What do Mattel, Banana Republic, Butcher Box, and Glossier all have in common? They power their businesses with Shopify. Shopify is the most innovative and scaled commerce platform on the planet that also happens to have the best converting checkout on the planet. And that's no industry secret. That's Shopify. Turn more at Shopify.com slash enterprise. Hurricane Helene has caused tens of billions of dollars in damages. Hurricane Milton could be even worse.
And insurance companies are jacking up prices all across the country. As someone in the insurance industry intellectually I understand it, but as a homeowner it's awful, it really sucks. This week on Unexplainable, as the world keeps getting warmer and storms keep getting worse, is our future uninsurable. There's a breaking point coming. Follow Unexplainable for new episodes every Wednesday. Shavon is in love. She and her boyfriend are maybe thinking wedding bells.
And even though there's no one she'd rather be with, she still wonders. What's the chance of this lasting? How do you get from that point to also being like, I cannot stand you to the point that I want to cut my ties with you? This week on Explanate To Me, we find out. Are we less likely to get divorced than our parents? Follow Explanate To Me wherever you get your podcasts. New episodes every Wednesday. Hello and welcome to Decoder. I'm Neil iPadell, editor and chief of the
Virgin. Decoder is my show about big ideas and other problems. Today I'm talking with Luis Fanon, the co-founder and CEO of Duolingo, the popular app that teaches languages. It's an interesting time to be in the business of languages. Now for authors, anything the current state of AI tech can do, it's babble away in different languages with people who can't quite understand what they're hearing. So there are lots of opportunities to enhance a product like Duolingo with AI.
And Luis and I talked about the new features and something called Duolingo Max, which now offers AI chat conversations with some characters and even video calls with an AI avatar called Lily. I was obviously curious about all that, but I also wanted to talk to Luis about learning generally. If you're like me, you've started and stopped Duolingo several times. If you're an overachiever, you've got the street going. You might even have a streak to maintain today.
And it's that streak that's key. You'll hear Luis come back to a big idea here several times. Engagement is the thing, he says, because simply showing up is the cornerstone of making progress with language learning over time. His point is that you can't teach someone who isn't there. And so over time, Duolingo has become more and more of a game because people like to play games.
But there are some real conflicts between games and actual learning. As you'll hear, Luis is happy to admit that conflict exists and he's given it a lot of thought. For him, the gamification is the most important thing because not only does it bring you back to Duolingo, keeping the business humming along nicely, but he says it also produces the results in language proficiency that Duolingo is aiming for. We talked about this in detail. I think you're
going to find it really interesting. We also talked about the money in detail. Luis got pretty deep into explaining where it comes from. As you might guess, the people who spend the most money in Duolingo are iPhone users in wealthier countries like the United States. And some technical decisions that Duolingo made very early on means the iOS version takes priority. It can take a year or more for features to roll out in the Android version of the app. But here's the hard part.
Duolingo is a global product. And the biggest chunk of learners in the platform are actually trying to learn English in poor countries. And those users are way more likely to have an Android phone and to want more need a free version of Duolingo. That's a lot of tensions to balance out. And you'll hear Luis talk about his own childhood in a poorer country and how that informs his decisions. I gotta say, this is one of my favorite Dakota conversations in quite a while.
Luis is the co-founder and he has seen the company through from being a startup to going public and now embracing a pretty big technology shift in AI that has a direct impact on the product he makes. And we talked about all of that in a pretty direct way. It's only a handful of jokes about founder. And of course, I asked him whether he approves of all the unhinged things that Duolingo Al says on social media. Okay, Duolingo co-founder and CEO Luis Fanon. Here we go.
Luis Fanon, you are the CEO and co-founder of Duolingo. Welcome to Dakota. Thank you for having me. I think a lot of people know what Duolingo is. I often start by asking CEO, what does Duolingo? But I feel like everybody knows what Duolingo is. How do you define Duolingo? Well, it's an app that teaches languages. That's what we're mostly well known for. As of the last couple of years, we also teach math and music. It's the most popular way to learn languages in the
world. A few fun facts, for example. There are more people learning languages on Duolingo in the US than there are people learning languages across all US high schools combined. And this is actually true in most countries in the world. We teach languages to more people than the public school systems. You have big announcements coming up at Duocon. That'll be public by the time this airs. One of them is chatting with characters like Lily and others. Video call with Lily.
Yeah. How does that work? Are you? Is that just more opening eye technology? How are you making that happen? We have a discussion of characters that are used as love. One of them is an E-motean with purple hair who is very unimpressed by you. And it turns out you can talk to her now and practice your conversation. And you're going to have really good conversations with her. And what's amazing about it? There's a lot of things that are basic about it. First of all, she adapts to your level.
And we know your level because you've been learning on Duolingo. So we have a pretty good idea of what your level is. She adapts to your level. The other thing is she has memory. So she knows that last time we talked about something, for example, I literally just had a conversation today where she remembered that last time we talked about the fact that I'd like Nirvana. And she was telling me that her favorite song is Smells like Teenspirant. So it's easy choice. I have to say.
Easy choice. By the way, we're dating ourselves on that one. But yes. And so these are pretty enjoyable conversations and you get to practice your language super well. And it is entirely spoken. And it just works really well. And we're very happy because it is the first time that you really think we really, really are not going to need humans for this. Yeah. The animations and stuff are those stock animations. Are they loops? How does this work?
No, no, no, we have we animate them. I mean, it's a comment. No, I bet like, is it in real time? Yeah, yeah, yeah. It's animated in real time. This is a three. We have a rig for her. And we have we have a really, it turns out we've bought these two animation studios in Detroit. This is why we have an office in Detroit. And they've done a really good job. So everything her mouth moves
just when she's speaking. And it is tied to what she's saying. She rolls her eyes at you, etc. When you think about that investment, we're going to start building rigs and animations for characters. We're going to do it all in real time. I'm just going back to cost. You've got a, that's a big investment. Do you think that's going to make your existing users pay more money or is it going to get you new users? I think it's both. We just see it as continuing to work on the app.
I mean, there's a lot of places where we use a lot of animation and we just see it as continuing to work on the app. And generally, as we continue working on the app, we both get more users and get more of them to pay. Yeah, no, I'm just the reason I said in this context specifically, it's just the economics of AI or so. Yeah, it's just a series of question marks right now. Yeah, yeah, yeah. So everybody is making the investments. I'm curious. Where do you see,
how do you see it coming out on the other end? I think for this feature, for this particular feature, I think it's just an excellent use of large language models. Yeah. And I think it's, you know, on our end, it's working pretty well. The other big announcement you have is called adventures. Sounds like a, sounds like a video game. It is basically, so some of the lessons in, you know, the way Duolingo works is the home screen is basically a path. And then you just kind of
doing lessons. Some of the lessons are now going to be this thing we call an adventure, which is really just like it is just like one of these video games where you move characters around. And what's cool about it is that you're learning how to solve kind of real world situations on Duolingo. So for example, it's just like a little video game where you're, you know, you are
one of the characters and you're told, okay, go buy a pizza. And you're like kind of move and have to ask around and you ask some people and then the people tell you, oh, the pizza place is over there, et cetera. So it's like, it's super fun. And it helps you learn kind of navigate the real world. So yeah, we've been working on that. What's cool about that feature is most of the scenarios or all the scenarios were mostly generated by AI. That feature in the past would have taken a
long time to scale, but we were able to scale it pretty quickly because of AI. Yeah. I played with Duolingo this morning. I have a long and complicated history of trying to learn Hindi. It's free. I was using it for free today. How does it make money? Well, it is free. You can use it entirely for free without ever having to pay, but if you don't pay, you may have to see some ads
and we make money from the ads. But also, if you want to turn off the ads, you can pay to subscribe and it turns off the ads and it gives you some extra features and we also make money from the subscription and actually the majority of the revenue comes from the subscription. The majority of the revenue comes from the subscription. Is Duolingo profitable as a company? Yes. As of relatively recent. Yes, yes, we are. Well, I'm curious about this. I hear about this
split from almost every one we talk to that we start. We want to grow our base of users. Ads help us do that. It helps us keep the product free and then the real money is going to come when we add value and we add paid, particularly with advertising lately, with app tracking, transparency on Apple platforms, with the massive influx of inventory from all the other platforms in the world with the desire to do that. All this stuff, it seems like ads are even harder to make money on than ever.
