Good morning, partners. Hello, everyone. Good morning. Welcome. Welcome to our Capital Markets Day. I'm very happy to be here with you. I am Fabricio. I am Prosus CEO. Do you notice the difference? Not the new CEO. It's one year. One year old. Everything changed. We had time to do everything we wanted. Let's stop talking about this new thing. I'm Prosus CEO. I'm here for one year, and it's almost a CEO birthday. And it's a perfect day to do a Capital Markets Day. We are celebrating one year in this new position. On Monday, we talked a lot about our past. We talked about the numbers, the results, the EBITDA, the EBIT. I'm happy about that. We grew a little. We grew to $440. I'm even happier that it was much more than my original guidance. But that's the past.
We are here today to talk about the future, to talk about what we are going to build. I am very excited to what we are building. I am really, really very excited. Not because of the results of the last year. This is just okay. Because the company is ready for a new phase of growth and innovation. That is what we are going to talk about today. Our agenda for today includes talking about—I will start here with who I am, just to quickly remind you about how we think big, but at the same time, we have a lot of discipline. We are going to talk about building ecosystems, lots of details about that, about innovation, and our numbers. The most exciting thing is that I am just the guy here making presentations. I am going to give you an overview in the first 40 minutes.
Everyone really delivering that is here today. A tech company is all about people, great people doing great things. They are here. We have people from—if you like the iFood results, do not believe what Fabricio said. You have Diego here. I can see him. Diego, where are you? Diego, and we also have Gonzalo from Despegar. You can ask him directly. Please, he is here, a little hidden there. Go talk to Diego. If you want to talk about innovation, we have Euro here to talk about that. We have lots of people from our European team, like the famous Christian. Where is Christian? Christian is a rock star, so probably you know him. He is here. You should meet him and ask everything about OLX. We have the people from eMAG and Iyzico. Many times we do not talk a lot about eMAG and Iyzico.
Very unfair. The companies are doing great. I can see Tudor here. Where is Orkun? Here. Everyone is hidden there. Hey, you should come here. They are shy because we just started. Go talk to them. We have lots of people here. Our Indian team, Harsha from Swiggy is here. Anirban too. That is the thing for our day. Talk to you all. I will be here until the end of the drinks. Today is the day to talk and to share and to drink together about what we are building. That is the agenda from my intro today. I invite you, please scan this Tokan QR code. You have the full day agenda, all the presentations. Vitti from People and Culture is going to talk today too. It is going to be an exciting day. Here you have the full agenda.
You also can talk to Tokan, so you can ask him questions about the presentations, the orders, the numbers. Tokan knows the numbers better than Nico, so you should ask Tokan all the numbers. We are going to use the Tokan data to share with you what you want to know more and to keep sharing more data with you. That is my agenda for today. Hope you are excited. I will be here the whole day. Let's start this conversation and let's work together to create a $100 billion bigger Prosus. Let's go for that. Who am I to start? I can ask Rox to put the slides full screen here. Who am I to start? I'm Brazilian. I look like a Dutch, I know. I live in Amsterdam, but I'm not Dutch. I am Brazilian. I'm from Bahia. Hope you know that.
If not, you should go to Bahia and Brazil. I started studying computer science many years ago. I'm 48 years old, married, and with four daughters. I started studying computer science many years ago in Brazil. I did management in FGV, Stanford, Harvard. I created this company called Movile that has more or less 10 businesses in Brazil, like content and payments and many businesses. We also had a food delivery business called iFood. I was the CEO of iFood for the last many years. You can see I could deliver your food if you are visiting Brazil a few years ago. iFood grew a lot. iFood is a very nice company. I heard it's even better today with the new CEO. Let's see. Let's see. He's the next speaker. I got a call from our board, our Mr.
Chairman, to say I'm partner for Prosus for 15 years. So I'm not new anymore. I'm CEO, but I work with Prosus for 15 years as partners. Our Chairman asked if I would be interested in coming here to run the whole business, the whole Prosus. My reaction is I'm not like a locator of a portfolio that just do the investments and keep doing. I would love the opportunity to create one of the biggest tech companies in the world outside the U.S.. I think this is an amazing opportunity, an amazing platform. I moved to Amsterdam, and here I'm in Prosus because that's how I see Prosus. Some of you see differently. That's how I see Prosus. I know Prosus from 15 years. Prosus started as a print company in South Africa.
It completely reinvented itself to a pay-TV, then a mobile network. It became globally known because it helped to create the social network vision of the world today through the Tencent investment. Look, we are not the biggest company in the world today yet. Think how many print companies from 100 years ago, you know, that reinvented like that and are successful as we are today. I don't think you know any of them. I don't think they exist. Pay-TV or mobile networks, you know. That's what I love in Prosus. This is a company that has innovation in its DNA that's reinvented itself many times. I came here with this big dream of we can keep reinventing ourselves to create the most amazing company, internet company outside the U.S., obviously. We have a trillion-dollar business in the U.S.. They are amazing.
I'm sure you all have hundreds of billion dollars invested in the U.S. company. Congratulations. They are amazing. You also need to invest outside the U.S.. There is a clear space and opportunity for real big tech outside the U.S.. That's my dream about Prosus. Prosus reinvented itself many times, and I'm very excited about our next stage. We are going to talk about that today. Moving on. The reason I believe so much in Prosus is this. It's not because of the $450 million EBIT. It's not because we have only cash, but it's because of this combination. We have 40,000 people all around the world, mostly outside the U.S., 100% of them outside the U.S.. We have 10,000 tech people with $11 billion in cash.
I know Microsoft and OpenAI and Google have more cash and more tech people, but I do not think outside the U.S. and China you have any platform with this capacity of creating, innovating, and pushing new products and services, starting with a 2 billion customer base. I think this platform is amazing. I think we are just getting started, and we have everything to do. Not because Fabricio is smart. I basically do presentations, but because we have hundreds of talented leaders, entrepreneurs that are very passionate about Prosus. You are going to listen to—you are going to listen from many of them today. Please make the difficult questions. That is why we are here today. We really want to share what we are doing with you. That is why I am excited about Prosus.
