Shopping what we see on our TV screens has long been an industry pipe dream — a great idea in theory that’s been difficult to execute.
Now, two former Google executives think they’ve cracked it. Josh Lanzet, who spent 13 years at Google leading media and entertainment partnerships for streaming, and Jason Fahlstrom, who spent 11 years working on the tech company’s AI strategy, have spent the last two years working with the latest AI and computer vision technology to build Silvr, a startup they’re hoping will bring us this dream of shoppable TV via an app that launches on February 24. Angel investors from Netflix, Disney, Google and LinkedIn backed the startup in a family and friends funding round last month, and the founders are now gearing up to raise $3 million in pre-seed investment.
Silvr’s consumer-facing app allows viewers to point their cameras at their TV or laptop screen to instantly identify and shop the exact item a character is wearing. They’re also building a white-label B2B platform to enable instant commerce for streamers like Netflix and HBO Max, so that viewers can pause and click on their screens to shop.
“For the longest time, the ‘where did you get that?’ problem has been a visual question answered with a text-based answer, but now our consumer app matches it with a visual answer,” Lanzet says. “There’s been object recognition but no fashion recognition, and that’s what Silvr brings.”
Fashion brands and streamers are taking note. Silvr has partnered with over 300 fashion brands ahead of launch, including Alexander McQueen, Isabel Marant, Etro, Alo Yoga, Nike, New Balance, and Ganni. It’s also got major retailers like Nordstrom, Macy’s and Amazon on board, resale and vintage platforms like The RealReal and 1stDibs, and has partnered with Impact.com, CJ and Rakuten, the affiliate networks behind e-commerce attribution. Lanzet and Fahlstrom also say they’re in “active discussions” with all the major streaming platforms for their B2B product, with one platform having already signed a letter of intent.
The founders say the launch is timed to the rapid recent developments in AI and computer vision, consumers’ growing readiness for shoppable entertainment like live shopping, and the accelerating adoption of visual search. AI models have rapidly developed over the last two years so that they can more accurately interpret visual context, with so-called “multimodal” search now possible via the main AI chat platforms like OpenAI, as well as Google Lens and Pinterest Lens — viewers can, technically, already point their cameras at their screens and use these features to identify products within film.
Younger generations, in particular, are driving adoption of visual search — 10% of Gen Zs now start their Google searches with a photo or image, for example. It’s also not the first time we’ve seen a tech platform try to master shoppable TV — Netflix partnered directly with Google Lens in late 2024 so that viewers could opt in for “pause ads” where they could directly shop certain pieces from the Emily in Paris series. Paramount has also partnered with Shopsense AI to allow viewers to scan QR codes and live shop some products when they’re watching awards shows like the VMAs. Amazon began experimenting with shoppable video as far back as 2021, while startups like Seek, TheTake and Beamly have all closed down after attempting to power shoppable TV.
These attempts have all struggled to reach widespread adoption, as they don’t always yield the most accurate results and represent a clunky multi-step user journey that filters out all but the most determined viewers. But Silvr’s founders say their tech is the most accurate yet for fashion.
“We stand on the shoulders of a lot of people who have tried to do this before — in fact, two of those companies informally advise us as founders — but I think we’re beneficiaries of where consumer behavior and technology are at the moment,” says Lanzet. “TV and movie viewing is now completely aligned to OTT apps [platforms that stream video and audio online, bypassing traditional cable and broadcast providers] via phones, laptops, or TVs; Instagram and TikTok have trained consumers — especially Gen Z — to know how to and want to shop the media they’re currently experiencing; and computer vision is better than it’s ever been before.”
Lanzet and Fahlstrom say they’ve built their model using the latest computer vision and AI, which — much like Daydream’s fashion discovery platform — is exclusively trained on fashion and accessories. They say it therefore understands garment details, style nuances, and accessory specifics, which can make or break accurate product identification.
They’ve also focused on partnering directly with brands and retailers, so that full product catalogs are available within the app. Users can see which products and shows are trending within their app homepage, and the more they interact with the app by searching products and shows, the more relevant their “curated” Silvr feed will become, the founders say. When users click on products that have been identified by Silvr to purchase, they’re sent directly to direct-to-consumer (DTC) sites, rather than a Silvr shopping platform intermediary — something Silvr’s founders say is a major pull for brands. In future, Silvr wants to incorporate an integrated checkout, so that consumers don’t need to click through from the Silvr interface onto a separate brand site.
Silvr will charge a platform licensing fee for its business-to-business (B2B) white-label technology, and for both its business-to-consumer (B2C) app and B2B platform, its main revenue stream will come from affiliate partnerships with brands. It will also charge streamers and brands to access its aggregated and anonymized data, so that they can learn how different consumer demographics are engaging with the platform. Silvr also plans to give brands and retailers the option to pay for priority placement within its app, much like Google shopping ads, so that their site becomes the first option in product search results. Lanzet and Fahlstrom say they’re mostly focused on scaling their B2C app in the near term, to reach the critical mass of photo uploads that Silvr’s AI model needs to improve its product recognition, and to encourage consumer behavior. But in the longer term, they’re hoping their B2B functionality proliferates the market, and they eventually become the tech layer enabling consumers to shop everything they see.
“In future, we’re likely going to see fewer people taking pictures of TV and movies specifically in the app, and focus more on social media and real-life outfits and fashion,” Fahlstrom says. “Beyond that, our future is not app-based. If you look at Silvr 10 years from now, our goal is to completely remove friction in the shopping journey.”
Fahlstrom says his dream for Silvr’s computer vision is a world where consumers can scan anything they see to instantly shop it. As tech founders race to develop AI-powered wearable devices that will do away with our need for separate devices, it’s a vision that ties in with the wider industry’s vision for e-commerce that’s incorporated into the real-world, beyond screens. “The idea is to create action and remove all boundaries for the consumer,” Fahlstrom says. “That’s where AI will be the big service for the future of commerce.”
