Operations
Fashion retailers face return rates of 30 to 40% in Europe. Here are the most effective ways to bring those rates down: better sizing data, accurate photography, honest descriptions, smarter review surfacing, and operational changes to the return process itself.
European fashion retailers face return rates of 30 to 40% on online purchases, substantially higher than any other e-commerce category and high enough that return processing has become one of the largest cost lines in the business. Reducing return rates by even a few percentage points is among the single most leveraged operational improvements available to most fashion brands.
This article covers practical, evidence-based methods for reducing online returns. We have drawn on industry research, retail operations literature, and our own work with fashion clients. There is no single fix. Meaningful reduction comes from addressing the multiple causes of returns systematically, not from any one silver bullet. The retailers who achieve the largest reductions tend to combine several of the interventions below rather than betting on one.
The most common reasons fashion items get returned, in roughly the order of frequency observed across European retailers:
A retailer who does not know which of these causes their returns cannot reduce them effectively. Before investing in any specific intervention, set up reason-coding in the return flow and review it monthly. The shape of the return-reasons distribution tells you which interventions will pay off. A retailer dominated by sizing complaints needs a different programme from one dominated by colour and fabric mismatches.
Sizing problems cause more returns than any other single factor. The most effective interventions, ranked roughly by ratio of impact to effort:
Detailed size charts with body measurements, not just garment measurements. Most fashion retailers publish garment dimensions (chest width, sleeve length). Leading retailers also publish recommended body measurements per size. The latter is much more useful to customers, who typically know their own body measurements better than they know how to interpret garment specs.
Fit predictor tools. Services like True Fit, Fit Analytics, and similar use customer purchase and return data to predict the right size by analysing what customers similar to a given shopper have kept versus returned. Effectiveness varies by category. Fit predictors work best on standardised items (jeans, t-shirts) and less well on highly cut items where personal preference dominates.
Customer reviews of fit. Allowing customers to leave fit feedback ("runs small", "true to size", "runs large") and surfacing this prominently on product pages is high-value, low-cost. Patagonia and Everlane do this well. The data is volunteered by customers and surfaces directly to future buyers.
Detailed model information. Telling customers what size the model in the photo is wearing, plus the model's height and measurements, helps shoppers visualise fit on a body close to their own. Many premium brands now do this and it is worth the small effort.
Do not oversize for vanity. Some retailers cut their sizes generously to flatter customers' size identity. This causes returns when customers buy their "usual size" expecting standard cut and find the item runs large. Consistency with industry-standard sizing reduces returns more than vanity sizing earns goodwill.
Adapt the approach by category. Sizing strategies that work for jeans (where standardised waist and inseam dimensions are dominant) do not translate to outerwear (where layering room, sleeve length, and silhouette matter as much as nominal size). A retailer with mixed assortment should not apply a single sizing programme uniformly. Category-specific size guidance, with relevant body measurements and fit notes per category, performs measurably better than a generic "size guide" page.
Items returned because "they did not look like the photos" account for a significant share of returns. The fix is photographic accuracy, not photographic flattery.
Multiple angles, including the back. Customers should be able to see the item from front, back, and sides. Lifestyle shots are useful but not sufficient. Clean studio shots showing the actual silhouette matter more.
Natural lighting and accurate colour reproduction. Colour cast is one of the most common photography problems. Items photographed in warm studio light look different from the same items in daylight. Calibrate cameras and screens, and consider customer-uploaded photos as a colour-reality check on top of the studio set.
Show the fabric. Close-ups of texture, weave, and detail let customers calibrate expectations. A wool that looks soft in a flat photo might feel scratchy in real life. Close-ups telegraph the texture honestly.
Honest descriptions. "Buttery soft", "incredibly comfortable", "slimming": superlatives without specifics fail customers. Describe the fabric weight, the cut, the construction. Customers calibrate expectations against the description, and if the description over-promises, returns follow.
Include video. Short product videos showing the item in motion help customers understand drape and movement. It does not need to be high-budget. A 15-second clip of the item being worn while walking is often enough.
Reviews are not just social proof. They are product feedback that helps future customers buy correctly.
Surface fit-related reviews prominently. Many retailers bury reviews below the fold or hide negative ones. The opposite approach (making fit feedback visible at the size selector) typically reduces returns more than the small loss in conversion costs.
