Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / When Marketplaces Become Toll Roads the Liquidity Trap
Transcript
- Lucas: So we've talked a lot on this show about how marketplaces build liquidity — the chicken and egg problem, the subsidies, the referral flywheels. But there's a phase that comes after you've achieved liquidity that is arguably trickier than getting there in the first place. Luna: The part where you start making real money and have to decide how much to charge for it. Lucas: Exactly. Once a marketplace has critical mass, the temptation is to raise the take rate — the cut you take from each transaction. And in many cases, that's the right business decision. But there's a point where you become a toll road, and the toll gets so high that drivers start finding alternative routes. Luna: The liquidity trap. You squeeze too hard and the whole thing starts to unravel. Lucas: Right. And the canonical example of this — the one that gets taught in business schools now — is Etsy's 2019 fee hike. In April of 2019, Etsy raised its transaction fee from 3.5 percent to 5 percent. On the surface, that's a 150-basis-point increase. Not outrageous. Amazon charges roughly 15 percent for third-party sellers. But Etsy's sellers are different — they're artisans, small-scale makers, people who operate on thin margins. Luna: And they revolted. I remember sellers organizing, creating alternative storefronts, pushing buyers to their own websites. Lucas: Within six months of the fee hike, Etsy saw a 4 percent drop in active listings. That might not sound catastrophic, but when you're a marketplace, listings are your inventory. And when inventory shrinks, buyer engagement drops, which makes sellers less happy, and you get a negative spiral. Etsy's stock actually continued to rise in the short term because revenue per transaction went up. But the underlying health of the network deteriorated. Luna: So the market rewarded the short-term revenue signal while the long-term damage was invisible in the quarterly numbers. Lucas: Precisely. And that's the liquidity trap in a nutshell. You have this delicate equilibrium where both sides need each other. If you raise take rates too much, you push the more price-sensitive side — usually the supply side, the sellers — to leave or to disintermediate. And once supply shrinks, demand follows. Luna: So how do you know where the line is? Is there a metric that tells you you're about to over-optimize? Lucas: The best marketplace operators I've spoken to look at something they call 'take rate elasticity.' Basically, you run experiments — small, incremental fee increases in specific categories or geographies — and measure the impact on listing growth, conversion rates, and seller retention. If a 1 percent increase in take rate causes more than a 1 percent drop in transaction volume, you've hit the elasticity threshold. Luna: So you need to know your seller's unit economics better than they do. Because the moment your take rate eats into their margin to the point where their business model breaks, they're gone. Lucas: Exactly. And that's why companies like Airbnb have been so disciplined. Airbnb's take rate has hovered around 14 to 16 percent for years. They've had opportunities to raise it — demand is strong, supply is relatively inelastic in popular destinations. But they've chosen not to, because they understand that hosts have options. They can list on Vrbo, they can build their own direct booking site, they can manage through a property manager. Every time Airbnb raises fees, they make those alternatives more attractive. Luna: Let's talk about that. Airbnb did increase fees in 2022, right? They added a fee for guests and also increased the host fee in some cases. And there was pushback. Lucas: There was pushback, but the key difference is that Airbnb did it incrementally and tied it to specific value adds — better search, better insurance, better payment processing. Etsy's 2019 increase, by contrast, felt like a pure tax. They didn't offer sellers anything new. It was just 'we're taking more because we can.' Luna: And that's the difference between a platform that sees itself as a partner versus one that sees itself as a toll collector. Lucas: Exactly. The best marketplaces treat take rate as a strategic lever, not a revenue lever. They ask: what is the minimum take rate we need to sustain the platform and invest in features that make both sides more money? And then they optimize for transaction volume, not per-transaction revenue. Luna: But wouldn't that logic suggest that all marketplaces should keep take rates low forever? At some point, investors want to see margin expansion. Lucas: That's the tension. And it's not wrong to want higher margins. But the smartest marketplaces find ways to expand margins without raising take rates. They do it by reducing their own cost to serve — automating customer support, improving fraud detection, using AI to match buyers and sellers more efficiently. Uber, for example, has kept its take rate relatively stable but improved margins by cutting driver acquisition costs and optimizing routing. Luna: So the take rate itself is almost a symptom of the underlying business model efficiency. Lucas: Right. If you have a 20 percent take rate but your cost to serve is 18 percent, your net margin is 2 percent. That's not a great business. Meanwhile, a marketplace with a 10 percent take rate and a 4 percent cost to serve has a 6 percent net margin. Which one would you rather own? Luna: The one with the lower take