You’ve probably heard the dream: build a marketplace, connect buyers and sellers, sit back and let the network effects take over. It sounds like passive income on steroids. But the reality is something few founders admit — and it’s a brutal, awkward, embarrassing grind. I’ve spent the last decade watching marketplace founders fail, and the one thing that separates the survivors from the quitters has nothing to do with code, UI, or features. It’s about whether you’re willing to do the work that doesn’t scale. The work that makes you feel like a glorified assistant instead of a founder.
Before you have liquidity, you are not a platform. You are a concierge. And that’s the only honest way to start. Every successful marketplace I’ve seen — from eBay to Airbnb to the weird niche ones nobody talks about — began with the founder manually connecting the first 10, 50, or 100 transactions. They didn’t write code to automate trust; they vouched for sellers themselves. They didn’t build a recommendation engine; they personally matched buyers with sellers on spreadsheets and phone calls. The ones who built features first? They built empty rooms. The silence of an empty marketplace feels like personal failure, but it’s actually the universal, unavoidable stage that separates those who fake it with polish from those who grind through it with manual intervention.
Let’s get specific. I remember a founder who wanted to build a marketplace for vintage camera gear. He spent six months building a beautiful search interface, a review system, and a payment flow. Then he launched. Zero buyers, zero sellers. He had a Ferrari with no engine. The problem wasn’t the product — it was that nobody trusted an empty room. He had to start over: he called every vintage camera store in his city, asked them to list 5 items manually, and then called 20 photographers he knew to buy them. He had to personally handle payment disputes, refunds, and shipping logistics. For three months, he was a concierge. Then, slowly, the transactions started happening without him. That’s the moment the marketplace actually became a marketplace.
Why do founders resist this? Because it’s uncomfortable. Building features is procrastination from the uncomfortable, unscalable work of forcing the first transactions to happen. You want to feel like a tech CEO, not a middleman. But the truth is that the hardest part of a marketplace isn’t the technology — it’s solving the chicken-and-egg problem between buyers and sellers. And the only way to break that cycle is to become the egg yourself. Your job is to manually stimulate both sides until the network effect kicks in. That means subsidizing one side (often the supply side), calling people personally, handling disputes, and even faking activity to build trust. Yes, faking it. Airbnb founders famously took photos of listings themselves to make the platform look professional. They were literally the photographers, the customer support, and the concierge.
So here’s my provocation: if you’re building a marketplace and you’re spending more than 20% of your time on code, you’re probably doing it wrong. The real work is in the messy, manual, people-driven part. The twist is that the more you control the platform in the early days, the more it feels like a closed, manipulative system — but that’s exactly what it takes to build the trust that makes it open and self-sustaining later. The magic of a marketplace is not in the software; it’s in the moment when both sides realize they need each other more than they need you. That moment doesn’t come from a feature. It comes from you getting your hands dirty.
I’ll leave you with this: the next time you obsess over your dashboard’s color scheme, ask yourself — are you building a platform, or are you building an empty room? The most successful marketplace founders I know always choose the concierge path. It’s slower, harder, and less glamorous. But it’s the only way that actually works.
FAQ
Q: Doesn't building a great product attract users automatically?
A: No, not in a two-sided marketplace. A great product with no participants is just a ghost town. Users don't join a platform; they join a place where other people already are. You need to manually create that initial crowd.
Q: How do I know when it's time to stop the manual concierge work?
A: When transactions start happening without your direct involvement — when sellers find buyers on their own, when disputes are rare, and when you can step back for a week without activity dying. That's the signal that liquidity has taken hold. Until then, keep your hands dirty.
Q: Isn't the concierge approach just a recipe for burnout?
A: It's temporary, not permanent. The goal is to build a system that eventually runs without you. But early on, burnout is a risk if you don't set boundaries. Focus on the manual actions that teach you what automation to build later. Every call you make is a data point for your future product.