Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / How Airbnb Built Its Take Rate Engine
Transcript
- Lucas: If these conversations are useful for what you're building or running, we want to say thanks right up front because listener support is genuinely what keeps this ad-free and allows us to dig into the granular details of business models like this. Luna: Yeah, and if you’ve gotten something out of the deep dives, buy me a coffee dot com slash fexingo helps keep the lights on and the research sharp. Lucas: Exactly. So let’s jump in. When we talk about marketplace take rates, we usually default to a simple percentage of transaction volume. But with Airbnb, the reality is far more nuanced than a flat cut. Today, I want to drill into how Airbnb constructed its take rate engine, specifically focusing on the structural shift from charging guests to charging hosts, and why that move was less about revenue maximization and more about solving a critical liquidity constraint. Luna: Right, so you’re saying the classic model of taking a slice from both sides didn’t work initially? Why not? Lucas: It did work, but it hit a ceiling. In the early days, Airbnb charged both guests and hosts. Guests paid a service fee around eight percent, and hosts paid a three percent processing fee. The problem wasn’t that people hated paying; it was that supply was too thin. Hosts were risk-averse. If you’re listing your spare room, seeing a three percent fee deducted from your payout creates friction. It makes you hesitate. And in a two-sided market, if supply doesn’t show up, demand evaporates instantly. Luna: So they moved the cost burden entirely to the guest to protect the supply side? Lucas: They didn’t move it, but they shifted the psychological weight significantly. Around twenty fourteen, they started experimenting with a one-sided fee model where only guests paid. Then, in twenty sixteen, they made a pivotal change: they introduced a separate service fee for hosts again, but structured it differently. They raised the host fee to ten percent for professional hosts while keeping the guest fee around twelve to fifteen percent depending on the region. The key insight here is that guests are price-insensitive compared to hosts when it comes to search ranking. A host who sees a ten percent fee might still list because the volume compensates. A guest who sees a higher total price might just scroll past. Luna: That makes sense. Supply is the scarcer resource in those early growth phases. You have to subsidize or incentivize the people bringing the inventory online. Lucas: Precisely. And this leads us to the current state, which is fascinating because Airbnb’s effective take rate has hovered around eighteen percent globally for several years now. That number feels static, but it’s actually a dynamic equilibrium. Let’s look at the components. The guest pays a service fee that covers booking support, insurance, and customer service. The host pays a flat ten percent. But here is the concrete detail most people miss: Airbnb also charges cleaning fees and extra guest fees, which are kept by the host, not Airbnb. This means Airbnb’s take rate is effectively subsidized by the host’s ability to manage their own unit economics through ancillary fees. Luna: Wait, so the host keeps the cleaning fee? That changes how we view the platform's margin. Lucas: It does. It means Airbnb isn’t capturing value from every aspect of the stay. They capture value from the transaction itself—the booking event. The cleaning fee is a pass-through cost that allows the host to maintain quality without raising the base nightly rate. If Airbnb tried to capture that, the nightly rate would look artificially high, hurting conversion. By letting hosts keep it, they keep the listed price competitive, which drives more bookings, which drives more transaction volume for Airbnb. It’s a clever way to hide the true cost of quality from the search algorithm. Luna: So it’s a deflection tactic. Lower headline prices get more clicks, even if the total trip cost is higher. Lucas: Exactly. And this brings us to the concept of liquidity density. Airbnb doesn’t just want transactions; they want transactions in specific geographic pockets. In New York or Paris, the take rate pressure is different than in rural Portugal. They use dynamic pricing algorithms to adjust service fees based on local competition and supply elasticity. In markets with abundant supply, they might lower the host fee to attract more listings. In markets with constrained supply, they can afford to raise the guest fee because there’s no alternative. Luna: But doesn’t that create inconsistency? Travelers expect transparency. Lucas: Transparency is relative. To the average traveler, the breakdown is clear enough: base price, service fee, taxes. They don’t see the backend negotiation between Airbnb and the host regarding the ten percent cut. What matters to the consumer is the total upfront cost. And here is the data point that anchors this episode: Airbnb’s revenue per available listing, or RevPAL, has grown consistently even as take rates stabilized. This proves that they aren’t relying on squeezing more percentage points out of each transaction. They’re relying on increasing the total volume of nights booked by optimizing the match between supply and demand. Luna: So the moat isn’t the