Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / How Marketplaces Use Subsidies to Jumpstart Liquidity
Transcript
- Lucas: Let's talk about the thing every marketplace founder gets told they have to do — and that almost none of them want to admit they're doing. Luna: Subsidies. Paying people to use your platform. Lucas: Exactly. And the conventional wisdom is that you subsidize the side that's harder to attract, because once you have enough of them, the other side shows up naturally. But the devil is in the details — which side, how much, and for how long. Luna: Right, because if you subsidize the wrong side, you burn cash and never get to liquidity. Or you train users to expect discounts forever. Lucas: Let's anchor with a textbook case: PayPal in the late 1990s. They were trying to build a person to person payment system, but nobody wanted to join because nobody else was on it. Classic cold start. So they offered new users twenty dollars just for signing up, and another twenty if they referred a friend. Luna: And that cost them something like fifty to seventy million dollars in the first year. But it worked — they hit a million users in months. Lucas: It worked because they subsidized the side that actually needed to be there: the senders. In a payment network, the value is having a critical mass of people who can receive money. But people only join if they think they can send money to someone. So PayPal paid the senders to join, and by doing so, they created a reason for receivers to sign up for free. Luna: They didn't subsidize the receivers. That's the key insight. They identified which side had the higher barrier to adoption and tackled that. Lucas: Right. And that's the first principle: correctly identifying which side faces the higher adoption friction. In ride-hailing, it's usually the drivers — they have upfront costs like cars, insurance, and time. So Uber subsidized drivers with guaranteed earnings and sign-up bonuses to ensure supply. Once riders saw short wait times, they adopted organically. Luna: But DoorDash and other food delivery platforms subsidized the demand side heavily — discounts and free delivery. Why the difference? Lucas: Because in food delivery, the supply side — restaurants — have a different calculus. They already exist and can list on multiple platforms. The friction is on the demand side: getting customers to try ordering food through an app instead of calling the restaurant directly. So DoorDash subsidized the customer, not the restaurant. Luna: Makes sense. But then you have the problem of what happens when subsidies end. If customers are only there for the discount, they leave. Lucas: That's the second principle: the subsidy must be a bridge to habit, not a permanent feature. Uber had to slowly reduce driver guarantees while improving the rider experience so that drivers stayed because of steady demand. DoorDash had to layer on subscription programs like DashPass to lock in high-value customers who would order even without discounts. Luna: The risk is that you create what's called 'subsidy-dependent liquidity' — where the marketplace only functions while you're paying. As soon as you stop, the network collapses. Lucas: Exactly. A famous example from a few years back was the on-demand laundry startup Washio. They subsidized heavily to acquire customers — ten dollars a wash, when the real cost was probably thirty — but they never got to unit economic breakeven. When the subsidies ran out, customers left, and the company folded. Luna: So how do you know if your subsidy is building a moat or just digging a hole? Lucas: You measure two things. First, the repeat purchase rate among subsidized users once the discount expires. If it's above a threshold — say forty percent — the subsidy is creating habit. Second, the cost of acquiring a user organically versus through a subsidy. If the organic channel starts to grow, you're building network effects. Luna: And you have to be disciplined about which users you subsidize. Some platforms have made the mistake of subsidizing everyone equally instead of targeting the high-potential users. Lucas: The best approach is often to subsidize the supply side first to ensure a core level of service, then use targeted demand subsidies to bring in specific segments. For example, Uber gave new riders their first ride free, but only in areas where driver density was already high enough to guarantee a good experience. They didn't blast subsidies everywhere. Luna: That aligns with what I've seen in the startup world. The smartest marketplace founders think of subsidies as a CAPEX, not an OPEX — they're investing in building a network asset, not buying revenue. Lucas: That's a great framing. And it ties into something we've discussed before on this show: the concept of take rate and how it evolves. Subsidies are essentially negative take rate on one side — you're paying to play. The goal is to flip that negative to positive over time as the network matures. Luna: And if you can't flip it, you don't have a sustainable business. Which brings us to the question: are there marketplaces that never needed subsidies at all? Lucas: Yes. The classic example is eBay. They didn't pay users to list items or to buy. They built a marketplace where the value was clear: access to a huge audience. But eBay had a unique advantage — they were first in the online auction space, and the internet was growing rapidly, providing organic traffic. Most marketplaces today face more competition and need a jumpstart. Luna: So subsidies are a tool, not a strategy. Use them to solve the chicken and egg problem, but design the exit from day one. Lucas: Exactly. And one more thing: the form of the subsidy matters. Cash is the most efficient but also the most likely to attract fraud and gaming. Many platforms now use credits, service bundles, or time-limited promotions to reduce abuse. For instance, some platforms offer a free month of premium features instead of cash. Luna: Another creative example is how some B2B marketplaces subsidize the demand side by providing free software tools — like a free CRM for buyers — and then charging suppliers for access to those buyers. Lucas: That's a great point. It's a form of subsidy that builds stickiness because the tool itself adds value beyond just the marketplace transaction. And it's harder to walk away from than a cash discount. Luna: Yeah. I think that's one of the most underused strategies: subsidizing with services that become part of the user's workflow, rather than just cutting prices. Lucas: Alright, before we wrap, I want to quickly step back. If these conversations about how marketplaces really work are helpful for what you're building or running, the show stays ad-free thanks to listener support. It's a small thing, but if you find value here, you can buy us a coffee — literally at buy me a coffee dot com slash fexingo. No pressure, just a way to keep this going. Luna: And it genuinely helps. We don't do sponsors, so listener contributions are what keep us independent. Lucas: Exactly. So back to subsidies. The final takeaway is that subsidies are a powerful but dangerous tool. The best marketplace founders think about them as a way to compress years of organic growth into months, but they always have a plan for turning off the spigot. Luna: And they measure relentlessly. I think the most successful ones treat subsidy experiments like scientific trials — they have clear hypotheses, control groups, and kill criteria. Lucas: Absolutely. If you're building a marketplace today, start by identifying which side has the highest adoption friction. Then design a subsidy that addresses that friction directly, with an exit plan baked in. Measure repeat behavior, not just signups. And if the unit economics don't improve over time, that's a red flag. Luna: Good advice. Thanks, Lucas. Lucas: Thanks, Luna. That's it for episode ninety-seven. We'll be back soon with another deep dive into the mechanics of marketplace businesses.