SCAYLE says AI can cut fashion returns at the source

SCAYLE’s Pascal Richter argues that fashion retailers can reduce costly returns by using AI before checkout, not just in reverse logistics. The Hamburg-based commerce platform points to size recommendations, virtual try-on, cart prompts and returns inspection as ways to protect margins and lower emissions.

Why it matters: - Fashion returns remain expensive for retailers because they tie up working capital, reduce product value and add transport emissions. - SCAYLE says many returns can be prevented before an order is placed, when shoppers are still uncertain about fit. - Size and fit are a top pain point for the sector, and return rates continue to pressure margins.

What happened: - Pascal Richter, Director Growth Continental Europe at SCAYLE, outlined five ways AI can reduce fashion return rates across the shopping journey. - SCAYLE published the commentary on October 28, 2026, from Hamburg, Germany. - The piece focuses on product pages, carts, checkout and returns processing.

The details: - More than half of senior executives at enterprise fashion brands, 54%, say size and fit make fashion more complex than other retail sectors. - Another 38% point to high return rates as a major challenge, according to SCAYLE's survey, Inside the Boardroom: Fashion Leaders on the Agentic AI Era. - In DHL's 2025 survey of 24,000 online shoppers across 24 countries, more than half said they had returned an item because it was the wrong size. - In the UK, Retail Economics and ZigZag forecast that 19.5% of non-food goods bought online will be returned in 2025, equal to £25.1 billion, or about $33.4 billion, in merchandise value. - AI-based size engines can use anonymized purchase histories, return reasons and garment measurements to predict fit for a specific shopper. - Those recommendations can be shown before an item reaches the cart, reducing the chance of multi-size orders. - Generative AI can show catalog items on body types, heights and skin tones that better match the shopper. - Virtual try-on can go further by showing outfits on an AI-generated likeness based on a shopper's own photos. - SCAYLE's survey found that 45% of fashion executives already use AI-generated models or visual content. - Retailers can flag bracketing in the cart when the same item appears in several sizes before payment. - An AI assistant can compare a shopper's fit history with garment measurements and suggest the most likely size. - Small incentives such as priority delivery or loyalty points can encourage shoppers to remove extra sizes. - Machine learning can estimate return risk from prior return behavior and cart contents. - Retailers can then adjust payment and delivery terms, including pay-later access, free shipping or upfront payment, based on that risk. - Björn Asdecker of the University of Bamberg has identified pay-later payments as one possible reason for Germany's high return rates. - Asos has added a transparency feature in the UK that shows shoppers their personal return rate in its app and warns them when they approach thresholds that could trigger return fees. - Computer vision at the returns station can inspect incoming items and spot recurring defects. - The system can identify issues such as faulty zippers, weak seams or sizing problems tied to a specific production batch. - When problems repeat on a newly launched item, retailers can add a fit note to the product page or pause the item until the issue is fixed.

Between the lines: - The argument is shifting returns management from a back-end logistics problem to a front-end conversion problem. - SCAYLE's framing suggests retailers can save money by reducing uncertainty before checkout instead of relying on faster reverse logistics after the sale. - The approach also links returns prevention to sustainability, since fewer returns mean less transport and handling.

What's next: - Retailers are likely to keep testing AI tools that predict fit, personalize product display and shape checkout terms. - SCAYLE's view is that the best return reduction comes from combining shopper-specific recommendations with feedback from returned products. - The company says retailers that reduce uncertainty before checkout will lose less margin to returns.

The bottom line: - AI can lower fashion returns best when it influences the purchase decision early, not when the package is already on its way back.