I'm excited because we are building something amazing, and we are going to explain to you what we are building today. The foundation for what we are building is culture, investment events. People don't talk a lot about culture. I'm sorry. You are going to hear about that today. I think it's a big mistake. That's the foundation about a company, how we have good people moving fast, innovating, taking risks, solving problems, and creating the future. That's the biggest transformation in Prosus over the last one year. We have a culture that is very entrepreneurial today. That is ambidextrous. It talks about results, but also about innovation that reinforces people, reinforces our impact. You are going to listen about that from Gustavo Vitti. He's our Chief People Officer. Hope you enjoy to know more about our culture.
If during the presentation you think, "Why they are talking about that?" that is because I'm confident we are going to keep growing, winning, and creating the next phase of e-commerce. We are going to talk a lot about culture today. We move on, and we are going to talk at the same time. You talk about culture, innovation, thinking big. I want to reinforce to you, our way to operate is thinking big, innovation, people, but also growth and discipline. Not because I am the new CEO, I have to say that. I'm not the new CEO anymore. That is what I learned in my 20 years running Movile and iFood. We have to think about the future, but we have to deliver today real results.
At the same time that we talk about disruption and thinking big, we have a very disciplined management model where we tell the whole 30,000 people, not include just ETFs, not closed. I can say that I can do that with them. We align 30,000 people to the same direction, to the same priorities, to think together how to create a winning company. That is very powerful. That is what makes our results improve. Because now I have the capacity to influence 30,000 people, direction to collaborate. That is what Prosus is doing. A detailed management model focusing growth and profitability, not only on innovation and thinking big. Our culture reinforces all the time. We call it internally face the brutal facts. Hope you read Jim Collins' book, Good to Great. He has many great books. He called it the Stockdale Paradox.
We talk about that all the time internally. The Stock Day of Paradox means keep optimistic that in the end you prevail. Every day face the most brutal facts of your current reality. That is our culture. Fabricio is optimistic? Yes. I am a tech founder. I am optimistic. Every day in the morning I wake up and say, "Fuck. We are not doing good enough. We are not moving good enough. We did not solve our problem in this county, not here. Let's talk about that. Let's point fingers to everything that is wrong, starting with myself, and let's fix it today." The culture is being optimistic, but very, I could say, direct, disciplined, brutal in talking about problems and solving problems fast.
Many of you actually say, "But are you going to confront the brutal facts and sold a few businesses and crystallized value, as you like to say?" We sold around $2.6 billion this year. More than that, we closed many markets where we are not winning. We exited businesses where we are not winning. We made trade-offs. We said, "We are going to win here. And if you are not winning in those other places, we are going to either change management or deprioritize or sell or exit." We are here to win, and we are going to make all the hard decisions, the difficult decisions to keep the focus in our winning objective. I have to tell you, Fabricio talks about that, but there are many leaders here that took many very difficult decisions, like closing markets, firing hundreds of people. They are here.
Please talk to them about that. Dreaming, but delivering. This ambidextrous culture is the base of what we do. We dream big, and it means we want to be the biggest company in Europe. We want to be the largest European company. Today, we value $122 billion. ISMILE is a little more today. I do not know how much it is, but it is like $130 billion. Good start. We want to be much more. Our big goal was to start at $170 billion. I am talking now much more than $200 billion. We want to build this company. We are going to move fast and empower people to get there. How are we going to do it? Two things: ecosystem and innovation. I am going to talk briefly because today I have all the owners of these projects to come here and talk.
Just to give you this overview about how everything fits together. First, we are building the number one lifestyle e-commerce company in Europe, India, and Latin America. Unlocking AI-first world, what it means. We are not investing in one single company, and we believe that they can do magic by themselves. We believe working in an ecosystem, we can do magic. With cross-sell sharing management model, people, talents, technology, AI, data, in a specific region, we can be the best, the number one in that region. That is what we mean about number one lifestyle ecosystem. We also do, specifically in eMAG, delivery of boxes, but our focus is not boxes. It is what we call lifestyle e-commerce. You can buy food. You can buy even boxes if you deliver it fast because it is more like a service.
I need a perfume to go to a party now or to give a gift, then I ask and I receive it now. You can pay your things quite fast. You can buy travels. You can buy events. It is more like digital items or items that you receive very fast instead of selling like a TV, for example. AI-first. The world changed completely. My master thesis is about innovation dynamics. I could talk about that for two hours. I am not getting details, but the world changed 300 times over the last 300 years with many disruptive waves from computers to internet to so many waves. The biggest wave is not the internet or smartphone. It is the next five years. It is AI, and its impact changing all the companies.
The only way we are going to be a best-in-class, a biggest company in the world is if we are leading that globally. I'm not talking about making a $10 billion check. I know some of you are U.S. investors. The U.S. investors are richer than us. We are not that rich. We are very disciplined in allocating capital. Winning in AI is the only way to dream about building a 10 times bigger company. We do not start from scratch. We start with 2 billion customers. This is so valuable. This is more valuable than you think because we are in a world where to train amazing AI models, you need a lot of data. A small startup can compete against me? Yes, but I think we are very fast in innovation.
A small startup does not have 10 trillion tokens of data that helps me to have the best models. That is what we are doing. Moving on, because I told you I will be fast in this opening, we are going to do that through ecosystems. What does it mean? In the foundation, food and fintech. Why is this the foundation? Because this is a very high-frequency business. This is the business that people use or buy one, two, three, four, five times per week. It is very different than regular e-commerce where people buy once every quarter or every four months. Even in China, and I think China was the best place in the world to have this kind of ecosystems where they call it super app there, where people can do many things in one place.
China needed the payments and the food to have even WeChat and the Tencent Alibaba ecosystem running faster. We are doing that. We are using food and payment as the foundation to build on top of that e-commerce and experience services. How do we leverage these synergies? User engagement. We can increase user engagement. We can reduce cost of acquisition. We can reduce costs. We can increase growth through cross-sell. We can use data to understand the customers, serve them better, and sell to them more often. I really believe in that. I think the big benchmark is China. We also can see that when Uber in the U.S. or Facebook or Microsoft or Google has a very complete ecosystem. We do not have similar big tech outside the U.S. and China. I think we are doing that.
If you think this is too abstract, we are going to give you more details about how we do that in Latin America. I want to spend one minute with you asking you to reflect on that. What we have in Latin America now, we completed the Despegar transaction. Very few regions in the world have one player with the market share and the scale we have doing all of that together. Obviously, in China, we have it. In a few markets, mostly smaller than what we have, but we have a very strong food delivery player, very, very, very strong with like 70 million customers, a payment infrastructure doing credit and payments. On top of that, e-commerce through OLX, Simpla, the number one events company, Despegar, the number one travel company. Very few regions in the world can implement this ecosystem.
