Ecommerce is about to get rebuilt by video, and most of the people who built the last version of it have no idea how much cheaper the next one already is. I ran a forty-person commerce operation at Woo, ITV's Gen Z media brand. At the same time, on the other side of London, Dominic Jacob and Mark Bage were building Sneak In Peace, solving the exact problem that was bleeding Woo dry. I didn't know either of them then. If I had known Dominic Jacob while Woo was still running, Woo would still exist. I joined Sneak In Peace later, once I'd seen what the infrastructure could actually do. This is the story of what curated commerce cost to build by hand, and what changes once shoppable video becomes the entry point to something far bigger sitting underneath it.
The Payroll Behind the Seamless
I pitched Woo into a very specific moment. It was post-Covid, developers were impossible to hire because every startup in London wanted the same small pool of engineers, and people had only just started speaking to each other in person again after two years behind a screen. The pitch spoke to that exact mental space, a generation relearning how to connect while also relearning how to buy. It was a cluster fuck of challenge and opportunity in equal measure, and we built into it anyway.
None of that touches the part that never makes it into a case study deck. To get 150 brands and 4,500 products live inside one marketplace, we had to build integrations with every retailer's own product feed, one at a time, because none of them spoke the same language and none of them were built to talk to each other. We built a finance stack from scratch to handle payouts, commission splits and reconciliation across 150 separate commercial agreements, each with its own rate, its own terms, its own approval chain, its own person who needed chasing on a Friday afternoon. And we built an account management function whose entire job was onboarding: getting brands live and keeping their stock feeds current. None of it was glamorous. All of it was the actual cost of curation, the plumbing nobody sees behind every "seamless" shopping experience a brand has ever put in a press release.
None of it was cheap either, and none of it was fast. The integrations alone took the better part of a year of engineering time before a single product page felt reliable. And even once it was built, we still couldn't answer the question that actually mattered: who was buying what, and what made them decide. Purchase data lived in 150 different systems that had no interest in talking to each other. Tracking a customer's actual journey meant pulling five dashboards into a spreadsheet by hand and guessing at the gaps in between. We had a marketplace. We didn't have a picture of the person shopping in it.
Forty people. To do what a content driven marketplace could now do with two.
Why It Didn't Work
We built a brand with real hype and real engagement, and the work itself held up: original enough to win awards for content, for brand and for social strategy from the most prestigious bodies in the industry, and covered by a press that loved us.
Woo ultimately didn't work commercially, and there were plenty of reasons why. Breaking even in 3 years was always going to be a tall task. But the biggest one was the cost of the people it took to keep the thing running while we were still figuring out conversion, still figuring out engagement, still fixing the tech infrastructure underneath all of it in real time. That's an expensive way to learn. Every month we spent tuning a broken feed or rebuilding a payment flow was a month of payroll that should have gone toward the content and the culture, the actual reason anyone was watching in the first place. The business never got to run at the speed the idea deserved, because too much of its runway was spent paying people to hold the plumbing together.
There was also the structural problem underneath all of it. We were trying to run a fast-moving startup inside a large corporation, and that was always going to be difficult. Decisions that would take a founder an afternoon took ITV weeks of sign-off. Budget that a standalone startup could redirect overnight had to move through a process built for a broadcaster, not a marketplace. None of that shows up in a case study either. It's just the tax on trying to move fast inside a structure built to move carefully.
Going from zero to forty people in six months meant we never got the culture right, either. By the time a team found its rhythm, three new people had joined, a process had changed, or the brief had moved again. It would start to feel right for a week or two, and then something underneath it would shift, and we'd be rebuilding trust and habits from scratch.
Getting the marketplace live at all, one person to forty and four thousand five hundred products across a hundred and fifty brands in six months, was genuinely fast by media standards. It still wasn't fast enough. We were building retailer integrations one at a time while the market we were trying to serve was already moving month to month. Speed built by hand has a ceiling. We hit it.
And we had to hit it while proving the model was already working. ITV wanted growth numbers every quarter. We were reporting real ones, a billion impressions against a target of 385 million, while the marketplace underneath those numbers was still being assembled in real time. Success had to be visible before the foundations were finished, which meant every fix, every integration, every finance reconciliation happened under a spotlight instead of in private.