Is that been the case for you? You know, it's probably true. We've ads have never been a priority for us. I don't know the exact number, but it's something like six or seven percent of our revenue comes from ads. For us, as long as they're there, they're a good reason for people to subscribe. But yeah, generally we make about 80% of our revenue from subscriptions. Even though, by the way, only about 10% of our monthly active users pay to subscribe. We're a little under 10% of our
monthly active users pay to subscribe. That 10% of our monthly active users give us more than 80% of our revenue. And all of that revenue is in languages or is math growing? It's the majority of languages. I mean, math and music are growing. They're just getting started. I mean, we really launched this about a year ago. And so they're just getting started. It's overwhelmingly languages. Yeah. What languages are the most popular?
By far, English is the most popular. 45% of our active users are learning English. The second is Spanish and the third is French. And then there's a big drop of after that. Yeah. And are the majority of your users guessing outside the United States? Are they inside the United States? Yeah. I mean, the US is about 20% of our users and 80% of us is international. And so 80% of users are trying to learn English.
About 45% of trying to learn English. It turns out in international. They also want to learn other languages. But yeah. It seems like there's a lot of languages offered in the app. And it seems like one way you could allocate resources by saying English is the most popular. The most resources there. But that doesn't feel like how the app works. How do you think about it? We definitely do some of that. We don't do it as I was going to say as much as we should,
but I don't know if that's the case. We don't do it commensurately with the number of users. Because then we would probably spend all our resources on English and Spanish and French. So certainly, you know, we spend the majority of our resources in the top eight languages to learn. And then we spend very little resource outside of that. The top eight are English, Spanish and French. Then there is German, Italian, Japanese, Korean, Portuguese, I think, Chinese. I think
those are the top eight. Hopefully that's eight. Mandarin and our Cantonese. Mandarin, Mandarin. And that's it. That's a top eight. And after that, there really is a huge drop off. So we have language, even large languages. For example, we have Arabic. Arabic is a large language. Not that many people are learning Arabic. So we do put some resources there, but it's much, much less than for the larger languages. How do you think about just that kind of demand?
Right? I opened Duolingo. I look at it and I should probably learn some Cantonese. That is a real thing that I thought many times. But I don't know if everybody feels that way. I opened Duolingo. I think, man, I should be much better at Hindi than I am. That's a real thing that I think all the time. I imagine there's a lot of people in my particular diaspora who feel the same way. But that's like latent demand. Do you ever go out and say to people, you should learn some Spanish.
We should market Spanish in the south of the United States. We have remained neutral about that. But it is an interesting thing that the man for learning languages is not as correlated as you would like with number of speakers or maybe even usefulness in a geopolitical world. So for example, Chinese, even though it is one of our top eight languages to learn, only about 2% of our users are learning Chinese. It's relatively small, even though as the most spoken language in the
world. So if I were to tell people, they're like, you know, maybe more of you should be learning Chinese. You know, one of the things that goes into people's calculus is how hard the language is to learn. Turns out Chinese, at least for English speakers, is just a lot harder. I mean, we have data to get to get to a pretty good point in Spanish for English speakers. Takes call it three to 400 hours.
That same level of knowledge for Chinese takes about 2000 hours. So the realities in the United States, if you're just going for pragmatism, return on investment, Spanish is probably a much better. I mean, in the US, you probably should learn Spanish. I mean, it's just it's quite an easy language to learn. In the US, you should probably learn Spanish. That is a marketing message. Yeah. Well, we don't say that. We've tried to remain neutral. We probably would get in trouble.
Or I would get in trouble inside the company. We started pushing people for certain languages. We try to remain neutral. Maybe not inside the app, but as a way to grow, right? As a way to capture new users. I'm just curious because that it seems like a lot of what Duolingo is right now is people know they should be multilingual or bilingual at least. And so Duolingo is there. But there's also a huge portion of population. It's like,
at least in this country, that's like, I'm speaking English. And the idea that there's value in learning a second language is foreign. Yeah, there is. Although I'm very happy with our results in the US. The US and the UK are pretty interesting because historically, there hasn't been a big desire to learn languages in the US and the UK. The thinking has been whatever you can learn in English. In the US, 80% of our users were not learning a language before Duolingo. So we're
growing the market in the US and the same number in the UK. So I'm very happy with that. Yeah. One of the things I think about just in terms of learning languages, I think, back to learning French when I was in high school in Wisconsin, was there's learning the language and then there's all of the culture that comes with a language, particularly some of the regional languages. High school French is a lot of like, look at a picture of a baguette. It's totally foreign
to whatever you're doing. Do you think about that instead of Duolingo that there's a huge cultural component here? That's more investment, right? We do. And we try to add the culture. We don't do it as much as maybe we should. We try to stick mostly with languages. It depends also on the language that we're doing. Some languages are quite tight to the culture. Some are less. Spanish is a good example. I mean, there are, I don't know the number, but 20 some countries that speak Spanish
and some of them are pretty different than others. So, we do a little bit of culture, but we try not to be like, oh, you're learning Spanish. You're a Mexican person with a sombrero. We try not to do that. I mean, we also have to be not offensive and stuff like that. But we try to add it a little bit. I would say it's not the primary goal. The reason I said is, Duolingo is instantiated for us people as a mascot. We should talk about the mascot's personality and its social media presence.
But it's fairly abstracted from a person teaching you the language. There's not someone on the other side. It's like, here's, I'm teaching you this. Here's the culture that comes with it. You might have other teachers who might teach you another way. There is an abstraction there that just feels interesting, especially as, you know, we're obviously going to talk about AI and how you're using that
and how you're expanding the platform. I just want to push on that a little bit. That abstraction, do you think it's resulting in people who've learned a language or people who've learned how to communicate? So let's see. It's been very much on purpose for us to not put humans in the app as in human teachers. There's nothing wrong with human teachers. It's just the case that, you know, from the beginning, we've been a technology company and we've wanted to make it so that technology
teaches you. There's a couple of reasons for that. One is just it's a lot cheaper to teach you with technology than with a human teacher. The other thing is it turns out somewhere between 80 and 90% of language learners. Just don't want to talk to another human. They may tell you they do, but they don't. And it's because when you're learning a language, you're pretty shy about it and only the extreme extroverts are okay talking to a stranger on video in a language that they're
not very good. The majority of people just won't do it. I mean, we've actually done these research studies over the years because over time, you know, we thought maybe we should add humans, maybe we should add humans. And we've done these research studies that are in some of the most amazing things that I've seen when you talk to a user, you ask them, what do you think could make Duolingo better? And, you know, historically in the past, they have said, well, more practice
conversation with a real person. They have said that. And then you asked the user, okay, so you're telling me, if I put a human on Duolingo, you would do that. And then they said, yes, yes, I would. And you can even ask them, would you pay for it? And then they'll say, yes, I'll pay for it. And then you tell them, okay, do you want to do it right now? And the answer invariably is, no, no, not right now, not right now. People just don't want to do that. So that's why we haven't
put the humans there. And I think it's been a good decision, especially now that with large language models, we can do a pretty good job getting you to practice conversation without a human. I want to ask what all this because I've been asking a lot of people on this show, what good are these large language models? What are the products you're going to make? I understand you're making the models. And it feels like Duolingo has a very natural solution, right, which is you can just
talk to it and talk back. And it kind of doesn't matter if everyone is hallucinating because all you're doing is practicing talking. That's exactly right. It is a really, really good application. I mean, there's a number of things you said it right. It doesn't matter if it says something that is like a little wrong because you're just practicing your language. Also, it doesn't matter if it makes
a small mistake. Sometimes it makes a small grammatical mistake. People can't even notice because they're usually kind of beginners that Spanish or beginners that French, you don't know. And it also can adapt to your level really well. Large language models are really good at adapting to your level. So we tell it, okay, adapt to a beginner Spanish. And by the way, we even tell it, hey, because we've seen this person learn on Duolingo, we actually know all the words they know.
