Virtual try on clothes: what Google, Zara and ASOS tools do, and what the return data shows

Virtual try on clothes tools are now built into Google Search, retailer sites and shopping apps, but the evidence that they cut returns is thinner than the marketing. Here is what the main tools do, what Google's own help page says they are good for, and what the return data does and does not show, all dated.

The short answer: they are decent at showing how a garment might look on you, and unproven at telling you whether it fits.

What Google's try-on actually does

Google first showed AI try-on with models in 2023, using a generative model that, in its words, can "take just one clothing image and accurately reflect how it would drape, fold, cling, stretch and form wrinkles and shadows on a diverse set of real models." At launch on June 14, 2023, it covered women's tops from brands including Anthropologie, Everlane, H&M and LOFT, shown on models sized XXS to 4XL.

The version most people meet now uploads your own photo. TechCrunch reported on December 11, 2025 that Google first launched the feature in July 2025 across Search, Google Shopping and Google Images, and that a December update let US users start from a selfie and generate a full-body digital version of themselves with its Nano Banana image model. Google's help page says you must be at least 18, signed in, in a supported country, and have Search history and personalized recommendations turned on. It covers tops, bottoms and dresses; lingerie, bathing suits and accessories are not supported, and sponsored shopping ads do not get the "Try it on" button.

Google's own framing is modest

The help page describes the result this way: "This offers a helpful vibe check, giving you a general sense of how the item might complement your personal style as you explore different looks." That is a style preview, not a fit guarantee. The same page tells users to wear fitted clothing in the photo and warns that "overly baggy outfits may distort the result."

That matters because Google's 2023 launch post cited a survey claim that 59% of online shoppers felt dissatisfied with an item because it looked different on them than expected. That is a Google-cited statistic, so treat it as a vendor figure. It describes the look problem the tool targets, not fit or size.

Who else is running it

PYMNTS, in an April 6, 2026 report, lists the field. Google added Australia, Canada and Japan in October, plus shoe try-on. Perplexity launched its own try-on tool in November for Pro and Max subscribers. Shopify has integrated the startup Genlook's try-on app. Zara rolled out a try-on tool in December alongside return fees for online orders. A startup called Catches went live on luxury brand Amiri's website in March, and projects a 10% conversion lift and a 20 to 30 times return on investment for brand partners, figures PYMNTS says the company itself supplied.

That last point is the pattern. The big numbers come from the people selling the tool.

The returns problem it is meant to solve

The National Retail Federation and Happy Returns, a UPS company, estimated in their 2025 Retail Returns Landscape report, released October 15, 2025, that retailers expect 15.8% of annual sales to be returned, or $849.9 billion. For online sales the estimate is 19.3%. Shoppers aged 18 to 30 made 7.7 online returns on average in the previous 12 months, and 82% say free returns are a major consideration when they buy. Those are all-category figures, not apparel only, so they show the size of the problem and not the clothing share.

PYMNTS relays Catches founder Ed Voyce telling CNBC that uncertainty over fit is the primary driver of both returns and abandoned carts. That is an executive's view, offered by a person selling a try-on product.

What the return data shows

Here is the careful part. PYMNTS reports that ASOS recently posted a 160 basis point reduction in its returns rate, partly from experimenting with virtual try-ons in a partnership with the startup AIUTA. That is 1.6 percentage points, and "partly" means the try-on tool is one of several causes. PYMNTS also reports that Zara paired its try-on tool with return fees, so any improvement there cannot be pinned on the tool alone.

PYMNTS states it directly: "no platform has published a definitive return rate reduction tied solely to virtual try-on." We found nothing in the sources we opened that contradicts that. Be wary of round figures in vendor blogs, such as returns "cut by 36%," that do not link to a study.

How to use one without being fooled

Use the preview for color, drape and overall look. Do not use it to pick between sizes, because the tool renders an image of a garment on a body, and the help page we read presents it as a style preview, not a size guide. Check the retailer's size chart and its stated fit notes, and read reviews that mention fit.

Be aware of the privacy trade-off. Google's tool asks you to upload a photo of yourself, saves it for convenience, and lets you delete it at any time. Decide whether that is worth a style preview before you upload.

Our take

Virtual try on clothes tools are worth a look for browsing and a poor substitute for a measuring tape. The strongest independent signal is ASOS's 160 basis points, and it is partial. We would use the tools for look and drape, ignore the vendor lift claims, and wait for a retailer to publish a result that isolates the tool. For the business model that makes returns so costly, see our explainer on what fast fashion is.