Virtual try-on 2.0: Will it change the way we shop?

The past 12 months have reshaped the consumer shopping stack. From conversational search to autonomous agents, AI is now steering discovery, decision-making and post-purchase service. Virtual try-on (VTO) technology, once a novelty for beauty and eyewear, is evolving fastest. A fresh cohort of startups, backed by significant venture rounds, is closing the realism gap that long held back adoption. Their digital twins look less like videogame avatars and more like real people — and the business models surrounding them are maturing.

Two Vogue Business Tech Innovators — Alta (https://alta.ai/) and Doji (https://doji.com/) — have become reference points for this new wave. Both lean on diffusion models to generate photorealistic avatars from a short selfie sequence, and both are positioning their products as much more than "fit tech." They see VTO as a top-of-funnel discovery engine where customers can play, curate wardrobes and form deeper ties with brands. Alta's $11 million seed round led by Menlo Ventures and Doji's $14 million raise from Thrive Capital underscore investor confidence that this is a durable consumer behaviour shift, not a passing experiment. The integration of ai try-on with augmented reality is pushing retail tech forward.

The second generation of digital twins

Alta’s pitch is a virtual stylist that lives inside a user’s wardrobe. Shoppers upload pieces from their own closet, layer in shoppable items, and refine recommendations with natural-language prompts. Founder Jenny Wang says the team has focused obsessively on tricky categories — multi-layer outerwear, detailed jewellery, textured footwear — and notes steady improvements in retaining graphics, embellishments and accessory positioning as the model trains on more data. Power users reportedly generate hundreds of looks a week, nudged by gamified streaks that keep them coming back.

Doji, still invite-only, prioritises sheer visual fidelity. Co-founder Dorian Dargan describes the company’s north star as “earning trust with the luxury community” by producing imagery people want to share publicly. The app’s first exclusive brand partnership, with Peter Do’s PD-168 collection, shows how designers can stage modular wardrobes inside the platform. Influencers like Jordan Grant cite Doji’s “world-building” potential — a sandbox where brands feel premium but playful.

Large platforms are watching closely. Google folded generative try-ons into Shopping earlier this year and continues to test Doppl, which turns AI imagery into motion. Zalando and other multi-brand retailers are inching forward with avatar-based fitting rooms tied to uploaded measurements. Matthew Drinkwater, who leads the London College of Fashion’s Innovation Agency, argues that as the core technology becomes commoditised, differentiation will hinge on partnerships, creative assets and emotional resonance rather than the underlying render engine.

Enter Own Every Look

Own Every Look (OEL) is staking out a different corner of the field. Rather than gating its experience behind a single marketplace or mobile app, OEL ships as a Chrome extension and responsive web app available at https://owneverylook.com/. The extension lets shoppers capture inspiration from anywhere on the open web — whether that is a runway livestream, a resale marketplace listing or an indie designer’s lookbook — snap or upload their image, and generate a refreshed outfit in seconds. Profiles, favorite looks and wardrobe histories sync back to the web interface, creating a persistent style archive that follows users from research to checkout.

That browser-first strategy yields a two-sided advantage. Consumers avoid the install friction of yet another app while getting faster feedback on real purchases they are already considering. Retailers, meanwhile, see affiliate-ready traffic arriving from audiences who have already validated a look in their own context. Styld, the team behind OEL, built its pipeline on top of a FastAPI backend with Google’s Gemini 2.5 Flash for rapid diffusion, layered with a facial preservation pass so that regenerated outfits keep users’ identities intact. The system supports hair transformations alongside clothing, nudging virtual try-on beyond single-category experiments.

OEL also leans into cross-platform data. Because the extension works across any e-commerce site, the platform aggregates insight on which garment cuts, fabrics and styling cues resonate for different body types. Those learnings help refine recommendations on subsequent sessions and open the door to white-label partnerships for brands that want to embed the experience on their own domains without surrendering customer relationships. For consumers wary of privacy trade-offs, OEL pitches a pragmatic stance: processing remains cloud-based for speed, but profiles are user-controlled, and galleries can be purged with a click.

A personalisation play — but make it practical

Investors backing the new VTO wave consistently emphasise personalisation over pure fit correction. Miles Grimshaw of Thrive Capital frames Doji as “the future of shopping” because it surfaces outfits a customer might never discover alone. Menlo Ventures partner Amy Wu Martin sees Alta’s AI-enhanced avatars as a retention lever — a feedback loop that keeps users styling as if they were playing a learning game. OEL, by contrast, is betting that practicality will win loyalty: if a shopper can try a piece directly on the website where they intend to buy, save it to a cross-device gallery, and revisit it later without re-uploading photos, the habit becomes sticky.

The friction point everyone acknowledges is onboarding. Creating a convincing avatar still requires several well-lit photos, and cataloguing an entire wardrobe takes time. Alta tackles this with streaks and stylist collaborations; Doji pairs exclusives with high-impact launches; OEL leans on utility, offering immediate value the moment a user clips a product page in their browser. As Drinkwater notes, with the diffusion engines themselves now widely available, the surrounding experience — onboarding flows, curation layers, community features — is what will differentiate winners from also-rans.

Winning trust from luxury and mass market alike

Luxury brands, long sceptical of virtual try-on’s visual fidelity, are testing the waters again. Peter Do’s PD-168 collaboration with Doji gives shoppers a modular, mix-and-match experience that mirrors the collection’s ethos. Alta is fielding requests from labels interested in private capsules, and celebrity stylists like Meredith Koop and Gab Waller are using the app as a behind-the-scenes assistant. OEL is courting both premium and mass-market partners with a lighter integration footprint: retailers can surface their catalogues within the platform or tap into its API for branded experiences without forcing shoppers off-site.

Skeptics still point to structural hurdles. Matt Powell at BCE Consulting reminds us that avatar confidence does not guarantee satisfaction when the package arrives, especially with inconsistent sizing. If the primary KPI is reducing returns, virtual try-on alone may disappoint. But when positioned as a discovery engine — a canvas for inspiration, storytelling and experimentation — the economics look different. A customer who self-styles a look, saves it, and shares it with friends is more likely to develop brand affinity, even if they cross-shop across platforms.

The next chapter

Meticulously rendered avatars are fast becoming table stakes. The next frontier will be how virtual try-on experiences plug into the broader commerce journey: stylist-guided edits, collaborative shopping, automated tailoring recommendations and sustainable merchandising cues. Startups like Alta and Doji are proving there is appetite for aspirational play; Own Every Look is demonstrating that there is equal appetite for practical, cross-site tooling that slots into everyday browsing. If the three approaches continue to steal mindshare from static product grids, the question may no longer be whether virtual try-on will change how we shop — but how quickly shoppers will expect every brand to offer it.