All-In Podcast
Adam Foroughi, CEO de Applovin: sobrevivir a una caída del 92%, los anuncios como ML 1.0 y el mercado de anuncios de videojuegos de $50B
Pregúntale al tutor sobre este vídeo
Entra con un email verificado y pregúntale al tutor lo que quieras sobre este vídeo.
TL;DR
Resumen de investigación
▶ AppLovin y el mercado de publicidad en juegos móviles
La conversación presenta a Adam Foroughi como fundador y CEO de AppLovin, y el entrevistador pregunta qué hace la compañía y cuál es su mercado. Foroughi explica que construyeron la empresa sin financiación de capital riesgo en sus primeras etapas y que eso contribuyó a operar discretamente. Describe AppLovin como una empresa de publicidad que ayuda a desarrolladores de juegos móviles a monetizar sus productos. Afirma que más de mil millones de personas juegan a diario a juegos casuales móviles y estima que la publicidad dentro de ese ecosistema representa unos 50.000 millones de dólares anuales: calcula unos 20.000 millones en su propia plataforma, a partir de 11.000 millones divulgados casi dos años antes y un crecimiento interanual aproximado del 60%, y más del doble al incluir otros participantes. Según él, los anuncios con recompensas generan atención e intención; la oportunidad creciente de la compañía es aplicar modelos de aprendizaje profundo para pasar de recomendar juegos a ofrecer experiencias de compra.
▶ Modelos de recomendación y anuncios más relevantes
Foroughi vincula la publicidad con el aprendizaje automático temprano y dice que algunas investigaciones sobre modelos de lenguaje pueden trasladarse a sistemas de recomendación, y viceversa, aunque reconoce que sus estructuras difieren. Explica que los modelos predicen resultados futuros y que, en publicidad, el valor de esa predicción puede traducirse de inmediato. Al comparar los anuncios de 2005 con los actuales, recuerda que entonces eran, en su opinión, de mala calidad y mayormente spam porque la tecnología era insuficiente. Atribuye a Facebook el mérito de combinar datos y tecnología para hacer anuncios relevantes. Sostiene que hoy las recomendaciones de compra en Instagram y los anuncios interactivos de juegos pueden sentirse como contenido, y afirma que observa mucha participación en minijuegos y avances porque los sistemas recomiendan opciones pertinentes.
▶ Publicidad, descubrimiento y privacidad percibida
Al hablar del efecto de la IA y los chatbots sobre la publicidad, Foroughi distingue la publicidad de intención de compra, como la búsqueda de Google, de la publicidad de descubrimiento en la que operan Meta y AppLovin. Considera que los modelos de lenguaje pueden competir con la búsqueda para cerrar transacciones que el consumidor ya pretendía hacer; en cambio, descubrir un producto que la persona no sabía que quería puede crear actividad económica adicional. Ante la preocupación de que anuncios aparezcan después de conversaciones privadas, dice que ciertas búsquedas o visitas previas pueden explicar anuncios relevantes, niega que AppLovin rastree ubicaciones precisas y considera poco realista que los micrófonos recojan y procesen conversaciones para orientar anuncios. Añade que las relaciones en redes sociales podrían influir en anuncios, por ejemplo, si alguien conectado busca un producto. Reconoce que la personalización puede resultar inquietante, pero afirma que los datos publicitarios están bastante controlados y que los anuncios relevantes aportan valor económico; sostiene que mejores tecnologías contribuyen a un crecimiento más rápido del PIB.
