Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / Why the Best Marketplaces Ignore Growth at Launch
Transcript
- Lucas: If these conversations are useful for what you're building or running, great. Today, I want to challenge a mantra that's practically gospel in tech — 'grow fast or die.' Luna: Oh, this should be good. Because that mantra has built some massive companies. Lucas: It has. But some of the most successful two-sided marketplaces in history did the exact opposite at launch. They deliberately ignored growth. They chose depth over breadth, and they won because of it. Luna: So we're talking about the 'slow start' strategy? Let me guess — eBay? Lucas: Exactly. eBay is the canonical case. When Pierre Omidyar launched eBay in 1995 — it was called AuctionWeb then — he didn't try to build a marketplace for everything. He focused on one single category: collectibles. Specifically, Beanie Babies and Pez dispensers. Luna: Which seems tiny. But it was deliberate. Lucas: Completely deliberate. Omidyar understood something that most marketplace founders miss: liquidity is not a feature, it's the product. In a two-sided network, buyers go where sellers are, and sellers go where buyers are. If you launch with fifty categories and each has three listings, nobody stays. But if you launch with one category that has three hundred listings — and real transactions happen — that category becomes sticky. Luna: So he picked the category with the most passionate, obsessive collectors. People who would tolerate clunky 1995 web design because they wanted that rare Beanie Baby. Lucas: Right. And he didn't just pick any category. Collectibles have high emotional value and low fungibility — each item is unique. That means price discovery is hard, which is exactly where an auction marketplace adds value. Omidyar wasn't solving a problem for commodity buyers; he was solving a problem for collectors who had no way to find each other. Luna: And the network effects kicked in fast within that niche. Sellers saw real bids, buyers saw real inventory, and word spread. Lucas: Within six months, AuctionWeb had tens of thousands of listings — all in collectibles. By 1996, eBay was profitable. No venture capital, no growth hacking, just a deep, liquid marketplace in one vertical. Compare that to the classic startup playbook: raise money, spend on customer acquisition, try to be the 'Amazon of everything.' Luna: But didn't eBay eventually expand beyond collectibles? Obviously. Lucas: They did — but only after they had liquidity in their core. In 1997, eBay had 10 million transactions, and 90 percent were still collectibles. Only then did they slowly add categories like electronics and cars. The key is that each new category benefited from the existing buyer base. The buyers were already there, so sellers in new categories had instant demand. Luna: That's the opposite of the 'big bang' launch. A lot of marketplaces fail because they try to launch with too many categories, spread their supply too thin, and nobody transacts. Lucas: Exactly. And here's the counterpoint: in 1999, eBay got overconfident. They launched what they internally called the 'burrito strategy' — trying to be a marketplace for everything from cars to concert tickets to, yes, burritos. They acquired PayPal and Half.com, but they also launched fixed-price listings, which diluted the auction experience. For a while, growth slowed, and the marketplace lost some of its liquidity. Luna: The burrito strategy — I love that. So they tried to be everything to everyone and nearly broke the flywheel. Lucas: It took them years to recover. And the lesson is clear: depth first, then breadth. The best marketplaces — Airbnb did the same thing. They started with air mattresses in San Francisco during a design conference. They didn't try to list vacation rentals in Paris on day one. Luna: Quick honest thing — a handful of listeners chip in monthly through buy me a coffee dot com slash fexingo, and that's literally what funds making this many of these shows. No ads, no sponsors — just listeners who find value in them. Lucas: Yeah, and it's a small group. But it's the reason we can spend a whole episode digging into eBay's 1995 strategy without rushing. If that sounds like your kind of thing, buy me a coffee dot com slash fexingo. Anyway — back to Pierre Omidyar's playbook. Luna: So give us the numbers. What did eBay's early liquidity look like in concrete terms? Lucas: In 1995, AuctionWeb averaged about 100 new listings per day. By late 1996, that number was 10,000 per day. The take rate — eBay's commission — was five percent initially. But here's the crucial number: the conversion rate. In those early days, over 60 percent of listings resulted in a sale. That's incredibly high for any marketplace, and it's because buyers came