Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / How Marketplace Fees Evolve as Liquidity Grows
Transcript
- Lucas: Let's talk about something that sounds almost contradictory on its face: the idea that a marketplace should raise its fees as it gets more successful. It feels like it goes against the whole 'network effects drive prices down' narrative, but actually, it's one of the most powerful levers these businesses have. Luna: And it's not just about greed — it's about capturing value that the platform has created. But when do you know you've earned the right to raise the take rate? Lucas: Exactly. And the canonical example here is eBay. When eBay launched in the mid-nineties, it charged zero insertion fees for most listings and only a small final value fee — around 1 to 2 percent. That was deliberate. They needed to get listings and buyers onto the platform. Liquidity was the priority, not monetization. Luna: Right, the classic chicken and egg problem. If you charge too much early, nobody shows up. Lucas: But by the early 2000s, eBay had massive liquidity — millions of listings, millions of daily active buyers. And they started raising fees. Insertion fees went up, final value fees crept to around 5 percent on average, and they introduced optional listing upgrades like bold or featured listings. Sellers grumbled, but most stayed because the buyer pool was too valuable to leave. Luna: So that's the liquidity wedge — the gap between what sellers would be willing to pay given the access to buyers, and what they're actually paying. As liquidity grows, that wedge widens. Lucas: Exactly. And a well-run marketplace systematically captures that wedge. The key is timing. Raise too soon, and you choke off liquidity before it's self-sustaining. Raise too late, and you leave money on the table — money that could fund better trust and safety features, or better matching algorithms. Luna: So what's the signal that you've reached sufficient liquidity? Is it a specific metric like conversion rate or seller retention? Lucas: There's no universal number, but I look at two things. First, seller retention: if sellers are staying on the platform for more than six months even after a fee increase, that's a strong signal. Second, buyer concentration: if no single buyer accounts for more than, say, 5 percent of a seller's revenue, the platform has genuine diversification. That's when the platform's value is structural, not dependent on a few power users. Luna: That makes sense. Though I think there's a subtlety — not all marketplaces have the same pricing power. A marketplace for rare collectibles, like what eBay started with, has more pricing power than a commodity marketplace like a ride-hailing app, where drivers and riders are more price-sensitive. Lucas: Absolutely. The elasticity of both sides matters. On eBay, a seller of vintage baseball cards knows that switching to a niche forum might get them fewer eyeballs. So eBay can charge more. But on Uber, a driver can switch to Lyft with one tap. So Uber's take rate is constrained by competition. That's why Uber's take rate has stayed around 20 to 25 percent, not 40 percent. Luna: And interestingly, eBay's take rate today is actually lower than it was a decade ago — around 10 to 12 percent, depending on the category. They've had to compete with Amazon and with fixed-price marketplaces that charge lower fees. Lucas: Right. So the liquidity wedge isn't static — it can shrink if competition emerges or if the value proposition erodes. That's why marketplace operators need to constantly reinvest in things like buyer trust, escrow, dispute resolution, and search relevance. If the platform stops improving, sellers will start to question the fee. Luna: Let's talk about a concrete example of a marketplace that timed its fee raises well. I think of Etsy. When Etsy started in 2005, it charged 20 cents per listing and 3.5 percent final value fee — very low. They focused on building a community of handmade sellers and buyers. By 2013, they had about 30 million items listed. That's when they started gradually raising fees. Lucas: And they were smart about it. Instead of one big hike, they did small increments. In 2018, they raised the transaction fee from 3.5 percent to 5 percent. Then in 2022, they added a 6.5 percent 'regulatory operating fee' in some markets. Sellers complained, but the platform kept growing. Etsy's gross merchandise sales went from under $2 billion in 2014 to over $13 billion in 2021. So the value was clearly there. Luna: But they also added services that justified the fee — like promoted listings, offsite ads, and shipping labels. So they weren't just extracting; they were providing more value. Lucas: That's crucial. If you raise fees without adding value, you're just inviting competitors. I think that's what happened to eBay in some categories — they raised fees without improving the buyer experience enough, and