We’re building an AI-native GTM Accelerator, and I’d already sold a bunch of solutions inside that same constellation, so the offer felt validated by my own sales numbers. I’d started building the marketing. And then I felt the hesitation in my chest.
I owed Jonathan a segment of work that day. Instead, I sat there and thought: even though I’ve already sold this, I need to talk to 20 founders and run real voice-of-customer interviews before I build another thing.
Then I got on a call with James Buck, co-founder of Roome, for this Growth Pilot feature. He’s spent six years mastering the two-sided marketplace cold start in student housing, and he handed me the exact playbook I want to walk you through.
The insider who cracked cold start from the inside
Every two-sided marketplace faces the same brutal cold start problem: you need supply to attract demand, and demand to attract supply, and neither wants to show up first. Roome had to bring students and university housing together at the same time. James solved it with an advantage most founders never think to use.
He was a student at the University of Surrey when he landed Roome’s first partnership. That insider position was more powerful than any cold outreach motion could ever be. He already had the trust factor. He already understood the internal dynamics. He was a member of societies, people knew who he was, and he had word-of-mouth distribution baked in before he ever asked for anything.
And yes, that advantage compounded into results most founders would envy. Surrey became Roome’s top-performing campus, with over 15% of the entire student base signed up to the app. For one campus, that’s around three and a half thousand students a year. And when James and his team got in front of the right university decision-maker, they closed at a rate of roughly 66% from presentation to contract. That’s an elite number in any B2B motion, let alone one selling into institutions.
The £10.5M win that shows how James reads a market
The clearest proof of James’s judgment came during clearing. In the UK, clearing is the scramble that opens when a student misses their grades or wants to switch universities last minute, with a one-month window between results and move-in. James lived that scramble himself once, which is exactly why he understood the stakes when a university came to Roome with more clearing students than they had beds for.
Roome matched and housed over 350 strangers in a two-week window. That generated a net saving for the university of over 10.5 million pounds. When the team called students afterward, 53% of the people they’d signed up said they would have chosen a different university if Roome hadn’t helped them.
Read that number again. More than half would have walked.
This is what a founder looks like when they read a market correctly. James saw a retention crisis the university itself had struggled to name, matched students into like-minded groups, and turned a one-month emergency into a permanent proof point.
The result was auditable, the mechanism was clear, and the value was measured in eight figures.
Is Your Startup Falling Behind Its AI-Native Competitors?
His cold-start sequencing, as a teachable playbook
James has a clear, hard-won point of view on how to sequence a two-sided marketplace, and it’s worth stealing.
Roome is a B2B2C marketplace, where universities bring distribution to students and students validate the platform back to universities.
The classic trap is the chicken-and-egg standoff: which side do you build first?
James went students-first, almost by accident of being one, and then discovered the university side held the real revenue. His honest take now is that you can run a marketplace cold start from either side, as long as you commit to one and get really good at it before you chase the other.
His deeper insight is about distribution. He puts it in his top three things that matter in any business, and describes it as the thing that puts the world at your feet once you crack it.
He had distribution at Surrey because he was inside the ecosystem. The frontier he’s working now is manufacturing that same insider trust at a cold university where he starts as a stranger, through testimonials, hard numbers, and a mix of bottom-up champions and top-down outreach.
The builder underneath the operator
James doesn’t just have market instincts. He builds the systems that make the instincts pay off.
Take his campus ambassador program. Ambassadors earn one pound per signup, everything is attributed through their own tracked links, and it turns into a real competition. Top performers drive 50 to 100 signups a week.
When James needed to measure all of it, he built his own attribution tracking in-house using synthetic data rather than bolting on someone else’s tool. He also brought on a strong chief product officer who introduced sprint planning, QA, and a real delivery lifecycle, so the growth engine has an operating spine beneath it.
He thinks past the vanity metric too. Signups are the top of the funnel. The real game, in his words, is activation, retention, and revenue, and he already maps more than twenty future monetizable actions he can switch on when the scale is right.
