A tempting belief runs through how fashion advertising gets judged. It says a healthy return on ad spend, the revenue an ad platform credits for each pound spent, shows that the ads are paying for themselves. Google's own help page tests it. The page says that a customer's conversion path does not always end at purchase. Customers return retail purchases, and you can adjust a conversion's value after it has been reported. Its example is a shopper who orders shoes in several sizes and returns the ones that do not fit. So the platform's figure starts as a record of what was ordered. It reflects what was kept only if someone sends the returns back. That last step is our reading of the page, not something the page says outright. The four numbers below are the ones to ask for instead, followed by one evidence check that is not a number at all.
1. Revenue after returns
Put the return rate first, because every other figure depends on it. The US National Retail Federation and Happy Returns, a UPS company, estimate that 19.3% of online sales in the US would be returned in 2025. In the UK, ZigZag says clothing (apparel in the US) return rates average 23.6% of ecommerce orders. Treat both as context, not targets. The US number covers online retail sales generally, and the UK number is a clothing average that counts orders rather than pound value. Your own rate is the only one that counts, and it moves by product.
Ask the agency whether the revenue in your report is before or after returns, and how often it updates. Then ask whether returned orders are sent back to the ad platform. Google's help page says you can retract a conversion for a returned purchase, which removes it from the conversion count, or restate its value after a partial return. It adds that retractions affect CPA and ROAS bid strategies. So an account that never receives its returns may be bidding on the first figure; the page implies this rather than stating it.
If you sell through Shopify, you can check the agency's numbers yourself. ShopifyQL's sales schema lets you show gross sales, discounts, returns and net sales together by month. Pull that view before the first meeting.
2. The break-even return on ad spend
Break-even return on ad spend (ROAS) is the revenue you need per pound of ad spend to cover what the sale costs you. Leave returns out and the answer flatters the ads. Here is an illustrative sum, built on assumed figures, not benchmarks.
Take £100 of checkout revenue. Assume 22% comes back and the stock is resold at full price, so you keep £78. Assume cost of goods is 35% of that, which is £27.30. Assume delivery, packing and payment fees come to £10, and return handling costs £5. That leaves £35.70. Break-even ROAS is £100 divided by £35.70, which is 2.8. Ignore the returns and the same sum leaves £55, which gives 1.8. A dashboard showing 2.2 looks comfortable on the second sum and loses money on the first.
Have the agency build this sum with your costs and name the break-even figure it has set for your account. Then ask it to show how the figure moves if your return rate goes from 20% to 30%. Our post on why ads stop scaling covers what happens to these numbers as spend rises.
3. First-order profit against the cost of a new customer
The third number compares what a new customer costs with what their first order leaves after returns. Use the sum above. If the first order leaves £35.70 per £100 and winning the customer costs £40, the first order loses £4.30. That can still be a sound choice if enough customers buy again. It should be a choice you make with the repeat figure in front of you, not a surprise in month four.
Ask for repeat purchases among customers who kept their first order. A blended repeat rate counts customers who returned everything. Also ask for new customers to be separated from existing ones in every cost figure. We cite no benchmark here. Your own history is the only comparison that matters.
4. Return rate by product and by campaign
A blended return rate can hide the lines that cost you. Loop, a returns software company, reported swimwear as its highest-returning vertical at 21.6%, out of ten verticals in its 2024 data from Shopify brands. That is one vendor's customer base, not the market. It is one reason to ask for rates by product.
Ask for the return rate by product group and size run, and by campaign and creative. With the sizing behaviour Google describes, an ad that sells several sizes of one item can look strong at checkout and weak after refunds. Ask what the account is optimised towards: orders placed, or orders kept.
Ask which returns policy the agency's forecasts assume. Retail Economics, working with ZigZag, found in its 2025 UK returns benchmark that 80% of the young fashion retailers in its sample charge for returns (excluding delivery fees), against 11% of the luxury retailers. A forecast built on someone else's policy is a forecast for someone else's brand.
The evidence check: environmental claims
One check sits outside the numbers. List the environmental claims running in your ads and product pages, and check how many have evidence behind them. The right answer is all of them.
In the UK, the Competition and Markets Authority (CMA) published compliance guidance for fashion retail on 18 September 2024, which explains how to follow the Green Claims Code. It is guidance, not legislation. It says broad claims such as green, sustainable or eco-friendly are much more likely to be inaccurate and to mislead consumers. It tells businesses to name a fabric by its objective property, such as recycled polyester, and not to imply a product is entirely one fabric when it is not. It also treats imagery as a claim: one worked example is a jacket called planet-loving, shown with an earth logo and no explanation. The guide covers claims about clothing, footwear and fashion accessories, and says each business in the supply chain is responsible for its own claims being accurate and substantiated. Agree in writing who checks wording before it goes live.
