Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / How Marketplaces Build Liquidity With Geographic Expansion
Transcript
- Lucas: So, Luna, we've talked a lot about liquidity inside a single market. But what happens when a marketplace that's thriving in one city tries to replicate that success in another? That's the geographic expansion problem, and it's arguably the hardest move a two-sided network can make. Luna: Yeah, because you can't just copy-paste your supply and demand. Every city starts with zero. And that's a cold start, but on repeat. Lucas: Exactly. And that's where most marketplaces fail. They assume that if the model works in San Francisco, it'll work in Cleveland. But each new city has its own local dynamics, its own competitors, its own regulatory quirks. The playbook that got you to a million users in one market doesn't automatically translate. Luna: So what's the actual playbook? I mean, Uber went from San Francisco to basically every city on Earth. How did they do it without bleeding cash on subsidies forever? Lucas: Well, let's start with the concept of a liquidity threshold. Every marketplace needs a minimum number of active supply and demand participants before the experience feels seamless. For ride-hailing, that might mean a driver can pick up a fare within five minutes. For food delivery, it's restaurants that can fulfill orders in under thirty minutes. Below that threshold, the platform feels broken, and users churn. Luna: And crossing that threshold in a new city is the whole game. So how do you get there without burning through all your capital? Lucas: Right, and that's where the Uber playbook comes in. They didn't just launch in a city and hope for the best. They had a launch team that went in with a very specific playbook. They'd identify a few dense neighborhoods, recruit drivers aggressively with guaranteed earnings, and then run referral campaigns to get riders. The goal was to hit that liquidity threshold in a small, contained area first, not the whole city at once. Luna: So it's about density over sprawl. You want a neighborhood where a driver can get you a ride in three minutes, not a whole metro where you're waiting ten. Lucas: Exactly. And once you hit that threshold in one pocket, you expand to the next pocket. That's how they built that spiderweb of coverage. It's a very deliberate, surgical approach. And it's why you see those initial surge pricing and driver guarantees — they're not just marketing, they're tools to engineer that initial density. Luna: And that's why we can keep doing these deep dives without any ads. We want these conversations to be useful for operators and builders, not interrupted by something that breaks the flow. If you find that valuable and want to support keeping it that way, the link is buy me a coffee dot com slash fexingo. No pressure at all. Lucas: Yeah, exactly. It's a simple way to keep the show ad-free if that's something you care about. But back to the playbook — the other key piece is supply subsidies. When you're launching in a new city, you're asking drivers to sit idle for a while. So you guarantee them a certain hourly rate. That's the cost of seeding liquidity. Luna: But doesn't that just attract drivers who are only in it for the guarantee, and then they leave the moment you cut it back? Lucas: That's the risk. And that's why a smart marketplace phases out subsidies gradually, tying them to performance. The drivers who stay are the ones who realize they can actually earn consistently once the network is dense enough. But if you cut the subsidies too early, you kill the network before it reaches critical mass. It's a very fine line. Luna: What about Airbnb? They didn't have the same driver problem, but they still had to crack the chicken and egg in each city. Lucas: Right, Airbnb's approach was different because the supply is fixed — it's apartments and homes. They had to convince hosts to list before there was any demand. So they used professional photography as a subsidy. They'd offer free photo shoots to hosts in new cities, which made the listings look better, which attracted guests. That was their version of a driver guarantee. Luna: That's a smart move. It's not a cash subsidy, it's a quality subsidy. It improves the product for both sides simultaneously. Lucas: Exactly. And they also focused on event-driven demand. When the Democratic National Convention was in Denver, they saw a spike in searches. So they went to local hosts and said, 'Hey, you have an opportunity here. List your place.' That gave them a core of supply that then became the foundation for ongoing demand. Luna: So it's about finding those anchor events, those moments where demand is naturally high, to bootstrap the network. That's a really different model from Uber's, where you're seeding supply and then driving demand with referral codes. Lucas: Right, and that's the key takeaway. The method depends on the type of marketplace you are. For a transaction-based marketplace like Uber, you need real-time matching, so you need supply and demand to hit a threshold simultaneously. For a listing-based marketplace like Airbnb, you can build supply ahead of demand because the inventory is durable. That's a big strategic difference. Luna: And what about the data moat angle? Once you've expanded to a bunch of cities, you have all this data on demand patterns, pricing elasticity, driver behavior. Does that make it harder for a new entrant to compete in each local market? Lucas: It does, but it's not insurmountable. The data helps you optimize pricing, predict demand, and allocate supply more efficiently. So in a new city, you can start with a much more refined playbook than your first city. But a local competitor can still win if they have better local knowledge or a stronger community connection. The data moat is real, but it's not a wall. Luna: So when you're thinking about expanding, is it better to go into a new city with a lean team and a small budget, or to go big with a huge marketing push? Lucas: Generally, the lean approach wins. You want to test the market's viability with minimal resources. If you can't hit the liquidity threshold in a small area with a modest budget, you're not going to do it with a million dollars in advertising. That's what a lot of startups get wrong — they throw money at a problem that's fundamentally about operational execution, not just brand awareness. Luna: So it's about finding that pocket of demand, proving the model, and then scaling outward. Almost like a series of mini-startups, each with its own cold start. Lucas: Exactly. And that's why the best marketplace operators think of themselves as running a portfolio of local businesses, not one global platform. Each city has its own lifecycle, its own competitive set, its own regulatory environment. You have to be willing to adapt your playbook city by city. Luna: Let's talk about a real example that people might not know. I was reading about how DoorDash entered new suburbs. They didn't just go after the big cities like Uber Eats. They focused on areas where there was high demand but low supply of delivery options. And they used a different approach to restaurant acquisition. Lucas: Right, DoorDash's strategy was to sign up local favorites, not just national chains. They realized that in the suburbs, people are loyal to their local pizza place or their local Thai spot. So they'd go to those independent restaurants and offer a lower commission rate to get them on board, and then they'd market to hungry residents who previously had no delivery option. That created a very sticky network. Luna: And that's a great example of how understanding local demand shapes your supply strategy. It's not just about flooding the market with listings; it's about knowing which listings matter to that specific community. Lucas: And that's where the data moat actually kicks in. Once DoorDash had data on ordering habits in those suburbs, they could predict demand for new restaurants before they even signed them. They'd go to a restaurant and say, 'Look, we can see that people in this area are searching for Thai food, and there's no Thai place on our platform. If you join, you'll be the only one.' That's a very compelling pitch. Luna: So the data from one market gives you an unfair advantage in the next. That's the flywheel at work. Lucas: Exactly. And that's why we're seeing this consolidation in the marketplace world. The players that have expanded successfully have these proprietary data sets that make it really hard for a newcomer to compete on equal footing. But it's not just about having data, it's about how you use it locally. Luna: So if you're an operator listening to this, what's the one thing you should take away for your own expansion? Lucas: I'd say, focus on finding one neighborhood, one campus, one small geographic area where you can hit that liquidity threshold, and do whatever it takes to make that work. Don't spread your subsidies too thin. Once you've cracked that code, you can replicate it. But if you can't make it work in a small pocket, you're not going to make it work in a bigger one. Luna: That's a good practical framework. And it's encouraging to know that you don't need to be a giant to start — you just need to be deliberate about where you start. Lucas: And that's the beauty of two-sided networks. They can be built block by block, city by city. The same principles that apply to a global player like Uber apply to a local marketplace starting in one neighborhood. It's all about finding that first spark of density and then nurturing it.