The Storefront We Built By Hand
That's not a humblebrag about how hard we worked. It's an indictment of the tooling we were working with. Everything at Woo that touched commerce was manual. A product went into a feed, a link went into a caption, and a viewer had to leave the video, open a new tab, and search for the item themselves, hoping it was still in stock. Every one of those steps was a chance for someone to give up. The real issue was we couldn't put product inside the video at all. We had automation pulling stock from our suppliers, but nothing that let a viewer act on a product while they were still watching, so everything we built sat next to the content instead of inside it. No platform existed that let video be the storefront. We built everything adjacent to it instead, out of people.
Video As the Storefront
This is exactly why I joined Dominic and Mark rather than starting something new. They had already solved the problem I was living through. Sneak TV takes shoppable video and does the thing Woo never had the infrastructure to do: it puts the product inside the frame itself. Watch a runway film, tap the trainer, check out, keep watching. No new tab, no search bar, no editor manually tagging four thousand SKUs at midnight because the CMS keeps timing out. Content and commerce were built as one system from day one, because splitting them into two separate teams is exactly what buried Woo's margins.
The Case for Never Being In a Hurry
This matters more than it sounds like it should, because everyone in this industry has spent the last decade obsessed with the wrong metric. We chase engagement, virality, reach, the noisy stuff that looks good in a case study deck. What actually converts is the opposite of noisy. It's slow. Sneak TV's own framing gets at this without quite saying it: "never in a hurry." That's not a brand tagline, it's an operating principle, and it's correct. A person sitting with a runway film for four minutes, half-watching, half-absorbing the way a garment moves and sits and catches light, is doing something closer to window shopping on Bond Street than scrolling TikTok. It's slow TV. Ambient, low-stakes, passive. And passive viewing converts better than active scrolling because nobody's defenses are up. You're not being sold to. You're just watching, and the product happens to be right there when you're ready.
The Hook Is Video. The Business Is Everything Underneath It.
Shoppable video is what gets someone to stop scrolling and watch. It's the format that makes a product feel worth four minutes of someone's attention instead of a skip. But video was never the actual business Dominic and Mark set out to build. The business is the affiliate ecosystem sitting underneath it, four hundred and fifty brands and thirty-four thousand products, all running through the same integration, payout and onboarding infrastructure that took Woo a year of engineering time to half-build for a fraction of that scale. Video earns the attention. Affiliates turn it into revenue once the attention has moved on to the next film.
The Infrastructure Nobody Sees
That affiliate ecosystem is available to anyone through Sneak TVx, our white-label version of the same integration, payout and onboarding infrastructure. A conversion rate that runs to 30% on luxury lines doesn't come from a well-shot video on its own. It comes from a brand or media owner being able to plug straight into retailer feeds, payout logic and onboarding that already exist, without building any of it themselves. That's the exact work I did from nothing at Woo, sitting underneath someone else's product now, ready for anyone to plug into. A brand doesn't need a finance team to reconcile commission splits or an account manager chasing stock feeds to make video sell. It needs a video and a feed, and the checkout happens where the attention already is.
Anyone With an Audience Can be a Retailer
Strip away the storefront and what's left is a data problem finally solved. Sneak TV captures emails and builds wishlists as a byproduct of watching, not as a separate ask bolted onto the end of a checkout flow, which means every viewer becomes a first-party record instead of a view count nobody can act on. At Woo, that record was the thing we never fully built, because we spent our engineering time on integrations and payouts instead. Here, the person watching is already inside the system before they've bought anything.
That's the part that should change how anyone with an audience thinks about commerce. A newsletter with forty thousand subscribers, a YouTube channel, a Substack people actually open: none of it needs a warehouse, a payment processor or a single line of code to become a fully elevated, multibrand retail proposition. The infrastructure that took Woo forty people and a year to half-build is now sitting underneath anyone willing to plug a feed into it.
What We Built Instead
This is the stack Dominic and Mark had already built by the time I joined Sneak In Peace: shoppable video as the default format, not an add-on; a product feed that lives inside the content instead of beside it; and a commerce layer that never asks a person to manually tag a product at midnight before a 7am call time. Woo's ambition was correct. It just needed infrastructure it never had, infrastructure that already existed a few miles away, built by people I hadn't met yet.
For any brand or media owner asking the same question I was asking three years ago, the answer isn't work harder, hire more, ship faster. It's find the infrastructure that already does the boring, expensive, headcount-heavy part, and put your people onto the ten things only humans can actually do well: the idea, the casting, the story, the eye. Everything else is a systems problem, and we built the system.
To explore the future of e-commerce check out Sneak In Peace.