So we tell the language model, like this person only knows these 200 words. So please, mostly use only these 200 words. So it works really well for that. How much investment into AI are you making? Right. This is a new product. It's very costly. Everyone is telling me about how much the NVIDIA GPUs cost all the time. We're only just profitable, you said. But this feels like the thing that will immediately make you not profitable again as you invest in AI. We're investing a lot.
Fortunately, it's actually being good for us in terms of profitability for two reasons. So there's two places where we invest in AI. The first is generating data that is going to be used in our lessons. That data used to be generated partly with humans. And now it's mostly generated with AI. And it turns out it's a lot cheaper to generate with AI than with humans. Also, it's a lot faster to generate it with AI. So we're very happy with that. And then the other big use is real-time
conversation. That one is expensive. It's actually expensive to provide a real-time conversation with a user. But what we had to do is we actually had to add a higher price plan. So we now have two subscription plans. We have Super Dual-Engle, which is our standard subscription. And we added a new one called Dual-Engle MAX, which is about twice the price of Super Dual-Engle. That one gives you the conversation practice. And it's expensive, but people pay twice as much. So actually, it really
doesn't cut into it. So it's actually worked out well for us. Let me dive into the economics of that. Because I'm really fascinated by, in general, whether any of this will result in profitable, sustainable companies. There's a lot of money flowing into this. So you charge twice the price to run inference. I'm assuming that someone else is large language model. You haven't trained one of your own. Yeah, we use the open AI. So you're buying some capacity from open AI. You're
buying some tokens from them. You're reselling them to users for twice the price of your standard plan. What's your margin on that resell? I don't know the percentages, but I do know that where, I mean, at some point probably knew the percentages, but I don't know the number of my head, but I do know that we are basically, you know, it is good for us in terms of margins. Yeah, that's the thing that I'm curious. I don't know if it's good for open AI. It's all the way
at the bottom of that chain. I don't know if that's profitable for them. But as you build products on this stuff, it seems like your economics depend on their economics in some way because you need to add a margin to that. Like that all seems very complicated and tenuous, especially if the AI features are what brings you new users. Yeah, the good news is at the moment, the AI features are not bringing those new users. What's going on is for new revenue. Yes, it is bringing us new revenue.
There's a good margin there. I mean, we could, so a couple of things, the price of the same exact call is going down over time. Whether you do it to an open AI or whether you do it through a Microsoft or whatever, the price is going down for a number of reasons. Everything's getting more efficient and also chips are getting cheaper over time, etc. So the price is going down. At the moment, there's a good amount of cushion, but we expect that over time, there will be even larger
amount of cushion. So at least for our application, I'm not particularly concerned in terms of margins. I think for our application, the margins work up pretty well. Yeah. Do you think the models are somewhat interchangeable? The thing that I've been hearing more and more is the model business isn't the thing, the product business is the thing. I think the answer is yes, but I think the operative term is somewhat, they are somewhat interchangeable. So we've tried to build our technology
stacks so that you can interchange them. But the reality is that you interchange them and you start getting wonky stuff because you probably spent a lot of time really testing your way into the right queries. You may have done some fine tuning, etc. So you can interchange them, but if you do, you probably need to spend a few months making sure that the wonkyness goes away.
Yeah. When you think about just this investment over time, does it feel like you need to put the money in up front and you'll get more efficient on the back end, or does it feel like, oh, this is going to be the future of the company. So we need to rebuild around the capabilities of large language model. It's somewhere in between. Yeah. Let's see, I do think that large language models are going to be very positive for dwelling already are. And I think they're going to continue being very
positive. What is not true is that, oh, my God, large language models just solve all our problems. I mean, one of the biggest issues that people don't seem to, or at least people are not talking about that, particularly with education when it comes to large language models is large language models are good at teaching new stuff. They're not good at engagement. And that's the hardest thing with education. I mean, the hardest thing about me trying to teach you something is just keeping you
engaged. And it's somehow people forget, you know, some people saying like, oh, my God, you can learn quantum physics with chat GPT. And yeah, sure, but that's some, you know, that's just not that impressive. You can learn quantum physics with a book. Like, it's the technology to learn this has been around for a long time. It's called a book and it works. It's just people don't really want to read a quantum physics book. And similarly, most people really probably don't want to go to
chat GPT and start asking questions about quantum physics. And it's the same thing for language learning. You know, large language models are very good at getting you to practice. But keeping you engaged is pretty hard. And so, you know, that part, I just don't know of large language models are going to help all that much. And in the end, this is a sad thing, but the reality is, you know, Dooling was very gamified. And I wholeheartedly believe most people would rather spend more time
playing Candy Crush than talking to others. And that's maybe a sad truth. And, you know, there's some exceptions, right? I mean, you love talking to somebody that you're in love with. And sure, that's nice. But their reality is for most in most of the time, most people would just rather spend time playing Candy Crush or scrolling on Instagram than talking to others. So all of that, I just don't
think large language models are going to help with all that much. We need to take a short break. We'll get back. Think scaling AI is hard. Think again. With Watson X, you can deploy AI across any environment. Above the clouds, helping pilots navigate flights, and on lots of clouds, helping employees automate tasks. On-prem, so designers can access proprietary data, and on the edge, so remote bank tellers can assist customers. Watson X works anywhere so you can scale AI everywhere.
Learn more at IBM.com slash Watson X. IBM, let's create. Your business deploys AI pilots everywhere. But are they going anywhere? Or are they stuck in silos? Exhausting resources, unable to scale. Maybe you don't need hundreds of AI pilots. You need a holistic strategy. IBM has 65,000 consultants with Gen AI expertise who can help you design, integrate, and optimize AI solutions. So you're not just deploying AI. You're scaling it across your business. Learn more at IBM.com slash
consulting. IBM, let's create. What do Mattel, Banana Republic, Butcherbox, and Glossier, all have in common? They power their businesses with Shopify. Shopify is the most innovative and scaled commerce platform on the planet. That also happens to have the best converting checkout on the planet. And that's no industry secret. That's Shopify. Learn more at Shopify.com slash enterprise. Welcome back. I'm talking with Duelingo CEO Luis Phanon. But why he sees the game of Duelingo
as the foundation of the education it's meant to provide. You actually, you have a long history in gamification, right? Your first project that I think he sold to Google was a gamified thing. You did recapture, which is essentially gamifying training data in a particular way. Do you think there's an evolution in Duelingo that the first thing that he worked on was the engagement and the bringing the people back and having the character and then the underlying
content was language lessons. There was a part of Duelingo when I started using it several years ago. I was like, oh, this is very familiar. It's just that this bird won't leave me alone. Right. And that's why I'm back again. And now you're talking about this whole other spectrum of things. We're going to use AI. We're going to have these natural language conversations. We're going to expand to mathematics. When did you feel like you were making a transition from
we're gamifying this very familiar thing. We're using this new engagement mechanism to this is now a wholly new thing. From the beginning, this is a central thesis that we believe here at Duelingo. The hardest thing about learning something by yourself is to be motivated. That is just it. That is in fact, that is probably the reason for the vast majority of our success is that we realized that early on. From the beginning, we have tried to have a thing that
is enjoyable to use and that keeps you coming back. We have probably spent more effort on that than anything else. Internally, our feeling is learning a language. It's a lot like working out in that. It doesn't matter all that much whether you're doing the elliptical or a peloton or a treadmill. It probably matters. But the reality is what matters by far the most is that you're doing it every day. Whatever the hell you're doing. It's kind of the same with Duelingo.