I believe there will be a lot of value you take from here. We are going to implement that. We are implementing that through our loyalty program, Club. I'll not talk more because Diego is going to come here in half an hour or in one hour and tell you how it's happening and what we are doing right now. We completed Despegar acquisition just one month ago. We are launching, I think, three to five. I don't know how much was already launched. Diego is going to tell details. Shared experience in one month. I am very confident we are moving very fast and we have amazing results here. This is the basis. The second big thing I really believe we are going to unlock Europe potential. Look, Europe is more or less the size of the U.S., okay?
10%-15% smaller in terms of population or GDP. Look, our tech infrastructure, our tech sector in Europe is 10-20 times smaller than the U.S.. The six biggest U.S. companies are $13 trillion. The six biggest European companies are like $500 billion. That is a very big opportunity. I think the market here has amazing potential. If you can move very fast, be aggressive, innovate, and invest to win here, that's our expectation. We are going to do that. I'm going to talk more about that soon. I deleted the third slide, but besides these two, we have also India. In India, we have again a food delivery leader, Swiggy, payment infrastructure, PayU and Vastu, Meesho, the leader in e-commerce, Urban Company, and Rapido. I think we have a very special set of companies, not assets, but amazing companies that working together can create something very unique.
If you don't believe me, okay. You are going to listen today to all those leaders talking about what they are going to do in these areas. Hope you get as excited as I am. Moving forward, investing and having customers and businesses that today are successful is good, but we only lead if we can create the next generation of services. We are going to do that through AI. After me, we will have Euro talking about what we are doing. I like this slide very much just to make you dream with me. Most of the value in most of the internet services like clouds or everything we have before are in the application players like Salesforce or all the application players. In AI, all the value today is on NVIDIA and one or two other infrastructure players.
I believe in a few years we will see $1 trillion or $2 trillion of value there in the application layer. That is where we are investing. I put there the DeepSeek moment to remember, to remind us all. When DeepSeek arrived, everyone thought, "Oh my God, this is not a one company show. Everyone can embed that in their services." There will be transformative things happening. We are doing that through the investment in our own company. We are very good in investing in Tokan internally. Euro is going to explain more, or to invest in our 800 models, or to invest in startups that can help us grow faster. I like it very much. We are also being, I would say, frontier, cutting edge in saying how we train, how we fine-tune open-source model to have knowledge embedded in the model of our 10 trillion data points.
You may say, "Fabricio, tell me more." No, I will let Euro tell you more. If you have a very difficult question, you should not ask me. Euro is here. You should ask Euro. Save the difficult questions for him. I'm kidding. I also can answer, but it's much more fun if he answers. Moving on. I really believe in the impact of innovation. I really believe in the impact of AI. I don't think we are overhyping AI. To be honest with you, I think some investors are overhyping the investment in AI. I think AI is going to create trillions of dollars of value. That doesn't mean we should invest trillions of dollars today. We should build the investments over time as new services are created. I think Prosus is very well positioned to explore that.
In Latin America, we are the big one that can invest there. In India, we are the big one that can invest there in Europe. And we can invest in local companies. We can invest in local talents. I'm very confident we are going to leverage this opportunity. Just to finish my quick introduction, also we are very focused on results. We did $6.2 billion last year. Fun. We are going to do between around $6.3 billion-$7.5 billion. We are giving you a little more, how can I say, guidance ahead. Our expectation is to double the revenue between 2025 and 2028. It means around $12.5 billion. I want to tell you a funny story. Someone told me, "Don't tell this story," but I'm sorry, I can't control myself. Originally, it was a range here from 2 to some other number.
We said, "Let's take the bigger one because sometimes we look too optimistic. Let's put the smaller number." This is the smaller number. Why to me is it important to say we expect it to grow? Because during this one year when I was a new CEO, I'm not anymore, every time I say we are doing good, many of you that write about us, you say they are doing good. Can they keep it? Or are they going to go down 50% by next quarter? I read it like 10 times. I said, "We are doing good," and the report is maybe it was one thing and it's going to go half next quarter. Because of that, we are saying, "Look, we expect it to keep growing substantially." As I told you, my real goal is bigger than that.
I'm trying to say we don't expect to go down by next quarter. We are confident the company can keep growing. Why? We'll have six hours about that in the next presentations. Hope you enjoy. If you have difficult questions, we are here to answer those questions. Hope you get as excited as we are. Moving forward, I know there is someone there. I think he is thinking, "Okay, everyone can do growth if you don't deliver profit." The guy is Brazilian and he's a tech founder. I'm sure he doesn't know what profit is. You are wrong. Yes, we know what profit is. We started last year. I'm showing here EBITDA also for the first time in the main presentation. A funny thing, you know, we always reported EBIT or trading profits and EBITDA. Usually, we put EBITDA in the attachment. The last slides are the reconciliation.
I'm showing here today both of them. Then you still have the last slide reconciling the two numbers. The $435 million is equivalent to $655 million in EBITDA this year. We expect it to go to around $1.15 billion, so $1.1 billion-$1.2 billion next year. This year that we are doing now. Our ambition is to keep increasing more than three times, more than three and a half times over the next three years, including some M&A. When I told you many times, forget measuring us in millions. We really expect that we are going to be talking about billions in EBITDA and also EBIT of Prosus over the next few years. We have a detailed plan that we are confident about how to get there. This is not the most aggressive number. This is what we really expect to deliver.
We do not want to show a year-by-year plan for the next few years. The reason is, I do not think the problem is Prosus, to be honest. The world is so uncertain now. Like, you know, even today we check the news to see what is happening. I do not think it is good for Prosus to give a detailed plan for a few years. What I want to show you is we are not playing to say we have $500,000,000 in profits. We are playing to have multiple billions and we are going to deliver that. If you say you do not believe, at least in the first year we deliver, let us go for the second one. In the first few months, we are very confident with the numbers. Let us keep reporting to you and writing letters about what is happening. We keep our buyback.
Our buyback is the biggest one of the tech companies. We already bought back 29% of our company. What means, in my opinion, we believe so much in our story that we keep buying ourselves because there is no better investment than Prosus. This is a message for you. We are on discount today, so you can buy a very good Prosus share with 30% discount. Great opportunity. Do not lose it. We are going to keep our buyback. I want to finish. My intention was to talk briefly because we are going to see everything I show in details now with this story. When I was a new CEO, I arrived here one year ago and I participated in the last meeting like that, not CMD, but the last results. People were a little, how can I say, nervous, upset.
