Cutting Return Rates on Fashion Shopify Stores: Fix the Product Page First

Fashion ecommerce returns in the UK run somewhere between 25% and 30% of online orders, depending on whose 2025 benchmark you read. The cross-category average is closer to 20%. Most brands file that gap under logistics: a line in the P&L, a returns portal, perhaps a courier renegotiation.

That is the wrong department. The returns that hurt a fashion Shopify store are mostly decided on the product page, weeks before anything reaches a warehouse. The customer ordered the wrong size because the size guide was generic. They ordered two sizes because they did not trust the guide at all. They sent it back because the colour was not the colour on screen. Reducing your return rate is a conversion and content problem first, and a logistics problem second.

Where fashion returns actually come from

A 2025 University of Liverpool study, built on returns data from one of the UK's largest pure-play fashion retailers, put fit and sizing behind roughly 70% of fashion returns, and estimated the cost to the UK industry at around £7 billion a year. Consumer surveys place colour and "not as pictured" next, at somewhere between 15% and 25% of returns. Damage and late delivery, the things a logistics team can fix, are a small slice.

Size guides built on garment measurements, not generic charts

Most Shopify size guides are a generic chart copied from a supplier or, worse, a theme default. They tell the customer that a UK 12 is a 34-inch bust. They do not tell them that this particular oversized shirt runs a full size big, or that the waist on these jeans is cut small.

The fix is unglamorous. Measure the garments. Publish flat measurements per size for every product, in centimetres and inches, alongside body measurements. Say plainly whether a style runs true, large or small, and by how much. If a product is cut differently from the rest of your range, say so at the top of the description, not in a tab nobody opens.

On Shopify this is a metafield job: garment measurements per size, a fit-note field, and a size guide block in the theme that pulls from both. It is a few days of theme development and then a discipline problem. Every new product needs its measurements before it goes live, and someone has to own that.

Fit feedback where the sizing decision happens

Your reviews already contain fit data. Customers write "size up", "tight across the shoulders" and "I'm 5ft 4 and it hits mid-calf" without being asked. Most stores bury that in a reviews widget below the fold, where nobody choosing a size will see it.

Collect it deliberately. Add a fit question to your review request (runs small, true to size, runs large) plus optional height and usual size. Then surface the aggregate next to the size selector: "8 in 10 customers say this runs true to size". That is the most trusted sizing information on the page, because it comes from people who bought the thing rather than from the brand selling it. Our guide to getting product reviews on Shopify covers how to get enough of them.

Model labelling belongs in the same category. Every image should carry the model's height and the size they are wearing. Two models of different sizes in the same garment is better still.

Fabric, colour and honest photography

Colour disappointment is the return reason nobody plans for. A "sage" that photographs as mint under studio lights, or a "navy" that reads as black on a phone, will come back. Colour-correct product photography against the physical garment, not against what looks tidy on the collection grid, and include at least one shot in natural light. If a colour is hard to capture, say so in the copy.

Then describe fabric like someone who has handled it. Weight in grams per square metre, how much stretch, whether it drapes or holds its shape, whether it is sheer. "Premium cotton blend" tells the customer nothing they can act on. "Mid-weight 220gsm cotton, no stretch, holds its shape" lets them predict how it will feel. Honest copy sells slightly fewer units to the wrong people and keeps far more of the revenue you do book.

Fit and sizing tools: what they cost and what they return

Virtual fit tools (True Fit, Fit Analytics and a long tail of Shopify apps) ask for a few body measurements or a reference brand and recommend a size. Vendors claim double-digit reductions in returns. Some of that is real. Some is selection bias: the customers who use the tool were always the most careful buyers.

They work best on stores with a broad size range and enough order volume for the model to learn. Costs range from a modest monthly app fee to enterprise contracts in the tens of thousands of pounds a year. Do the cheap things first. If return rates on your best sellers are still high after that, run a measured trial, comparing tool users against a control group in the same period rather than against last year.

Bracketing, and whether to discourage it at all

Surveys consistently find around six in ten online shoppers admit to ordering multiple sizes intending to return some. Bracketing is expensive: two units picked, packed and shipped, one returned, inspected and restocked, all for a single sale.

It also lifts revenue. A customer who would have abandoned rather than guess a size now buys. Brands that punish bracketing with fees or blocked accounts often find conversion falls faster than return costs do. The better options are to make bracketing unnecessary through the fit information above, to run an exchange-first returns flow so a size swap never becomes a refund, and to shape the incentive rather than ban the behaviour: free exchanges and a small fee on refunds, for example. We went into the exchange-first approach in how to turn your Shopify returns process into a retention tool.

The data work: reason codes, SKUs and repeat returners

None of the above tells you where to start. Your returns data does, provided it is specific enough to act on.

Reason codes that mean something. "Didn't like it" is useless. "Too small in the chest", "colour different from photos" and "fabric thinner than expected" tell you what to fix.Return rate by SKU and by size. A small number of styles usually create a disproportionate share of returns, often one size of one product. Fix or delist those first.Return rate by traffic source. Paid social typically returns at a higher rate than email or organic, because the customer bought on impulse from a single image.Repeat returners. A small cohort can account for a startling share of return volume. Know who they are and what they cost before deciding what, if anything, to do about them.

Shopify's native reports give you the basics. A returns app, or a plain export into a spreadsheet, gets you to SKU and size level. This is the same analysis we run at the start of any conversion rate optimisation engagement on a fashion store, because the products returned most are usually the products whose pages convert worst.

What UK consumer law lets you do, and what it does not

Design the policy around the law, not around wishful thinking. Under the Consumer Contracts Regulations 2013, an online customer has 14 days from delivery to tell you they are cancelling, then a further 14 days to send the goods back. You must refund within 14 days of receiving them. You can ask the customer to pay return postage, but only if your terms said so clearly before purchase. You can deduct for handling beyond what a shop would allow, such as worn items or removed tags. Personalised items and sealed hygiene goods, like swimwear with the seal removed, are exempt. Faulty goods are a separate matter under the Consumer Rights Act 2015, and that cost is yours.

Within those limits, a restrictive policy almost always costs more than it saves. The customer reads "no refunds on sale items" and simply does not buy. State a policy you can enforce, state it plainly, and then spend your effort reducing the reasons anyone needs it.

A return is a conversion that went wrong. Treat it as data about the product page, not as a cost of doing business.

Where to start this quarter

Pull twelve months of returns by SKU, size and reason. Take the ten worst products by return rate. For each, check garment measurements, fit notes, model labelling, colour accuracy and fabric copy. Fix what is missing, measure the return rate on those products over the next 90 days, then roll out across the range. For the wider product page picture, see our guide to conversion rate optimisation for fashion Shopify stores.

Most fashion brands we speak to, in Glasgow, across the UK and further afield, do not have a traffic problem. They have a conversion problem, and returns are the part of it that shows up after the sale. If you would rather a senior team ran this alongside a continuous testing programme, that is what our Conversion Growth Retainer is for. Get in touch and bring your returns export.