From ChatGPT and Google’s shopping-focused search updates this summer, to AI agents becoming a reality, the tech that’s shaping online shopping today looks completely different from this time last year. Virtual try-on is no exception.
After a few fallow years on the investment front, venture capitalists are betting on a new wave of virtual try-on startups that promise to transform the way we shop. The apps’ digital avatar technology makes previous experiments look rudimentary — digital twins, created from user selfies, are more photorealistic than ever. Celebrities, influencers and stylists are buying in, and this weekend, Doji launched its first exclusive in-app brand partnership with Peter Do. Will more brands follow?
Virtual try-on 2.0
Two virtual try-on apps featured on this year’s Vogue Business tech innovators list for the strides they made with the AI technology underpinning the digital twins in their apps: Alta and Doji. Alta, which launched in public beta mode in May, raised an $11 million seed round led by Uber and Poshmark backer Menlo Ventures and others including Aglae Ventures, the VC firm backed by the Arnault family of LVMH. Doji, which also launched in May and is currently invite-only, raised a $14 million seed round led by OpenAI and Skims backer Thrive Capital.
Both are competing with Big Tech companies like Google, which announced integrated virtual try-on features as part of its wider Google Shopping update in May, and unveiled experimental virtual try-on app Doppl in June that creates AI-generated videos so users can see how an outfit might feel. Some multi-brand e-tailers like Zalando also offer customers the ability to upload their body measurements to virtually try on outfits on an avatar that resembles their body shape.
The key difference between these startups and the experiments of larger tech companies lies within the photorealism of the virtual avatars. The AI models powering virtual avatars like Google s, Doji s and Alta s are diffusion models — the latest, most advanced type of generative AI for creating novel, hyperrealistic images — to create digital twins from a handful of selfies and full-body images that users are required to upload.
Alta is pitching itself more as a virtual personal stylist and is predominantly focused on letting users visualise outfits from their own closets, which they can upload to their virtual wardrobe inventory while occasionally trying on outfits that include shoppable pieces. Users can also create wishlists, specify their favourite brands and ask Alta’s AI agent what they should wear for a specific occasion, while giving it feedback when it throws up a suggestion they dislike.
“Diffusion model technology improves every day — looking at my own Alta avatar, I’ve seen a significant evolution from January to May to today,” says Jenny Wang, Alta CEO and founder. “Some of the nuances where we’ve seen the most improvement include the ability to retain words and graphics on shirts, capture stripes or embellishments on shoes, and properly position fun accessories.”
According to Wang, the complexity of the try-on depends on the garment category — her team of engineers are particularly focused on more complex try-ons, such as multi-outerwear layering and detailed jewellery stacking. My Alta avatar isn’t as photorealistic as Doji’s, but upon initial testing, I find the former’s styling recommendations are more accurate, thanks to the ability to prompt in natural language — a feature Doji’s founders say they may add in future.
Like Alta, Doji’s AI model learns more about the customer with time as they select their favourite brands and style more looks. Users can view item pricing, try an item on and click to be redirected to brands’ stores. Co-founder Dorian Dargan tells Vogue Business Doji has prioritised image quality to gain respect among the luxury fashion community and to maximise social shareability.
“We see the future of shopping as being fun, but also deeply personal. There’s a utility to fit tech, yes, but there’s a deeper build at play here — trust,” says Dargan. “We feel like we’re inheriting and respecting the tradition of image making, because that’s really what I think the fashion industry is built upon.”
A personalisation play
Where previous virtual try-on attempts were pitched more as pure-play “fit tech”, touted for their potential to fix fashion’s mammoth returns rate, investors and stylists are backing these new apps for their potential to transform how we shop online. While Big Tech companies like OpenAI and Google race to connect with consumers on a more personal level through AI-powered shopping search, investors are backing these virtual try-on apps for their reverse potential as a top-of-the-funnel sales tool.
“I wouldn’t call this fit tech, it’s much bigger. I’d call it the future of shopping,” says Miles Grimshaw, Doji’s lead investor at Thrive Capital. “Right now, when you shop online, it’s like you’re an amorphous blob, at best a cookie — there’s no you. But this is bigger than just, ‘I might not return this.’ It’s: ‘I might not ever have discovered this.’”
Rather than harnessing consumers’ intent-based search to drive purchases, Doji and Alta offer consumers a gamified discovery tool.