They told it to me very directly, let's say this way. I heard a lot of constructive feedback from you. I told Eoin, "Why is the discount so big? Why are they so unhappy?" They said to me, "Yeah, some people make this joke. We are 10- cents ." I think it's a bad joke because I don't believe that. I see us as, "Wow, we are going to build an amazing company." One year after that, I made a slide about this bad joke. That's what Eoin explained to me one year ago. Some investors say we take 10 cent dividends, $560 two years ago. We have our losses around $200. We have our corporate costs, $160. There is something there in the end, and then we try to distribute that if we don't spend everything.
They say that's why there is some discount because we are expanding. Also, he told me that is also because of the tax impact. Actually, we have like $120 billion of tax shield. I don't think that's because of the tax inefficiency on taxes. That's what I heard one year ago. Yes, I am optimistic about the future. Don't forget the stock DAYO paradox. In the end, we will prevail. That's how I see Prosus. Not like that. These are the current numbers that we released this week. Tencent dividend is 32% up last year, right? Which is great, in my opinion. Our e-commerce EBITDA is 100% up, which, in my opinion, again, is great. Our corporate costs are more or less the same, a little smaller. Our dividends increased by 100%.
In my opinion, instead of saying Tencent minus , you should say at least Tencent plus. Because look, where is the minus here? Everything is up, up, up, up. This is thinking small. What I'm building is Prosus Plus. Yes, we have a Tencent. It's nice. It's good to be a big Tencent shareholder. Just to remind you, we made this slide also looking for three years ahead. I said, "They are not good." You are a little too pessimistic. You're not going to believe in three years ahead. Let me show you what we are doing this year. This year, Tencent dividend is up 24%. I think Tencent is an amazing company. It's the leader in the biggest markets in the world, innovating a lot.
I am very bullish that Tencent is going to keep playing very well. It's up 24% this year. Now that they just released the data, it's public data. What I'm announcing today is we expect at least 83% improvement in EBITDA. Some people believe in more, but let's say at least. Our corporate costs, we are going to reduce at least by 10%. Because we are a big company, big companies are always too rich, and we have to keep reviewing that all the time. Because of that, we already announced dividend. I already told about that. The dividend increase last year was 100%. In my opinion, we have a lot of opportunity to grow. Can be dividends, can be buybacks, can be investments, smart investment with good IRR.
What I'm trying to build is we are Prosus, and Prosus is a great investment. Plus, it's big shareholding on Tencent. If some of you don't believe, come back here next year in the next capital markets day and let's keep keeping track of our results. Some of you are thinking he's optimistic. Yes, I don't know very successful people that are not optimistic, but we also deliver. You are going now to listen from lots of our leaders why the vision I just showed you is going to become reality. Thank you, everyone, for coming. I come back in one hour for a Q&A. I come back in the afternoon for Europe. We have many Q&As during the day, like five Q&As or six. Hope you enjoy the day as much as me.
Don't forget, we are just getting started and we have a lot of amazing things to do. Thank you very much. Let's go. I'd like to come to ask here, Mr. Euro. Euro, please come here. Thank you. Let's go.
AI at Prosus isn't an experiment. It's live, embedded, operational. Every product, every platform, every move smarter than the last. This is intelligence as infrastructure. Tech that connects across markets, across brands, across teams, from lifecycle management to personalization, from fraud signals to real-time routing. Every click sharpens the system because innovation doesn't sit in one corner. It runs across the entire ecosystem. What a customer triggers in Brazil can optimize a system in India. A learning in Romania improves a product in South Africa. That's the power of connected intelligence. The power of full loop feedback systems in motion. Not just insight. Operational action. Shared. Fast. Intelligent.
This isn't siloed AI. It's a learning system. And it's already igniting growth. Powered by Prosus.
Good morning, everyone. Good morning. My name is Euro, like the money. It's a real name. But I'm not going to talk about money. Nico is going to do that. I'm going to talk about how AI powers our next wave of innovation. I'm going to talk, in fact, about three things. The first thing is how we got here. We started investing in AI seriously in about 2018. So how we got here. And the second part is going to be how are we going to reinvent e-commerce and how we are going to reinvent our workforce. These are the two pillars of our AI strategy. But let me start first how we got here. So back in 2018, we were a small team, maybe 60 AI engineers globally.
Now it's 10 times that. Over time, we have increased capacity. We increased the number of models in production. We prepared, I think, very well for the next wave of innovation. At this moment, we invest close to $100 million in AI. It's in talent, modern development, technology infrastructure. That enables us to bring in production about 800 models. Each one of them either reducing cost or increasing profitability market share. We estimate that collectively, we have a return between 2-5 times. For every dollar we invest in AI, the return is 2-5 times. That is one of the reasons why the results we've seen recently are those they are. This is a contribution to that. We also invest in our internal, the way that we work, how we can make AI improve the efficiency of our operations.
We released an agent, an assistant for everybody in our group. Everybody uses Tokan. That's the name. You also have a version of that in the CMD application. And that's powered by Tokan. About 15,000 colleagues use it now, out of 24,000 at this moment before Jet and Despegar. We are closing to saturation about everybody. We estimate that Tokan has saved about 1.7 million hours in the last 12 months, which is about 1,000 FTEs. It means that we have done the work of 1,000 FTEs extra without hiring 1,000 FTEs extra. The way we operate is as a tightly connected network of AI specialists around the globe. Our hub is Amsterdam. We have development centers for AI everywhere in the ecosystem where we operate: Latin America, South Africa, and India. They tend to specialize on their particular business. São Paulo is for food delivery.
Buenos Aires is for travel. We work very close together as if we were one. The best practices in one company are shared as fast as we can to other companies so they can get things done faster or make fewer errors in that process. It is a tightly connected network of tech centers with over 750 specialists. Just a few more details of what that means. Typically, everything that we do either saves costs or increases some form of profitability. Here, you got four examples. iFood uses AI to reduce the cost of customer acquisition or reacquisition, about 32%. Meesho uses AI to automate as much as possible customer support. You see they automate it from about 90%. By the way, it is more or less the same everywhere. Customer support is one of the areas where there is a significant degree of automation.
