Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / How Marketplaces Use Geographic Expansion to Solve Liquidity
Transcript
- Lucas: So you’re a marketplace founder. You’ve got liquidity in one city — let’s say you’ve cracked the chicken and egg problem in Austin. Buyers and sellers are transacting, take rates are healthy, unit economics look good. Now you want to go national. How do you replicate that magic in a new city where nobody knows your name? Luna: And the obvious answer is: you can’t just clone your playbook. Each new market is its own mini chicken and egg problem. Lucas: Exactly. Geographic expansion is one of the hardest moves a marketplace can make, and we see it all the time — companies that had great local liquidity flame out when they tried to scale. Today I want to look at two cases that handle it very differently: Uber Eats and Vinted. Luna: Uber Eats — the on-demand food delivery giant. And Vinted, the second-hand fashion marketplace that started in Lithuania. These are almost opposite models. Lucas: Right. Uber Eats is hyper-local — you order food from restaurants within a few miles. Vinted is a shipping-based marketplace where a seller in Paris can sell to a buyer in Berlin. The geographic expansion playbook for each is totally different, but they both solved a core liquidity problem. Luna: Let’s start with Uber Eats. When they launched a new city, what was the first move? Lucas: They used Uber’s existing rider network as a supply-side wedge. In a new city, Uber already had drivers. Those drivers, during slow meal periods, could be repurposed as delivery couriers. So the supply side — couriers — was seeded instantly. The hard part was demand: getting enough people to open the app and order. Luna: And they solved demand with subsidies. Heavy discounts, free delivery, that sort of thing. Lucas: Yeah, but the interesting part is how they timed it. Uber Eats didn’t launch 200 cities at once. They launched city by city, over about four years, focusing on density. They’d saturate one metro area — get restaurant density high, delivery times low — then expand to the next. The economics only worked if they achieved a certain orders per square mile threshold. Luna: So it’s a density play, not a breadth play. Makes sense for delivery. Lucas: Exactly. And that’s where many copycats failed. They saw Uber Eats’ model and said, let’s launch nationally day one. But without local density, delivery times stretched to 50 minutes, restaurants got unhappy, customers churned. Uber Eats’ geographic expansion was really about replicating a density unit over and over. Luna: Now Vinted — completely different. They don’t have a driver network. Their supply is just people with old clothes in their closets. Lucas: Right. Vinted started in Lithuania, a small market. If they only served Lithuania, they’d hit a supply ceiling fast. So they expanded across Europe by enabling cross-border transactions. A seller in Poland could list an item, and a buyer in France could buy it and pay for shipping. That unlocked liquidity because supply in one country could satisfy demand in another. Luna: That only works if shipping costs are low enough and trust is high enough. Lucas: Exactly. Vinted invested heavily in a shipping integration — they negotiated discounted rates with national postal services and created a simple label system. They also built an escrow-like payment hold: the buyer pays, Vinted holds the money until the buyer confirms receipt. That solved the trust problem across borders. Luna: So Vinted’s geographic expansion was really about enabling supply to travel to demand, rather than building demand in every country from scratch. Lucas: Exactly. And it worked. Today, about 40% of Vinted’s transactions are cross-border. That’s massive for a marketplace. Compare that to, say, eBay, where cross-border is maybe 10 to 15 percent. Vinted essentially turned geography into a liquidity multiplier. Luna: But there’s a risk — if you enable cross-border supply, local sellers might get crowded out by cheaper international listings. Did Vinted have that problem? Lucas: They did. In some countries, local sellers complained that buyers preferred buying from wealthier Western European countries where the clothes were perceived as higher quality. Vinted’s response was to keep local discovery alive — they still show local listings first, and only surface cross-border items if the buyer explicitly searches or if local supply is thin. Luna: So it’s a blended model: local-first, cross-border as a liquidity backstop. Lucas: Right. And that’s the key lesson — geographic expansion isn’t one-size-fits-all. For a local-delivery marketplace like Uber Eats, you need density. For a shipping marketplace like Vinted, you need cross-border infrastructure. But both rely on the same principle: liquidity in a new geography doesn’t happen magically. You need to seed one side deliberately. Luna: What about failures? Can you think of a marketplace that expanded geographically and just couldn’t get liquidity? Lucas: Sure — lots. One classic example is the original Webvan, which tried to launch grocery delivery nationally in the late 90s. They built massive warehouses in multiple cities simultaneously, but they didn’t have the demand density in any single city to make the economics work. They burned through a billion dollars in venture capital and went bankrupt. Luna: And that’s the cautionary tale — you can’t outspend the chicken and egg problem. You have to sequence it. Lucas: Right. Modern marketplaces like DoorDash learned from that. DoorDash actually used a very targeted approach — they launched in a few suburbs of a city first, tested density, then expanded outward. They called it the 'test and learn' strategy. It’s slower but more capital-efficient. Luna: So if I’m a founder today looking to expand my marketplace to a new region, what’s the first thing I should ask? Lucas: First question: is my marketplace local or shippable? If it’s local — delivery, services, ride-hailing — you need to build density in one city at a time. If it’s shippable — goods that can be mailed — you might be able to enable cross-border from day one, but you need the trust and logistics infrastructure to support it. Luna: Second question? Lucas: Second: which side is easier to seed? For Uber Eats, it was supply — drivers already existed. For Vinted, it was actually demand — they found that buyers in larger markets like France and Germany had pent-up demand for affordable second-hand clothes, and supply from Eastern Europe was abundant. So they seeded demand first, then supply followed. Luna: And third: what’s the unit of liquidity? Is it a neighborhood? A city? A country? That determines your expansion granularity. Lucas: Exactly. If your liquidity unit is a neighborhood — like for a dog-walking marketplace — expanding to a new city means expanding to dozens of neighborhoods. That’s a much slower process than if your liquidity unit is a country, like for a cross-border goods marketplace. Luna: And that’s why some marketplaces never go global. They’re optimized for local density and just stay local. Lucas: Right. And that’s fine. Not every marketplace needs to be global. The best ones understand their liquidity geometry and expand only when it makes sense. Luna: Quick thing — and this ties in. We keep talking about what makes these marketplaces work, and honestly, the reason we can dig into this stuff is because a small group of listeners chips in monthly via buy me a coffee dot com slash fexingo. No ads, no sponsors — just listener support. It’s what lets us make this many episodes, and we really appreciate it. Lucas: Yeah, it’s true. And that support lets us stay focused on the strategy, not on selling ads. So thank you to everyone who does that. Luna: Alright, back to geographic expansion. One more case I want to touch on: how did Airbnb handle this? Lucas: Airbnb is interesting because they’re a shippable marketplace — guests travel to the listing — but the liquidity unit is a city, sometimes a neighborhood. They used a 'city launch' playbook: they’d send a small team to a new city, photograph listings themselves to ensure quality, and run local events to build community. That human touch seeded both supply and trust. Luna: So they invested in local authenticity rather than just throwing money at ads. Lucas: Exactly. And it worked because travel is emotional. People want to feel like they’re staying in a real home, not a generic hotel. That local curation created a network effect that scaled city by city. Luna: So the takeaway: geographic expansion is about understanding your liquidity unit, seeding the easier side, and being patient. There’s no shortcut. Lucas: Right. And if you get it right, you turn a local liquidity moat into a regional or global one. That’s the real prize.