Filter reviews by reviewer body type. A fit comment from someone whose body matches yours is more useful than the average review. Sephora's "skin tone like mine" filter on beauty reviews is the model. Equivalent fashion implementations exist (Asos's "fit assistant" surfaces reviews from similarly sized buyers, for example).
Use return data to fix the listing. If a specific item has a high return rate flagged "runs small", update the listing to say "we have heard this runs small, consider sizing up". Some retailers fear this dampens conversion. The data shows the opposite: accurate sizing guidance increases trust and reduces both returns and customer-service contacts.
Choose review tools that surface fit, not just star ratings. Generic five-star review widgets are weak signals for fashion. Tools like Yotpo, Stamped, and Okendo support structured fit attributes (size purchased, body type, age range, height) and surface filtered reviews to similar shoppers. The structured data is also useful for product-page SEO and for internal product-quality monitoring. The implementation cost is small relative to the return-rate impact, particularly for retailers above a few thousand orders per month where the volunteered review volume is enough to be useful.
Return policies sit on a spectrum. Generous policies (free returns, long windows, no questions) drive purchase confidence but also drive return volume, particularly bracketing behaviour where customers order multiple sizes intending to keep one.
Consider charging for returns. This was contentious for years, but as of 2025 a substantial majority of European fashion retailers either charge for returns or have introduced incentives to discourage them (store credit instead of refund, for example). The conversion impact is typically smaller than feared, and the return-rate reduction is measurable. Zara, H&M, and many smaller retailers have moved in this direction in the last two years.
Shorten return windows to industry-typical 14 to 30 days. Long return windows (60 to 90 days) increase the likelihood of customers forgetting items and then returning them when they finally remember. They also increase the likelihood of items being returned out of season, when they cannot be resold at full price.
Identify and act on serial returners. A small percentage of customers account for a disproportionate share of returns and refund complications. Most major fashion retailers now flag accounts with sustained high return rates and adjust policies accordingly. In extreme cases this includes declining future orders. This is no longer controversial; it is standard practice across the industry.
Do not refund items not in good condition. "Worn once and returned" is a category of return that has been normalised but is actually a form of abuse. Clear policies on what counts as returnable condition, with photo evidence required where appropriate, reduce this category meaningfully.
When returns do happen, the operational design of the return process determines a large share of the cost. There are two ways to think about this.
Reduce the cost of each individual return. Better packaging that ships and returns in the same wrap. Label-free returns where the courier prints labels in store. Integration with return portals (Returnless, Bleckmann) to automate processing.
Route returns more efficiently. Conventional returns go from the customer back to the warehouse, get inspected, then ship to the next buyer when one materialises. This warehouse round-trip is expensive in money and emissions.
A more recent approach is peer-to-peer return forwarding. When a customer returns an item, it ships directly to the next customer buying the same product, skipping the warehouse entirely. The matching is automated, the consumer experience is unchanged, and the saving per return is typically €5 to €12. The methodology was independently validated and published in Omega 128 (Elsevier, 2024) and is now in production at several European fashion retailers, including Kuyichi.
Peer-to-peer routing is not always the right answer. It works best where return volume is high enough that incoming returns frequently match outgoing demand for the same SKU, where items are in good resaleable condition, and where the retailer is comfortable letting consumers ship to consumers under defined quality controls. Retailers with under 50 returns per month, very long-tail catalogues with little SKU repetition, or assortment dominated by items that show wear after a single try-on are not ideal candidates. Where conditions are right, the operational and emissions savings are significant.
At It Goes Forward, this is what we build, but it is one option among several. The point is not that this is the only solution. It is that the conventional warehouse-routing model is no longer the only option, and retailers facing 2026 EU sustainability regulations should evaluate alternatives.
The best return-reduction programmes work systematically across these areas rather than focusing on one silver bullet. Most retailers have low-hanging fruit in three or four of the categories above. Identifying which ones starts with looking at your own return-reason data first.
For more on the broader returns problem, including environmental and regulatory dimensions, see our returns problem overview. For specific guidance on EU sustainability regulations, see our comprehensive guide to EU sustainability regulations for fashion retailers.
Return policies have shifted from a race to the most generous toward thoughtful designs that reduce abuse without alienating good customers. Done well, return policy becomes a brand signal and competitive moat, not just a cost line.
Last updated:
Operations teams that have run returns the same way for years now face higher volumes, regulatory pressure, capacity constraints, and environmental accountability simultaneously. The teams that adapt fastest gain real advantage.
Last updated:
