Charging everyone for returns saves you postage and costs you some sales. Making returns free for everyone wins sales, but the best study we could find suggests it can lose money once the parcels come back. So we'd stop asking which policy is right and start asking which orders deserve free returns, a question that only your own numbers can answer.
What brands charge now
Free returns used to be the price of admission for selling clothes online, and that has changed quickly, especially at the cheaper end of the market. When Retail Economics and ZigZag looked at 100 UK clothing and footwear retailers for their 2025 benchmark, 42% charged for returns. The range is the interesting part: 11% of luxury retailers charged, against 80% in young fashion.
The fees themselves are small. Which? checked 20 big fashion retailers in August 2024 and found 12 without a free postal return, and these are some of the fees it reported. They're two years old and some will have moved, so please don't quote them as today's prices.
| Retailer | Postal return fee, August 2024 |
|---|---|
| New Look | £2.95 |
| Zara | £2.95 |
| Next | £2.50 |
| River Island | £2.00 |
| Boohoo | £1.99 |
| PrettyLittleThing | £1.99 |
| H&M | £1.99, non-members |
The case against charging
Ask shoppers and they'll tell you they care. In the US, 82% of people who had returned something online called free returns a major consideration when buying, and in the UK 49% said they had abandoned a purchase over a returns policy they didn't like. We'd treat both numbers with some suspicion, because people are poor witnesses to their own shopping, and a survey question about free returns isn't the same as a shopper standing at a checkout with a £3 fee in front of them.
Experiments are better evidence. In a field experiment published in 2025, offering free returns to shoppers who had left without buying raised the purchase rate by 31.6% and the profit from that targeting by 37.8%. That's a big number, and it's easy to over-read. The researchers tested a late nudge aimed at people who were already interested, which tells you rather little about what would happen if you changed the policy across a whole store. It worked best on younger, newer customers and on dearer products.
The case against free
The most useful study we know of followed an online fashion retailer that introduced free returns in one country while the others carried on as before, which gave the researchers something to compare it with. Order value rose 9.15% and returns rose 7.86%. When the authors did the sums (roughly, and they say so), the extra cost of the returns came to more than the extra gross profit.
So the sales were real and the profit wasn't. It's one retailer and a back-of-the-envelope estimate, so it isn't a law of nature, but we'd still trust it over any survey because it counted what happened after the parcels came back.
Why would that happen? Take two customers, whom we've made up but you'll recognise. One orders a £60 dress on a Friday, wears it to a wedding and never thinks about it again. The other orders five dresses, keeps one and sends the rest back in a bag that's been sitting in the hall for three weeks. Under a single free policy they get exactly the same deal, and the first is quietly paying for the second. ZigZag's research puts serial and slow returners at 22% of shoppers and nearly half of returns. It sells returns services, so it isn't a neutral party, but your own order history will show you whether your split looks anything like that.
What happened at L.L.Bean
For about a century L.L.Bean, the Maine outfitter, took almost anything back at any time, and the promise was as much a part of the brand as the boots. In February 2018 it moved to a one-year limit, saying that abuse by a small number of people had become too expensive. It still covered faults, and it honoured everything bought before the change, and it got a lawsuit and plenty of unhappy press anyway.
We take two things from this. A generous policy will eventually attract the people it wasn't meant for, however old and well-loved the company. And when you take that policy away from everyone, the customers who never abused it are the ones who feel accused.
Charging only some people
ASOS is the best-known attempt to split the difference. From October 2024 it charged frequent returners £3.95 on some orders, and in January 2026 it went further and began showing customers a personal return rate in its app. If more than 70% of what you've ordered by value over the past 12 months has gone back, charges apply, and if you bring that down they stop. Whatever you think of the idea, at least everyone can see the rule.
The weakness of any fee based on past behaviour is that it only works on customers you already know. It can't tell you anything about a first order, and a first order is where a lot of the hesitation lives.
| Policy | What you get | What it costs you |
|---|---|---|
| Fee for everyone | Lower return volume. Simple to run. | Sales from hesitant first-time buyers, many of whom would have kept the item. |
| Free for everyone | More sales. An easy message. | You pay for your most expensive orders too. |
| Fee for frequent returners | Aims the cost at the people causing it. | Needs history, so it does nothing for new customers. Some will feel singled out. |
| Decided per basket, before payment | Free returns only where they pay. | More to build and to explain. Shoppers must see their terms before paying, every time. |
The last row is what Kept does, so you should read our enthusiasm for it with that in mind and test it against a control group, as you would anything else.
What UK law allows
This is our summary and it isn't legal advice. For most online sales a customer can cancel within 14 days of delivery, and then has a further 14 days to send the goods back.
- Postage. The customer bears the direct cost of returning the goods unless you've agreed to pay, or you failed to tell them before the sale that they'd have to. If you leave it out of your terms, the postage is yours.
- Fees. You must not charge a fee for the refund itself. Passing on the postage is one thing, but a restocking fee taken from a cancellation inside the 14 days is another, so take advice before you add one.
- Original delivery. You have to refund it, up to the cost of your cheapest standard option.
- Worn goods. You can reduce a refund for handling that goes beyond what a shopper could do in a shop.
- Faulty goods. There's no return fee, and the delivery charge comes back too.
How to decide
- Work out what a return costs you. Include the postage, the handling, any payment fees you don't get back, and the margin you lose when the item goes back on sale at a discount. Our returns cost calculator will do the arithmetic.
- Look at who returns. Sort last year's customers by the share of their orders they sent back. If a small group accounts for a large share of your returns, then one policy for everyone is costing you money.
- Test it, and hold some customers back. Change the policy for a random half and leave the other half alone, then compare profit per visitor after the return window has closed. Comparing this year with last year will tell you as much about the weather as it does about your policy.