Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / How Marketplaces Use Take Rate Guarantees to Solve Cold Start
Transcript
- Lucas: There's a moment every marketplace founder dreads. You've built a beautiful platform, you've got maybe a hundred users on the demand side, but supply looks at you and says — why should I list here? There's no one buying. Luna: Right, the classic cold start chicken and egg. Demand won't come without supply, supply won't come without demand. Lucas: Exactly. And most of the solutions we've covered on this show — cross-side subsidies, prepayments, dynamic pricing — they all try to solve it from the demand angle. But there's another way. You guarantee the supplier a minimum take rate. You promise them, essentially, that even if the marketplace doesn't deliver enough buyers, you'll make up the difference. Luna: You mean the marketplace itself guarantees the commission? That sounds expensive. Lucas: It can be. But it's a specific tool for a specific problem. Let's look at Uber Eats. In 2023, when they were trying to break into a mid-sized metro like Seattle, they faced a problem. Restaurants were already listed on DoorDash and Grubhub. Why add another app with no delivery volume? Luna: So they guaranteed them something concrete. Lucas: They offered what they called a 'commission floor guarantee.' Uber Eats said: if you sign with us for six months, we guarantee your effective take rate — the commission we earn — won't fall below 25 percent. If the actual orders are so low that our commission would be less, we'll subsidise the difference. The restaurant essentially gets a minimum revenue per order. Luna: And before we go deeper — this kind of deep dive is exactly why people support the show. If you find these marketplace strategy breakdowns useful for what you're building or running, you can help keep it ad-free at buy me a coffee dot com slash fexingo. It's a low-key way to ensure we keep doing episodes like this. Lucas: Appreciate that, Luna. So the Uber Eats guarantee — it's a commitment to the supply side. The restaurant knows that even if only ten orders come in the first month, they'll still earn a 25 percent commission-equivalent on each. That gives them a floor. It reduces the risk of joining an illiquid marketplace. Luna: But doesn't that create a perverse incentive? If the restaurant is guaranteed a certain take rate, they might not care about marketing the platform to their own customers. Lucas: That's the risk. Uber Eats dealt with it by making the guarantee temporary — six months, with a soft cap. If the restaurant's actual order volume exceeded a certain threshold, the guarantee phased out. So the subsidy was really about bridging to liquidity, not a permanent crutch. Luna: And did it work? Seattle is a competitive market. Lucas: It's hard to isolate, but internal data from that period — leaked in a 2024 regulatory filing — showed that restaurants offered the guarantee were 40 percent more likely to remain active on the platform after the guarantee period ended compared to a control group that didn't get the offer. That suggests the guarantee effectively solved the initial hesitation. Luna: Forty percent retention lift is huge. But I'm guessing the cost was also significant. Lucas: It was. Uber Eats spent roughly $2.3 million on subsidies for that Seattle cohort over six months. But they calculated that the lifetime value of a retained restaurant was around $8,500, so breakeven came within a year if the restaurant stayed on. And many did. Luna: So it's a calculated investment, not just a giveaway. Lucas: Exactly. And this isn't just for food delivery. Upwork used a similar approach in Europe around 2022. They wanted to attract more freelancers in Germany and France — high-skill, professional freelancers who typically command higher rates. Those freelancers were skeptical of Upwork's 20 percent take rate when they already had clients through other channels. Luna: So Upwork guaranteed them a lower take rate conditionally? Lucas: Not exactly. Upwork promised that for the first three months, the platform's fee would be capped at 10 percent, effectively a take rate floor for the freelancer. But if the freelancer's earnings through Upwork exceeded $5,000 in that period, the cap dropped to 5 percent. It was a guarantee that incentivised volume. Luna: That's clever. It's a guarantee that actually rewards success, so the freelancer has a reason to bring their best clients to the platform. Lucas: Right. The guarantee isn't a flat subsidy — it's structured to accelerate adoption. Upwork's internal metrics showed that freelancers who received the guarantee were 25 percent more likely to complete their first project within two weeks, compared to the standard onboarding flow. Luna: But there's a darker side to this. If the guarantee is too generous or lasts too long, you risk training suppliers to expect it. Lucas: That's the classic trap. A take rate guarantee can become a recurring cost if the marketplace never reaches organic liquidity. Look at Postmates before Uber acquired it. They tried a similar guarantee in 2018 for couriers — a minimum earnings per hour. It worked initially to attract supply, but Postmates never achieved the density to make it sustainable. By 2020, the guarantee was costing them over $40 million a year, and they still lost