Amazon ate their lunch in fixed-price goods. Luna: So let's get practical. If I'm building a marketplace today, what's the playbook for setting and raising take rates? Start low, then increase in small steps, tied to new features? Lucas: I'd say start with a take rate that's roughly half of what you think the long-term target is. For most marketplaces, that long-term target is somewhere between 10 and 30 percent, depending on the category. Then you measure liquidity using metrics like time to first transaction, match rate, and repeat usage. Once those stabilize — say, 70 percent of listings get a transaction within a week — you can start testing fee increases. Luna: And you test them on a subset of users first, right? Like, raise fees only for new sellers, or only for high-volume categories. Lucas: Exactly. A/B test the fee increase. If the control group has 5 percent higher seller churn, you know you've gone too far. Also, watch the buyer side — if buyers start leaving because sellers raise prices to compensate, that's a red flag. The fee ultimately comes out of both sides' surplus. Luna: One thing I've seen work well is what I call a 'fee for service' model. Instead of just raising the transaction fee, you introduce optional paid features that sellers actually want — like analytics, promoted listings, or priority customer support. That's how Etsy and eBay both raised their effective take rate without calling it a fee hike. Lucas: That's a great point. It's less painful because the seller feels they're getting something in return. And it aligns incentives — if the feature doesn't drive more sales, the seller stops using it. The platform has to keep making features that actually work. Luna: There's also the risk of over-monetization. If you push the take rate too high, you create an incentive for sellers to transact off-platform. That's a huge problem for marketplaces — especially in services, where the transaction isn't a physical good. TaskRabbit had to deal with that early on. Lucas: Right. Off-platform leakage is a direct signal that your take rate exceeds the value you provide. That's why marketplace design has to include friction that makes off-platform transactions harder — like escrow, messaging filters, or insurance that only applies to on-platform bookings. But those are topics for another episode. Luna: Absolutely. So to tie it together: the ideal take rate isn't a fixed number; it's a function of liquidity, competition, and the value the platform adds. And the best marketplaces constantly adjust it as those factors change. Lucas: Speaking of adjustments — this seems like a good moment to mention something about how we keep this show going. We don't run ads, and we don't sell your data. A small group of listeners supports us directly through buy me a coffee dot com slash fexingo. That's what lets us keep digging into these niche marketplace mechanics without worrying about sponsorship quotas. Luna: Yeah, and it's genuinely appreciated. It's not a big ask — just if you find these episodes useful and have a few bucks to spare, it helps us keep the lights on. No pressure at all. Lucas: Exactly. And back to our topic — I think one final point worth making is that the liquidity wedge is not just about fees. It's also about what you do with the revenue. Smart marketplaces reinvest a chunk of their take rate into R&D, trust and safety, and expanding the network. That creates a virtuous cycle: better platform, more liquidity, higher willingness to pay, higher take rate, and more resources to improve. Luna: So the take rate is really a strategic variable, not just a pricing decision. It's about how fast you want to grow versus how much profit you want to show. And different marketplaces optimize differently at different stages. Lucas: Right. Early-stage, you optimize for liquidity and growth, so take rate is low. Growth stage, you start capturing value. Mature stage, you balance take rate with competitive pressures. And if you're lucky enough to have a strong moat — like network effects — you can sustain a higher take rate for longer. Luna: Which brings us back to eBay. They had a massive moat in the early 2000s — the largest online auction audience. But they didn't invest enough in the buyer experience, and the moat eroded. Their take rate eventually came down. It's a cautionary tale. Lucas: It is. The take rate is a privilege, not a right. You earn it by consistently delivering value to both sides. And the moment you forget that, the wedge starts to close. Luna: Great way to put it. So for anyone building a marketplace, the key takeaway is: start low, measure liquidity, raise fees gradually, tie increases to value, and never stop improving. That's the recipe. Lucas: And remember that the wedge is dynamic. Keep an eye on your competition, your churn, and your match rates. If you do that consistently, you'll capture the value you've created without destroying it.