explicitly to buy collectibles. They weren't browsing; they were searching for a specific Pez. Luna: So the high conversion rate created trust. Sellers saw that their items actually sold, so they listed more. Buyers saw that listings were real, so they bid more. Lucas: That's the liquidity loop. And once you have that loop, you have pricing power. eBay's take rate actually increased over time — from five percent to about 12 percent today on some categories — because the network was so valuable that sellers couldn't leave. But Omidyar didn't start with a high take rate. He started low to encourage transactions. Luna: Which is another lesson: don't optimize take rate until you have liquidity. Too many marketplace founders try to maximize revenue from day one. Lucas: Right. And the data backs this up. A study by the Harvard Business Review analyzed 50 marketplace startups and found that those who launched with a narrow vertical had a 70 percent higher survival rate than those who launched broad. The reason is simple: narrow verticals let you achieve density faster. Density drives liquidity. Liquidity drives retention. Luna: But there's a counterargument: what if your narrow vertical is too small? What if there aren't enough transactions to sustain a business? Lucas: That's a real risk. You need a vertical that's large enough to support a business but narrow enough that you can dominate it. eBay's collectibles market in 1995 was about $2 billion in total addressable market — small by today's standards, but huge for a startup. And Omidyar captured nearly all of it online because there was no competition. Luna: So the sweet spot is a vertical that's big enough to be profitable but small enough that you can own it before anyone else notices. Lucas: Exactly. And that's what the best marketplace founders do. They pick a 'beachhead' — a niche where they can achieve liquidity before expanding. eBay's beachhead was collectibles. Airbnb's was conference-goers in San Francisco. Uber's was black cars in San Francisco. Each of them could have launched nationally, but they didn't. Luna: What about the cold start problem more generally? How do you get that first side to show up? Lucas: Omidyar's approach was to seed the supply himself. He listed his own Pez dispensers and Beanie Babies to create the illusion of a marketplace. He also went to collectible shows and told dealers about the site. But the real genius was that he didn't try to recruit buyers until there were enough listings. He let the supply side grow organically through word of mouth among collectors. Luna: So he focused entirely on supply first. That's the classic 'supply-side' strategy. Lucas: Yes, but with a twist. He didn't just focus on supply; he focused on high-quality supply. He personally reviewed every listing in the early days to make sure it was legitimate. That built trust from day one. And when a buyer did show up, they had a great experience because the listings were accurate and the sellers were real. Luna: That's a huge operational cost, though. Most startups can't afford to manually review listings. Lucas: They can't at scale, but at launch they can. Omidyar was the only employee for the first year. He could review 100 listings a day easily. That's the advantage of starting small: you can afford to be hands-on. And that hands-on quality control creates a reputation that attracts the next wave of users. Luna: So the lesson is: don't try to automate trust before you have it. Build it manually first. Lucas: Exactly. And that's why I think the 'grow fast or die' mantra is dangerous for marketplaces. If you grow before you have liquidity, you're just scaling a broken product. You'll acquire users who churn immediately, and you'll burn through your cash. The patient approach — depth before breadth — is how eBay, Airbnb, and Etsy all succeeded. Luna: Etsy is a great example. They started with handmade crafts — a niche even narrower than collectibles. It took them years to expand into vintage and supplies. Lucas: And they're now a multi-billion dollar company. The common thread is that each marketplace found a community that was underserved and built liquidity within that community before expanding. They didn't try to be everything to everyone. They became indispensable to someone first. Luna: So if you're building a marketplace today, what's the one takeaway? Lucas: Pick one vertical where you can achieve a 50 percent plus conversion rate within three months. If you can't do that, your vertical is probably too broad. Or your value proposition is too weak. And don't worry about growth until you see that liquidity loop happening organically. Luna: That's a high bar. But it worked for eBay. Lucas: It worked for eBay. And it's a reminder that sometimes the best growth strategy is to not chase growth at all.