▶ Caída bursátil, recompras y recuperación del modelo
Foroughi sitúa la sede en Los Ángeles y las oficinas de ingeniería en Palo Alto, Pekín y Singapur. AppLovin salió a bolsa en abril de 2021 con una capitalización de unos 28.000 millones de dólares y 600 millones de EBITDA; llegó a 40.000 millones, pero en 2022 su valoración cayó a unos 3.800 millones pese a generar 1.000 millones de EBITDA ese año. Foroughi atribuye el desplome a una oferta elevada de acciones de inversores previos y poca demanda de inversores institucionales que no habían investigado la compañía; el múltiplo pasó, según él, de niveles que describe como altos a menos de cuatro veces EBITDA. Como respuesta, dejó de dedicar tiempo a inversores y lanzó recompras agresivas: afirma que desde entonces compraron cerca de 6.000 millones de dólares en acciones y retiraron entre el 20% y el 25% de las acciones en circulación; en su máximo, esas recompras llegaron a valer más de 50.000 millones. Para sostener la cultura durante la caída, describe una mentalidad de equipo y un plan de acciones ligado al desempeño para personas clave. En abril de 2023 lanzaron un modelo de aprendizaje profundo que mejoró el rendimiento para anunciantes y aceleró el crecimiento. Foroughi dice que comenzó a hablar con inversores cuando la acción rondaba los 80 dólares y que, durante esa semana, subió a 150 dólares; más tarde describe una subida de 9 a 750 dólares por acción en dos años y medio y una capitalización máxima de 250.000 millones. Advierte que los inversores públicos siguen tendencias y pueden reaccionar tarde, y describe extremos de pesimismo y optimismo.
▶ Regulación de privacidad y venta de estudios de juegos
Sobre las restricciones de Apple y la privacidad, Foroughi dice que unas reglas claras permiten que la tecnología se adapte. Según su explicación, cuando ya no se puede orientar con precisión a una persona, se la agrupa y el anuncio puede ser menos pertinente; afirma que después de cambios de Apple recibió quejas de usuarios que querían anuncios más relevantes y menos spam. Sostiene que los consumidores valoran la relevancia cuando ven un anuncio durante 30 segundos para obtener una vida gratis en un juego, y que las empresas tecnológicas adoptaron las reglas mientras los modelos de aprendizaje profundo siguieron siendo capaces de ofrecer recomendaciones. Preguntado por compras de juegos, aclara que AppLovin vendió todos esos estudios: los había adquirido para obtener datos de entrenamiento, ya que los estudios no solían compartir datos con terceros. Dice que esos datos sirvieron para entrenar el primer modelo profundo; tras su éxito y la llegada de terceros a la plataforma, la compañía se deshizo de los estudios.
▶ Agentes y hábitos de compra
Ante la posibilidad de que agentes y nuevos dispositivos cambien el comercio, Foroughi prevé que algunas personas usarán agentes para optimizar compras repetitivas, como renovar mensualmente una suscripción de suplementos y recibirla a tiempo. Distingue ese uso de las plataformas de descubrimiento y considera que el comprador típico no es necesariamente un usuario avanzado de tecnología. Compara la audiencia de AppLovin con la de The New York Times y señala que muchas personas siguen usando propiedades de Yahoo. Afirma que los compradores habituales quieren encontrar productos, mirar opciones, comparar, seguir el pedido y recorrer la experiencia de compra. Opina que, en una transacción de 50 dólares, ahorrar un 20% mediante un agente quizá importe menos que disfrutar el proceso de compra. Concluye que algunos usuarios tecnológicamente avanzados adoptarán agentes, pero cree que se sobreestima a ese grupo cuando se toma como representativo al entorno tecnológico de Twitter.
▶ Enfoque operativo y competencia con las grandes plataformas
Preguntado por cómo una empresa pequeña pudo competir con Meta y Alphabet, Foroughi responde que AppLovin no considera haber ganado y que trabaja con la expectativa de tener que esforzarse para no quedarse atrás. Atribuye su posición a una organización ligera, especialistas enfocados en la experiencia de juegos móviles y su traducción en comportamiento transaccional, y la capacidad de moverse más rápido que empresas grandes. Su respuesta plantea el enfoque y la velocidad como ventajas para seguir compitiendo, sin afirmar que la competencia haya desaparecido.
▶ Márgenes, ventaja tecnológica y equipo
Al hablar de oportunidades de expansión y fugas económicas, Foroughi afirma que el margen de EBITDA de AppLovin es del 84% y lo describe como el más alto del mercado; por ello dice que no ve mucha fuga en esa métrica. Explica el modelo con un anunciante que compra clientes potenciales para vender un producto: si el cliente compra, el anunciante recupera el coste de adquisición en la transacción y puede escalar el gasto. Define la fuga como no controlar toda la cadena porque AppLovin no es el anunciante, aunque dice que la compañía busca conectar a anunciantes y consumidores y que su enfoque algorítmico y automatización mantienen pocos puntos de fuga. Ante el argumento de que márgenes cercanos al 85% podrían atraer competidores dispuestos a reducirlos al 60% o al 50%, atribuye la persistencia de los márgenes a la complejidad tecnológica, los datos diferenciados, la innovación y la escala del modelo, que puede formar una barrera difícil de superar. Sobre el equipo en China, lo califica de humilde, trabajador y capaz, y afirma que trabaja con algunas de las mentes más brillantes; dice que le motiva rodearse de gente de la que puede aprender.
