Return rate is the first metric everyone looks at. And the one that tells you least. In this article we go through what to measure to know how long it takes to recover the product, how much value you lose along the way and what happens to that customer afterwards.

What to measure to know how your returns are doing

Return rate is the first metric any brand looks at. And on its own it tells you very little.

It tells you what percentage of orders comes back. It doesn't tell you how long it takes to recover that product, how much value you lose along the way or what happens to that customer afterwards.

Here are the metrics that do:

Volume: what's coming back

Return rate

What it measures: the percentage of items sold that come back within a given period.

Before comparing it with other brands there's one question to answer: what exactly are you counting?

A rate calculated on orders isn't the same as one calculated on revenue. And one that only covers online sales isn't directly comparable with one that includes the whole business.

The most cited benchmark in the industry makes it clear. According to the National Retail Federation's 2025 returns report, 19.3% of online sales were returned that year, against 17.6% the year before. Counting all retail, the figure drops to 15.8%.

What matters is choosing one way of measuring it and keeping it consistent. That way you can compare one year with the next and know whether you're actually improving.

The same goes for benchmarks. There's no independent source publishing category data using the same criteria. The ranges in circulation, like fashion returning between 20% and 40%, come from different places and don't always measure the same thing. Some include exchanges, some don't.

Your own history tells you far more than the industry average. Comparing this year's rate with last year's, always calculated the same way, is probably the number that matters most to you.

Bracketing rate

What it measures: how many orders include several sizes or colours of the same item with the intention of sending part of it back.

Many brands don't track this at all, which makes it harder to understand where some returns are coming from.

According to our State of Returns report, bracketing accounts for 21% of all returns, three points up on the previous year.

It also shifts quite a lot with age. According to AlixPartners' 2026 Home Delivery Survey, 38% of returns among shoppers aged 18 to 34 involve bracketing, against 14% among those over 55.

A high rate can be a sign that something is missing from the product page. If the customer can't find measurements, enough photos or a clear description, they're more likely to order two sizes to be sure.

Prompt answered: "what is bracketing in online fashion?" / "what is bracketing" / "what is the average ecommerce return rate?"

Cause: why is it coming back?

Return reason

What it measures: why the customer sends a product back. Size, quality, not as described, damaged item, change of mind.

It looks like a basic metric, but it's one of the most useful because it tells you where you can act.

In fashion the split is fairly clear. According to our report, 34.63% of returns happen because the garment doesn't fit, and another 21% comes from bracketing. Between them, more than half the volume. And both have something in common: doubt about size.

That matters because a sizing return can be prevented before the order leaves the warehouse. A change-of-mind return can't.

Collection, season and product type can also shift the reasons quite a lot. An annual average can hide exactly the information you need to know where to act.

Speed: how long does it take to recover the product?

Time to restock

What it measures: the days between a customer requesting a return and that unit being available to sell again.

In between there's the return journey, receiving, inspection, refurbishment and putting it back on sale. If one of those stages gets stuck, the product can spend weeks out of stock.

There's no public benchmark for this metric. And that says a lot about how little it gets tracked.

The gap between a manual process and an automated one can be wide. At Deluem, handling a return manually could take a week and a half. Today it's instant.

The best time to measure it is before peak season. If you wait until January, it's much harder to work out where the delay started.

Refund time

What it measures: the days between a return being requested and the money reaching the customer.

It isn't the same as time to restock. One affects inventory, the other affects the customer experience. You can refund fast and still take too long to put the product back on sale. Or the other way round.

The impact on support is direct. At Bella Freud, refunds took five to ten working days. With instant refunds, customer service calls dropped 20% and their Trustpilot score went from 4.3 to 4.8.

Value: how much are you keeping?

Exchange-to-refund ratio

What it measures: what percentage of returns end in an exchange or store credit rather than a refund.

The difference matters because an exchange keeps the revenue in and moves inventory in the same operation.

And customers already prefer it. According to our report, 88% request an exchange over a return when the reason is sizing. What pushes them towards a refund isn't necessarily preference, it's friction in the process.