It's just one example. OLX uses artificial intelligence to get more value from every dollar of advertisement. Here you can see that for every dollar of advertisement spent with AI, we get 17% more. It is actually higher than that, but let's say 17% more. iFood, again, increases conversions on the homepage 23% before and after AI. These are two four of 800 models. Each one of them is in production because it improves on some metrics. Each one of them is there because it adds value. We also invest in AI-native startups. What's interesting here, independent of the name of the startups, which are exceptional for what they do, but they are in these two areas. On the one hand, AI workforce, and the other hand, improving or changing or reinventing e-commerce.
The way that we look at these startups as an engine for accelerating what are our pillars. On the one hand, e-commerce reinvention, the other one, AI workforce. That is what comes to the rest of my thought is where are we going to do? Where are we going to be? We believe that we can reinvent e-commerce through two things: AI agents and large commerce models. That is going to go into these two directions. On the one hand, e-commerce reinvention, and the other one, reinventing our own workforce. Let me start with e-commerce. This is customer-facing, outwards facing. We believe that the future of e-commerce is going to be predicated on agents and large commerce models. Agents are technologies that not only answer; you ask something, you get an answer. They do things for you.
This is something that was not viable until two years ago, business viable about one year ago. Now it is getting adopted at scale. Because of that, a few things are going to change. First of all, new interfaces are going to be possible. WhatsApp, voice, any type of interface, it's going to be viable. It can be these glasses that eventually are going to be the ones from which you initiate a transaction. That's because agents and LCM empower that. It is going to be everywhere because once you have that, it doesn't have to be a browser, which is the place where you buy. Any digital surface is an e-commerce surface. It can be, again, the glasses. It can be a kiosk. It can be the car. It can be a video. Anything which is digital becomes an AI interface. Extremely important is the ability of personalization.
Everything that you have seen in e-commerce until now was based on similarity. You buy this because somebody like you likes it too. But we never really understood intent properly. I think we are at the point where we start understanding intent and the reason why we buy, which is a core component to make the experience hyper-personalized. Of course, once you understand intent, the ecosystem makes more sense because you can serve that intent through a portfolio of offering. Intent understanding, it's powered by the ecosystem, and you power back the ecosystem. Finally, it's going to be much more autonomous with entire tasks that are going to be executed in the background for you. You do not have to go and search for something. An agent will do it for you. Let me focus first on the large commerce models.
Fabrício already stole my thunder, but just a bit. He anticipated this structure here. Let's go back a few years. We, and when I say we, I mean the AI ecosystem globally, have taken everything we could find on the internet, books, everywhere. We use that data to train models for language, and these models predict text, predict the next word. That's how ChatGPT and Gemini and all the others have been created. It's only simple text prediction, but extremely powerful. You have, I assume, everybody of you have the same experience. Now, a couple of years ago, people figured out that systems that reason performed better. What does it mean, reasoning? Instead of just giving an answer, it gives you the chain of thought that leads to the answer. That reasoning part makes systems better. How do you train for reasoning? Many ways.
Through coding, because coding has a lot of logics inside there. Through acquiring data explicitly for that, through human judgment. Then we train with that, and you have seen an example in DeepSeek, and you have agents. Agents are predicated on the ability of this model to think. What the agents do, they take your questions, they break it down in pieces, and for each piece, decide which tool to use to execute the task and then give you the answer. This is the state of the art for everybody. This is what we do. We take the same technologies that we have used in the past year, and we train them with e-commerce data, with transactions, and we make these models understand e-commerce. They are the brains of e-commerce. We use our own data, which is terabytes of data.
We use open-source models in this case, and we call these tools large commerce models. Let me give you a little bit of additional, let's say, insights about how this works. This is the strategy that we use to train the models. 99% of what matters is not in this slide because these are all the specific technical problems that we solve to get to that place. The logic is like this. We take transaction data and we create user insights and product insights. This is the same data that we have for a long time, but with intelligence, we are able to transform that data into something that makes more sense. It is the same as telling you temperature, humidity, and wind speed. Or I tell you good weather. It's the same thing. There's no difference. But temperature, humidity, and wind speed may mean nothing to you.
I can transform that in something that matters, that means something for you. This is exactly what we do with terabytes of data. We create user insights, which are transformation. You will see it in a second of what the user likes, does not like, and the way it behaves. We do product insights because in the same way, assume that you are a vegetarian and you want to order some vegetarian food. You will never get a pizza margarita. I am Italian. A good pizza margarita is a great choice. The reason why is that unless the restaurant has put the tag vegetarian on the dish, the machines do not know that. These machines understand that the ingredients of these things have to be such that this dish is vegetarian. Simple example, but it is the same idea, product insights.
We transform the data into something that makes more sense. Then we do something else. With that data and with the past transactions, we put a model in between and we ask the model, why did the user do that? You have the things that happen, you have the results, but why? We call that rationals. That is where the model starts thinking. It starts learning to think. Once we have connected this information, this information, we train our model. Our model is based on open source and is trained on all these product insights, user insights, rationals, and we call it LCM. LCM is designed to understand e-commerce and the behavior of users. How? One of the items that the LCM does, it understands users 10 times more in depth.
For every user, we have clicks, payments, searches, anything that the user does, and the LCM is capable of telling you preferences, routines, eating patterns, health preferences, social habits, and something which is beyond what you have trained the system for. It's probably easier if you can try it out. If you could show my screen here. Thank you very much. What I have here, a chat message. What I'm going to do, I'm going to talk to the LCM. I'll do it in a very slow manner. I'll do it one message at a time. You have to consider that this tool, in fact, is used in the backend of our systems, and it does it millions of times every minute. We do one question at a time very slowly. First of all, let me go here. I have a user here.
I tell you what the user insights are about. About this user, this is what we can get from the data. Lives in a high-income neighborhood with a likely household of two and a cat. During weekdays, lunch tends—sorry. During weekdays, lunch tends towards practical dishes and so on. Looking at delivery analysis, we see the dietary preferences, the top categories of dishes and so on. In analytics, we can see time when the user orders, payment methods, order location, and so on. This is what I mean with user insights. Let me try one thing now. Let's ask the same based on what we know about this user. What are they like? What are they more likely to order for lunch on Monday? Pasta or sushi. Let's find out. What happens now is start reasoning about what he knows about the user.
