A return is a second fulfilment you never sold. The refund is the visible part and the smallest part of the story. Everything around it, the two delivery legs, the handling, the working capital and the calendar, is paid quietly out of margin and rarely lands in one report.
Here is what a returned garment actually touches, and the figures worth writing next to each line for your own store. Easy returns sell clothes, and a shopper who trusts the policy buys the second item, so the case here is for knowing the number. A cost you have measured can be designed against. A cost you have not measured becomes the reason margin never quite arrives where the forecast said it would.
The scale, in the industry's own numbers
The National Retail Federation and Happy Returns released their 2025 Retail Returns Landscape on 15 October 2025. Retailers in that research expected 15.8 per cent of annual sales to be returned during 2025, worth 849.9 billion dollars, with an estimated 19.3 per cent of online sales coming back. Those are US figures and the shape of the problem travels.
Fashion sits above the all-category line, and UK fashion sits above it comfortably. Fulfilment provider fulfilmentcrowd's research, reported by Just Style in January 2025, suggested around 30 per cent of fashion items bought online in the UK are returned each year. Plan for a fifth to a third of what you ship coming back, then find out which third of your catalogue is producing it.
Eight lines a single return touches
- The outbound delivery, already spent. The picking, the packing, the packaging and the courier leg were paid on the day the order shipped, and none of it comes back with the parcel. On a cancellation within the statutory period you also refund the customer's standard delivery charge: GOV.UK states you must refund the cost of standard delivery if the customer paid for it, and that where they chose a more expensive option you refund the standard cost. Two delivery costs, one of them recovered from nobody.
- The inbound leg. Free returns are a marketing decision with a logistics invoice attached. The NRF and Happy Returns 2025 Retail Returns Landscape reports that 82 per cent of consumers say free returns matter when they shop online, which is the reason most fashion retailers absorb the cost. Whether you charge or absorb it, price the leg per parcel and know the number.
- Handling and inspection. Somebody opens the bag, checks the item against the order, decides whether it can be resold, steams or presses it, re-tags it, re-bags it and puts it back into stock. This is minutes of skilled attention per unit, at a peak that arrives after your sales peak, when the same team is already stretched.
- The refund clock. GOV.UK sets the timings: the customer has 14 days from receiving the item to tell you they are cancelling, another 14 days to send it back, and you must refund within 14 days of receiving it. The cash leaves before the item is back on sale, and during peak the gap between the two events is when a returns backlog turns into a cash flow question.
- The season that closed while the parcel travelled. Fashion is dated stock. An item that comes back near the end of its season has lost the part of its price the season was carrying. The loss is real and it lands as markdown, and most reporting books that markdown as a buying decision.
- Return fraud, at a measurable rate. The same NRF and Happy Returns research found that 9 per cent of all returns are fraudulent. Wardrobing, the worn-once return, sits in the same territory and is harder to prove. Your inspection standard and your policy wording are the controls, and they need to be applied the same way by everyone on the returns bench.
- Customer service time. Every return that goes wrong generates contact: where is my refund, why was a deduction applied, the parcel was signed for and the refund has not appeared. The handling cost of a smooth return is small. The handling cost of a disputed one is a conversation, an investigation and often a goodwill credit.
- The environmental line, which is becoming a commercial one. A 2025 study of UK fashion e-commerce returns in Transportation Research Part E, by researchers at the University of Liverpool, frames the problem with an estimated 7 billion pounds of cost to the UK industry in 2022 and an estimated 750,000 tonnes of CO2 associated with discarded apparel. When a customer, a buyer or a lender asks what your own returns cost the business and the planet, a measured figure of your own is the only answer worth giving.
Where the returns actually come from
Fashion returns cluster, and they cluster around fit. Poor fit was the reason customers gave most often in the fulfilmentcrowd research cited above, which is the useful part of that study for anyone running a catalogue.
Bracketing follows from it. When a shopper cannot tell which size will work, ordering two or three and returning the rest is the rational move, and free returns make it a sensible one. That behaviour is a signal about your product information.
The pattern is rarely spread evenly across a catalogue. In most stores a small group of lines produces a disproportionate share of returns: a dress that runs small, a trouser whose measurements were copied from a supplier sheet nobody checked, a colour that photographs differently from how it arrives.
The rules you cannot design around
UK distance selling rules set the floor, and knowing them precisely stops policies drifting into positions you cannot defend.
Customers have 14 days from receiving the goods to cancel, then 14 days to return them, and you refund within 14 days of getting the item back. You refund the standard outbound delivery cost. You may pass on the direct cost of the return where you told the customer so before the contract was made, and regulation 35 of the Consumer Contracts (Information, Cancellation and Additional Charges) Regulations 2013 places that cost on the trader where no such notice was given.
GOV.UK also states that customers are allowed to handle an item in the same way they would in a shop, and gives trying it on as the example, and that you cannot deduct fees from a refund unless the item has been used or damaged. Write your policy against those sentences, and check that your inspection standard matches what they permit.
Read together, those rules mean the trying-on has to happen somewhere. Your store either hosts it at home, with two delivery legs attached, or gives shoppers enough information to do it in their heads before they order.
What to measure before you change anything
Return rate by product, with a reason code, is the metric that leads to action. Alongside it, keep four more.
Net margin after returns, per line, which is the only version of margin that tells the truth in fashion. Days from return receipt to resaleable, which is where your peak backlog shows up. Cost per return, built from the inbound leg, the handling minutes and the average markdown taken on returned stock. Repeat-return rate by customer, which separates a fit problem from a policy problem.
Most of these are already in your platform's data and simply are not joined up. Joining them is a reporting job measured in days, and it is the cheapest step in this article.
Building a cost per return takes an afternoon. Take a month of returns. Add the inbound carrier cost you actually paid, the refunded outbound delivery, the packaging you cannot reuse, and the handling time at your real hourly cost. Divide by the number of returns processed. Then add a second figure beside it: the average markdown taken on returned stock that missed its window. The two numbers together are what a return costs you, and they are usually a surprise to everyone who has only seen the refund total.
What reduces returns, in the order worth trying
Accurate garment measurements per size, published on the page, do more than any other single change, because they answer the question the customer is actually asking. Photography on more than one body, with the size worn stated, comes next. Then reviews that carry a fit question, so shoppers see whether people found it small or large. Then a size guide written for the line itself.
A virtual try-on step belongs in this list as a way to let a shopper see a garment presented on a body before they order. Our own Try-On Tool sits in that category, as a WooCommerce product built for stores that want the step on their own site. Treat any effect on your return rate as something to measure on your own catalogue, with a before and after on the same lines, and take no one's word for it, ours included.
Start with the twenty worst lines
You do not need a returns programme. You need a list, ranked by returned units, cut to the top twenty, with a reason code against each.
Many of them will have a cause you can fix in an afternoon: a measurement table that was never filled in, an image that misrepresents a colour, a size label copied wrongly at import. The rest will need a supplier conversation or a buying decision. That list is where the money in this article actually sits.
Run it again a quarter later on the same lines. A returns programme that cannot show a before and after on named products is a set of opinions with a budget attached.