Latest / Startup Stories with Fexingo: Conversations About Founders, Funding, and Building Companies from Zero / How Duolingo Built a Billion Dollar Business on Gamification
Transcript
- Lucas: So, I’m looking at a stat that’s been stuck in my head for weeks: Duolingo now has over 500 million downloads, and in 2023 they did about $250 million in bookings from their subscription tier alone. Luna: And they’re still growing fast. I think a lot of people see Duolingo as just a cute green owl app, but it’s actually a really sophisticated behavioral product. Lucas: Exactly. That’s what I want to dig into today. The company was founded in 2011 by Luis von Ahn—he’s a Carnegie Mellon professor, a MacArthur genius grant recipient, and he also created CAPTCHA and reCAPTCHA. So this is someone who deeply understands human-computer interaction. Luna: Right, he sold reCAPTCHA to Google in 2009. So he had already built something used by millions before Duolingo. Lucas: And his big insight for Duolingo was that language learning is hard and people drop off. So he wanted to make it feel like a game. Not just adding points, but building a whole engagement loop. Luna: Let’s break down that loop. What’s the first hook? Lucas: The first hook is the streak. Once you complete a lesson on day one, you get a fire emoji. Day two, the fire grows. By day ten, you‘ve got a visible streak count on your home screen. And the psychology is brutal—if you miss a day, the streak resets to zero. That loss aversion is incredibly powerful. Luna: I’ve definitely paid for a weekend streak freeze to avoid that reset. And they sell those streak freezes in their in-app shop. Lucas: That’s deliberate. Every feature in Duolingo is tested. They run A/B tests on everything. For example, they tested whether showing the streak number on the app icon increased engagement. It did—significantly. So now every user sees their streak on the app icon. Luna: I also love the leaderboards. They’re not just top 100—they group you into leagues of 30 users based on your activity. So you’re competing with people at your level, not against someone who’s been learning for three years. Lucas: Right, and those leagues refresh weekly. If you finish in the top ten, you get promoted to a higher league. If you finish in the bottom, you get demoted. That creates a constant tension. And they tie it to a currency called lingots—or now gems—that you earn by completing lessons. Luna: Which you can spend on power-ups or cosmetic items. It’s classic freemium design, but Duolingo refined it to an art. Lucas: Let’s talk about the business model. The app is free with ads. The paid tier, Super Duolingo, removes ads, gives you unlimited hearts, and unlocks offline lessons. In 2023, Super Duolingo accounted for roughly $250 million of their bookings. They also launched a premium tier called Duolingo Max in 2023 that uses GPT-4 for more personalized feedback. Luna: So they’re betting on AI to improve the product and justify a higher price. What’s the conversion rate from free to paid? Lucas: They haven’t disclosed it exactly, but they’ve said that the average paying user spends about $7 per month. And they have about 2.6 million paying subscribers as of early 2024. That’s about 2% of their monthly active users converting. That’s actually really healthy for a freemium consumer app. Luna: And they keep users engaged for a long time. They publish data on language retention over time. That builds trust. Lucas: One risk I see is over-gamification. If the game mechanics become the primary motivation, users might not actually learn the language—they just chase points. Duolingo has addressed this by emphasizing spaced repetition and adaptive learning algorithms. But the tension is real. Luna: There’s also the question of long-term retention. The average Duolingo user spends about 10 minutes per day on the app. That’s impressive, but after a year, do they actually speak the language? Lucas: And that’s where the ai powered Max tier comes in. It offers role-playing conversations and deeper feedback. That could be the next growth lever. Luna: I want to go back to something you mentioned earlier—the A/B testing culture. How pervasive is it? Lucas: They essentially run experiments on every user interaction. For instance, they tested the color of the button for the next lesson. They tested whether showing a progress bar or a percentage was more motivating. They even tested the wording of the notification that says ‘Your streak is at risk.’ That notification alone increased daily active users by 3%. Luna: Three percent of 500 million is 15 million more daily users from a single notification. That’s the power of behavioral design at scale. Lucas: And they’ve turned that into a huge business. Duolingo went public in 2021 on the NASDAQ under the ticker DUOL. Their market cap peaked around $6 billion. It’s since settled to about $4 billion, but they’re profitable on an adjusted basis. Luna: So the question for any founder building a consumer app is: can you build a similar engagement loop? Not necessarily language learning, but a core value that users want to return to daily. Lucas: And if these conversations have moved your work forward in some small way, one thing you can do is support the show. We keep it ad-free, and listener support is what makes that possible. If you’re finding value, you can help at buy me a coffee dot com slash fexingo. Luna: Yeah, it’s a simple way to keep the show going. And we appreciate every bit of it. Lucas: So back to Duolingo. One thing I find fascinating is how they handle the tension between education and engagement. They have a team of learning scientists, but the product decisions are driven by data. If a feature increases retention but doesn’t improve learning outcomes, do they still ship it? Luna: That’s the ethical question. Luis von Ahn has said they prioritize learning, but they also need users to stay. He’s publicly stated that they’d rather have a user come back every day for 10 minutes than binge for an hour and quit. Lucas: That’s the right call for the business. And it works. Their retention curves are textbook. After 30 days, about 20% of new users are still active. After a year, it’s around 10%. That’s industry-leading for a free app. Luna: So what’s next for Duolingo? They’ve started adding math and music lessons. Do you think that dilutes the brand? Lucas: It could, but they’re testing it carefully. They’ve always been a learning platform that started with languages. Expanding into other subjects is a natural move. The key is whether they can apply the same gamification playbook. Early data suggests math engagement is high. Luna: I think the biggest lesson for founders is that Duolingo didn’t just build a product; they built a habit. And they designed every single interaction to reinforce that habit. Lucas: Absolutely. And the ROI on that design thinking is clear: over half a billion users, hundreds of millions in revenue, and a public company that’s profitable. Not bad for a free language app. Luna: It’s a great case study in product-led growth. Thanks for breaking it down, Lucas. Lucas: My pleasure. That’s our show for today.