The one lesson that clarity bought him
Here’s the belief that ties James’s whole approach together, and it’s the one that stopped me cold on that Monday.
Early in Roome’s life, before the university model, the team spent a quarter of a million pounds building a piece of proptech that no longer exists. They ran it free, asked customers whether they’d pay once it became real technology, and got glowing answers. Then revenue time arrived and those same customers went quiet. James names the failure mode with a precision that took me aback. They asked questions designed to return the answer they hoped for, and learned how convincing a false positive can be. As he puts it, people build awesome things all the time, and a good idea doesn’t make a good business.
That’s the SpaceX point, and James draws it beautifully. NASA would spend three years iterating on a hypothetical engine. SpaceX understood that you can’t answer a hypothetical question on a whiteboard. You get the answer by sending a rocket up and trying to land it on a pad. The founder’s job is to sort which assumptions can be reasoned through and which ones require you to launch the rocket.
“Build your own roadblocks, and decide before you ask a question which answer you’ll ignore.”
Most founders take a decade to articulate that. James had it down cold, and he turned the bill into a discipline he now lives by: build your own roadblocks, and decide before you ask a question which answer you’ll ignore.
The validation discipline I run for clients
Here’s where James’s lesson lands for me, and why I stopped building that Monday.
When I diagnose a stalled GTM motion for a founder, I treat it like a system and look for root cause before I touch tactics. The root cause is almost always upstream of the marketing: a fuzzy ICP, a problem nobody is urgently paying to solve, or an offer that was ego-validated instead of customer-validated. You can’t out-market a thing people will tell you they love and then decline to buy.
So I run voice-of-customer interviews, and I’ve stripped them down to something a founder can actually execute. For a 30-minute VOC interview, you only need these five core questions.
- What stage are you at?
- What is the biggest growth bottleneck?
- What have you already tried, and what disappointed you?
- If you disappeared for 30 days, what breaks first?
- If this were solved in 90 days, what would change, and what would make you buy?
Everything else is optional. Question three surfaces what people have already rejected, and question four finds the load-bearing pain. Question five separates a nice-to-have from a budget line.
Here’s what you should notice as missing. I never describe my offer and ask if they’d want it. That’s the trap James paid a quarter million pounds to learn.
Steal this: the founder’s pre-build checklist
Whether you’re launching a two-sided marketplace cold start or any other offer, run this before you build.
Find the people who have budget now and urgency now. Those two together define your real ICP, and you might talk to 100 people before you find them.
Ask the boring questions about their problem, the ones that feel like a waste of time. Conversion is the only metric that proves demand, so look for paid agreements from more than one person.
Decide in advance which answers you’ll ignore. Write them down before the call.
Sort your assumptions into whiteboard questions and rocket questions. Reason through the first kind, and go test the second kind in the real world.
James built Roome into a six-year company that beats the odds because he pairs sharp market instincts with the discipline to test them. He turned an insider advantage into elite penetration, a crisis into an eight-figure win, and an expensive early lesson into a framework he can hand to the next founder. That’s the operator you want to learn from.
James is actively working to bring Roome into new cities and new universities, and this problem reaches well beyond the UK. If you’ve experienced the student-housing problem yourself or through your children, or if you can introduce Roome to a university or student housing team anywhere in the world, he’d love to connect. Find James Buck on LinkedIn and tell him Lillian Pierson sent you.
P.S. I never did send Jonathan that segment on the Monday I’m telling you about. I sent him something better a few days later: a plan to interview 20 founders before we build another line of the Accelerator.
If you’re a founder staring at an offer you’re about to build, and a small part of you suspects you’re validating your own ego instead of the market, that’s the signal. If it would help to pressure-test your growth assumptions with someone who runs this discipline for a living, grab a 30-minute conversation with me here. Come with your riskiest assumption, and we’ll find out together which one is a rocket.
