Virtual Try-On Tools Cut Fashion Returns by Over 40 Percent, Early Data Shows

Online fashion return rates have become one of retail’s most intractable problems. Fashion e-commerce returns ran 24.4 percent of sales in a 2023 study, well above the 19.3 percent average NRF found across all online categories in 2025. That translates to roughly $38 billion in clothing shipped back to retailers in 2023 alone. Size and fit is the leading cause, named by 53 percent of apparel sellers, creating a cycle in which customers order multiple sizes hoping one will work—a practice the industry calls bracketing—knowing many will be sent back.

Virtual try-on technology powered by AI body scanning aims to collapse that uncertainty. By letting shoppers visualize how a garment fits their specific body before purchase, these tools reduce the guesswork that drives returns. Early adopters report return reductions ranging from over 40 percent to 160 basis points, though the data remains fragmented across different retailers and implementations. The technology has begun spreading beyond standalone retailers to search engines and commerce platforms, signaling that the industry sees the experiment as worth running at scale.

How body scanning and virtual fitting work

Virtual try-on starts with body capture. Shoppers take a photo or video of themselves on their phone. The customer then selects a garment and watches it rendered on their own captured body in real time, often adjustable for color and fit options. Some platforms, like the tool ASOS deployed in partnership with the startup AIUTA, let shoppers upload their own photo or choose an AI-generated avatar matching their proportions. The rendering typically completes within four to seven seconds.

The technology extends beyond retail. Google expanded its virtual try-on feature to Australia, Canada, and Japan in October 2025, adding shoe fitting to apparel capabilities. Perplexity launched a try-on tool in November 2025 for Pro and Max subscribers, with results generated in under a minute. Shopify has integrated Genlook’s AI try-on app directly into its commerce platform, which the company says removes sizing doubts and drives higher conversion rates.

The return rate math
Fashion e-commerce returns ran 24.4 percent of sales in 2023, versus 19.3 percent across all online retail in 2025. Each returned item costs retailers $25 to $30 to process and handle, with 58 percent of shoppers ordering multiple sizes of the same item.

Why sizing accuracy matters to margins

The economics of returns weigh heavily on fashion retailers. Each returned item costs between $25 and $30 to handle—accounting for shipping, product damage, customer support, and other expenses. With billions in returns each year, cutting returns even slightly reshapes profitability.

Bracketing alone inflates return volume. Research found that 58 percent of online shoppers order multiple sizes of the same item with the plan to keep one and return the rest. A customer ordering a dress in three sizes is generating three shipments and two returns, each hitting the retailer’s cost structure. Virtual try-on targets that behavior directly by providing confidence in fit before checkout.

ASOS has tied the technology to financial incentives. The retailer introduced a transparency tool in January 2026 that shows customers their individual return rates in real time. Customers maintaining return rates below 70 percent enjoy free returns. Those exceeding 70 percent face a £3.95 fee if keeping under £40 worth of items, with an additional £3.95 restocking charge above 80 percent. The virtual try-on feature, deployed across approximately 10,000 products on its iOS app, works in concert with these policy changes and improved product videos and sizing information.

What retailers have documented so far

ASOS recently reported a 160-basis-point reduction in its returns rate, partly attributable to experiments with virtual try-on in partnership with AIUTA. However, the retailer attributed much of that improvement to other factors including clearer sizing information, product videos, and new return policies that charge fees to high-return customers.

Other retailers cite broader improvements. Fytted, which provides virtual fitting room software across more than 600 brands including Free People, J. Crew, and Lululemon, claims to have reduced returns by over 40 percent for its users. The platform analyzes multiple factors when recommending fit, considering color season, body shape, and personal style. Amazon introduced virtual try-on for shoes on its iOS app in the U.S. and Canada, with thousands of sneaker styles from eight brands including New Balance, Adidas, Reebok, and Puma. The company has not published data on whether the feature reduces returns or increases conversion.

A more controlled study offers credible baseline data. Zalando, the European e-commerce retailer, conducted an A/B test with 720,000 customers per group and found that size advice—recommendations on which size to select—reduced size-related returns by 3.8 percent. The finding, based on a controlled sample at scale, suggests that sizing tools deliver measurable but modest improvements on their own.

Luxury brands are also experimenting. Amiri, a high-end apparel brand, went live with Catches’ virtual fitting platform in March 2026, enabling customers to see how pieces fit before purchase. Research from Perfect Corp. found that brands using try-on technology experienced sales increases of 2.5 times, and 38 percent of respondents identified virtual fitting rooms as the most effective method for bridging digital and physical retail experiences.

No major retailer has published a definitive return rate reduction tied solely to virtual try-on, separate from return policies, size guides, and other concurrent changes.

Widespread adoption, uneven documentation

Zara deployed a virtual try-on feature in December 2025 alongside new return fees for online orders. The retailer has not isolated the try-on feature’s impact on returns from the effect of the fee structure itself. ASOS’s rollout across approximately 10,000 products on its iOS app reached select UK and US customers initially, with plans for broader expansion.

Yet a critical gap remains: no major retailer has published a definitive return rate reduction tied solely to virtual try-on, separate from return policies, size guides, and other concurrent changes. Most published improvements represent the combined effect of multiple interventions. Simeon Siegel, senior managing director at Guggenheim, notes that AI-generated visuals can now be delivered in the cloud cheaply enough for brands to generate a return on the investment, making margin improvement feasible at scale.

What the technology still can’t solve

Virtual try-on addresses sizing but not style. A garment may fit perfectly according to AI measurements but disappoint a customer who doesn’t like the drape, color accuracy online, or how it looks in natural light. McKinsey research suggests about 70 percent of apparel returns trace back to poor fit or style, meaning even perfect sizing captures only part of the problem.

Accuracy varies across body types and lighting conditions.

Adoption remains uneven. Some brands and retailers have integrated the technology; many have not.

Photo: Rob Olivera · CC BY 2.0 · via Wikimedia Commons