He starts reasoning about this dish, food, or pasta and starts looking at the various processes. Based on the user profile, they are more likely to order sushi for Monday, which for an Italian is disappointing. Still, this is the best option for this user. Why is that? Let's go back a second. It looks at the weekday patterns. It compares the characteristics of sushi and pasta, and then it looks at Monday patterns in particular, and then it reasons on that. Based on this analysis, the user is most likely to do this. Okay, let's continue. Another thing that we do frequently, and it's one of the cornerstones of advertisement and marketing in iFood, is we send push notifications. These are very short messages that the user receives on the phone, and it's supposed to trigger some activity.
Now, these messages work if they are very targeted at the right time with the right message. Let's do it. Prepare a push notification for sushi, I think, on iPhone, suggesting restaurant and dish. Make it short and compact. Make it engaging. All right, so what did I do? It looks at the preferences, it looks at Monday lunch, it looks at what it means to do something on the iPhone, and it gives me the notification. Notification is in Portuguese here because it is designed for a Portuguese user. Notice one thing, let's say sushi is very maintenance and so on. I just noticed that without wasabi. Why is that? It is without wasabi. It has to be somewhere in the thinking process. There you go. Sushi preferences, they consistently decline wasabi. All right.
That makes a lot of sense because think about 20% of users in food have specific restrictions, preferences, and all these kinds of things. The ability to pick in that specific one makes a difference. Let's do another thing now. I'm going to take a restaurant. Let's see, what do I do? I pick one of those available here. I don't know them, but it's just to pick a restaurant. I'm going to ask if the user is going to like this restaurant on a scale between zero, do not like, and five, like a lot. How much do you think this user will like this restaurant? This user will like this restaurant. All right, what it does, it has to go now in the restaurant. It has a bunch of facts about the restaurant.
It looks at the specific menu of this restaurant, and it says, well, it's not five, but it's a good bet. You probably should recommend this restaurant. The reason why is that, because given the characteristics of the user, given the characteristics of the restaurant itself, it does a number of these things. The user never ordered here. However, they do empanadas, which seems coherent with the other things. The premium pricing might be acceptable. The user tends to do combos. You can see where it goes. It understood the characteristics of the user as it matches the characteristics of the rest. Now, let me try something which is maybe strange here. This model has been trained on iFood data. It does not see anything else. In Brazil, we have Simpla. Simpla is our event company. Thousands of events of any type, concerts, local events, big events.
It would be nice if I could say something about preferences for events. Let me try. Based on what you know about this user, could you make an educated guess on which of these two events the user could like? Sorry, typing mistake. A jazz improv show or a local Brazilian band? I need to be gentle in asking this because otherwise it comes back, hey, be gentle here. I have not seen anything about concerts. It is going to say, let's start thinking. First, I need to understand the user overall preference and lifestyle. It tries to extrapolate from the behavior on, let's say, the food preferences and what can be generalized about that. After thinking quite a while here, it is coming back with, see, it has to go through a lot of thinking. There we go.
The user profile, they are likely to enjoy both, but jazz might be more appealing for a Monday lunch. Hey, right? However, they also appreciate traditional cuisine and so on and so on. Now you have seen, all right, here I push the user to go for an event in lunch. Makes no sense, right? But you know the meaning. What this tool has done has taken characteristics of the food behavior and extrapolated something which is beyond that. This is what we're really looking at, exploiting, using, because the more you train across properties, the more you know about both. The more you extend what you know from one property to outside, that is going to impact the others, it's going to be beneficial for the other side as well. Let's go back to the slides now.
As I said before, we use proprietary data, our own data to train LCM at high quality and low cost. We use open source for that. Why do we use open source? Because we want to have total control of our data, and because open source, it's only about 3% on average inferior to closed source. We can get an open source model pre-trained. It's about 3% inferior to the best models out there. The other thing which is really important is this. We made this test. We trained our own LCM based on Qwen. This is an open source model. It has 32 billion parameters. We trained on our insights and our rationals. Then we took the best model out there, O3 from OpenAI. It's the biggest reasoning model available at this point in time. We exposed this model to our own data.
We didn't give it the rationals. You can see this. They are the same. This is a massive model, extremely expensive. This is a small model that we trained. Besides, the small model here, it is 20 times less expensive and five times faster. That's the reason we try LCM. Let me now see how this is used in practice. I want to look at three things. First of all, LCM, when we apply it in operations, we just started this process. We expect it's going to improve the way we do e-commerce in three main ways. It improves the way we do it now. The notification that I've shown you before has already been tested, and these are the results. It doubles the efficiency. We are testing, we are deploying the same application on iFood in concierge.
That means what people see, and we see an increase in conversion. The next one is going to be this one, and you will see the results over summer. Let's do a very quick check on ordering some food on iFood. I'm not sure if I'm going to receive a phone. I think I will.
A nine to time, please. Thank you very much. I'm live now. It is 6:00 A.M. in the morning in Brazil, and I had a long night. I would really love a gourmet burger. The reason why I'm on WhatsApp is because in many geographies, WhatsApp is the platform of choice for everybody. It's not the web, it's not the application, it's WhatsApp. The second thing is the reason why I'm here is because I want to show an agent. In WhatsApp, since it's an open interface, you can't really know what's the next thing the user is going to be. Therefore, I need an agent behind that listens to what I'm doing and decides what to tell you as I'm doing it. All right, let's see what it says.
It says, "You are in the breakfast sheet right now, but yes, I already found the burger P6." That sounds good. Show me options. Show me options. At every stage of this conversation, the agent has to find out, first of all, it has already found out who I am and what my preferences are, but now it has to find out all the other things that it could offer me. Through conversation, I am going to go into the options that I have, into the preferences that I want to do. Oops, sorry, it gave me something different. Oh, it gave me breakfast things. All right, all right, that's fine. Would you try searching for burger? I could not find dishes matching your search for good burger right now. That might be a problem because it is 6:00 A.M. I do not have breakfasts which are open.
I say, "That's all right. Why don't you suggest something that I still like?" I go with voice, and it's going to understand what I'm saying, and it's going to look for something that I probably like. The role of the agent here is that of connecting what they know about me, the user insights. It can still give me something which is close to my needs: chicken segment, wrap, omelet, and bacon. It went away from suggesting something which is sweet to something which is savory. At this moment, I could continue the conversation, find out what are the things that I want to do in toppings and all this kind of stuff, put an order, and continue. There are two things that I want to emphasize here. First of all, your experiences in WhatsApp.