Transcripción completa
[0:31] >> Please welcome Adam Foroughi. >> Welcome Adam. >> Hey man. Hey Ben. How you doing bro? Good to see you. >> Likewise. >> Adam, thanks for being here. Um, we thought it'd be really great to chat because a lot of people don't talk as much about you're not in the headlines all the time with your business. You're operating your business almost like absent media. You don't do a lot of
[1:01] press. You don't get out there a lot talking about the the company. But it's such an incredible business. Can you just tell the audience kind of what AppLovin is? And maybe also frame up the market a little bit for us? >> Yeah, totally. I think the the fact that we were able to build a very big company without having VC funding at the early stage created this world where we just had to build quietly. And and then obviously the goofy name didn't help us all that much as well. But what we are ultimately is an advertising company that's helping mobile game developers monetize that
[1:31] space. Now what people don't realize is just how big the mobile gaming universe has become. You've got over a billion people a day playing mobile casual games. These are all adults, heads of households, and the scale of the opportunity is just humongous. We disclosed last January, so nearly two years ago, that on our own platform there was $11 billion a year of ad spend. Since then we've grown 60% year-over-year roughly. And so, if you gross it up to a nice round number today, you get 20 billion. Now, we're
[2:02] not the only player in this marketplace. This is a market that's monetized by a lot of other ad companies as well. So, then you'd probably more than double that again, and round it off and say there's probably about 50 billion dollars of advertising being spent every single year in this mobile gaming ecosystem. It was not very long ago that social was a 50 billion dollar opportunity. Space is growing really quickly. Lot of audience. These people watch ads. They a lot of times they watch the ads to get rewards. And so, that dynamic creates this possibility to
[2:32] create intent. For the most of the life of companies' life, we've been creating that intent to drive a user to take one game's experience and go to the next game's experience. And what's really gotten investors excited about our company and just us excited about the opportunity that we we have in front of us is that deep learning models have gotten so so powerful now that you can take that same space and try to take that adult and give them a shopper behavior experience. And that allows us to tap into much larger economies, make
[3:02] more of an economic impact in the world. And that's why our team's just really really pumped up on what we're doing. >> Um the first wave of internet advertising was in many ways the spark for a lot of critical technologies that then sort of diffused out into the world. So, like if you think what Google was able to do with AdWords, AdSense, applied semantics, all of that whole range of technology. Is that true in this generation of internet advertising? Are there are there technologies and things that are being birthed here that are consequential and foundational now to
[3:33] the rest of the internet? >> Yeah, I mean, advertising is like ML 1.0, but really it was the first implementation of all these technologies that now are are driving AI today. And the the economic value of a large language model and what it's doing in our society today is much greater than advertising. But advertising is a very profitable implementation of a deep learning model. Now, recommendation systems are structured differently than large language models, but a lot in a lot of ways they follow the same trajectory. So, a lot of the research
[4:04] that's being done in the space in the large language model space can port to recommendation systems. And vice versa, a lot of the researchers in the large language model space might have started early in their careers looking at advertising systems. So, these two spaces are really related. The nice thing about our business and any advertising business is that when you build a model, you're predicting a future outcome, an advertisement, or if you're building a social network, an engagement post, or a sequence of them. But you can translate the value of that prediction immediately.
