Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / How Thumbtack Solved the Chicken-and-Egg Problem Twice
Transcript
- Lucas: If you've ever tried to start a two-sided marketplace from scratch, you know the chicken and egg problem is the first real test. You need supply to attract demand, but you need demand to attract supply. Most marketplaces fail at this stage. But some solve it once and then have to solve it again, years later, when the model changes. Thumbtack is one of those companies. Luna: I've used Thumbtack to find a plumber and a photographer. I remember the old model — you'd put in a request and get multiple quotes. Then they switched to a booking system. That felt like a totally different product. Lucas: Exactly. And that transition is what we're drilling into today. Thumbtack launched in 2008 as a lead-generation marketplace. You post a job, local pros bid on it, you pick one. Thumbtack charged the pros a fee per lead — usually between three and fifteen dollars, depending on the category. Luna: So the classic lead-fee model. Like Angi or HomeAdvisor. Lucas: Right. And it worked for a long time. By 2019, Thumbtack had over 250,000 service professionals on the platform and was generating about 250 million dollars in annual revenue. But the unit economics were getting worse. Pros were complaining about lead quality — they'd pay for a lead, and the customer would never respond, or they'd get three leads and only one converted. Churn among pros was climbing. Luna: I remember reading that some pros were spending hundreds of dollars a month on leads and only booking a fraction of that. That's a terrible return on investment. Lucas: It was. And that's the fundamental tension in a lead-fee marketplace: the platform makes money by selling leads, but it has no direct incentive to make sure those leads convert into bookings. In fact, if a pro buys ten leads and only one converts, Thumbtack still collects on all ten. That's a misalignment. Luna: So they had to solve the chicken and egg problem all over again, but this time with a different egg. Lucas: Exactly. In 2020, Thumbtack announced a major shift: instead of charging per lead, they would charge a percentage of the booking value — typically around fifteen to twenty-five percent, depending on the category. They called it 'Instant Match.' The customer submits a request, and Thumbtack's algorithm matches them with one pro, instantly. No bidding, no multiple quotes. Luna: That's a massive change. You're changing the core transaction model. You're also changing the value proposition for both sides. Lucas: And you're risking your entire liquidity engine. Because pros who were used to paying per lead suddenly had to accept a percentage model. Some hated it. Thumbtack lost about 20 percent of its active pros in the first year of the transition. That's supply leaving the platform, which means fewer matches, which means worse customer experience, which means fewer customers. The classic death spiral. Luna: So how did they avoid the death spiral? They must have had a plan for keeping enough supply on the platform to serve demand. Lucas: They did. The plan was to gradually roll out the new model by category and by geography. They started in a few service categories — like moving and cleaning — where the lead to booking conversion was already high. In those categories, the pros actually earned more under the booking-fee model because they weren't paying for leads that didn't convert. Thumbtack could show them the data: 'You were spending 200 dollars on leads per booking. Under the new model, you pay 40 dollars per booking. You're better off.' Luna: That's a powerful argument. But it only works if the algorithm is good at matching. If the customer doesn't book, the pro pays nothing. So Thumbtack's incentive completely flips. Lucas: Exactly. Now Thumbtack only makes money when a booking happens. So they're highly motivated to make sure the match is good. That led them to invest heavily in machine learning — building a recommendation system that considers price, location, availability, past reviews, even the time of day. They also added features like instant scheduling and upfront pricing. Luna: It sounds like they moved from being a lead-generation platform to a managed marketplace. That's a much higher-touch business. Lucas: It is. And it requires more operational complexity. But it also allows for a higher take rate. Fifteen to twenty-five percent of a booking is much more lucrative than three to fifteen dollars per lead, especially if the average booking value is in the hundreds of dollars. By 2023, Thumbtack reported 820 million dollars in marketplace revenue, up from about 450 million in 2020. The take rate went from roughly 12 percent under the lead model to over 20 percent under the booking model. Luna: So the take rate nearly doubled. That's a huge jump. But was there any pushback from customers? Suddenly they're paying a booking fee instead of just a flat price from the pro. Lucas: Customers saw a small service fee added to their booking — usually around 5 to 10 percent. But Thumbtack was careful to frame it as a convenience fee. The value prop for customers was: you get a pre-vetted pro, instant booking, no back and forth. And in many categories, customers were already paying a premium for that convenience on other platforms. So it stuck. Luna: It's interesting that they solved the chicken and egg problem twice. The first time, they built supply by offering pros a low-cost way to get leads. The second time, they had to rebuild supply by offering pros a higher-conversion model, even though it meant a share of revenue. Both times, the key was finding the right incentive for the supply side. Lucas: And that's the lesson: the chicken and egg problem isn't a one-time hurdle. It reappears every time you change your pricing model, your matching logic, or your value proposition. Thumbtack was willing to accept short-term supply loss for long-term alignment. That takes conviction, especially when you're a private company — Thumbtack had raised over 400 million dollars in venture funding — and investors want growth. Luna: But it paid off. They've been profitable since 2022, I think? Lucas: They reported positive EBITDA in 2022 and 2023. Not huge margins, but sustainable. And they're now poised to expand into new categories like legal services and financial advisors — other high-intent, local service markets where the same model could work. Luna: That's a big ambition. Legal services is a completely different beast. The trust requirements are higher, the regulatory environment is stricter. Lucas: It is. And Thumbtack will have to solve the chicken and egg problem a third time. But they've shown they can do it. The playbook is the same: start in a narrow category where conversion is high, prove the model with data, and then expand. Luna: You know, this kind of deep dive reminds me why these conversations are so valuable for anyone building or running a marketplace. If this episode was useful for what you're working on, that's actually the cue to mention something. Lucas: Yeah. If today's discussion was worth a coffee to you, you can find us at buy me a coffee dot com slash fexingo. It's the simplest way to support the show and keep it ad-free. No pressure, just an option if it felt worth it. Luna: Really, that small gesture goes a long way in helping us keep doing these episodes. Now, back to Thumbtack — one thing I want to ask: how did they manage the transition without alienating their best pros? The ones who were generating the most revenue? Lucas: Great question. Thumbtack actually offered a hybrid option for a while. Pros could choose to stay on the lead-fee model for some categories and switch to the booking model for others. That gave them flexibility. Over time, as the booking model proved its value, most pros migrated voluntarily. By 2023, over 90 percent of transactions were on the booking model. Luna: So they used a gradual transition, data-driven persuasion, and a hybrid period. That's a textbook case for any marketplace considering a pricing model change. Lucas: It really is. And the key metric they tracked was 'pro net revenue per booking' — how much money the pro made after platform fees. Under the old model, that number was declining because of rising lead costs. Under the new model, it went up by about 15 percent on average. That was the story they told. Luna: Alright, so what's the one takeaway for our listeners? If you're running a marketplace and thinking about a pricing overhaul, what should you do first? Lucas: First, pick a segment where conversion is already high. Don't roll out a new pricing model across the entire marketplace at once. Second, have a clear metric that shows pros they'll be better off — and be ready to prove it with their own data. Third, give them a safety net: a hybrid option, a grandfather period, something that reduces the fear of change. Thumbtack did all three. Luna: And they ended up with a stronger marketplace, a higher take rate, and a more aligned incentive structure. Not bad for a company that started as a lead-gen site. Lucas: Not bad at all. And it's a reminder that the chicken and egg problem isn't just for startups. It's for any marketplace that wants to evolve.