Tilt's Snap: What AI Is Actually Doing for Live Commerce
Tilt launched Snap in April 2026, an AI feature that creates product listings from live video in under a second. Early testing shows a 47% uplift in sell-through rate. The mechanism is less glamorous than it sounds, and more interesting than the headline suggests.
There is a version of live commerce that looks effortless: the seller is relaxed, the products are flowing, the chat is active, someone buys the lamp. What you don't see is the tab switching. The paused stream. The seller typing a product title while their audience waits. The 8 to 12 minutes it takes to create a single listing manually, in real time, in front of people who came to be entertained.
That friction is where most Western live commerce ambitions have quietly stalled. The format has never really had a consumer problem. People will watch someone demonstrate a product and buy it if the show is engaging. The problem has been operational, on the seller's side, and it's specific: running a live stream and running a shop at the same time is a genuinely difficult two-job role.
Tilt's Snap feature, launched in late April 2026, addresses exactly that. And the approach is more interesting than the usual AI launch framing suggests.
What Snap Actually Does
The function is specific: Snap watches and listens to a live stream in real time, identifies products as they're held up or described, and automatically generates a complete listing (title, description, category, price estimate) the moment a seller presents an item to camera. The buying link is live while the product is still in the seller's hands. No tab switching, no typing mid-broadcast, no audience waiting.
Tilt says Snap identifies products with 94.8% accuracy, a vendor-reported figure with methodology not disclosed, but consistent across the independent trade coverage. In early testing, items listed via Snap achieved a 47% increase in both sell-through rate and sales efficiency, compared to 32% for manually created listings. These are Tilt's own internal figures from a limited early-testing cohort, not independently audited — but the mechanism that would produce them is straightforward, and worth examining.
The launch positions Tilt as what the company is calling "The World's First Agentic Live Commerce Platform", a superlative that is essentially unverifiable but signals clearly that Snap is phase one of something larger, not a standalone feature.
"Live selling works best when sellers can just be themselves," said Abhi Thanendran, Tilt's co-founder and CEO. "Right now they're trying to run a shop and host a show at the same time. We want the only job a seller has on Tilt to be talking to their audience. Everything else should be the AI's problem."
The Mechanism of the Uplift
The 47% improvement deserves some examination, because it's coming from a specific intervention and the mechanism matters for understanding how generalisable the result is.
If a product listing goes live in under a second rather than 8 to 12 minutes later, the purchase link is available while the seller is still demonstrating the item and the audience is at peak attention. In live commerce, timing is everything. The impulse to buy happens during the demo, not after it. Any system that keeps the commercial window open through that moment, rather than closing it while the seller does admin, is going to improve conversion. That's the core of the mechanism, and it's not magical.
The listing quality improvement is a secondary effect. Better titles, more accurate categories, more complete descriptions make listings more discoverable to viewers who browse after a stream ends, and improve the post-session catalogue that buyers return to. Neither of these is a trick. They're operational improvements with a direct and predictable conversion impact.
What makes this more interesting than a standard AI feature announcement is precisely that mundanity. This is not a generative AI experience for the buyer. It's friction removal for the seller. And friction removal that's this targeted tends to scale across seller types and session contexts in ways that a clever consumer-facing feature often doesn't.
Tilt sellers Kameleon, who sell vintage fashion and trading cards, described the practical effect: "Snap has basically cut our workload in half. One of us can run the whole stream now while the other runs our second account," said Tom. "The listings come out looking proper now, better titles, right categories," added Anna.
The UK Live Commerce Context
Live commerce in the UK has been growing but remains behind both China and the US TikTok Shop-led market. The format has established genuine traction in specific categories: fashion, beauty, collectibles, trading cards. These are the contexts where demonstration and personality make a real difference to conversion, and where independent sellers with personality and stock are the natural supply.
TikTok Shop is now the UK's fourth-largest beauty retailer and hosts thousands of live shopping events daily. Tilt operates in a related but distinct space: a live auction platform for fashion, collectibles and more, founded in 2021 by Abhi Thanendran and Neil Shah, both early Revolut employees, London-headquartered, active across the UK, Italy, Spain and Poland, with $27M raised from Balderton Capital, TQ Ventures, Earlybird and Seedcamp.
The sellers already building on platforms like Tilt, Depop, and Vinted have the content muscle that live commerce requires. What they have lacked is infrastructure to make the workflow manageable solo. The gap has been operational, not creative.
The more interesting question is whether tools like Snap can extend participation to the mid-tier of would-be live sellers: people who understand the format and have suitable products, but find the current workflow too demanding to sustain. If the sell-through improvement is reproducible across seller types and session contexts, that's a meaningful capability. Live commerce is the format that most closely replicates the discovery and persuasion dynamics of physical retail. Anything that makes it operationally viable for more sellers strengthens the format overall.
The harder question for the UK market is whether live commerce can reach beyond its current audience, which skews younger, fashion-forward and already fluent in TikTok, into the broader British shopping population that still does most of its online spending on Amazon, on supermarket sites, and via search. That audience is not going to watch a grocery livestream. But it might buy a piece of jewellery, or a lamp, from someone it has just spent four minutes watching. That is the market worth building infrastructure for.
Data Sources
Stay Connected
Follow LLCommerce on LinkedIn
Get the latest AI commerce insights, analysis, and industry news delivered to your feed.
Large Language CommerceAbout the Author
Senior Editor
Sarah covers the intersection of AI and retail, with over a decade of experience in technology journalism. Based in Bangkok, Thailand — and will explain at length why that's actually the best place to cover e-commerce if you'll let her.