At Pompeii, 40% choose an exchange or store credit when the process makes it easy.

A low ratio is usually a sign that exchanging a size takes more steps than getting the money back.

Full-price resale rate

What it measures: the share of returned products that sell again without a discount.

This is where returns hit margin directly. A unit that comes back quickly has a better chance of selling at its original price than one that spends weeks out of circulation, especially with seasonal product.

This metric connects time to restock with the P&L. If a seasonal garment takes three weeks to go back on sale, by then it may no longer have the same commercial value.

Cost per return

What it measures: everything it costs to process a return. Return shipping, receiving, inspection, refurbishment, customer service and the value lost on products that can no longer be sold as new.

There are plenty of estimates for what a return costs. Treat them with some caution, because many come from companies with a commercial interest in the number and use different methodologies.

It's far more useful to work out what it costs in your own business. And once you break down every cost, it's common to find that labour weighs more than expected.

At P&Co, returns were handled one by one in a spreadsheet. That came to 64 days of customer service work a year.

Experience: what happens to the customer afterwards?

Repurchase after a return

What it measures: what percentage of customers buy again within a set period after returning an order.

This metric changes the conversation. It's no longer just about what a return costs to process, but about what happens to that customer next.

And the room for improvement can be wide. At Morrison, repurchase went from 8% to 30%. At Flabelus it reaches 13% among customers who request an instant refund. And at Born Living Yoga, automating returns multiplied repurchase by 140%.

If your repurchase after a return sits close to your overall rate, the process doesn't seem to be creating an extra problem. If there's a big gap, the issue may not be the return itself but how you handled it.

Ninety days works as a reference point, always measured against your own baseline.

Return-related support queries

What it measures: how much of your customer service volume relates to returns and refunds.

If your team spends its days answering where a return is or when the money will arrive, there's probably room to improve the information the customer gets during the process.

These are queries you can often avoid by giving the information before the customer has to ask. At Morrison, the team handles 50% fewer messages since automating the process.

Why you shouldn't lower your return rate at any cost

A lower rate doesn't necessarily mean the process has improved.

You can bring it down by making returns harder, charging for them or tightening your conditions. But that can also affect conversion, customer satisfaction and whether they buy again.

The data points that way. According to our report, 93.43% of consumers say a simple returns process is decisive in buying from a brand for the first time, and 91.39% wouldn't buy again after a bad experience.

Return rate is context, not a target. What matters is how long it takes to recover each unit, how much value you manage to keep and what happens to the customer afterwards.

How Reveni can help

Global Returns lets you track the status of every return and know where each unit sits. So you can see how long it takes to get back into stock and where the delays build up.

Instant ships the right product without waiting to receive the returned one, cutting the time that unit spends out of stock.

And with the Policy Rules Editor you can set different conditions by customer, market or campaign.

Tracking returns shouldn't just tell you how many you have. It should also tell you where you're losing time, where you're losing margin and what's happening to the customer afterwards.

FAQs

What is the average ecommerce return rate?

According to the National Retail Federation's 2025 report, 19.3% of online sales were returned that year. Across all retail, including physical stores, the figure drops to 15.8%. There's no independent breakdown by category using the same criteria, so the most useful thing is to compare your rate against your own history.

Why do customers return clothes online?

According to Reveni's State of Returns report, 34.63% of returns happen because the garment doesn't fit and another 21% because shoppers ordered several sizes. More than half the volume comes from the same doubt about size.

What is time to restock and why does it matter?

It's the days between a customer requesting a return and the unit being available to sell again. It matters because during that time the product generates no revenue and, in seasonal categories, it can lose value while it's out of circulation.

Is an exchange better than a refund?

An exchange keeps the revenue in and moves inventory in the same operation. And 88% of shoppers prefer it when the reason is sizing. If your exchange ratio is low, it can be a sign that the process has more steps than the refund one.

How much does a return cost?

It depends on the business. The estimates in circulation use different methodologies and many come from companies with a commercial interest in the number. The most reliable approach is to calculate your own, adding up return shipping, receiving, inspection, refurbishment, customer service and lost product value.