Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / How Marketplaces Use Loss Leaders to Build Liquidity
Transcript
- Lucas: A marketplace without liquidity is just a website. Two-sided networks only work when enough buyers and sellers show up at the same time. And the fastest way to manufacture that simultaneity is to lose money on purpose. Luna: You're talking about the loss leader strategy — subsidizing one side of the market to get the flywheel spinning. Lucas: Exactly. And it's one of the oldest tricks in the marketplace playbook, but most founders get the timing wrong. They either never pull the subsidy, or they pull it too early and the whole thing collapses. Today I want to walk through how the smartest marketplaces have used loss leaders — and the hard math of knowing when to stop. Luna: Before we dive into the numbers — if these conversations are useful for what you're building or running, consider supporting the show at buy me a coffee dot com slash fexingo. Listener support keeps us ad-free and focused on topics like this. Lucas, you were saying about the timing? Lucas: Right. So the canonical example is eBay in the late 90s. When eBay launched, it faced a brutal chicken and egg problem. No sellers meant no buyers, no buyers meant no sellers. So eBay did something that seemed crazy — it let sellers list items for free. No insertion fees, no final value fees. They lost money on every transaction. Luna: And that worked because they had a clear exit plan. They knew exactly when to start charging. Lucas: Exactly. eBay's data showed that once a seller had completed three transactions, their likelihood of returning for a fourth was over 80 percent. So they subsidized the first three listings, then started charging on the fourth. That's a loss leader with a hard stop. The subsidy wasn't open-ended — it was a customer acquisition cost disguised as a free service. Luna: So the key isn't just subsidizing — it's knowing which users to subsidize and for how long. Lucas: That's the nuance. Uber did a similar thing on the demand side. Early on, Uber subsidized rides heavily — giving new users their first ride free or deeply discounted. But they didn't just hand out coupons randomly. They geo-targeted the subsidies to neighborhoods where driver supply was already dense. That way, the free ride was actually a good experience — short wait time, no surge pricing — and the rider got hooked. The subsidy was wasted if the rider waited twenty minutes. Luna: That's smart. Most people think of loss leaders as a blunt instrument — just throw money at one side. But Uber was using data to make sure the subsidy actually built a habit. Lucas: Right. And the most controversial loss leader right now is DoorDash on delivery fees. DoorDash has been subsidizing delivery fees for years to keep order volume high. The argument is that they're buying market share and building habit. But the question is whether they'll ever be able to raise delivery fees to a profitable level without losing customers. That's the risk of a loss leader that becomes a permanent crutch. Luna: That's the Groupon problem, isn't it? Groupon trained customers to expect 50 percent off everything, and when they tried to pull back, the merchants revolted and customers left. Lucas: Exactly. Groupon is the cautionary tale. They subsidized discounts so aggressively that the discount became the product. Merchants hated it because they lost money on every deal, but they felt trapped. When Groupon tried to raise its take rate or reduce the discount, the whole model cracked. The loss leader had become the permanent value prop. Luna: So how do you avoid that? What's the right way to structure a loss leader so it doesn't become a permanent subsidy? Lucas: Three principles. First, time-box the subsidy. eBay's three-transaction rule. Second, tie the subsidy to a behavior that builds long-term value — like Uber tying free rides to neighborhoods with driver density, so the rider learns that Uber is reliable. Third, have a clear metric for when the network is liquid enough to stand on its own. For a marketplace like Airbnb, that meant reaching a certain density of listings in a city before they could start charging hosts a percentage. Luna: And what about marketplaces that are trying to use loss leaders today, in 2026? Customer acquisition costs are higher than ever, and venture capital is more disciplined about unit economics. Lucas: That's the reality. In this environment, a loss leader has to be surgical. You can't just blast money at both sides and hope it works. The best example I've seen recently is from a smaller marketplace called Zeel — it's for on-demand massage therapy. They subsidized the supply side heavily at launch, paying massage therapists above-market rates to ensure availability. But they only did that in cities where they had pre-sold a certain number of first-time appointments. So the subsidy was capped by demand. Luna: That's a nice hedge. You're only losing money on supply if demand is actually there. It forces you to validate demand before you burn cash. Lucas: Exactly. And that brings us to the most important lesson: the loss leader is not a growth hack. It's a strategic investment in network liquidity. If you don't know exactly when and how you'll stop subsidizing, you're not building a marketplace — you're running a charity. Luna: So let's talk about the exit ramp. What metrics do you watch to know it's time to pull the subsidy? Lucas: The classic metric is the liquidity ratio — the number of completed transactions per active seller per month. Once that crosses a threshold, typically around 10 transactions per seller per month, the platform starts to feel essential to the seller. They can't afford to leave. That's when you can start charging. Uber, for example, watched driver utilization rates. Once drivers were occupied more than 60 percent of the time they were logged in, Uber knew the platform had enough demand to justify taking a larger commission. Luna: That makes sense. But what about the demand side? When do you stop subsidizing buyers? Lucas: You look at repeat purchase rate. If a cohort of subsidized buyers returns within 30 days without a subsidy, you know you've built a habit. That's your signal to reduce or remove the subsidy for future cohorts. The key is to test this on small cohorts before rolling out changes platform-wide. A B test is your friend here. Luna: So it's not a one-time decision. It's a continuous process of testing and adjusting. Lucas: Exactly. And that's the difference between a marketplace that uses loss leaders as a strategic tool and one that uses them as a crutch. The strategic ones have a plan. The others just keep burning cash. Luna: Alright, so to sum up: loss leaders can be a powerful way to build liquidity, but you need a clear exit strategy, you need to time-box the subsidy, and you need to tie it to behaviors that create long-term value. Otherwise, you end up like Groupon. Lucas: Well said. And if you're building a marketplace today, start with a really small test. Pick one city, one category, and one cohort. Run the experiment. Measure the repeat rate. Then decide if the loss leader is worth it. Luna: Good advice. Thanks, Lucas. Lucas: Thanks, Luna.