Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / How the Best Marketplaces Solve the Chicken-and-Egg Problem
Transcript
- Lucas: So every marketplace founder wakes up to the same nightmare: you need sellers to attract buyers, but you need buyers to attract sellers. That's the chicken and egg problem, and it kills more marketplaces than any competitor or regulation ever will. Luna: Right, and there's a whole graveyard of platforms that never got past that first cycle. But a few did — Uber, OpenTable, Etsy — and they didn't just get lucky. Lucas: Exactly. And the key insight, which Andrew Chen wrote about in his book 'The Cold Start Problem', is what he calls the 'atomic network'. The idea is you don't need a whole city. You need the smallest possible unit of supply and demand that's actually useful. Luna: So instead of trying to launch across San Francisco, you launch one neighbourhood? Or one type of restaurant? Lucas: Precisely. Uber famously started with just a handful of black cars in a small area of San Francisco. They didn't even cover the whole city at first. They made sure that in that one pocket, the wait time was under five minutes. Once that pocket worked, they expanded it like a fractal. Luna: And that's the opposite of what most failed marketplaces do — they blast a press release and hope both sides show up. Lucas: Right. The mistake is thinking you need critical mass from day one. What you actually need is what Simon Rothman, the former eBay Motors executive, calls 'minimum viable liquidity'. That's the smallest number of transactions that proves the model works for both sides. Luna: And that number is often shockingly small. I remember reading that for a local services marketplace, it might be ten transactions a week in one zip code. Lucas: Exactly. But getting those ten transactions requires a deliberate strategy. You have to subsidise one side — almost always the harder side to acquire. For Uber, that was drivers. They guaranteed a minimum hourly rate so drivers would sit and wait, even when there were no passengers yet. Luna: But subsidies can be a trap. Homejoy, the home-cleaning marketplace, subsidised heavily on both sides — deep discounts for customers and guaranteed pay for cleaners. Once they pulled back the subsidies, both sides evaporated. Lucas: That's the cautionary tale. Homejoy raised a lot of money, grew fast, but never built genuine loyalty. The cleaners would go to whoever paid them more, and customers would go to whoever was cheaper. It was a race to the bottom. Luna: So how do you subsidise without creating a dependency? OpenTable did it well, I think. Lucas: Great example. OpenTable gave restaurants free tablets and software for the first year. That was a big upfront cost, but it solved the supply side problem. Once restaurants were on the system, diners found it valuable because they could see real-time availability. The subsidy was time-limited and it built a habit. Luna: And once the restaurants saw bookings coming through the platform, they were happy to pay the monthly subscription fee. The subsidy turned into a business. Lucas: Exactly. The trick is designing the subsidy so that the subsidised side experiences the value of the network before the subsidy runs out. For Uber drivers, that meant enough rides within the first two weeks that they didn't need the guarantee anymore. Luna: But what about marketplaces where the supply side is professional or has high switching costs? Like lawyers or doctors? Lucas: That's harder. Professional services marketplaces often fail because the suppliers don't need the platform — they already have clients. So you have to offer something beyond leads, like scheduling software or billing tools. That's what Thumbtack did initially — they gave pros a free profile and lead generation, but the real value was the back-end tools. Luna: So the chicken and egg problem is really about finding the right bait for the right side, and then making the hook invisible. Lucas: Absolutely. And one more tactic that works: start with a manual, concierge model. That's what Airbnb did in New York — they actually went door to door, took professional photos of listings, and uploaded them themselves. They faked the supply side until the network was real. Luna: And that's totally fine to do. It's not cheating — it's proving the concept manually before you automate it. Lucas: Right. The goal is to get those first ten transactions to happen, even if you have to hand-hold every step. Once you have proof that the marketplace works, you can build tools to scale. But without those first transactions, you're just theorising. Luna: So for someone listening who's thinking about starting a marketplace, what's the one concrete takeaway? Lucas: Design your first ten transactions before you write a line of code. Literally, write down who the first ten suppliers will be, who the first ten customers will be, and how you'll get them to transact. If you can't do that, you don't have a marketplace yet — you have a hope. Luna: That's really good advice. And it ties back to something we've talked about before — marketplaces are hard because they're not just products, they're ecosystems. Lucas: Yeah, and that's why I love talking about them. There's no one-size-fits-all solution. The chicken and egg problem forces every founder to be creative, to really understand their users, and to be willing to do things that don't scale at first. Luna: If you're building or running a business, these conversations are meant to be useful. And if they are, listener support is what keeps Fexingo ad-free. You can find us at buy me a coffee dot com slash fexingo. Lucas: Yeah, that kind of support makes a real difference. So, back to the topic — I think there's one more tactic worth mentioning: the 'come for the tool, stay for the network' strategy. Luna: Oh, that's a good one. Like Instagram started as a photo filter app, not a social network. People came for the filters, then stayed for the community. Lucas: Exactly. Marketplaces can do the same thing. Offer a standalone tool that's valuable even without a network. For example, a marketplace for wedding photographers could start with a portfolio builder tool. Photographers use it because it's useful, and then eventually they want to get booked through it. Now you have a supply side ready to go. Luna: So the tool solves the chicken and egg because it gives the supply side a reason to be there before any demand exists. Lucas: Right. And then when you bring demand, the supply is already there. That's a much easier cold start than trying to build both sides from scratch simultaneously. Luna: Alright, so to wrap up: find your atomic network, subsidise strategically but temporarily, consider a concierge model, and maybe start with a standalone tool. That's a solid playbook for the chicken and egg problem. Lucas: It is. And the common thread through all of them is: don't wait for the network to happen — make it happen manually first. Then scale what works.