[4:34] >> Is it Is it true that there's just a broad-based behavior around humans' reaction to ads in 2026 versus 2006? Like has there been an evolutionary arc that's very predictive? >> Yeah, it's it's interesting. I started my career in 2005, so I saw the ads back then. Complete garbage. It was all spam. And the technologies just weren't powerful enough. And your old company, Facebook, did a really good job of realizing if you can take all the data we have in front of us and pair it with good technology, the ads
[5:04] can become really relevant. And if you talk to most people who shop today, most of their shopping recommendations are coming from Instagram. The ads have become very much like content. And in our domain as well, people love the ads that we show. And you would think people wouldn't like them, but we see tons of engagement on little mini-games that are appearing in other games, and people are playing these previews because the technologies have gotten so good at recommending something relevant to someone. >> There's been a lot of hand-wringing about the impact AI will have on
[5:35] um the ad networks, specifically Google's interface. And OpenAI has an ad product now. I'm sure you've been monitoring it uh and trying to learn from it. What is advertising going to look like when people are doing five or six queries with a chatbot? Because it's pretty obvious 95% of the world are not going to pay 20 bucks a month for this technology. They're going to expect it to be free. ChatGPT has already said they're going to make it free. Tell us what they're doing in advertising and is it going to be less effective each time,
[6:06] but in aggregate people are going to use it more or is it going to just be even better than Google Search's franchise? >> Yeah, I mean there's two sides of advertising. One One part of it is bottom of funnel advertising where a consumer sort of knows what they want to buy, but they're doing research to go complete the transaction. And that's Google Search business. If I wanted to buy a pair of dress shoes, go to Google historically, do some research, and they direct me to where I need to go based on the ads. And today you can go to a large language model and close the loop on that same thing. So that ads model is
[6:37] almost going to exclusively compete with the Google Search business. What we operate in is a world where we're showing a user an ad and we don't know what their intent is. So we're trying to create something that that didn't exist before. Show them a recommendation and get them to go, "Wow, that looks really cool. Let me go transact on that and do it really quickly." That That's what drives Facebook's ad business, too. And so the reason that's interesting to me is that the transaction via search or LM was going to happen anyways. If the LM didn't exist and Google Ads had never
[7:08] come to existence, but Google Search existed, that transaction to close loop would have happened. So there's not actually a whole lot of economic expansion that happens from that. But when you show a consumer an ad for something that they had no idea existed, they didn't know they needed to buy >> Discovery, basically. >> Totally. Complete discovery. And this is what makes Meta so amazing in their ad business and what we aspire to do. You create that discovery moment. Not only is it a really fun moment for the consumer cuz then they're excited about what they bought, they wait for the package, they're excited to open it up,
[7:38] but you create economic expansion. >> the >> [clears throat] >> the arms race that develops over time where some people say, you know, I mentioned something with my friends at lunch and all of a sudden I show up and there are these ads on Meta or wherever. Is that just us overreacting or is that is that actually happening? And is there a push not to be more um not intrusive, but you know, like the tendency to want to sell more and
[8:09] >> That's creepy. >> Yeah. >> It feels creepy when it does happen. >> to push the boundaries. And I'm like, what what is actually happening when people say I say something at lunch and all of a sudden an ad for that same thing appears. >> I mean, I I think you've done other actions that are [snorts] trackable. Like do a search, browse a website, do a product search and you don't realize it and then you say something related to and you start seeing ads that are relevant. >> like the mic is on or there's an app that has actually taken snippets. >> though that if we were all at lunch, especially with these apps, you know our geolocation, you've kind of put us into
[8:39] a group. So we might be talking about this new car we're all interested in or watch and then Friedberg, when he's leaving, searches for the watch to bookmark it after the conversation. But you're tracking all of our locations and then you say, "Okay, let's give all four of them the ads for the watch." And you're mix and matching >> Is that is that what's happening? >> That's what I'm told is happening. >> I don't think advertising companies can track locations. So we don't track location at all. It's a really heavy concept to track people's precise location to then run run an ad. And then
