Fashion returns cost €140 billion annually. Fit causes most of them. Measurement infrastructure is the fix.
TL;DR: Fashion e-commerce return rates run at 30–40%. Fit causes roughly 70% of those returns. Processing a single returned luxury garment costs €400–€600 in direct losses before markdown risk is counted. Across the industry, total return-related costs reach approximately €140 billion annually. The root cause is not consumer behaviour — it is a sizing system built on labels rather than measurements. When garments are specified to actual body measurements, return rates fall below 5%.
The economics of fashion returns are rarely discussed openly, because the numbers are uncomfortable. Fashion returns cost the industry approximately 20–30% of the retail value of each returned item — including reverse logistics, inspection, repackaging, and the discount required to re-sell it. For brands operating luxury e-commerce channels, where return rates consistently exceed 30%, this is not a margin erosion event. It is a structural business problem with a precise, measurable cause.
Fashion e-commerce returns cost the global industry approximately €140 billion annually in direct and indirect losses, according to aggregated estimates from Invesp, McKinsey, and industry-wide logistics data. That figure encompasses reverse logistics, processing labour, inventory holding during transit, markdown losses, and destroyed unsaleable units — not just the courier cost brands typically report internally.
Online fashion return rates have grown in direct proportion to online sales. Industry data from Statista and Invesp consistently places overall fashion e-commerce return rates between 30% and 40%, with fit cited as the primary reason in roughly 70% of cases. For luxury online retail — where average order values are higher and consumer expectations of precision are elevated — both the frequency and the financial impact per transaction are compounded. A brand seeing a 35% return rate on a €2,000 jacket category is losing €400–€600 per returned unit before markdown risk enters the calculation.
Brick-and-mortar return rates for the same categories run at 8–10%. The gap is not explained by consumer dishonesty. It is explained by the information asymmetry between a shopper who can try a garment on and a shopper who cannot.
Fit accounts for approximately 70% of fashion e-commerce returns, according to multiple independent studies including ScienceDirect research on digital product fitting. The conventional narrative blames consumer behaviour — buyers ordering multiple sizes to try at home, keeping one and returning the rest. That practice exists and costs brands money. But it is the secondary driver, not the primary one.
The more consequential driver is simpler: shoppers buy garments they expect to fit, and the garments do not fit. This is not deliberate arbitrage. It is the failure of a sizing system that gives buyers no reliable foundation for predicting fit across brands, seasons, or even product categories within the same brand. A size M shirt from one Italian manufacturer runs 3–4 cm narrower in the chest than a size M from another. Neither label is wrong — they are just not standardised.
In practice, a buyer selects a size based on one of three unreliable references: their experience with a different brand, a size chart they measured against approximately, or a previous order from the same brand in a different cut. Any of these references can be wrong for the specific garment ordered. When the garment arrives and the fit is off, the return follows. This is a measurement infrastructure failure, not a consumer behaviour problem.
Most brands measure return cost too narrowly — tracking only outbound and return shipping costs. The full cost of a fashion return includes five distinct loss categories, each individually material and collectively capable of turning a profitable sale into a net-negative transaction.
Definition
Direct logistics cost
Return courier cost (frequently brand-absorbed in premium e-commerce), inbound processing labour at the warehouse, inspection, quality assessment, and repackaging. For luxury items, this alone can reach €30–€60 per unit.
Definition
Inventory deadstock cost
A returned item is removed from saleable stock during the processing cycle — typically two to six weeks for luxury brands with quality inspection requirements. The capital it represents earns nothing during this period, and the item is unavailable for sale at full price during peak demand windows.
Definition
Markdown and liquidation loss
Returned luxury items that cannot be re-sold as new — because they show wear, carry odour, or were damaged during return transit — require markdown, outlet channel sale, or destruction. Markdown typically realises 40–60% of RRP; outlet sale 30–50%; destruction 0%. Each step carries a progressively greater margin hit.
Definition
Customer lifetime value erosion
Research cited by McKinsey in its 2024 State of Fashion report links poor fit experiences directly to reduced repeat purchase probability. A customer who receives an ill-fitting garment is statistically less likely to repurchase within 12 months — even when the return itself was frictionless. The failed transaction degrades the brand experience regardless of how smoothly the logistics ran.