“This is much more impactful down the funnel, too,” Grimshaw adds. “From an investment perspective, I’m not as interested in doing the best volumetric scan of a piece if it means the return rate of a retailer improves by 2 per cent. That’s helpful, but it’s small. Here, the breakthrough in technology is massive. It allows everyone to feel that shopping is personal, and commerce becomes uniquely fun and inspirational. That’s a really big default behaviour change we think this technology will make happen.”
Getting users to upload items from their closets remains a high-friction moment. Amy Wu Martin, Alta’s lead investor at Menlo Ventures, argues that giving them an idealised avatar of themselves — using AI enhancements — drives users to keep generating more outfits. She says Alta’s so-called “power users” generate around 300 looks a week.
“It’s kind of like gamified learning — Duolingo showed us that the daily streak is a super powerful feature of retention,” Wu Martin says. “These avatars have the same power. Virtual try-on isn’t the product itself, it’s a question of what can the apps build around that daily streak?”
These kinds of experiences will be a key differentiator in the race to develop virtual try-on technology people actually use. “There’s a bigger shift underway that both sides will have to contend with: the commoditisation of the technology itself,” adds Matthew Drinkwater, head of the London College of Fashion’s Innovation Agency. “The underlying infrastructure for virtual try-ons — 3D body models, AI garment simulation, real-time rendering — is rapidly becoming more accessible. As barriers to entry fall, the core value is shifting away from raw tech and towards what surrounds it: compelling digital assets, strong partnerships with brands and creators, and the ability to deliver emotionally resonant experiences.”
Can they win over luxury brands and consumers?
Previous iterations of virtual try-on struggled to meet the high visual standards that luxury brands require for adoption. But there are some early signs that this fresh wave of AI-powered apps is finally closing the gap.
This week, Doji launched an exclusive in-app partnership with Peter Do, to showcase the designer’s new PD-168 collection. Users can purchase PD-168 items by tapping “buy” links within Doji, while Peter Do’s website links to the app so that shoppers can virtually try on the collection.
For Do, the ability to let customers try on PD-168 pieces virtually reflects the idea behind the collection as a “modular system” that integrates into the wearer’s lifestyle, he tells Vogue Business. “Collaborating with Doji allows customers to mix and match pieces to create their ideal uniforms while seeing their individuality represented in the e-commerce experience,” Do says. “This partnership enables a unique expression of the brand’s ethos.”
Jordan Grant, a fashion influencer and co-founder of members-only luxury shopping platform Mile Club, says she invested and collaborated with Doji because of its cultural relevance and accessibility, as well as the “world-building” potential the app offers luxury brands.
“It’s not just another sales channel, it’s a space where consumers are actively experimenting, discovering and building outfits. This means brands can show up in a much more organic, playful way, rather than through static ads,” Grant says. “What excites me most is how it deepens engagement with the aspirational customer. Instead of just following a brand on Instagram or window shopping, they can now style themselves head to toe in brands they admire and feel part of the brand’s world. That level of interaction is incredibly valuable.”
While Alta is yet to announce any brand partnerships, Wang says she wants to explore deeper integrations with labels in the future, where exclusive capsule collections and drops could also be showcased in-app. Alta has so far collaborated with celebrity stylists including Meredith Koop, as well as Gab Waller, who partnered with the startup in August to give its members access to her Sourced By platform’s luxury sourcers, including her own services. Waller tells Vogue Business that Alta serves as a “silent partner” for her work. “It allows stylists to focus on what matters most: the human relationship with their client,” she says.
“At the same time, there’s an opportunity for brands to tap into collaborations with stylists in this capacity, especially for up-and-coming designers looking to reach a new market and audience,” she says. “With wholesale in a strong disruption phase, this presents an opportunity for brands to reach new customers.” Wang says she’s received requests from brands to white label Alta’s technology and integrate it into their own digital platforms, and that luxury brands have been coming to her for updates on how the underlying tech is evolving.
While Alta and Doji’s backers are bullish on their potential to transform the way people shop, sceptics in the retail industry say consumers and brands’ product discovery habits may be much more difficult to change.
“Visualising how I would look in a large blazer doesn’t get me to like how the blazer fits me when it arrives at my house,” says Matt Powell, retail analyst at BCE Consulting. He points to the lack of standard sizing at brands, and the fact that even if a consumer discovers and buys more products through one of these apps, the return rate could remain just as high.
“At some point, the consumer will want to try it on in-person,” Powell says. “So the bigger question is: is this really a problem that needs this solution, or is the consumer going to say, ‘I’m just not interested in shopping this way?’”
Correction: This article was updated to rephrase Google s use of diffusion models. (22/09/2025)
Comments, questions or feedback? Email us at feedback@voguebusiness.com.
