We tried Google’s new Doppl app, which promises to turn your phone into an AI fitting room. Between startling successes and hilarious failures, is this the future of retail?
Introduction: The “Clueless” Dream Becomes Reality
I wasn’t planning on trying on Kate Hudson’s iconic yellow dress from “How to Lose a Guy in 10 Days” from my office this week. But that’s what happened when I downloaded Doppl, Google’s new AI-powered fashion experiment that lets users try on any outfit virtually through an AI fitting room. Think of Alicia Silverstone’s digital closet in “Clueless”—but with AI, and on your phone.
All you have to do is take a full-body photo, upload the outfit you want to try, and within 30 to 60 seconds, your digital twin appears wearing it. It’s supposed to replace the traditional fitting room. So, naturally, I decided to give it a try.
Next-generation technology enables full virtual clothing try-ons and boosts shopper confidence
The Shocking and Convincing Moment
My “twin”—who looked disturbingly like me, but with slightly different proportions and longer hair—stood in the iconic yellow dress I had been obsessed with since middle school. Then she waved. Every graphic movement is different; the app can generate short clips of your digital twin moving in the outfit, usually with a slow spin or a stiff pose. In this case, my twin raised her arm and struck a pose as if she were headed to the Academy Awards.
It was shocking. But I couldn’t stop watching. The fit wasn’t entirely accurate, but it was much more precise than I expected, and enough to make me genuinely want the dress. Maybe I need it.
How Does Doppl Work?
The Doppl app, launched last week through Google Labs, is both a clothing try-on tool and a tech experiment. Users can upload photos of outfits—whether from Pinterest, your favorite store’s website, or a jacket you saw at a thrift store—and the app generates a virtual version of you wearing the outfit. You can also skip using your own photo and choose from 20 pre-made virtual models of different ages, ethnicities, and body types.
Right now, Google says Doppl “may not always work.” The app only supports shirts, pants, and dresses—no shoes, bags, or accessories—and does not offer sizing advice or fit assistance. Even so, I wanted to see what it could do.
Between Success and Failure: Real-World Tests
One of the outfits I tested was from my Pinterest board titled “Shopaholic Life.” I chose a Saturday night look: a black top and a long flowy skirt. Doppl gave me a short black dress and black shoes that looked nothing like it. In some photos, it even added a few inches to my hair.
Other outfits worked much better. I uploaded a picture of Zara jeans that had been sitting in my shopping cart, and Doppl surprised me by generating an image that included the belt from the product photo, even though Google said accessories aren’t supported yet. The rendering wasn’t perfect, but as someone who is 178 cm tall and struggles to find jeans long enough online, it looked good enough. I bought them.
From what I’ve seen, simple outfits work best. The AI fitting room struggles with complex designs—multi-layered looks, blurry photos, and difficult fabrics—and sometimes invents new clothing from scratch if it can’t figure things out. When it works, it’s convincing. And when it fails, you find yourself watching a distorted digital twin wearing something you never asked for.
Conclusion: A Useful Tool or a Radical Shift?
Sucharita Kodali, a retail analyst at Forrester, said: “This is generative AI in augmented reality form. I can’t imagine it won’t be useful. Will it be transformative and double anyone’s business? No. But it will be useful.” The app isn’t perfect yet; it skips personal questions like your height or measurements, which could make the trials more precise. But while it may not replace store fitting rooms anytime soon, for a free app on your phone, it comes surprisingly close. And it just might convince you to buy something you wanted anyway.