[9:10] to imagine the amount of data that's transferring if your mic on to then parse the mic on content to try to translate to an ad, not realistic. >> about us being friends and being connected together? Like groups. >> So we wouldn't have that data, but if you're on a social network, of course, your relationships together might drive an ad experience. If Chamath searches for something, then you might see something relevant to it. There's nothing wrong with that. I mean, the one thing that people lose, there there's a creepy factor that scares people somewhat, but all of the data collected at this point given the scale of
[9:40] advertising across all these companies is pretty much controlled in a lot of ways. What people then forget is the economic value that's created from these ads becoming that relevant. That ad that you saw, you recognize that ad. 20 years ago you would not have recognized the ad. And there's a big part of GDP that's now coming from this digital ad economy. The better these technologies get, faster GDP growth. >> Adam, let's just go back to the cuz what I find so fascinating about the business is the
[10:10] the way you've operated it. You're you're based in LA. Is that right? >> I'm based in LA. Company started in Silicon Valley. We're in Palo Alto. >> Palo Alto, but you're here. And then you have a lot of developers in China, is that right? >> We have our engineering offices are in Palo Alto, Beijing, and Singapore. >> And then the company didn't raise a lot of venture money. You take the company public 2021, it went public. Like 20 billion market cap out the gate? >> Yeah, we were COVID IPO. We went out in April 2021. It was about 28 billion
[10:42] dollars. >> 28 billion. And then in 2023, what did the market cap collapse to? >> Well, this is the funny thing about the public market. So, we went out in 2021 $600 million of EBITDA, $28 billion market cap. We got as high as 40 billion. And then in '22 the stock went down literally every day. We got to about a $3.8 billion market cap. In that year we did a billion dollars in EBITDA. >> That's incredible. So hold on. So so so so so let's just go through this. So the market's in disbelief for some reason
[11:12] about the business. And what do you do? >> Yeah, what what you learn pretty quickly and I'm a finance background, so I had good education on this, is that your price in the markets is determined by the quality of your investors. And we had private market investors both private equity and and ex-cofounders and other team members that were going to sell when we went public. And because there were so many companies going public during COVID, by the time we went out, blue-chip investors weren't doing the research to figure out what is this goofy-named company. So, we ended up with no demand and a lot of supply. And
[11:44] that construct created this world where we just tanked. And multiple went from fairly high. I mean, I wouldn't really value companies on 50 times EBITDA, but to something that was absurdly low, sub four times. So, being that finance-minded person, you have to remember with that kind of a bashing, you do have an opportunity on the other side of it. And I turned internal to the team and said, "I'm not going to talk to investors at all anymore. They're not buying our stock. It's a waste of time. But guess what? We generate a ton of cash. Let's start
[12:15] buying our own stock. Let's become our best investor." So, we kicked off a super aggressive buyback program. And in over the Since then, I think we bought roughly $6 billion of the company's stock. Um retired 20 to 25% of the shares outstanding. At peak, that $6 billion was worth over $50 billion. And so, you can take that moment, which does feel super depressing, and turn it into a huge opportunity. >> you manage Sorry, did you feel that way from the like the whole time? Or was there this period of depression where you're like,
[12:46] "Oh my gosh, what are we going to do?" >> what I was going to ask. How do you manage the internal culture when the stock is off 92%? >> It's tough. I mean, like I'll tell you like I would get phone calls from family members, friends, "Are you suicidal?" And I'm like, "Look, uh we we got stock at a penny. The stock's still like 10 bucks. It's still up a lot." But it's very tough then cuz you realize as a CEO, your team is getting those same phone calls from their family members. >> And they don't have the the gravitas that you do, nor the ownership. >> So, so we built it by just saying,
[13:17] "Look, it's an us against the world mentality." Like, everyone's turned against us. We're going to buy back shares. And we implemented a performance stock plan, which typically goes to CEOs, but we did it across key people in the company and said, you know, we understand it's tough right now. We understand you thought you had a house and now you don't. But if you dig in and we recover, you're going to make a ton on the upside. >> And then what happened? Investors started showing up and saying, "Hey, we're paying attention again?" >> It was interesting cuz for us what happened was we went from ML 1.0, like
[13:48] we talked about a couple minutes ago, to ML 2.0. We we went from a regression model to a deep learning model and the outcome was we're driven by our advertising algorithm. The better it works, the better advertiser return is on our platform and everything is performance based. So, we're selling revenue to advertisers the more they scale. And so, the company just started growing really quickly. Now, we turned into 23 and launched that model in April. We still weren't talking to investors. So, people hadn't found out and then it was