Whether you're doing this method or that method, maybe some methods are more efficient than others, but what matters is that you're doing it every day. We got very good at that. Once we got very good at that, we started trying to add more sophistication and what we teach. We've been doing that for the last few years. But always primarily, we are a motivation engine.
Is that the core of it still? When you say always primarily, there's a thing where everyone starts to take the initial innovation for granted and then focus on the rest of it. I'm going to end up asking you about founder mode. How do you keep the focus on that part instead of everything else? I do. I spend effort on that, but it's not just me. I think at the company, it is pretty well understood that if it's not fun, it's not going to work.
We spend a lot of effort trying to keep Duelingo enjoyable. This is why we, for example, when we did this thing where you can talk to an AI to practice conversation, you're not just talking to random AI. We have a cast of characters. You're talking to one of our cast of characters. It has a personality. We've just really everything we do. Every time we put something out,
it's ingrained in our thinking that this has to be enjoyable. I spend effort pushing that agenda, but I don't have to do all that much because it's just varying ingrained in the company. How big is Duelingo now? About 850 people. It's everywhere. Is it all in New York? Where are you all living? It's the largest offices in Pittsburgh, Pennsylvania. We have about 400 people there. We have about 250 New York. Then we have offices in a few cities. We have one in Detroit. We have one in Seattle,
one in Berlin, and one in Beijing. All of those have like 30 people in them. One key thing is we are not remote. We got to come to one of those offices. Yeah. Why is that? I just wholeheartedly believe that you can work better that way, especially when it comes to most of what we do. Not 100%, but most of what we do is creative stuff. It's just a lot harder to do so over Slack and Zoom. That worked out for about nine months during the pandemic,
but it is actually impressive how when the pandemic started, we all had to go remote. We execute it pretty well. Towards the end of it, our ideas had kind of run out. We were executing the ideas, but we kind of had run out of new ideas. It's pretty amazing. As soon as we came back to the office, within like three months, you would just see all these ideas popping up. It's because my God, first of all, you can't sit in front of a whiteboard and talk about stuff. Also,
we have lunch together here every day. In the lunch line, you hear people being like, hey, I've been seeing you a while. I thought of saying this to you. It's just something that you would never send a Slack for. I think the combination of all this just makes it a better company. I don't have that much proof, but I am extremely convinced of this. Sundar Pachayat, Google, told me at the very beginning of the pandemic that he was worried
that the company would run out of ideas if they stayed too much along. We know what we need to do for the next turn. I'm worried about what happens in the next turn. Did you have a controversy when you re-implemented returned office ever since? No, fortunately not. Do your employees think that? No, they agree. You know why? We've done a lot of dumb things at Duelingo, but this was not one of the dumb things we did in retrospect. From my first message saying everybody's going to go
work remote, I, in that message, I said, but we're going to come back to the office. I do not want Duelingo to turn into a remote company. We are not a remote company. We kept saying that the whole time. In a lot of companies did this thing during the pandemic, they would hire people all over the world because whatever your remote, we never did that. Whenever we hired people, we would say, I get that you're not coming to the office right now, but your job is in New York.
And we expect you to be in New York. And because at some point, so hopefully soon, we will be back in the office. We just never stopped repeating that. And so by the time we said, okay, time to go back to the office, this was not a surprise for anybody. And I don't think we lost a single
employee for that. Do you think that the markets here in help you with that being in Pittsburgh and Detroit and these swan, like if you were in San Francisco, I think a lot of people would disagree you like that is probably that's probably true that we are not, we are not in San Francisco and that that's probably true. Although the New York office is, you know, is now the second largest and we also didn't lose people in New York. Yeah. Do you, do you find that
people are demanding more flexibility even with a full return to office? Sure. I mean, compared to before, I mean, for example, we do, we're not here five days a week, we're on the office. So we, the way we work is Tuesday, Wednesday and Thursday, you have to be in the office. Monday and Friday, it's optional. What happens in practice is Monday or half the people come in and Fridays is like 20% of the people come in. We are talking on a Friday. And I know what the
top of this. I am here in the office though. I am here in the office. I come. But I would say, I don't feel like I have the political power inside this company to say, all right, people, you got to come in five days a week. I feel like that would, that would not go over well. Yeah. One of the other pieces of the pandemic puzzle return to office is there was a suppression of demand to travel and explore. And I have friends who, for at least from our Instagram,
they just haven't set foot back in the United States in like two or three years. Has that, have you connected to that group of people who want to learn languages on the go? Has the re-explosion of travel had an impact on their business? You know, travel is interesting. But you know, now that we're publicly trying to company people have hypothesis, all kinds of stuff about travel with us. Oh my god, travels opening up that may be good for dwelling or
travels drying that may be bad for dwelling. The reality is travel does not affect us all that much. I can have hypotheses for why that is, but we just have not seen, you know, traveling, closing down or opening up, affecting us all that much. The reason is probably because the majority of our learners, we have, we have a lot of learners that have different motivations. But the two big buckets are not travel. One of them is a hobby. And that's, that's the biggest
bucket in the US. So, you know, you ask people in the US like, why are you using the Olingo? And the most common answer is like, well, you know, I used to play a lot of candy crush or I used to, or I used to do a lot of Instagram. And now I'm doing the Olingo. At least I'm learning some Spanish. Like, that's, it's just a hobby. And then the other huge group of people is people learning English.
And for them, it's not about travel. They just actually honestly got the need to learn English, either because they, you know, it's for educational opportunities or for job opportunities, etc. So, and that, those two big buckets account for like 90, 95% of our users. So travel is just, doesn't affect it very much. Yeah. When you think about the structure of the company and an opportunity like that, you know, we started talking about like, what is the latent demand?
What are people coming to you for? And then there's growth, which is how do we go create some demand? Would you ever set up like, okay, we have, we've got to go do marketing to make travel happen? Or is that just like, not how you think about it? No, I mean, our marketing, we haven't thought about that. I mean, most of our growth is like 90% of our users come in from word of mouth. And that just will keep happening, I think. We also spend very little and marketing comparatively.
I mean, our marketing budget, our entire marketing budget for the whole world. And we really do operate in every single country in the world is $50 million a year, which is quite small for a company with our revenue. But we just, you know, whatever we're doing with marketing seems to be working pretty well. And we don't spend a lot of money on it. Yeah. Actually, I feel like I get to ask you
the Dakota question. So as long as we're talking about your marketing spend, how is the company structured? Ah, yeah. So we have functions. There's like the marketing function. There's the engineering function. There is the, you know, product management function, the sign itself. So we have functions, and each function has a function head to give you a relative idea of sizes of functions. Engineering,
product management and design combined account for about 70% of our employees. The sign is weirdly large for a company. We have a number of employees total is 850. We have about 130 people in the design department. So the sign's large, but we have, you know, engineering product and design account for about 70%. So that's the people working on the product. And now if you look at that group, the people working on the product, that is structured into these things we call areas.