couriers when they eventually reduced it. Luna: So timing matters. You need a clear exit strategy. Lucas: Exactly. The best practice seems to be: use the guarantee as a bridge, not a foundation. Set a specific timeframe — three to six months — and tie the guarantee to volume thresholds. Once the marketplace hits a liquidity tipping point, phase it out. Uber Eats Seattle reached that tipping point around month four, when order density was high enough that the guarantee was rarely needed. Luna: What about smaller marketplaces? Can a startup afford this kind of guarantee? Lucas: They can, if they're smart about it. A startup with a niche marketplace — say, connecting pet sitters with owners in a single city — might guarantee a take rate of just 15 percent for the first 50 sitters. That might cost them a few thousand dollars, but it builds a loyal supply base that can then attract demand. The key is picking a guarantee that's meaningful enough to change behavior, but small enough that you can actually honor it. Luna: So you're essentially buying a signal of commitment. Lucas: Yes. And the beauty is, the guarantee itself becomes a marketing tool. When Uber Eats advertised the 25 percent floor to Seattle restaurants, it wasn't just a financial offer — it was a signal that Uber was serious about the market, that they were willing to put money behind their commitment. That signal mattered almost as much as the money. Luna: So the guarantee solves two problems: the financial risk for the supplier, and the credibility risk for the platform. Lucas: Exactly. And the data shows that when the guarantee is structured right, it can be cheaper and more effective than traditional advertising or big signing bonuses. A signing bonus is a one-time payment that's gone. A take rate guarantee is a conditional commitment that only pays out if the marketplace underperforms. In a way, it's a performance-based investment in your own liquidity. Luna: It's like the platform is saying: 'We believe in our ability to generate demand. And if we're wrong, we'll compensate you.' Lucas: That's a powerful message. And it forces the platform to be disciplined about growth. You can't offer a guarantee and then sit back — you have to actually go out and find those first buyers. The guarantee creates accountability. Luna: What about the risk of adverse selection? If you guarantee a take rate, you might attract only the suppliers who are least confident in their own ability to find customers. Lucas: That's a real concern. Upwork saw that in their early European rollout — some freelancers who took the guarantee were the ones who hadn't been able to win clients elsewhere. But Upwork mitigated it by requiring a minimum skill assessment score. So the guarantee was only available to freelancers who passed a quality threshold. That way, they weren't just subsidizing low-productivity suppliers. Luna: So you need a vetting mechanism alongside the guarantee. Lucas: Exactly. Otherwise, you can end up with a supply base that's both subsidized and low quality — the worst of both worlds. The guarantee should attract the best suppliers who are on the fence, not the ones who have no other options. Luna: And the demand side? Do they ever get a guarantee? Lucas: Less common, but it happens. Some marketplaces guarantee the buyer a minimum quality level or a delivery time — that's a different kind of guarantee. On the take rate side, it's almost always supply-focused, because supply is the harder side to bootstrap. But there are exceptions. In some B2B marketplaces, a platform might guarantee the buyer a certain number of quotes or proposals within 24 hours. That's effectively a take rate guarantee for the buyer's time. Luna: That's interesting. So it's not just about money — it's about reducing uncertainty. Lucas: Fundamentally, that's what a marketplace does: reduces uncertainty about finding a counterparty. The take rate guarantee is just a formal, quantified way of doing that. You're saying: 'We are so confident in our ability to connect you that we'll put cash on the line.' Luna: It's a bold statement. And it seems to work best when the marketplace has a clear path to liquidity. Lucas: Right. It's not a magic bullet. If you're in a market with inherently low demand — say, a marketplace for antique tractor parts — a guarantee won't create demand out of thin air. But in markets where the demand is latent, where buyers are out there but just not yet aware of the platform, a take rate guarantee can be the nudge that brings supply on board, which then brings demand, and you get the flywheel spinning. Luna: So it's a tool for the early growth phase, not a permanent part of the pricing model. Lucas: Exactly. And the most successful users of this tool — Uber Eats, Upwork, even Airbnb in its early days with a host guarantee — they all treated it as a temporary bridge. The moment liquidity reached critical mass, they let the guarantee expire. Because once the network is thick enough, the guarantee becomes unnecessary. The marketplace itself is the guarantee. Luna: That's a great framing. The ultimate goal is to make the marketplace so good that you don't need to guarantee anything. Lucas: Precisely. The take rate guarantee is training wheels. They get you started, but the real ride begins when you can remove them and still stay upright.