A CMA press release of 27 March 2024 announced that ASOS, Boohoo and George at Asda had signed formal undertakings with the CMA. They bind those three companies, not the whole sector, but the CMA issued an open letter alongside advising all fashion retailers to review their claims. The same release says the undertakings require statements about materials in green ranges to be specific, such as organic or recycled, and the percentage of recycled or organic fibre to be clearly displayed. Ask your agency whether its copy for you would meet that standard.
The penalty regime is newer. In a press release of 7 April 2025, the CMA said that it can now fine a company that infringes consumer protection law up to 10% of its global turnover. That is a ceiling, not a fixed charge. For advertising, section 11 of the UK's advertising code, the CAP Code, as published on the Advertising Standards Authority's site, says absolute claims must be supported by a high level of substantiation.
In the US, the Federal Trade Commission's Green Guides are guidance rather than legislation. They set out the FTC's current views and do not bind the public. The FTC can still act under the FTC Act if a marketer makes a claim inconsistent with them. They also advise marketers not to make unqualified general environmental benefit claims.
Ask the agency which words in your live ads it would need to prove, and where the proof sits. Ask for the evidence behind every fabric percentage, range label and icon. Include unpaid captions and influencer briefs, since the CMA guide's examples include a social media post. A claim without a file behind it should not run.
What the report contains, and how often
Ask for one page, with the same four numbers every month and the claims count beside them. Put the definitions and date range at the top, and add one line on what changed and why. A fashion specialist is not automatically better here than a general ecommerce agency; what matters is whether it can produce the four numbers.
Returns arrive late, so a month's net figures keep moving after the month ends. UK government guidance says online customers can cancel within 14 days of receiving an item. They then have another 14 days to return it, and the seller must refund them within 14 days of getting it back. Adding those periods, a UK order can still be refunded around six weeks after delivery. The page gives the periods; the six-week total is simple addition. So ask for the previous two months to be restated every month, with launches and sales checked weekly while returns are still arriving.
What should happen when a number turns
Agree in advance what happens when a number moves the wrong way. Write down three triggers: net ROAS below break-even for 14 days, a return rate above a ceiling you set for a product group, and a claim nobody can evidence. For each, name who can pause spend, how fast the agency tells you, and when you decide.
When returns rise, ask for a diagnosis by product and creative before any change to budget. Cutting spend on the whole account punishes the lines that are working.
Who should not hire an agency yet
Wait if your returns are concentrated in a few lines. Pull the return rate by product group and size run first. If two or three lines carry most of the refunds, the fix sits with you: fit notes, size guides, product photography and, sometimes, dropping the line. An agency's media plan cannot repair a line customers keep sending back, and more spend on it buys more refunds.
Wait, too, if your orders are too few to read a test by product. Split a small month of orders across sizes, colours and creatives, and each group holds too few kept orders to tell a good ad from a lucky week. Refunds that can land weeks later thin each group again. That is our reading of the arithmetic, not a published threshold. Until volume allows a readable test, put the money into the product page and the size guide.
Hire an agency once your return rate by line is steady enough to plan around and you want help moving all four numbers, whether that help is our paid media for fashion brands or another agency's.
Pull your gross sales, returns and net sales by month and bring them along: tell us about your business. A break-even ROAS after returns is the first thing we can work out with you.
↳ Frequently asked
01What break-even ROAS should a clothing brand set when a quarter of orders come back?
Divide your checkout revenue by what is left after returns, cost of goods, delivery, payment fees and return handling. A brand that keeps 75p of every pound of checkout revenue needs a higher figure than one that keeps 90p. Use your own costs and recalculate whenever your return rate moves.
02Can an agency call a range sustainable if only some of the fabrics are recycled?
Probably not in the UK. The CMA's fashion guidance says to name each fabric by its objective property, such as recycled polyester, and not to imply a product is entirely one fabric when it is not. The undertakings signed by ASOS, Boohoo and George at Asda also require the percentage of recycled or organic fibre to be displayed. Ask for the percentage and the evidence before any ad uses the word.
03When do fashion ad results settle if a UK shopper orders two sizes and sends one back?
Not for several weeks. A UK online customer can cancel within 14 days of receiving the item and then has 14 days to send it back, so the refund for the size they did not keep can arrive weeks after the ad that sold the order. That is why last month's net revenue needs restating next month, and why launches need a weekly check while returns arrive.
04Should an agency report swimwear returns separately from the rest of a fashion range?
Yes, if swimwear is a meaningful share of your sales. A blended return rate can hide lines that behave differently, and Loop's 2024 data from Shopify brands put swimwear at 21.6%, the highest of the ten verticals it analysed. Ask for return rate by product group and size run, so a weak line is fixed or dropped before spend scales it.