And each area is related to one of the things we're trying to optimize. So for example, with language learning, the main three things we're trying to optimize are engagement. So how fun, dual, and voice teaching better and how much money we make. And so we have an area for each one of these. And then in each area, there are teams and in each team, there are
people that's kind of how we're organized. It's a little bit of a matrix structure. But one important thing that I think has worked really well for us are areas and our teams are not feature based. And what I mean by that in most software companies or most app companies or whatever, you usually have a team for each feature. So like this is the login team we own login. Or this is the, you know, if you have a leaderboard, this is the leaderboard team we own the leaderboard,
we do not have that. Our teams don't own features. Our team own our team's own metrics. So we have a team for subscription revenue. We have a team for daily active users. And they can touch whatever they want in the app. All they have to do is continually increase the metrics. They're positives to this, which is where there is very aligned to metrics. There are negatives in that no team own certain features. So when something breaks, there's a lot of people being like, not my feature. I
don't know. So there's positive and negatives. But in all, this has actually worked out really well for us. This sounds one like a reaction to working at Google where teams do own things like the login screen. And they endlessly communicate how they're going to change the login screen. But it also seems like this might work for a small company where one person can see the whole
product or understand the whole product and how it all works together. And then you're going to get inevitable collisions as two people try to change something to increase two different metrics and different directions. How do you resolve those collisions? Yeah, we're already, we pass the point that there are collisions a while ago. I mean, there's definitely collisions. Sure. Yeah. There's a couple of things that really that help us here. One is every change to the app
passes through this review process called product review, which is not just one person. There's a group of people that have a lot of knowledge about kind of how the whole thing works. So they, you know, they serve us a little bit of a semiforour, a little bit of like, no, you should not do that, et cetera. And then the other thing that is really important is we have guardrail metrics. So here's how it works. So if you're in the team that's trying to improve subscription revenue,
your goal is to improve that metric. But we tell you, you cannot mess up any of the other metrics. So for example, if you do an experiment that improves subscription revenue by a million bucks or whatever, but it decreases daily active users, you cannot launch it. So that has really helped, you know, teams kind of police themselves. They at least won't go do anything that really messes something else up. The combination of these two has helped. You are right that if we've had
probably a hundred thousand employees, I don't think this structure would work. That said, I just don't think that a company like Duolingo, at least with the products that we have, I don't think we need a hundred thousand employees. I mean, I think we're, you know, we'll grow and we'll continue growing. Maybe we'll get to, I don't know, five thousand employees, but I doubt we'll ever get to something like a hundred thousand employees. Yeah. How often do these collisions come all the way
up to you? How often do you have to make a trade-off? Not that often. Teams police themselves a lot. I mean, I do see every single change that passes through the app. I do see that. But I usually am not making, you know, trade-off costs. I'm just the main thing that I'm looking for is making sure that everything's high quality. Yeah. You started the show. You were joking about Founder Road because I called it the co-founder. That's Brian Chesky. He's been on the show. He's
talked a lot about how he refactored Airbnb. So he was the conductor of the orchestra. But do you see yourself in that kind of role that you're the person who can see the whole app that you are the person who understands how all these trade-offs are getting made now? Yeah. It's definitely true. The good news is I'm pleased to come around for a very long time. And our leadership, particularly in the product areas, it's been the same for the last, I don't know, eight years. It's
the same people for the last eight years. So yes, I, you know, I kind of have a view of everything, but the reality is that our head of product, Gem, has a view of everything too. Our head of design, Simmy also has a view of pretty much everything. So yes, I am in that mode, but we have a number of people that could probably play that role as well. How do you, when the three of you disagree, how do you resolve those? You know, the good news is there's little disagreement for a few
reasons. The first one is that we're a very metrics-based company. So usually we just let the metrics speak. You know, if we run an A, B test and the metrics say something, my opinion doesn't matter all that much. Unless it's something that we think is, you know, really like a dark pattern or something, but generally my opinion or their opinion doesn't matter all that much. That's one reason. The other one is we've been working together for so long that we're just pretty aligned
on everything. And then the last thing, if, you know, I have this saying, if we're going to go by opinion, let's go by mine. So, you know, generally when we have disagreements, I kind of see how how deeply they believe in their thing. And sometimes I just disagree and commit, but if we both believe with equal strength on something, you know, I will just go with my thing. That might be the most succinct definition I found about. I've heard yet. What? If we're going to go by opinion,
let's go with mine. Just do what I say is really the answer to what founder about it. I mean, that if that's, but I'm telling you, the majority of things we don't really go by opinion, the majority of things is just by data. So here's the decoder question. It's a good foundation for it. How do you make decisions? What's your framework to make decisions? For the company or for me, and they're similar, but they're not identical. I'm actually curious. Some people don't think
they're different. And some people think they are very different. So answer however, whichever way you want. For the company, the decisions are very much tied to return on investment. Most things are, there's a return on investment calculation. Even if we don't sit there and write the numbers down, there's a, you know, how much how much effort you're going to put in something and how much
you're going to get back. That drives most of our decision making. There's another thing that is not unique to do a lingo, but I think it is not the norm in most companies, but it is what happens to lingo, but it's not unique to do lingo, which is, you know, usually when you're doing a project, there's kind of three things that matter, which is how much does it cost, how fast are you going to do it, and what's the quality. And at do a lingo, and usually the straight-offs
between these, at do a lingo, the most important thing is quality. The second most important thing is speed and the third most important thing is whether we're on budget or not. That is different than many companies and many companies. It's the other way around where it's just the most important thing is budget, then speed, then quality here. Just quality is the most important thing. And so that's another thing in our decision making. That's for the company. For me, it's very gut feeling.
And I have found myself, you know, I used to try to justify that. I have stopped at this point. I'm like, look, this is what I think we should do. I can tell you reasons that I can probably coming up with after the fact, but the reality is my gut says we should do that. And at this point, because I've been at this for, you know, I've been working on do a lingo for like 13 years, my gut's pretty good. It's not 100% correct. Like, surely I make mistakes, but it's pretty good.
And so I mainly do things with gut feeling, and then I tell people, uh, I justification afterwards, but everybody around me knows that these justifications are after the fact they're not, you know, rational thoughts. That obviously works for a startup founder, for a private company. You've been a public company CEO for three years now, three three and a half years. Yeah. Is that working for you as a public company CEO? Yeah, yeah, yeah, because, you know,
that's again, the majority of decisions that we have to make this clear answer. Like the majority of things is just like, well, look, this is going to lose those money. Let's not do that. Um, you know, I've heard this from a handful of CEOs now who've taken our company's public, and now they're on the quarterly reports, cadence, and they have investors on Elliott capital management might show up on their doorstep and be unhappy that you're not marketing to more shiny speakers or
whatever. Like have you had to change that attitude now that you run a public company? No, it's been very fortunate. First of all, we hired an amazing CFO before we went public Matt's Karupa. And fortunately, he kind of deals with most public company stuff. I mean, I, I don't do a lot of that. And I'm very thankful for that because that's just, you know, I, I don't have a finance background. I don't know. I, I have a PhD in computer science. That's what I'm good at. Not, not finance.
So there's that. The other thing is I think we've been fortunate that so far as a public company, we've executed well. So I think that has given us a little bit of latitude in that, you know, we basically don't get asked very tough questions because we've executed very well. I am sure this won't be the case forever. I'm sure at some point we're going to miss a quarter we haven't so far, but I'm sure we will. And when that happens, I probably have to answer some questions that I probably
have to tell people that, you know, sorry, we'll, we'll, we'll be more buttoned up from now on. But so far, you know, I show up to our new schools on a t-shirt. The day you see me show up to our new schools on a suit, you'll know that we fucked up. Yeah. Sorry to get out. You know, I'm so sorry. The other thing I hear from public companies, CEOs, and I'm very curious because it relates to your emphasis on quality first, right? You have a lot of metrics. That means your investors can see probably
a lot of your metrics or demand a lot of your metrics in different ways. Quality is not measurable in that way, right? It's not. And so if you're saying, okay, we're going to invest a bunch of money in AI, right? And we think in this use case, it's going to be really successful for us or we can charge more for it. But we have to spend a bunch of money upfront and we have to wait to make it good. That's not at least in the current market, like a great story for investors. How are you managing
that now? Yeah. I mean, your right quality is not measurable. The way we measure it, not measure it, but the way we make decisions about that is we just have particular in our design department. We just have people who are very much sticklers about quality. And that's we just like, nope, that's just not good enough. That's just not good enough. And so we just do that a lot. And in terms of investment, I mean, honestly, with the public markets, we kind of just don't talk
much about that. You know, we talk about the metrics. We kind of just don't talk too much about how the, you know, turns out we spent an extra month working on this feature just because we didn't really like the way they always animated. We just kind of don't talk too much about that. And I think that's fine. I don't think we're, you know, I think that's fine. But my guess is if we went on earnings calls and spent all our time talking about how, how much effort we put into animating
or how I don't think people would like that. I think I honestly think more companies should spend more time talking about how much time they spend making things good. That would be I think a significant upgrade to American capitalists. Well, I would like that. Yeah, I would like that. But yeah, I mean, but the reality is we do spend, I mean, we spend an inordinate amount of time. You know, if you look at our app, it's over time has just become very animated. We spend an inordinate amount of time
looking at the precise frames of the animation. We're just like, Oh, no, no, see, this is not smooth enough. Yeah. And I'm not just claim that our app is perfect. It's not perfect. But we at least tried really hard for it to get as close a perfect as possible. We have to take another quick break. We'll be back in a minute. This week on Profty Markets, we speak with Mark Zandee, Chief Economist of Moody's
Analytics. We discuss its take on the housing market, the recent jobs data, and how the US can reduce the federal deficit. You allow immigrants into the country, let them do their thing. You do it in a rational way. We need to make sure we have a rational immigration policy and system. But you allow that to happen. We're going to get one to two tens percent more GDP growth every year. And I assure you that makes that the deficit debt problem look a lot less daunting
going forward. You can find that conversation and many others exclusively on the Profty Markets podcast. Hey, I'm Sam Sanders and I'm making a new show with KCRW all about the stuff we obsess over in our free time. We're calling it the same Sanders show. Every week I'll be talking with smart people about the TV and movies and books and music and memes we're loving right now. Join me and get caught up on the fun stuff. I guarantee you a good time. Listen now wherever you get your
podcast. No LV election is nigh. It sure is. Can you name all the swing states? Michigan, Wisconsin, Nevada, Arizona, Pennsylvania, Georgia, and is it Virginia? It is not. It's North Carolina, but you got the others. I mean, I think it's arguable. Let's just say I'm kind of right. Let's give you partial credit. Partial credit it is and we are doing episodes about all seven of the real swing states. Oh, it sounds like I should listen. What are we doing?
