Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / How Marketplaces Use Take Rate Guarantees to Win Supply
Transcript
- Lucas: Uber and DoorDash both offer earnings guarantees to new drivers — a promise that you'll make at least $26 an hour for your first two hundred trips. Instacart does something similar for shoppers in busy zones. And these guarantees are getting more aggressive. Luna: So the marketplace is basically saying, 'If the demand isn't there, we'll make up the difference ourselves.' That feels like a big bet on supply. Lucas: Exactly. And it's a specific strategic move. A marketplace's biggest problem is always liquidity — having enough supply so that demand-side users get a fast, reliable experience. One way to solve that, especially in a new market or a peak period, is to guarantee the supplier a minimum. Luna: But that's expensive. If a driver does ten trips in a week and earns $200, but the guarantee promised $300, the platform eats the $100. That's a direct cost. Lucas: Right. So the question is: when does that make financial sense? And when does it become a leaky bucket that hides bad unit economics? Let's look at a specific scenario. Lucas: Say a marketplace charges a 25 percent take rate. The typical driver on the platform earns $30 an hour before commission. That means the platform keeps $7.50 per hour. But if the platform guarantees the driver $20 an hour net — after the take rate — the driver actually needs to earn about $26.67 an hour gross for the take rate to cover the guarantee. Luna: But if the driver only earns $25 an hour gross, the platform gets 25 percent of that — $6.25 — and the driver takes home $18.75. The guarantee kicks in and the platform pays an extra $1.25. So now the effective take rate drops. Lucas: That's the math. In that case, the guarantee effectively subsidizes the supply. The platform is betting that either the driver gets better with time — faster, more efficient — or that demand density improves so that the driver earns more per hour naturally. Luna: So it's a kind of customer acquisition cost for supply. You're paying to get drivers on the road, hoping they become profitable later. Lucas: Exactly. And this is analogous to how marketplaces think about demand-side acquisition. You might subsidize a first ride or a first delivery to get a user to try the service. Here, you're subsidizing the supplier's first few hours. Lucas: But there's a risk: if the guarantee is too generous, you attract suppliers who are only there for the guarantee. They churn as soon as the guarantee expires. DoorDash saw this with some markets where driver retention after the guarantee period was significantly lower than for drivers who joined without a guarantee. Luna: So it becomes a selection problem. The guarantee attracts the wrong type of supplier — the one who wants a safety net, not the one who wants to build a business on the platform. Lucas: Right. And that's the tension. A well-designed guarantee should attract suppliers who will stay because the platform works for them, not because the guarantee is a crutch. Some marketplaces try to solve this by making the guarantee contingent on behavior — completing a minimum number of trips, maintaining high ratings, or working during peak hours. Luna: Interesting. So the guarantee isn't unconditional. It's a tool to shape supplier behavior. Lucas: Exactly. And some marketplaces have gone even further. Instead of a time-limited guarantee, they offer a permanent earnings floor in exchange for a higher take rate. This is essentially an insurance product for suppliers. Luna: That's a different model. You're saying the platform says, 'We'll take a bigger cut, but we promise you'll never earn less than X.' That flips the incentive alignment. Lucas: Yes. And a few smaller marketplaces are experimenting with this. For example, Faire, the wholesale marketplace, has a program where retailers can opt into a higher commission tier — say 20 percent instead of 15 percent — in exchange for a guaranteed minimum order volume from the platform. If Faire doesn't deliver that volume, they refund the difference. Luna: That's essentially a performance guarantee. The marketplace is putting its money where its mouth is, saying, 'We're so confident in our demand generation that we'll bet on it.' Lucas: Right. And that's a powerful signal to suppliers. It reduces the risk of joining the platform. But it also puts pressure on the marketplace to actually deliver demand. If the guarantee is too generous, the marketplace could end up paying out a lot. Luna: So how do you set the right guarantee level? It can't be too low to attract anyone, and not too high to destroy unit economics. Lucas: This is where data matters. The best marketplaces use historical data to understand the distribution of supplier earnings. They look at the 25th percentile, the median, and the 75th percentile. Then they set the guarantee at a level that covers, say, the bottom 30 percent of earners — enough to be meaningful, but not so high that it covers everyone. Lucas: They also segment. A new driver in a low-density market might get a higher guarantee than a veteran driver in a dense city. The guarantee is a tool to solve specific liquidity gaps, not a blanket policy. Luna: That's smart. Use the guarantee to fix the edges, not the core. Lucas: Exactly. And the other key is duration. Most guarantees are time-limited, with a clear expiration. That creates urgency for the supplier to learn the platform and become self-sufficient. It's a ramp, not a crutch. Luna: But what about the risk of gaming? Can't suppliers just create multiple accounts to keep getting the guarantee? Lucas: That's a real problem. Marketplaces invest heavily in fraud detection to prevent that. They use identity verification, device fingerprinting, and behavioral analytics. Uber, for example, requires a driver's license and background check, and they track device IDs. It's not foolproof, but it raises the cost of abuse. Luna: So the guarantee strategy is really a balancing act. You want to attract the right suppliers, give them a good experience, and then wean them off the guarantee before it becomes a permanent subsidy. Lucas: Exactly. And the best marketplaces treat the guarantee as part of a broader onboarding funnel. The guarantee gets the supplier in the door, but the platform's core value — consistent demand, good earnings, low friction — is what keeps them. Lucas: There's another angle here: the guarantee as a signal to investors. When a marketplace is raising capital, a generous guarantee can signal confidence in future demand. But it can also signal desperation — 'We have to pay people to use our platform.' The narrative matters. Luna: That's a good point. And it ties into the broader question of whether marketplaces should ever subsidize supply permanently. Some would argue that if you can't attract suppliers without a guarantee, maybe your value proposition isn't strong enough. Lucas: I think that's the right framing. The guarantee is a tool, not a strategy. It's a way to accelerate liquidity in the early stages or during seasonal spikes. But if you find yourself needing a permanent guarantee, something is broken in the core marketplace. Luna: Yeah. And that's where the 'if it was worth it' test comes in — if today's episode helped you think about your own marketplace or business, maybe it's worth the price of a coffee. You can find us at buy me a coffee dot com slash fexingo. That's where listener support keeps this show ad-free and independent. Lucas: Absolutely. And it's a small way to say 'this is useful' — we appreciate it. So back to the main thread: the most successful marketplaces use guarantees surgically, not as a blanket. They're a scalpel, not a sledgehammer. Luna: And that's the takeaway. If you're building a marketplace, think about where your liquidity gaps actually are. Don't just throw money at suppliers. Use a guarantee to solve a specific problem, measure the impact, and iterate. Lucas: Right. And if you do it right, the guarantee becomes a temporary stepping stone to a self-sustaining network. That's the goal.