outbound shipping you already paid, which is rarely recovered. The return shipping, if you pay it. The labour to receive, inspect, repackage and restock. The payment processing on the original order, which is often not fully refunded to you. The cost of any item that cannot be resold at full price, which for opened cosmetics, worn clothing or damaged packaging can be the whole value. And the working capital tied up while the item is in transit.

Add those and a return typically costs a multiple of the refund itself in margin terms. That is why a small reduction in return rate on your highest volume products is often worth more than an equivalent effort spent on acquiring more traffic.

It also changes how you read channel performance. A campaign with a high return rate is spending your money twice, and its real contribution should be calculated on net revenue after returns rather than on gross orders.

Build that adjustment into your reporting rather than doing it once as an exercise.

The point of measuring is to change something, so collect the reason at the moment of the return and keep the list short enough that people pick honestly. Five or six options with a free text box beats twenty categories nobody reads.

Then look at reasons by product rather than in total. A store level chart showing wrong size as the top reason is not actionable. A list showing that three specific styles account for most of the sizing returns is.

Work the top five products by returned units, not by rate, because a poor rate on something you sell twice a month costs nothing. For each one, read the free text comments. They usually name the problem in the customer's own words within ten entries.

Then fix at the source: better measurements and fit notes, photography that shows scale and true colour, clearer material descriptions, a note about what the product is not suitable for, and packaging changes where damage is the reason.

Re-measure the same products after a full cycle. Returns lag orders by weeks, so a change made this month shows up in the data next month, and reading it too early will tell you it did not work.