rate and higher efficiency, obviously. Lucas: And yet, if you look at public marketplace companies, investors obsess over gross take rate — they want to see it going up. They don't always dig into cost to serve. That creates perverse incentives. Management teams raise fees to boost the headline number, and the market rewards them, even if they're destroying long-term value. Luna: That's a governance problem. Who pushes back? The board? The founders? Lucas: Ideally, both. But it's hard to push back when your stock is up 20 percent after a fee increase. The discipline has to come from a deep understanding of network dynamics. And frankly, most boards don't have that expertise. They see a comp set of other marketplaces charging higher fees and think 'we're leaving money on the table.' Luna: It reminds me of the classic Innovator's Dilemma — incumbents optimize the current business model until it breaks. Lucas: Exactly. And the liquidity trap is the marketplace version of that. You extract too much today and you hollow out the network. Then a competitor comes along with a lower take rate and a better value proposition and eats your lunch. Luna: Let's talk about a positive example. Which marketplace do you think has managed take rate the best over a long period? Lucas: I'd point to the ASX — the Australian Securities Exchange. It's a two-sided marketplace between buyers and sellers of stocks. Their take rate is essentially the clearing fee, and it's remarkably stable. They haven't dramatically raised fees in decades, because they understand that liquidity is their moat. If they raised fees, trading volume would move to alternative venues or offshore. So they keep fees low and make money on volume and data services. Luna: So they treat the exchange business almost like a utility. Lucas: Exactly. And that's the right framework for any marketplace where switching costs are low. If your users can easily transact elsewhere, you have to behave like a utility — low margins, high volume, constant efficiency improvements. If you have high switching costs — like a SaaS marketplace where users have years of data and workflows — you can charge more. But even then, you have to be careful. Luna: Because eventually, technology reduces switching costs. Look at what happened with Etsy — sellers built their own Shopify stores. Lucas: Exactly. And that's the thing. The liquidity trap isn't just about current competition. It's about the fact that the same technology that enabled you to build a marketplace also enables your users to leave. You have to keep earning the right to be the middleman. Luna: This is exactly the kind of conversation that keeps me coming back to this show. And it reminds me — we have a small group of listeners who make this possible. I think it's worth mentioning that this podcast is ad-free, and it stays that way because a handful of listeners chip in monthly through buy me a coffee dot com slash fexingo. Lucas: Yeah, and we don't say that often, but it's true. We've never run a sponsor spot, and we don't plan to. That keeps the conversation honest. So if this show is useful for what you're building or running, and you want to keep it going, that's how. Buy me a coffee dot com slash fexingo. Luna: And honestly, it's not a big ask. Even a small monthly contribution adds up. We just wanted to put that out there. Lucas: Anyway, back to the liquidity trap. There's one more dimension I want to cover, which is how marketplaces can use data to detect when they're approaching the trap. Luna: What are the early warning signs? Lucas: The first is a decline in repeat usage on the side that pays the fee. If sellers or hosts are completing fewer transactions per month, that's a red flag. The second is an increase in direct outreach — buyers and sellers finding each other off-platform. You can detect that through communication patterns. If messages contain email addresses or phone numbers, that's a signal that your take rate is too high. Luna: So you can actually measure the disintermediation risk. Lucas: Yes. And the third sign is a shift in the mix of transactions toward high-value items. Because if your fee is a percentage, high-value transactions generate more absolute fee. If sellers start focusing on high-ticket items to make the fee worthwhile, that's a workaround — and it can distort your marketplace. You might end up with fewer, larger transactions, which reduces liquidity. Luna: So the marketplace becomes less useful for the majority of users who want everyday items. Lucas: Right. And that's the trap. You optimize for revenue per transaction, and in doing so, you reduce the number of transactions. The network becomes thinner, and eventually, a competitor with a lower take rate can come in and build a thicker network. Luna: It's almost like a tragedy of the commons — except the commons is your own marketplace. Lucas: Exactly. And that's the key insight. The liquidity that you spent years building is a shared resource between you and your users. If you extract too much, you deplete it. And once it's gone, it's really hard to rebuild. Luna: So the lesson for marketplace operators: don't treat take rate as a lever you can pull whenever you need more revenue. Treat it as a structural decision that defines your relationship with your users. Lucas: And if you do need more revenue, look at your cost structure first. Automate, optimize, find efficiencies. Only then, if you must, consider a small, tested fee increase. And always, always listen to the signals your network is sending you.