Company

Resources

Our Channels

Our solutions

Article

ChatGPT Will Now Show You Wearing the Shoes Before You Buy Them

18 October 2026
chatgpt-virtual-try-on-fashion-shopping

James Murphy – Head of Digital Marketing

OpenAI has added a virtual try-on feature to ChatGPT. When you are browsing clothing or accessories, you can hit a “try on” button, upload a photo of yourself, and it generates an image of you wearing the product. You can then save the items to Favourites or sort them into folders in your Library. It is live on both the website and the mobile app.

Now, anyone who knows me knows that fashion and I are not close. My wardrobe is three jumpers on rotation and a pair of shoes I have owned since roughly the last World Cup. So an AI that shows me what I would look like in something before I commit is, honestly, doing the nation a favour. In my case it mostly just confirms that the jumper was the right call again. For people with actual taste, it goes a good deal further than that.

And that is the point worth sitting with. This is another tool in the armoury for anyone selling fashion online, and it lands on the single biggest problem the category has.

Fashion is the most returned category in ecommerce, and footwear sits right at the top of it. The pattern is always the same. Someone orders two or three sizes of the same shoe, keeps one, sends the rest back, and the retailer eats the shipping, the handling, and the repackaging on every return. It is the tax everyone in apparel has quietly accepted for years.

Anything that helps a shopper buy the right thing first time goes straight to that line. If a try-on image means one fewer size ordered “just in case”, the saving is real and it is immediate. Perfect Corp found in a 2024 survey that retailers using virtual try-on saw their conversion rate more than double, and 38.4% of shoppers said virtual fitting rooms were the closest thing to recreating the in-store experience online. More confident buyers, fewer hedged orders, fewer parcels coming back.

None of this is brand new, which is exactly why it matters. Google has been building the same capability for years. Women’s tops in 2023, dresses in 2024, shoes in 2025. OpenAI is now catching up, and it is doing it inside the place where more and more people already start their shopping.

That tells you where this goes next. Once try-on becomes something shoppers expect from a chat assistant, the marketplaces will not sit still. Amazon, with the catalogue and the fashion ambition to match, is the obvious one to bring something similar to its own storefront, and the other platforms chasing agentic shopping will follow. The retailers who already have clean product imagery and accurate sizing data will be the ones ready to plug in when they do. The ones who do not will be scrambling.

This is the kind of shift we spend our time watching for clients at Luzern. Getting product data, imagery and listings in the right shape so that when a new discovery surface opens up, whether that is ChatGPT, Google or the marketplaces themselves, the brand is ready to appear and ready to convert. The try-on button is a small feature on its own. What sits behind it is where the returns get cut and the sales get made.