You have a platform that people use for everything: talking to the doctor, paying at the bank, and so on. It's the platform of choice. Second, it is agentic. Therefore, it understands what you need to do. All right, let's go back to the slides. Once you have that in place, this is what we expect is going to happen. We are going to train the LCM here on billion-rich data from our own platforms. It's going to be food delivery, initial, instant commerce, cars, everything. By the way, I'm going to show examples about food because everybody relates easily with food, but this is applied everywhere. It's not a food platform. Once you have this platform there, then agentic applications are going to be available for everybody. All right, that's about the first part. Let me change completely gears and talk about an AI workforce.
The AI workforce, it's about empowering our people to enhance the productivity. We started a while back with an assistant. Back in 2022, actually back in 2019, we started testing with OpenAI some of the models that they had at that time, GPT-2 and GPT-3. They were really bad, and they were really promising at the same time. We said, "This thing is going to improve so fast. We need to pay a lot of attention." We started investing. In 2022, we released Tokan to everybody in our group, and in 2024, we made it agentic. At this point in time, we are looking at how to introduce this AI workforce. What are the results so far? It is used by 15,000 employees. 24,000 is the benchmark, if you want. We are growing towards that point. 15,000, 60,000 use it 20 times per day.
It means that they change the way they work. Tokan is like ChatGPT or Gemini and so on, but it is designed for us. It has access to our data. It knows the way we work. And those that are the super users, they create 1,500 versions of these tools for them, for their workflows, leading to this amount of hours reclaimed. There are three top use cases: software, customer support, and market analytics. One of the use cases that we use most, if not most, in any case, which is very common, it is data analytics. If you think about this, everybody in your organization will love to have, will need to have access to data. Not everybody in the organization can code or write SQL. What we want is to give access to data in English. That's the data analyst.
How does this work? If you go here, we need to choose a space, which means a specific colleague, and that colleague is this one. This colleague knows about food and is going to give us answers about everything it has in the database for food. Since part of that information is proprietary, I'm going to ask something which doesn't have to be proprietary. Let me ask this one. Visualize the % traffic. Oh, sorry, I'm not doing it. Visualize the % traffic share of the top five food delivery companies in the U.S. over the last 12 months. The first thing it has to do, it has to go out in our data assets and needs to figure out which database do we even start looking at.
Once it starts looking at and finding the right databases, it figures out which tables and which columns. The first thing after that, it interprets the 12 months. After it interprets the 12 months, here it says, "I've got full data until May." It's going to create the query and it's going to execute that query. The reason why this is interesting is because it gives access to everybody to data in English. I have written something which is only English. I haven't asked anything. Now it is doing a bunch of stuff out there. It might take a while because it has to do the calculations, but this is what it comes back with. It's a diagram with the market share of these five companies. We use this a lot.
We use it for acquisition, M&A, analyzing, let's say, performance of various organizations. I analyze our own performance. What's important, you do it. I say, if you can speak English or Portuguese, you can do it. That's what makes a big difference. Let's go back to the slides. Can I have the slides back, please? Anyway, I can anticipate what is going to happen next. We plan to use these tools and become our own helpers, their own senior colleagues, which are AI employees. We try, we try, we plan to hire the largest AI workforce in the industry: AI data analyst, HR office, AI Excel expert, and so on. We do that before the end of this year. If there are three things that I would love you to take back, are these.
The first one, we know how to create value with AI. We have prepared for this for years, and every 800 models do this every day. We are developing a unique foundation for the next type of e-commerce, and we call it large commerce models. The third one, we believe that the AI workforce is going to supercharge our productivity. With that, thank you very much. I'm now going to leave the floor to Gustavo Vitti, and he's going to talk about our Prosus way. Thank you.
Culture is our edge, our engine. It's real. It's operational. It's evolution at full speed. Prosus, are you ready? I'm ready. I'm ready. I'm ready. I'm ready. I'm ready. Very much ready. I feel ready. I'm ready to go. I'm ready. This is the Prosus way, like Darwin on a jet ski, powered by Prosus.
Go od morning. Good morning, everyone.
My name is Gustavo Vitti. I'm the Chief People Officer of Prosus, completing one year right now. I arrived at your gathering with Fabricio. Prior to that, I was the CPO for iFood since 2019 and 2020. I'm here to talk about and share more about the Prosus way. Fabricio spoke a little bit about it. First time that a CPO is presenting here at CMD for Prosus. I am very excited to share more with you. If there's one thing I wish everyone here can remember after my presentation, there are three things, in fact. The first one is we firmly believe that the right set of behaviors, which we call right now the Prosus way, are our secret recipe of success. The second thing is a strong culture can connect people and can connect our companies. Therefore, it enables our ecosystem strategy.
The third thing, it aims also to change people's mindset from + 10% increase to 10X increase. Let me bring you to this journey and how we crafted that, how we implemented that, and some results and the next steps. When you look for most great companies that are there for many years, including Prosus itself, they refreshed themselves. They reinvented themselves or they reinvigorated themselves many times. When I look for Prosus, we started as a printed media, and then we changed for something, and then we changed it again, and then we changed it again, and right now we have this new strategy. The Prosus way is exactly this refresh we're having in our culture, reinvigorating a few behaviors and traits that we lost in the past and adding a few new recipes as well. Moving along, how was this journey created?
First thing, we look internally for our companies and we benchmark it. Which companies we have in our portfolio that are succeeding for a very long period of time and how do they behave? We look at externally as well. What are the best companies out there that we really admire? We put all of those behaviors in a culture. We sat together with the CEOs of the largest company we have in our portfolio, and we were polishing and crafting, polishing and crafting in a way that is very simple and easy for our employees to understand. It is very, very actionable. Then we came with a culture tsunami across the company, communicating over and over again and embedding this on our mechanisms. The question is, what is the Prosus way?
Fabricio mentioned briefly, I'm going to spend two minutes here, but basically, it's a set of five behaviors that we firmly believe that if our employees, they have that, this is the blueprint for our future success. The first one is entrepreneurship. We firmly believe that companies that are out there and they are creating lasting values, they have this entrepreneurial spirit. It's scrappy. People are easy to work with in ambiguous environment. They're not looking for stability. They're dealing well with discomfort, which is a rare trait for human beings. The second thing is focus on results. You might be a great speaker. You might be very good at PowerPoint. You might come like in a very nice background, very great pedigree. What really matters is your results.
