Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / How Marketplaces Use Referral Programs to Jumpstart Liquidity
Transcript
- Lucas: So, Luna, you remember the PayPal story, right? In the early days, they were handing out twenty dollars just for signing up, and another twenty if you referred a friend. Luna: Yeah, that was the famous viral loop. But wasn't that insanely expensive? I mean, forty bucks per new user, and that was back in the late nineties. Lucas: It was, but it worked. And it's the classic example of why referral programs are such a powerful tool for marketplaces. They solve the chicken and egg problem by turning your early users into your sales force. Luna: So today we're talking referral programs as a liquidity lever. And I think we should start by defining what a referral program actually is, because it's not just 'tell a friend, get a discount.' Lucas: Right, it's a structured incentive that rewards existing users for bringing in new ones. And the key is that it's tied to the network effect. In a marketplace, every new user makes the platform more valuable for everyone else. So referrals aren't just about acquisition, they're about deepening liquidity on both sides. Luna: And that's why marketplaces love them. Because unlike a random ad, a referral comes with built-in trust. Your friend is basically vouching for the platform. But the tricky part is designing the incentive so that it actually attracts the right users, not just anyone looking for a freebie. Lucas: Exactly. And that's where a lot of programs go wrong. Let's take the classic mistake: offering a reward to the referrer but not the referee. That can lead to spammy behavior, where people just blast their link to everyone they know, hoping for a commission. Luna: But PayPal was interesting because they rewarded both sides. You got cash for signing up, and your friend got cash too. It was symmetrical. But that's also why they burned through so much money. They were paying for both sides of every transaction. Lucas: And yet, it worked because they were buying a payments network. Every new user made the platform more valuable, and the cost of acquisition was justified by the long-term value of a customer who would stick around and transact repeatedly. That's the calculus you have to run. Luna: So what's the right metric? Is it the cost per new user, or something like the payback period? Lucas: It's really about the ratio of customer lifetime value to customer acquisition cost. If a referred user generates more revenue over time because they're more engaged, then you can afford to pay more upfront. And there's some data that suggests referred users are more loyal and churn less. Luna: That makes sense, because they came in with a social connection to the platform. They're not just random traffic, they're pre-vetted. But how do you measure that? Do you actually see a difference in behavior? Lucas: Some studies show referred customers have a higher retention rate and a higher lifetime value. For example, a study by the University of Chicago found that referred customers were 18 percent more likely to stay with a company after two years. So the quality is real. Luna: Okay, so the theory is solid. But let's talk about the practical design. What are the key decisions a marketplace has to make when setting up a referral program? Lucas: First, you need to decide what the incentive is. Cash is straightforward, but it can attract arbitrage seekers. Credits or platform currency keep the user inside your ecosystem, which can be smarter. Then you have to decide whether to reward the referrer, the referee, or both. Luna: And that's where you can get into trouble. If you only reward the referrer, you might get a lot of spam. If you only reward the referee, you might get free-riders who sign up but never become active. So you have to align incentives. Lucas: And that's why many marketplaces use a two-sided incentive, but with different values. For example, Uber gives ride credits to both the driver and the rider, but the amount might vary based on the market or the user's history. They also use geotargeting and timing to make the offer more relevant. Luna: So it's not just about the amount, it's about the context. But what about the risk of fraud? People creating fake accounts to game the system. That's a huge concern. Lucas: Absolutely. Some marketplaces have seen as much as 20 percent of their referral traffic being fraudulent. That's why you need to have strong verification measures in place. PayPal had to crack down on people opening multiple accounts just to collect the sign-up bonus. They started requiring bank account verification before you could get the cash. Luna: And that's a great lesson: the friction you add to the referral process has to balance the risk of fraud against the ease of conversion. Too much friction kills the viral loop, but too little invites abuse. Lucas: And there's another layer: the type of marketplace. A referral program for a ride-hailing app might work differently than one for a marketplace for vintage furniture. In ride-hailing, you need liquidity in a specific geographic area at a specific time. So Uber often runs localized referral campaigns, targeting users in areas where they need more drivers or riders. Luna: That's a smart point. So it's not just about growing the network overall, it's about growing it in the right places. And that's where data comes in. You have to track which users come from referrals, where they are, and how active they become. Lucas: And that leads to another critical component: the