How AI is trying to solve retail’s returns problem
Memory Makers Group | Istock | Getty Images
It’s getting tight here; drags there; Covering is wrong. These are some examples of the feedback that a new set of AI applications could enable a potential customer to try on clothing before purchasing, reducing the chances of an item being returned to the store in the process.
Fashion retailers are increasingly turning to AI to solve a problem of increasing product returns, a persistent drag on profitability, and what many in the industry are calling the industry’s “silent killer.”
A growing number of AI start-ups have emerged to provide virtual try-on technology that allows potential customers to visualize fit and style before purchasing.
While tech companies have been trying to solve online compliance issues since the 2010s, the rapid development of generative AI has finally made these applications good enough to meaningfully impact retailers’ bottom lines.
The US National Retail Federation predicted late last year that: 15.8% annually retail sales total $849.9 billion in 2025. For online sales, this figure increased to 19.3%. According to NRF’s findings, Generation Z continues this trend; Shoppers aged 18 to 30 made an average of eight online purchases per person last year.
Most returned products never make it back to the shelves, and often cost the retailer more in processing costs than the value of the refund. This is a multibillion-dollar problem for the industry that directly eats into companies’ margins.
“Finding out how to proactively use returns and then minimize them can be a meaningful driver of business and profitability,” Guggenheim Senior Managing Director Simeon Siegel told CNBC.
Siegel said that while fit technology is never as good as trying something out in person, it’s a great way to bridge the gap. “It’s going to continue to get better, I think it’s going to continue to reduce returns.”
Mirror-like realism?
The main reason for returns and cart abandonment is uncertainty around compliance, Ed Voyce, founder and CEO of AI startup Catches, told CNBC in an interview.
Catches has developed a platform that allows users to create a “digital twin” where they can virtually try on clothes using what it calls “mirror-like realism.” The app went live on luxury brand Amiri’s website last month for a select range of clothing.
Unlike other models that Voyce says “just look nice,” the Catches platform combines the physics of fabric texture and how the material interacts with a moving body.
“The reason we built Catches was to leverage the kind of confluence of technologies that are happening right now to solve this problem effectively,” says Voyce, who founded the startup backed by . LVMH’s Antoine Arnault and what was built on him Nvidia’s CUDA platform.
“The reason it’s now solvable from a scheduling standpoint is that you need to be able to run images on bare metal in the cloud for end users cheaply enough to build a project. [return on investment] for brands,” says Voyce.
“This technology has the potential to impact the entire industry and will truly usher in the new wave end users have been waiting for.”
Margin protection
These AI tools not only reduce returns but also help increase purchases.
While e-commerce has grown rapidly in recent years, with online shopping driving retail sales, current U.S. trade policy under President Donald Trump It will have a dampening effect on the industry, which is largely reliant on manufacturing in Southeast Asia. Across the price spectrum, retailers are struggling to maintain margins as costs rise and consumers become increasingly price sensitive due to inflationary pressures.
While returns have a meaningful impact on profit margins, they are also a critical factor in consumers’ purchasing decisions. NRF data shows that 82% of consumers think free returns are necessary, but the cost of providing them is becoming unsustainable for many brands.
Retailers are now testing a mix of technology and policies to protect margins.
Strategies to reduce returns range from charging for return shipping to providing more detailed size information to encouraging exchanges over refunds.
Owned by Zara Inditexit was one of the first to introduce returns fees for online orders, and although it was a controversial change for some customers, it helped the Spanish retailer maintain gross margins and curb the practice of “bracketing”, the practice of buying more than one size to try on at home.
The retailer also launched a virtual try-on tool called “Zara try-on” in December.
Meanwhile, ASOS It recently highlighted a sharp improvement in profitability, partly due to a 160 basis point drop in the rate of return.
Online fast fashion gamer I’m doing virtual experiments The trials, in partnership with deep tech startup AIUTA, allow potential customers to see an outfit on a variety of body types, heights and skin tones. However, ASOS warns that the tool is designed to provide general guidance and customers should still check size guides before purchasing.
ShopifyMeanwhile, the startup has integrated Genlook’s AI virtual try-on app into its commerce platform, which it says “eliminates measurement doubts, increases buyer confidence, and encourages higher conversion rates while reducing costly returns.”
Technology giants like it Amazon, AdobeAnd Google We have also created virtual trials in various shapes and forms by partnering with big brands to popularize the technology.
Starting April 30, Google’s virtual try-on technology will be accessible directly from product search results on Google platforms, according to Google Labs’ website.
As for Catches, it predicts that the app can deliver a 10% increase in conversions and a 20 to 30x return on investment for brand partners. It focuses on luxury brands because of their higher prices. The startup hasn’t yet put a number on how much returns could drop using the platform, but it’s aiming for “large reductions.”
Not a fix-all solution
“There are definitely companies that are seeing benefits; it’s harder to figure out how to measure those,” Siegel said.
While the benefits are clear, the analyst warns that AI is not a magic bullet. Beyond compliance, retailers are looking to AI for inventory management, customer targeting and fraud prevention.
“These are all really interesting use cases, as long as companies don’t abandon who they are,” Siegel says.
“What you sell will always be more important than how you sell it, and so I think remembering that will help determine who will gain from AI, who will benefit from it, and who will be consumed by AI.”