[14:18] somewhere around, I think September of 23 that I went to New York and I said, "The stock's now like 80 bucks." And we'd recovered quite a bit cuz the performance was good, but I said, "I'm going to start talking to investors cuz market cap's getting high enough and we can't really buy back all that aggressively anymore." And in that week, the stock went from 80 to 150. And I think it was like 28 billion to 55 billion. >> in New York? >> From me just going out and saying, "Hey, our company still exists. We survived this >> EBITDA grow. >> Yeah, and then people like I'd sit in
[14:49] the meetings and do like it's pretty easy to read the other side of the room if you you you do that kind of thing to sell your company. I sit in the meetings and I'm like, these people are literally calling their friends in the room going, "Buy buy buy buy buy." And I was like, "Oh, this is >> And then what's the what's the opposite side of that? Once they're long, are they now asking you, "Okay, Adam, how do we expand? How do we grow faster? Why just games? Why not e-commerce? Why not this? Why not that?" >> Absolutely. >> Damned if you do, damned if you don't. >> So, unfortunately not a lot of people are contrarians. So, you can go to the extreme down and then on the other side of it you can go extreme up, too. We
[15:20] ended up going from nine to $750 a share in a matter of two and a half years. And so like it was like a $3.8 billion market cap. Some people bought some options back then and were probably living in some massive homes. And we got to $250 billion market cap. So, extreme on both ends. We've now settled into a place where we have a lot of excitement about our growth opportunities. But I found public market investors are not all that different than private market investors. They follow trends, but a lot of times later than you'd want. The most sophisticated
[15:51] hit those trends early. And that's why you have you have really good VCs and you have average VCs. You have really good public market investors. You have average public market investors. >> could talk a little bit about privacy, Apple, the EU really are look at companies like yours, um and they this is a little too aggressive in terms of the data you're collecting. Some video game developers don't like having data collected on their users. And And they've tightened the screws a bit. Zuckerberg had to deal with it specifically. So, what's the
[16:21] headwind on this business? And And how do you manage privacy when Apple really is trying to let's call it what it is. They're trying to neuter your business. >> Well, like in any of these spaces, you want the regulations to be clear. So, once they're clear, technology can deal with them. And so if you could precisely target a user five years ago on iOS and today someone says, "I don't want you to precisely target me." You group them in a in a bunch and you serve them a worse advertisement. Now, the funny outcome of that is we'll get a lot of complaints
[16:51] after that change that Apple made from users that say, "Serve me more relevant ads. You're showing me a bunch of spam." So, there is this notion that you need privacy regulation so that technology companies can do exactly what's expected of them. On the other side, consumers do want relevant ads. It helps them discover products. If you're sitting there in a game and you're watching an ad for 30 seconds to get a free life, you're getting something that has monetary value. Now, if you're doing that, do you want to sit and watch garbage for 30 seconds or do you want to watch something relevant?
[17:21] And so, what's what's happened since a lot of the privacy noise is a lot of calm. The rules were written, technology companies have adopted, and deep learning networks are really powerful. They can do it. >> with the consumer? I just give me one quick follow-up. Yeah. With this amazingly profitable business, I think you dabbled in buying some of the games, and we have um Bending Spoons coming on today uh to talk about their aggressive acquisition of not bad businesses, but let's call them slower growth businesses that maybe venture isn't um
[17:52] interested in. Is that going to be a sustainable plan for you to become a game studio, and does that put you in conflict with the with the partners? >> Yeah, we we sold all those games. We bought them originally as a data play. We when we built our first deep learning model, we needed to have data to train it, and game studios don't tend to want to share data to third-party companies. So, we bought our own studios, we seeded the training data in our first model, we built a model that was really successful in market, we started growing really quickly. Once we started doing that,
[18:22] third parties were coming in, and I was and we divested that. >> What's the world of advertising look like where um there's agents everywhere, agents are servicing you, maybe the human interface to compute changes, so it's not necessarily a computer that you're typing on or a phone that you're browsing, maybe it's the meta glasses or some other device. Um what role does the ad play, and what happens in this agentic commerce that so many other people are trying to now push into existence?