Like history lessons? No, the big issues. Abortion, the economy, election, security, all of it. Okay, so you're saying if you listen to today explains seven episodes on the seven swing states between now and the election, you are going to be ready for whatever comes on November 5th. There you go. Today explained wherever you listen, we drop in the afternoons Monday to Friday.
Welcome back. I'm talking with Duolingo CEO Luis Fanon. Right before the break, we are chatting on Duolingo's goal of product quality over quick revenue and how the quote, inordinate amount of time they spend on things like perfecting animation is something they don't necessarily focus, shareholder attention on. But now we need to talk about some other important metrics. One thing that also seems hard to measure or a metric that might lead you in different
directions is how successful Duolingo is. Like maybe the most important metric of all is, are people getting good at Spanish? Are they leaving this experience with the ability to communicate in Spanish? Not just not just know the language but actually communicate. Can you measure that? Yeah, we can. We can, but not as not as effectively as you would like to measure it. The answer is yes, Duolingo works. We have measurements. And I'll tell you how we measure it. Unfortunately, this is the
only way we know how to measure it reliably. It's not that great of a way, but it is it. It is, you take somebody who's just starting Duolingo, you ask them a bunch of questions about their previous knowledge. You also give them a test to measure how much they know. Then you have them use Duolingo for a long time because it takes a while for you to actually learn stuff in the language, like use it for a year, use it for two years, something. And then at the end of that period, you ask them
questions about whether they use other resources. And then you also give them a test to figure out how much better they get. And it turns out that people who knew nothing before and used Duolingo and did not use any other resources. And then take a test at the end, learn about as much, in some cases, it varies in the study, but either the same or more than in a classroom. So we're pretty happy with that. The results actually work. The problem with this way of measuring is that it's
very slow. I mean, you take one or two years for us to get a new measurement. Yeah, I really don't like that, but we have not been able to come up with a better way, despite the fact that we have tried a lot of like, you know, kind of doing things in the app where we're like, Oh, okay, we can do kind of micro measurements of whether you've learned a little bit, sort, et cetera. It's been super complicated to do that and never given great results. So we just rely on these old school,
give you a test. And then it's a pre-test post test method. That's it. That's it. I mean, this is where you veer right into lengthy society level debates about education and how we measure performance of schools and teachers. Do you feel like you're participating in that system? Like you're using their measurements, right? This is how schools do it. They test you. We are using their measurements and efficacy is really important. We spend a lot of effort
trying to make sure that we're efficacious. And you know, the other good news, even though we do this only, you know, the time scale here is years, you can plot how effective dual linguists. And if you look at it over the last 10 years, dual linguists clearly more every year, it is more effective than the previous year, for sure. On this test base. Yeah. Yeah. Basically more people are passing the test. More people are getting higher scores in the test. I mean, basically people
are learning more on dual linguists every year. And there's a number of reasons for that. I mean, because we work a lot to try to teach you better, but it is definitely true. And at this point, you know, when we compare ourselves, we know we are as good or better than a classroom. We know we're not yet as good as a one-on-one human tutor, a good one-on-one human tutor. We know we're not yet as good at that. And but our goal, and I think we can do that over the next three to five years,
I think will be as good as a one-on-one human tutor in terms of efficacy. We're way better in terms of getting you to stick around. Getting you to stick around, we're just way better than that. But efficacy, it just turns out, if you have the money and the strength to continue going, a one-on-one human tutor does better. Yeah. Do you think that there's a conflict between gamification and engagement, the things that you're historically successful at, and education?
Yes, there is. How do you manage that conflict? Very easily. Okay. Always go with engagement. Really? Yes. I mean, presumably you've heard both sides of the argument. Why have you made this decision? I'll give you many arguments, but one that works the most is this. It doesn't matter how effective you are. Can't teach somebody that's not here. And that's it. That's it. That's it. People leave, and then the reality is, it's not always true that engagement and learning
outcomes are at odds. But when they are, we usually prefer going for engagement, and it's just, I'll give you an example. There are some things that are frustrating, and frustration makes you leave. What we do with that is we actually just smooth them. So there may be, and by that I mean, I may be able to teach it to you. If I could force you to sit there, I may be able to teach it to you in five minutes, but it'd be a very frustrating five minutes. Instead, what we do is,
fine, we'll teach you two hours. Just way slower. But at the whole time, things are animating on the screen, and you're getting dopamine hits or whatever, it may have taken you two hours, even though a really good teacher may have taught it to you in five minutes, watching you make mistakes, and it would be frustrating. But we much prefer to keep you around. And part of the reason is because we're
in an app setting, as opposed to a school setting. You're in a school setting. The truth is, the kids are hell-hosted. They sort of can't leave. With an app setting, the tiniest frustration, people are like, you know what, we're going to go to TikTok now. So we just can't lose those users. So we always opt for engagement. But that doesn't mean we won't teach it to you. We'll just smooth it. We'll just take it a little slower. It's clear that you have thought about this a lot.
