In online fashion, a large share of returns has nothing to do with product quality: the shopper simply received the wrong size. For an e-commerce manager, that means direct costs (shipping, handling, restocking), eroded margin and, often, a customer who doesn’t come back. The good news is that these returns are decided before the purchase, on the product page. This guide brings together practical ways to reduce size-related returns, from diagnosis to rolling out a size recommendation tool.
Why size is still the number one reason for fashion returns
Recent data points the same way. In the US, the National Retail Federation estimates that 19.3% of online sales were returned in 2025, versus 15.8% across all retail (NRF and Happy Returns, October 2025). Apparel is higher still: 23.4% of clothing bought online was returned in 2025, and nearly 70% of shoppers who sent clothing back cited size and fit (Coresight Research and Alvanon, May 2026). Europe tells the same story. In France, data from BaBack, which handles returns for several hundred online stores, attributes 64% of returns to size (47%) or fit (17%) issues (October 2024). In Germany, Austria and Switzerland, the EHI Retail Institute puts fashion return rates at 26 to 50%, the highest of any category (EHI, December 2023).
Three things explain why size weighs so heavily:
- Sizes mean different things from brand to brand: your M isn’t your competitor’s M, and shoppers know it.
- Fit is hard to judge from a photo: slim, regular, oversized, sleeve length, drape… a screen doesn’t show everything.
- Bracketing: unsure of their size, shoppers order two or three and send back the ones that don’t fit. The NRF lists it among the costly return behaviours retailers face.
What a size-related return really costs
A return costs far more than the shipping label. To estimate it for your store, add up:
- reverse logistics: label, pickup, receiving;
- processing: inspection, steaming or repackaging, restocking;
- markdowns: an item that comes back late in the season, or damaged, sells at a discount or not at all;
- customer service: sizing questions, refund follow-ups, exchanges;
- lost acquisition cost: the ad spend that won the order is never refunded.
Once transport, receiving and inspection, repackaging, restocking and admin are added up, a return costs a French e-commerce brand €8 to €20 on average (Logistique E-commerce, June 2026). When the item can no longer be sold at full price, the markdown pushes the bill higher: €12 to €15 for an unsellable item, versus €5 to €8 for a standard return (Logistique Mode, July 2026). A single return can therefore exceed €15 before you even count lost acquisition spend.
Free returns amplify the problem, as we explain in the 4 drawbacks of free returns in e-commerce. According to the NRF study, retailers that charge for returns point first to higher processing costs (40%) and higher carrier costs (40%).
Step 1: measure your size-related returns
You can only reduce what you measure. Before picking a solution, get a clear picture:
- Usable return reasons: replace “doesn’t fit” with specific options such as “too small”, “too large”, “fit not as expected”, “colour”, “quality”, “changed my mind”. Everything else builds on this.
- Size-related return rate by category, product and size. A style that mostly comes back as “too small” points to a pattern or description issue, not a customer issue.
- Multi-size order rate (same product, several sizes in one basket): your bracketing indicator.
- Exchange rate: a size exchange beats a refund, but it still costs two shipments.
Track these monthly: they will show you the impact of everything that follows.
Step 2: make fit easy to read on the product page
Many returns come from missing information rather than customer error. On every product page:
- Describe the fit in plain words: “slim fit, take your usual size”, “oversized, size down for a closer fit”.
- Give garment measurements (length, chest width, inseam) and fabric details (stretch, shrinkage).
- State the model’s height and size: “Model is 5’8″ and wears a S.”
- Show customer fit feedback (“runs small”, “true to size”, “runs large”) next to the size selector.
A size chart still helps some shoppers, but it is no longer a must: in our experience of the market, most shoppers never take their measurements. Keep yours accurate, but don’t make it your only answer.
Step 3: recommend the right size to every shopper
This is the most direct lever: instead of asking shoppers to guess, you tell them their size in seconds, right on the product page. There are two main approaches:
- Size-chart-based recommendation: the shopper enters a few details (height, weight, age, fit preference…) that the tool matches against the merchant’s size charts. Accurate when your charts are reliable, but you have to build and maintain them.
- Fit quiz: a few questions about body shape and preferences, with no size chart to prepare. Faster to set up.
