Utterly Clueless, Google Virtual Try-on

Reading Time: 4 minutes

The search for the right fashion merch went up just a perch. Belated as it may be, Google’s latest experiment with Virtual Try-on (VTO) is at once a turn on for seasoned fashionistas and creators of online shopping vistas. Set aside the neural networks and cross attentions, the image diffusions and geometric warpings. A training data set of 35 billion+ products alone is unmatched canvas, a frontal assault that takes the eCommerce battle right to the TikToks and Snaps of the AI-verse.

Game on

VTO is mighty ammo for spartan brands and retailers short on resources to hire and photograph models. With one glimpse of the garment it tells you how it would realistically drape, fold, cling, stretch and form wrinkles and shadows on a diverse set of real models in various poses. On the trail of enabling advertisers to tweak advertising imagery to make it more compelling, to visualize lipstick and foundation on a range of models and to respond to product searches with additional details such as top considerations and product summaries, VTO is part of the effort to ringfence Google’s market-leading 28.4% share of the digital ads space. Google is first about access to information and discovery and secondarily about direct eCommerce. In terms of monetization, the advertising model is much more lucrative and will always be primary. What’s unique about Google VTO though is the ability to visualize a piece of clothing on a real model who looks like you across a wide range of retailers, at the same time

Will Google, given its massive reach and Gen AI capabilities, be able to fully disrupt the VTO space? Rephrase that to: Does it really matter? As we have said earlier, tech companies have lots of bets that don’t pan out —but they collect a lot of data from these bets, which can be applied to other products in the works. “These bets are not always meant to scale,” said Ashwini Asokan, co-founder and CEO, Vue.ai, adding that the try-on service is also a way for Google, one of the pioneers of generative AI, to show off its AI muscle.

For over two years now, Vue.ai has deployed the Virtual Try-on tech as part of its Virtual Dressing Room offering at about half a dozen fashion brands and retailers like Lane Crawford, Picard and Showpo. On clothing, for both men and women. On non apparel categories such as bags, sunglasses and accessories. Client testimonials attest to the tech’s versatility both as a revenue generator and product returns reducer. 

One up on Google Virtual Try-On

Vue.ai’s Virtual Dressing Room solution provides a much richer experience for both sellers and shoppers. 

• While Google provides VTO on 80 pre-determined models, Virtual Dressing Room actively works with and licenses more than 150 real models of different ethnicities, sizes, and genders to create expansive options for representation within the Virtual Dressing Room experience.  This 150-model collaboration is only a starting point, and can be supplemented or exchanged with virtually any model a brand may want to cast and employ for the Virtual Dressing Room experience.
• Participating models are licensed and employed for strategic photoshoots that amplify the capabilities of the Virtual Dressing Room experience, including being shown in stylistically fashion-logical footwear and hairstyles that align best with the products being visualized.
• Virtual Dressing Room enables styling options beyond a single-product try-on while Google VTO currently only visualizes women’s tops, with men’s tops slated for later this year. Vue.ai also allows for visualization of entire outfits, as well as styling (shirt tucked in / out; pairing and layering of top, bottom wear, dresses and jackets; Shop the Look feature).  This allows customers to mix-and-match complete looks and organically creates opportunities to cross-merchandize and increase basket size.
• Personalization of recommended products based on the shopper’s purchase / session history is also a distinct and decorated feature of Virtual Dressing Room
• Virtual Dressing Room also provides customers a plush dashboard of precious analytics for backend mining. Design, manufacture, sampling, buying, merchandising, inventory management and marketing would all benefit from data-driven decision making.

Cognizant and considerate of the ethical dimensions of deploying human models in Generative AI solutions, Vue.ai adorns a very sensitive and equitable approach. “We pay models for photoshoots like they would a traditional photoshoot. We brief them on how their images will be used prior so they’re well-informed, which we find works well,” says Brian Harris, director of R&D, new initiatives and digital transformation at Vue.ai.

At the AI Transformation retreat in Jamaica, German fashion brand PICARD extolled its experiences deploying Vue.ai, reaping 65% increase in overall conversions, 12% increase in AOV, 22% reduction in product returns and 75% reduction in photo shoot costs.

The revenue-generating advantages of Virtual Dressing Room includes greater personalization, improved customer experience, location independence, higher customer engagement, and increased competitiveness. Cost-reducing perks are fewer returns, minimized losses, and the prevention of wasted time and resources. With Google VTO, additional guided refinements allow shoppers to use color, pattern, and style to find items similar to what they are browsing. Incremental improvements based on data such as these from previous shopping enhancements have earlier been available from third-parties in order to improve shopping but companies like Google raise the baseline across eCommerce by providing foundational tools, models, and algorithms. 

Sure, brands, retailers and marketplaces need feature parity with competitors and Google in order to keep consumers in the purchase funnel. Integrating independent Virtual Dressing Room solutions into the digital storefront enables them to not only prevent shoppers from going offsite to use try-on tools, but also monetize the rich first-party data for incremental revenues. A Retail Media Network for a sidekick, if you will. Unthread Google VTO’s seams a little and do we see this as yet another opportunity for data grab by Google, riding on Shopping Graph’s legions of sellers across the world? First-party data about body sizes, ethnicity and gender are on the platter, empowering Google to make incredibly definitive inferences about lifestyle, health and more. As Biden says, God save the fashion queen, man.

Cher Horowitz would have cherished this VTO moment so very dearly, in the 1995 fashion flick Clueless.