We spent years thinking about this. I want to round this out a little bit, because you have a very clear answer at a very clear point of view. What do you think the specific tension between gamification and education is? What are you losing when you always pick gamification? Probably the thing you're losing is efficiency. By that, I mean, a amount of content learned per unit time. You're probably losing efficiency. I mean, the truth of the matter is,
I grew up, first of all, I grew up in the third world a while ago. Some of the stuff that I grew up being taught, my teachers were hitting me. I'm not kidding. They would hit me. And their reality is that I probably learned really fast. Because when you were learning penmanship, if you did the wrong thing, they hit you without ruler. You have a real incentive to get that done very fast. You just learn really fast. They're like, whoa,
I think so. I think it's true. You can learn faster if you're in an environment where you're forced to just do so. Nobody cares about whether you're feeling good about it or not. You can probably learn faster. But in our case, I'm okay with slightly slower learning as long as you're still engaged. So efficiency, I understand that one. I had some pretty strict teachers in my time. But I was really good at taking the tests. My wife and I went to college here. She's
much smarter than me. But you're a good testing. But I was very clearly. She was always mad at me. Because I could just show up at the end and take the test. This is like truly, probably why she didn't date me for years. Because of that core frustration. This is what I mean by education. I thought human teacher can evaluate whether or not you're good at taking a test or whether you actually learn something. And that's the trade-off that I was
actually getting at. If it's all a gamification engagement, people might just learn to play the game. They might not have learned anything. Yeah. And there's probably a little bit of that. Very hard to measure, of course. But the reality is, I mean, ultimately, it works. Do a lingo works. I mean, just an example. For me, I've been for the last few years only using the lingo to learn French. At this point, I can watch Netflix shows in French. And with no subtitles,
I just watch it. And it works. So you're right. There's probably a trade-off. It's probably pretty hard to measure. But what we're looking for here is that people are actually using their time well. I want to try to tie all of these kind of themes and ideas together. You have a big vision for doing a lingo. You've talked about it a lot, right? Being available to teach everybody all languages around the world, being a number of countries. And then there's the fact that you're
a public company. If you've got to make money, you're still shown up in T-shirts. The first thing that comes immediately to my mind is you're launching new things like math and music. And they're not available on Android, which is the single most popular operating system in the world. It's used by the majority of lower income people in the world. That feels like an immediate tension, right? The best experience of dual lingo is on the platform that wealthier people use. How do
you resolve that? Yeah, it's a good point. I mean, well, first of all, math and music are about to be available on Android. By the time this airs, they will be available on Android. We are about a year behind on Android. The reality is this has been true on dual lingo almost since the beginning. Android has been six months to a year behind iOS. There's a number of reasons for that, but probably the biggest one is actually it has been harder to find really good Android developers
when compared to iOS developers. There's just more really good iOS developers. We have more of them at dual lingo. The way we work is most new features. We experiment them on iOS. So, because our new feature usually is usually not that great off the bat. You usually kind of have to do trial and error and stuff and try to make it better. By the time it's good, we port it to Android. That's kind of how we operate. We understand the importance of Android. You are
right. There's more people with Android phones than iPhones. But we feel like it's okay that we're six months behind. Generally, all features are going to make it to Android. It'll just be, it's just six months behind. And we feel okay about that. It also, it turns out to be easier to develop on iOS for a number of reasons. Not just that there's more developers. So, that's it. We just ahead up in retrospect, given the technology that there's today, maybe we would be doing something
that is cross platform, we'll redevelop on Web Platforms at the same time. But we're locked in to being native on both. So, our apps are native. We have a native app for iPhones and we have a native AppRendered phones. That was the best thing we could do, you know, whatever, 10 years ago. We're locked into that. Yeah. When you think about growing the company, right, supporting multiple platforms, that's just double the effort all the time. Yeah.
Is that on your mind? Okay, we're going to intentionally slow down development here so we can get this team smaller. It is. And we're, you know, the way we don't have one huge project where we're like going to stop all development and be like, you know what, we're going to now be single platform kind of thing. Yeah. But we are slowly getting there. I don't know how long it'll take. The hard part with this is, you know, you are right. If we were to start from scratch right now, the decision
would be clear. But you also have to, you know, keep the plane going. And it is such a big investment to do this that we will probably have to stop all development for, I don't know, a year and a half or something. I don't even know the timeline because it's just so big. So, yeah. So far, we have not, we have decided to do this piecemeal rather than all at once. The premium subscription tier, the max tier, only just arrived on Android. Literally, like, you know, in the last few days.
Yeah. One thing I've heard over and over again since the dawn of the modern smartphone, is that iOS users spend more money. Yeah. Four times as much. So it's four times as much. At least for dualingo, a given user spends four times as much as per capital. It's four times as much. The majority of your money is coming for iOS users. Is what I'm getting from this? Yes. More of our money comes from iOS than for my, even though we have more Android users than
iOS users. It's just hard to overcome that 4x. Wait, is that, is that demographically there for Xeancome or demographically there's iOS users are spending four times the money? No, it's not for Xeancome. It's a user spends four times as much. Yeah. We have more Android users, but it doesn't balance out in the end. We make more money from iPhone, but that I'm going to give you a number here that is this is very rough and it's not true, but it's just to give you an
idea. The split of users is like 60 40 so 60 percent Android 40 percent iOS. The split of revenue is the other way around. It's like basically 60 percent iOS 40 percent Android. These are very rough, but yeah. So when you think about expansion again, you're a public company. If the most of your users are on Android and Android is the biggest operating system in the world, but all of
your money is on iOS. How do you resolve that tension? Right? Is it just we're going to we're going to make our money in iOS and not because the big mission right is to bring free language education to all of those other people. Yeah. So the iOS users subsidizing the free mission. It's a funny, I mean, it's one way of thinking about it. Well, first of all, I mean, it's not quite true that all our money is on iOS. It's just more money is on iOS. That's for sure. But it is a little
bit true regardless of Android versus iOS. If you just look at our payers, who pays for Duolingo at the moment? There's usually people who are well off and now they may not be millionaires, but there you know, people that live in countries like the US that are wealthy countries that have salaries like a hundred thousand bucks a year kind of thing. Like good, good stable job in a wealthy country. That is who pays for Duolingo. The people who use Duolingo for free are usually
in poor countries. They may not have a stable job, etc. So it is true that we're getting the wealthy people to subsidize the education for everybody. That is that is the case and that'll probably always be the case. Now, on our end, we also need to get better as a business to get more people in some of these developing countries to pay. I mean, a really good example. Netflix has gotten a really good job at getting people in a Brazil or India to pay. We have not done as good of a job. And part
of the issue is that we're free-mium. And again, I grew up in a poor country, in a poor country, even if the price is scaled down to match the GDP per capita, which is much cheaper. The problem that you have is that in a poor country, the attitude is I won't pay unless I have to. That's just the attitude. Like it doesn't matter if it's just a dollar and I do happen to have a dollar. I just won't pay unless I have to. So what you see is extreme tolerance for, for example, we can put
10 ads at the end of a lesson. They still won't pay. And so this is why, for example, Netflix does so well in some of these countries because in Netflix, there's just no free, they're just like, look, whatever, you got to pay. And people are like, fine, fine, I'll pay. So we have to figure out what to do as a freemium product in these countries. And we have some ideas, but the reality is we have not really succeeded at strong monetization in countries like India. We haven't.
Yeah. Do you think of that as the next logical place for you to grow? For sure. Is it easy to think about English education? For sure. We are spending a lot of effort on that. And it is growing, which is nice. But it is, and it's a massive opportunity. I mean, it just language learning is another funny thing where, you know, most markets, not language learning, but most markets, the largest
market is, you know, country like the US, like rich countries. Turns out language learning as a whole, not do a lingable language learning as a whole, the largest market is actually developing countries. Yeah. I mean, the India's Vietnam's, Brazil's, Mexico's of the world. They're learning English and that's the largest language language learning market. But we have not cracked it. We have cracked the smaller one, which is like US and Western Europe and
richer countries. We've cracked that in terms of monetization. In terms of users, we have users, we have a lot of users in India. They just don't pay us. Yeah. I feel like I have to ask you about the owl. It's very important to everyone to ask you about the owl. The owl is like a very, at least as expressed in this country, is like a very online, very culturally defined character. Like, if you took the owl outside of United States
social networks and dropped it anywhere else, it wouldn't make any sense. Is the owl expressed culturally in all the different markets, or is it just one owl? It is, I don't know how to answer the question. It's in between. We started using social media with the owl a while ago. It grew mostly in the US through TikTok because the owl does unhinge stuff on TikTok. Wait, wait. The owl doesn't do anything. How big is the team that writes and performs the owl?
Five people. Okay. And they work at Nualingo. Yes. I'm assuming they're in New York City. No, really. Actually, no. I'm mostly in Pittsburgh. Okay. I didn't realize if it's where you have this money, terminally online people, Godspeed. Yes. It started out with TikTok and it was mainly US. Okay. That was several years ago.