Definition
Carbon and regulatory cost
Under the EU Ecodesign for Sustainable Products Regulation 2024/1781, reverse logistics emissions will increasingly form part of mandatory sustainability reporting. The European Environment Agency estimates that textile transport generates 0.5–1.2 kg CO₂ per return shipment at current European network averages. At scale, this is a material compliance exposure, not merely a reputational one.
The margin mathematics of luxury returns are significantly worse than those of mass-market returns because the economics compound in both directions. Luxury items carry higher absolute values, meaning each return event produces a larger absolute loss. They also carry brand equity — an expectation of precision that makes a sizing failure more damaging to customer trust than the same failure from a mass-market brand.
When measuring this effect operationally: a luxury brand with a €1,500 average order value, a 30% return rate, and a 25% processing cost per return is writing off €112.50 in direct cost per order placed — before markdown risk. At 10,000 annual orders, that is €1.125 million in annual direct return cost from fit-related failures alone. The equivalent mass-market brand at €60 AOV writes off €4.50 per order placed. The absolute loss is 25x larger for the luxury brand, while the operational complexity of handling luxury returns — inspection protocols, authentication, boutique-grade repackaging — is significantly higher.
A luxury house selling a jacket at €2,000 online, accepting a return, and recovering it at a 25% markdown has lost €500 on that single transaction — before reverse logistics, inspection, or the opportunity cost of idle inventory.
Reducing fit-related returns does not require changing consumer behaviour. It requires replacing the information available to the buyer at the point of purchase — specifically, replacing size-label-based selection with measurement-based specification. When a buyer's body measurements are known precisely and matched against garment specifications rather than size labels, the probability of fit error falls from roughly 35% to below 5% in made-to-order contexts, per ScienceDirect research on digital product fitting.
Three structural interventions reduce fit-related returns in sequence of implementation complexity: first, enriching product data with full garment measurements (not just size labels); second, deploying digital fitting tools that compare buyer measurements against garment specifications at the PDP level; third, shifting high-value SKUs to made-to-measure production where garments are produced to individual body measurements and the fit question is answered before production rather than after delivery.
Related analysis in this cluster examines the specific operational cost breakdown in detail: see the article on the hidden operational cost of poor sizing for the warehouse and markdown loss model, and the real cost of fashion returns analysis for the customer lifetime value dimension. The fit-related returns margin killer piece covers the luxury brand case specifically.
Size Passport's made-to-order infrastructure applies the third and most complete intervention. Garments produced to individual body measurements return at below 5% in practice — because the fit question has been answered during specification, not discovered upon delivery. The measurement profile persists across orders, meaning each subsequent purchase carries the same precision without requiring a repeat measurement process.
Fit accounts for approximately 70% of fashion e-commerce returns, based on consumer survey data from Invesp and operational data cited in ScienceDirect research on digital product fitting. The remaining 30% includes changed mind, product not matching description, quality issues, and the deliberate multi-size ordering practice. Fit is the single largest and most structurally addressable cause.
Direct processing costs for a returned fashion item range from €15–€60 depending on the brand tier, inspection requirements, and whether the item requires repackaging. For luxury brands absorbing return shipping, costs reach the upper end of that range before any markdown or liquidation loss is counted. When markdown risk is included — representing 20–40% losses on items that cannot be re-sold as new — the all-in cost per return commonly reaches €400–€600 for luxury garments at €1,500–€2,500 RRP.
The €140 billion estimate applies to the global fashion and apparel category specifically, derived from aggregated e-commerce return rate data (30–40% of online sales), industry revenue figures, and per-return cost models from McKinsey and Invesp research. Total global e-commerce returns across all categories are estimated above €500 billion annually; fashion's share is disproportionately large because return rates in apparel are three to four times higher than in electronics or home goods.
A size label (S, M, L, or a numeric equivalent) is a marketing convention with no standardised relationship to body or garment dimensions. A garment measurement is a physical specification: chest circumference 102 cm, shoulder width 47 cm, sleeve length 65 cm. Two size-L shirts from different brands can differ by 6 cm or more in chest circumference. Two shirts specified at 102 cm chest circumference will be identical on that dimension regardless of brand. Measurement-based selection eliminates the primary information gap that causes fit-related returns.
Yes. A persistent measurement profile — a record of verified body measurements associated with a buyer — allows every subsequent purchase to be specified against the same data without a repeated measurement process. This is the core function of the Size Passport: a single measurement event that informs every future garment order, reducing the per-order fit uncertainty that drives returns to near zero for made-to-order items.
Sources
Related concepts