Fitle AI Reco combines both approaches: a fit quiz and a size-chart-based recommendation engine. Fitle even offers a third model, based on the non-returned purchase history of a store’s customers (social proof); this engine is not available on the Shopify plans. Since 2013, Fitle has helped more than 300 fashion brands, including major retailers, reduce size-related returns and improve conversion. According to Fitle, size recommendation cuts returns by 40% and lifts conversion by 4 to 6%.
For Shopify stores, the Fitle Size Recommender app (launched in May 2026) is a native app block: you add it to your product page from the theme editor, with no code. The widget is available in English and French from the free plan. Plans scale with your volume:
- Bronze: $0/month + $0.20 per order;
- Silver: $55/month + $0.16 per order, with a 14-day trial;
- Gold: $223/month + $0.11 per order, with a 14-day trial, size-chart-based recommendation and a sizing expert from the Fitle team who reviews the tool’s performance with you, in English or French.
To compare the options on the same criteria (method, size chart required or not, support, pricing), see our comparison of Shopify size recommendation apps. To understand how the models work, read AI and size recommendation: how does it work?
Step 4: show the product worn, on a range of body types
Shoppers don’t just return a size: they return a product that didn’t look the way they imagined. Realistic on-model photos help them picture drape, length and volume. The better they can picture it, the more likely they are to buy something they will keep.
The usual obstacle is the cost and lead time of a photoshoot for every style and colourway. Fitle AI Photoshoot creates on-model product photos with AI models, without a studio, at 70% lower cost than a traditional shoot. It is also an easy way to show the same garment on several body types.
Step 5: curb bracketing without putting shoppers off
- Detect multi-size baskets and show a helpful message: “Between two sizes? Find yours in a few questions.”
- Favour exchanges over refunds in the returns flow, with a size recommendation at the point of exchange.
- Write a clear returns policy that doesn’t blame the customer: time limits, conditions, any fees.
Step 6: feed returns data back to product and design
Return reasons are a goldmine for your product teams. A style that consistently comes back “too small” in one size should be flagged to design or to the supplier. In the meantime, update the description (“runs small, size up”) and the size recommendation for that product.
Category-specific tips
Footwear
Shoe sizes vary by brand and last (width, instep). Show the fit width and customer feedback on sizing.
Workwear and PPE
Orders are often placed for whole teams: one sizing mistake is multiplied by the number of employees. Individual size recommendations before ordering pay off quickly.
Plus size
Fit differences between brands are often pronounced, and many shoppers have been disappointed before. Visuals on representative body types and a size recommendation build confidence on the product page.
A 30-day action plan
- Week 1: review return reasons and calculate your size-related return rate by category.
- Week 2: improve the product pages of your 20 most-returned items (fit, measurements, model size).
- Week 3: install a size recommendation tool on those categories (on Shopify, the 14-day trial on the Silver and Gold plans lets you start testing).
- Week 4: compare return rate, multi-size order rate and conversion before and after, then roll out to the rest of the catalogue.
The bottom line
Reducing size-related returns comes down to giving the right information at the right moment: a clearly described fit, a recommended size, a visual shoppers can relate to. Measure, fix your most-returned product pages, then equip yourself. To see how Fitle can help, discover Fitle AI Reco and Fitle AI Photoshoot.
Sources
- BaBack, return-reason data from several hundred online stores in France, presented in October 2024 (An-Hacking, episode 8 #BFCM).
- National Retail Federation and Happy Returns, “Consumers Expected to Return Nearly $850 Billion in Merchandise in 2025”, October 15, 2025: nrf.com
- Coresight Research and Alvanon, “Shifting the Size and Fit Paradigm”, May 2026, reported by FashionUnited, June 30, 2026: fashionunited.uk
- EHI Retail Institute, study on e-commerce returns (Germany, Austria, Switzerland), reported by FashionUnited, December 5, 2023: fashionunited.de
- Logistique E-commerce, “Retours e-commerce : politique, coûts et gestion logistique” (in French), June 13, 2026: logistique-ecommerce.fr
- Logistique Mode, “Coût d’un retour e-commerce mode” (in French), July 2026: logistique-mode.fr
- Fitle: conversion and returns figures from Fitle customer data; Fitle Size Recommender pricing (Shopify App Store, November 2026).