What has changed in the time is, first of all, we're no longer just relying on TikTok. It is now, you know, it's kind of YouTube, YouTube shorts, Instagram, etc. So everywhere on social media, that's one big thing. The other one is that we learn how to localize these two different markets. So, you know, we started a dual-ingo accounts for a bunch of countries. Mexico, well, Spanish speaking, Japan, Brazil, Germany, France, China, etc. And it turns out,
we have figured out how to make all of them succeed. And I was, you know, I was dubious at first. When somebody told me, we're going to open an account in Germany. I thought, you know, no offense to Germans, but I thought these people don't have a sense of humor. But it turned out. They do. They do. And it is, in fact, one of our most successful accounts. They are a little different. So if you look at it, they are, it's not that different, but they are a little different.
And it's, this is the same five people. No, no, no, no, no. So the, we have global team, which is these five people. And in each country, we have a small number of people. It's probably one or two people that localize this stuff. And now, localize doesn't mean we take the exact same videos. And, you know, in Mexico, put a sombrero on that's not that. It's just, we kind of have themes. And we have figured out what themes work globally. And kind of also what things, themes work in certain
countries. And so, for example, the German one, we had a really big thing on October fest. And also, at some point, because, you know, there's this kind of dance club scene in Berlin, I guess, the Al went to one of these 24-hour dance clubs. So, you know, in each country, does different stuff. And it's worked out pretty well. Yeah. What's the hiring process like to be the writer for the duo, Lingo? Do you, do you just make people, do you just repeat
people's Twitter accounts and say, you're, you're unhinged enough to do this? It's a lot of that. It's a lot of that. I mean, it's, and it's a, it's so relatively by now, because we are such a presence online. By the way, I didn't know this until recently. There are weeks when our video on TikTok is the most watched video in all of TikTok that week that happens. And so, by now, our account is so well-known. And so, or our accounts are so well-known. It's said that we have,
we have our pick in terms of, you know, a lot of people want to work for that team. And typically, we just look at what they've done before. And so, we're able, there's, it's a small group of really good creators. And so, we, you know, we hire, we hire from that group. And usually these are pretty funny people that are even funnier online. But when they're offline, they're not as funny. I mean, they're still funny. But when they're offline, you're like, it's you,
it's you who came up with that. And you measure everything, it sounds like, is this working? Are you getting lots of new users because of the out? Oh, yeah, yeah, yeah, this works. About, by the way, this is not paid. So, we make all that, all that, all that social marketing is not paid. It's, we, it's free. It's like, we make our videos and it just kind of go viral. If you look at a number of users that are, or fraction of users that are coming in from social media,
it's about 15% of our users are coming in from social media. Now, if you look at social media views, which we measure in the billions of, of our content, there's an equal number of social media, roughly equal number of social media views of our content versus the content that is about us, but not made by us. So, there's a lot of, also, there's a lot of other people just making content about dualingo, but they're not us. And they combined, how about as many views as we do? Yeah.
Do you, is there, have you ever told the team to pump the brakes? Have you ever looked at something they've made? And so we just can't do this? Yeah. There's some review, there's an approval process. You know, we're, we're close to the line in, in some of the stuff that we put out. And we have, in fact, on across the line and published things that we shouldn't have, we have. And since the, since we did that, we now have a pretty strict approval process. And you know,
this is a whole layer in, and the last step is basically me. But the, basically, stuff doesn't come to me because there's usually, before me, there's the CMO, before, so there's a lot of steps, but what's the last one the CMO was like, I don't know, Louis has to approve this one? I'm trying to remember what that one was there. What's the last one they convinced you to do, even though you were skeptical? I also, I mean, it's been such a long, I'll tell you a couple of
things. I don't remember the exact video what it was, but I know that the last one that I approved I was wrong. As in, I shouldn't have approved it in retrospect. And a lot of this is you don't know in retrospect. You don't know until it happens because you put it out and then you just see this reaction. And I just, I don't remember what it was, but I know I approved it and I know I was wrong. Because I just, I just didn't imagine that it was going to have that reaction. We haven't had that
many full-poss. I mean, it's been like three or four videos that just like, yeah, probably shouldn't have done that. I'll tell you the other thing was we, about a year ago, we had made this crazy video. It was insane. We were, we were a little hesitant about it. And we ended up cutting it. We cut it. The, you know, there's all these memes online about how the owl really wants you to learn a language. And it goes through great lengths, including kidnapping your family.
This was a video about kidnapping. And we were a little hesitant about it. And then the October 7 attack happened. And then we cut it. And then we cut it last year. And when we thought, well, you know what, we may use it next year. This year came along. Again, we cut it. It just, you know, and then we came up with the, you know, internal thing that like a year when we can play that, it's probably been a good year for humanity. Yeah. Yeah. We're probably never
going to play that. The world context of that one needs to be substantially improved, I think. All right, I got to end with a feature request. You've given us a lot of time and then I'll let you get out of here. We talked a lot about India. We talked a lot of it. Emerging languages. Can you put good drop the in the SAP? Oh, man. This is the language that I can understand and speak like a baby, but I can't read or write. And I love to just close the loop. You're asking for
languages. That's a hard one. Yeah. We are. It's a huge language. I know. I know. Native language of Gandhi, the current Prime Minister of India, you know, this is, you know, it is, there is this unfortunate thing about being a huge language versus desire to learn it. It's a pretty big difference. Hindi is probably the one that has the most desire to learn it
in terms of Indian languages. It's a tiny number of people that are learning it. It's got to be, I don't know exactly off the top of my head, but it's certainly well below 1% of our learners are learning Hindi. I'm going to guess 0.1% of our learners are learning Hindi. That's a guess that I have. So that's the hard part about adding languages that we have to maintain them. We have to do a really good job with them. And then in the end, we just don't get a lot of usage.
So sorry. That's a hard now. It's one of the first times the CEO is going to be a hard move. That's the, again, founder bed. Well, it's just really hard to say yes to, you know, in the past, I used to say yes to this stuff. And we made a lot of mistakes adding languages that in retrospect, we probably shouldn't have added. Have you ever caught languages? We have. We caught, I think it was African's. But yeah, we have. And it's, but the cut, it was, and part was because there was very
little demand. But the biggest reason was it was just a low quality course. And at some point, we thought this was a bigger brand risk than anything else. And we made the decision, you know, we're like, well, could we improve it? Or what? And we made the decision that it was not worth improving. Do you think AI is going to help you add languages or keep the quality high? Unfortunately, maybe, but unfortunately, at the moment, AI is really good for big languages
and really bad for smaller languages. Yeah. So there's a pretty high correlation with languages we have. I mean, AI is very good at the languages we have, like the Spanish and the French, not super good at your, you know, Esperanto or Navajo or, you know, kind of smaller languages. Yeah. AI is notoriously bad at math. Is it going to help you? Or at least the current elements are pretty bad math. Are they going to help you with that? The good news is that it
in the constrained environment that we have, it can help quite a bit. It's been helping quite a bit. I mean, a lot of the data that we generate for it are math course is with AI. The other thing is some of it is without AI, but it turns out just computers are good at math. It is funny how many times I ask this question, someone fails to bring up the idea that they're so computer.
I mean, computers are good at math. And, you know, I understand AI is not so good at like, you know, they're like, whatever, follow this pattern or whatever, it may not be so good at that. But the data that we generate for a math course is a lot of stuff like fractions and multiplications of computers are pretty good at generating that data. Yeah. Yeah. All right. Well, Luis, you've given us a lot of time. Thank you so much for being on the code. Well, thank you. Thank you for
for having me and great questions. I'd like to think Luis found on for taking the time to join me on Dakota and thank you for listening. I hope you enjoyed it. If you'd like to let us know what you thought about this episode or really anything else, drop a slide. You can email us at decoderethroverge.com. We really do read all the emails where you can hit me up directly on threads. I'm at reckless 1280. We also have a TikTok. Check it out. It's at decoder pod. It's a lot of fun.
If you like decoder, please share it with your friends and subscribe over here to your podcast. Decoder is a production of the verge and part of the voxmodeodcast network. Our producers are Kate Cox. Next up, this episode was edited by Xander Adams, our supervising producer is Liam James, with the Conan music its web break master cylinder. We'll see you next time.