viral coefficient. It's the number of new users each existing user brings in. If it's above one, you get exponential growth. If it's below one, you need to supplement with paid marketing. And a well-designed referral program can push that number up significantly. Luna: But there's a difference between a viral coefficient and a sustainable growth loop. You don't want to just spike new sign-ups; you want them to become active, transacting users. So you need to think about activation as part of the referral funnel. Lucas: Right. And that's where the reward timing comes in. If you give the reward immediately, you might get a lot of people who sign up and never come back. If you delay the reward until the referee makes their first transaction, you're incentivizing both sides to actually use the platform. Some marketplaces even require the referee to complete a certain number of transactions before the referrer gets the full reward. Luna: That's a good way to ensure quality, but it also adds complexity. You have to communicate that clearly to your users, or they'll get frustrated and abandon the program. Lucas: And that's why you need to test and iterate. What works for one marketplace might not work for another. The key is to run experiments, measure the results, and adjust the incentives based on the data. Luna: So let's talk about a specific case that went wrong. I remember that 'Placester' or some other platform had a referral program that backfired because the rewards were too easy to game. Do you have a better example? Lucas: There's a classic example from the early days of Groupon. They had a referral program that offered a ten-dollar credit for each friend you brought in, but they didn't have strong fraud detection. People started creating multiple email accounts to refer themselves. Groupon had to retroactively cancel a lot of those credits and tighten their verification, but it left a bad taste for some users. Luna: So the lesson there is that you need to build fraud detection into the design from the start, not as an afterthought. And that means tracking things like device IDs, IP addresses, and payment methods. Lucas: Exactly. And that's where machine learning can help. You can build models that flag suspicious referral patterns and automatically hold rewards for review. But that's a technical solution to a structural problem. You also need to think about the user experience. If you make the referral process too complex, people won't bother. Luna: So what's the ideal referral flow? You want it to be as simple as possible: a unique link, a prefilled message, and a clear call to action. And the reward should be clearly stated upfront. Lucas: And you want to make it easy for the referee to redeem the reward. If they have to enter a promo code, that's friction. If it's automatically applied to their first transaction, that's better. Some platforms even let the referrer send a gift directly to the referee's account. Luna: Let's shift gears for a second. We talk a lot on this show about the mechanics of marketplaces, and a lot of our listeners are actually building or running these kinds of businesses themselves. If you're finding these conversations useful, and you want to help keep this show ad-free, you can support us at buy me a coffee dot com slash fexingo. It's a simple way to keep the lights on. Lucas: Yeah, we really appreciate it. Every little bit helps us keep digging into these topics. And we're back to the question of whether referral programs are a sustainable growth strategy or just a short-term hack. Luna: Right. So we've talked about the design and the risks. What about the long-term? Do referral programs stop working as the marketplace matures? Lucas: They can, if you don't evolve them. When you have a critical mass of users, the marginal value of a referral might be lower, but the cost of the incentive stays the same. So you might need to reduce the reward over time, or shift from cash to credits, or target specific segments where you still need liquidity. Luna: And you also have to consider the competitive landscape. If a competitor launches a more generous referral program, you might need to respond, but that can lead to a race to the bottom. Lucas: That's why it's important to focus on lifetime value, not just acquisition. A referral program that brings in low-quality users can actually hurt your network by increasing spam and reducing trust. So you need to measure the quality of referred users, not just the volume. Luna: And that brings us back to the core principle: referrals are a tool to build liquidity, but they have to be aligned with the overall health of the marketplace. Lucas: So, what's the one thing our listeners should take away from this episode? I'd say it's this: before you launch a referral program, think carefully about the incentive structure, the fraud risk, and the long-term value of the users you're bringing in. And then test, measure, and iterate. Luna: And remember that referrals are just one lever. They work best when combined with other liquidity strategies, like content marketing, partnerships, or even offline events. It's not a silver bullet, but it's a powerful tool in the right hands. Lucas: Absolutely. And as we've seen, the best marketplaces are the ones that constantly tinker with their growth levers. So next time you're looking at your numbers, ask yourself: are our referral users becoming our most loyal users? If not, it might be time to rethink the program. Luna: That's a great place to end. Thanks for joining us, and we'll see you in the next episode.