[18:52] >> Yeah, I mean I I think the reality is part of the world will start using things like agents to optimize certain shopper behavior that's consistent. For instance, I might put my supplement subscription into an agent and have it optimized every single month and deliver on time. But, these discovery platforms aren't that, and the typical shopper is not the person who's deep into agents and sitting on Twitter and and adopting the latest technology. I sort of say like our audience is a New York Times audience. There's still a ton
[19:24] of people using Yahoo properties every single day. The typical shopper wants to find a product and wants to actually go through that shopper behavior. They want to window shop, they want to go through the transaction experience. >> to compare probably. >> Totally. They want to track it. And if you told them after the fact, "Hey, an agent could have done this for you and saved you 20%." I don't think that matters on a $50 transaction because the dopamine hit from going through it is what they enjoy. So, I I think there is this part of the world that is technologically advanced that's going to adopt these technologies. I just don't
[19:54] think I think we really over index on the Twitter verse and forget that the average shopper is not that. >> Yeah, let me just try and understand how you won because um the two of the smartest companies in the world with the best engineers, Meta, Alphabet, Google make most of their revenue from advertising. They they've built their own models. They've been doing it now for one to two decades. Um how did a small company compete in this particular domain and win? And what's the the
[20:25] operating model that you think gives you an advantage to continue winning? >> Yeah, I mean here's here's something that that helped us get to this point. We never think we won. We think every day we wake up and we're probably going to get screwed right now and we better work hard. And so, uh you've got a company that's lean with a lot of subject matter experts who are really really focused on this thing, this mobile gaming experience and translated it into transactional behavior on the other side. And so, I I I think there's this ability to take on giants if you're
[20:58] very focused, you remain lean, and you can just move faster than them. And >> What's the leakage in the business then? So, meaning when you look at a P&L, you know, we did this I did this thing with Amazon a decade ago where it's like you look at all of these places in which they were leaking and our big insight is well, they just absorb these things and they'll become the new businesses and that was our long thesis for Amazon. >> Yeah. >> Um what's that version for you? Like there there must be is it payment infrastructure? Is it uh other kinds of things Jason asked you about apps, but I guess you've divested that. So, where is
[21:29] the leakage or said differently, where's the opportunity for margin expansion so that people underwrite this thing? >> Well, our EBITDA margins I think are number one in the markets 84%. So, I don't know how much leakage we have yeah given given the metric, but the the way to think about it is advertiser comes into our platform and they have a transactional model. Let's say they're selling lipstick. We give them an arbitrage. They buy the they from us are buying the consumer. That consumer transacts and they cover the cost of the consumer immediately. So, the consumer buys a lipstick for 20 bucks. They pay us less than the 20 bucks minus cost of
[22:00] goods sold. They're happy. They scale up. And that performance model is very scalable. Now, our leakage is we're not the full chain. We're not the advertiser in the equation, but we want to power the advertisers to meet the consumer and we've run extremely lean and been so algorithmically focused and automation focused that we haven't had a lot of points of leakage. >> The other side then is when you have 85% EBITDA margins, people say wow, they could be over earning, right? That's the classic phrase and then you have competition that says I can compete
[22:30] Adam's margins away. I'm willing to do this at 60% or 50%, but sort of maybe as a corollary to David's question, that hasn't really happened and it's been incredibly sustained. >> Yeah. >> And why do you think that is? >> Because these technologies are really complex and if you can innovate and you have differentiated data, you can build an advantage. I mean, by that token Anthropic shouldn't be running away with the large language model space, but the power of a model that then reaches a point of scale and gets adopted by a large scale community becomes something
[23:00] that is a a moat that is hard for other people to overcome. >> And uh talked to us about the the team in China and how big of an edge these folks are. >> I mean, Chinese people are very humble. They're very very hard working. They're very sharp. And if you can work with them, whether out of China or United States or any other part of the world, you're working with some of the brightest minds in the world. And so uh when I started the business, one of my goals of this company was just work with great people and figure things out. And so I'm when I sit in a room with some of
[23:31] the people on my team, I know I'm probably the dumbest person in that room. And that gets me excited to show up. >> Mhm. >> Got him. >> It gets me to show up for the pod every week. >> that? >> Yeah. >> Works for me. >> That was me. >> All right, let's give it up for Adam. >> Yeah, Adam. Thank you. Thanks, bro. That was great. Thanks, man. >> [music] >> Thanks, man. It was great to see you.