fee structure itself, it’s the volume efficiency that the fee structure enables. Lucas: Spot on. Now, let’s talk about the recent pushback. Regulatory bodies in Europe and the US have been scrutinizing Airbnb’s impact on housing markets. Some cities have capped short-term rentals or increased taxes. How does Airbnb respond to that without breaking their take rate model? Luna: Do they absorb the tax, or do they pass it to the guest? Lucas: They almost always pass it to the guest. But they frame it carefully. When new regulations hit, Airbnb updates their fee structure to include 'regulatory fees' or increases the service fee slightly to offset compliance costs. However, the core ten percent host fee remains stable. Why? Because changing the host fee triggers churn. If you raise the host fee from ten to twelve percent, many marginal hosts will leave the platform or go underground. Once they go underground, Airbnb loses visibility and control. They lose the data. And without data, their recommendation algorithm gets worse, and liquidity drops. Luna: So the ten percent is a hard floor. It’s the cost of maintaining platform integrity. Lucas: Yes. And that’s why the take rate is resilient. It’s not just a pricing decision; it’s an ecosystem management decision. Let’s look at another layer: the difference between experienced hosts and occasional hosts. Airbnb segments them. Occasional hosts, who rent out their home a few times a year, often pay zero commission or a reduced fee. Experienced hosts, who operate like small hotels, pay the full ten percent. This segmentation is crucial. It allows Airbnb to onboard casual suppliers without threatening their core revenue base. It’s a funnel strategy disguised as a fee structure. Luna: That’s smart. You bring them in cheap, then scale them up as they become professional. Lucas: Exactly. But there’s a catch. As these occasional hosts become professional, they start negotiating. They ask for lower fees. Airbnb sometimes offers discounts to high-volume hosts to prevent them from switching to direct booking channels or competitors like Vrbo. So the take rate isn’t a monolith. It’s a negotiated variable for the top tier of supply. For the long tail of casual hosts, it’s fixed. This duality allows Airbnb to maximize revenue from the heavy users while keeping the barrier to entry low for everyone else. Luna: It sounds like they’re constantly balancing between extraction and retention. Lucas: Always. And this balance is reflected in their financial reports. Look at the gross bookings versus net revenue ratio. It’s remarkably consistent. Even during periods of high inflation or travel boom, the ratio holds. This suggests that their pricing engine is highly automated and responsive to market conditions. They don’t need to manually adjust fees every quarter. The algorithm handles the micro-adjustments based on real-time demand signals. Luna: So the technology is part of the take rate. The algorithm is doing the pricing. Lucas: Absolutely. And that’s the final piece of this puzzle. Traditional businesses set prices based on cost-plus or competitor-based models. Airbnb sets prices based on predicted willingness to pay and predicted supply elasticity. It’s a feedback loop. Higher fees might reduce bookings slightly, but if the remaining bookings are higher value, the net effect is positive. Lower fees might increase volume, but if the volume is low-margin, the net effect is negative. The algorithm finds the sweet spot dynamically. Luna: It’s basically yield management, like airlines, but applied to homes. Lucas: A perfect analogy. Airlines use it for seats; Airbnb uses it for listings. But there’s a key difference. Airlines sell standardized products. Every seat is roughly the same. Airbnb sells heterogeneous products. Every home is unique. So the pricing complexity is exponentially higher. They have to estimate the value of a specific cabin in Vermont versus a loft in Brooklyn in real-time. That’s why their tech stack is so central to their business model. The take rate is only possible because of the pricing intelligence underneath it. Luna: So if you’re building a marketplace, you can’t just pick a percentage. You need the data infrastructure to support it. Lucas: Correct. You need to know your elasticity curves. You need to know your churn thresholds. And you need to understand the psychological differences between your supply and demand sides. Most startups fail because they treat fees as a static line item. They should treat them as a dynamic lever. Luna: It really highlights why operational depth matters more than just having a good idea. Lucas: It does. And as we wrap up, I think the lesson for entrepreneurs is to look at your fees not as revenue generators, but as liquidity regulators. If your supply is lagging, lower the cost to supply. If your demand is outstripping supply, raise the cost to demand. It’s counterintuitive, but it’s the core logic of two-sided networks. Luna: So the next time you book a stay, remember that the price you see is the result of a complex algorithm trying to keep both the host and the guest happy enough to transact. Lucas: Exactly. And that balance is what keeps the engine running. Thanks for listening to Marketplace Businesses with Fexingo. We’ll be back tomorrow with another deep dive into the mechanics of modern business.