If you are a nice person, but you do not deliver results, you are going to be our friends, but it is going to be out there, not in the company. Third thing is innovation. We want people to innovate regardless of the function that they are. We want people to innovate over and over again. This creates a kind of momentum for the company where everyone is innovating. They are not only innovating, like trying to create a new plan, six months planning, and then implementing something new. This is what we call idea implementation. Innovation is testing, failing, testing again, adjusting, iterating over and over and over again, and being easy to talk about, "Okay, this is not working right now." Yeah, but it is going to be working. It is going to be better in three months. It is going to be better in six months.
Having this confidence that innovation is there. When you combine results and innovation, this creates an ambidextrous organization. This is what Prosus is all about right now. The fourth value is people. We want people to be more comfortable in dealing with failures, success, collaborating, working together, sharing more information. We want to have the best A players out there. Last but not least, we want to make sure that our business, they are not only profitable, as you saw here, but we also are great society members in the communities we are around. We have a lot of drivers that depend on us. We have a lot of merchants that depend on us, a lot of customers. How can we use education to help them to get the best versions of themselves? This is pretty much the Prosus way.
The next step is, how do you bring this culture to life? Our brains, when they're thinking about something, usually we spend a lot of brain power to think about something and do something. We thought about having some mechanisms and the right mechanisms that facilitate our leaders to really live those values. Because in the end of the day, a culture is not what you have on the wall, but how your leaders behave in the organization. We have four main mechanisms that we spend a lot of time and we put a lot of effort to make sure that they were solid and they were easy for our leaders to follow and to implement and to live them. I'm going to go in the next six slides a little bit. I'm going to deep dive in some of those initiatives that we implemented.
The first one is ways of working. This is something that we took a lot of time and we put a lot of effort. If we want everybody to go in the same direction, you need to be clear with that. We communicate over and over and over again. We have monthly firesides right now where we discuss what is going well, what is not going so well, what are the problems that we're facing, and we keep track of things that are our priorities right now. The second thing is something that we call strategy share. Strategy share is an event that we do with every single employee that we have in the company and the leadership of the companies in our portfolio. It happens twice a year, and we spend the whole day discussing what are our main priorities.
Fabricio mentioned before, one slide with our main priorities. What are our main priorities? What is doing well? What do we learn in the last six months? What are our priorities for the next six months? What are the challenges that we have? We have brutal transparency to all of our employees. We also have the Prosus Way Awards where we recognize and we celebrate people that really live our values because people are going to mimic that. This creates a ripple and lasting effect throughout the organization. Who you recognize? Okay, I understand what is the kind of profile that we recognize inside of our organization.
We also have the Leadership Summit where we put more than 50 CEOs of the companies in our portfolio and the companies that we invest altogether, not only to connect, but to talk about culture, to talk about innovation, and to talk about AI. We improved our management model. Instead of having a long period of time planning and then implementing for a year, what we did was we split it into two halves. For one month, we discussed what are our priorities, what should not be our priorities, what are the projects that are going to be dropping, what are the projects that are going to be investing 80% of our energy. With that, we created a strategic map. In one page, we have every single priority within the organization. What we do with that, every month we track the results.
Every single initiative right now in the company, they have an owner, and this owner comes every month and says, "This is what I did last month. This is what is working. This is what's working." It can be red, yellow, or red. This creates full accountability within the organization. We track that in a monthly results meeting, monthly operational meeting, monthly investment meeting. We have a lot of discipline to make sure that the strategic that we discuss, it's being fully implemented and is on track. We also help our employees to achieve their best version. We firmly believe that if our company wants to achieve a next level, our employees, they need to achieve the next level as well.
We invested heavily in our employees' development with more than 11,000 employees on our Prosus Academy, which is our online academy that we have within the organization. We brought 120 leaders to Stanford. Right now, next week, we're going to bring another 66 people to Stanford. In November, we're going to have another cohort. Stanford is not only a nice place to go, but it's the source of a lot of the behaviors that helped us to craft the Prosus way. Those leaders, they not only go there to learn something, but they also connect each other in a meaningful way. This is ways of working. The next one is incentive. Once you are aligned on the direction to go, you need to make sure that your incentives are the tailwind for that.
In other words, if we want to create an entrepreneurship spirit, our incentives need to mimic. Right now, our incentives, they mimic value creation for employees the same way that a shareholder, an investor, they create value as well. It was a massive change, massive communication across all the companies in our portfolio. Right now, our incentives are fully aligned with the behaviors that we want to recognize and we want to reemphasize. The next one is how you recognize people. In a company, when people see other people being promoted and getting more recognition, they mimic themselves on that. Also, when you recognize someone or when you terminate someone, you're giving a company a message because culture is not only what you recognize, but also what you tolerate as well. In the past, our performance review was more focused on what someone delivered.
Right now, it's what this person delivers, how he delivers as well in terms of culture and in terms of leadership as well. We changed completely our performance model, and this is what we have right now. Last but not least, if you want to create the best football team, you're not going to hire someone that is great in baseball and teach this person to play football. You're going to try to find the best football player. Right now, we launched and developed the hiring the Prosus way to support and help our leaders on the recruitment process to make sure that they bring people that are more aligned with our culture. With those mechanisms, this is pretty much the cultural tsunami we have been over the last 12 months. I think the next and fair question is, okay, how can we track results?
Because all of this, this seems a little bit abstracted, but we have a few data points that can prove that we're going in the right direction. I'm going to bring three data points here. The first one is when we look for the engagement survey, which is, by the way, not something that we develop internally, something that you have a third party to look to us. We were able to increase the engagement from 64% to a whooping 87%. This was, sorry, there's something here on the slides here, the screen. Yeah. The engagement score came from 64 to 69, and right now, 87%. This 87% brings us closer to the top 10% best tech company to work for.
This is a message that we're receiving back from our employees saying, although we had all of these changes, this is still like a great place to work. This is an amazing place to work. When we look for clarity on strategy and direction, we increase it 10%. People right now, they have more clarity on what is our direction, where we should be going, what we should be prioritizing, and also what we should be dropping as well. Last but not least, people are overly proud and working to Prosus. With this, the results increasing from 86% to 94%. Not so sure if we're going to be able to check to adjust the screen here. I will finish my remarks bringing something back that I mentioned in the beginning. Right now, the Prosus way for us is three things again.
The first thing is it's a set of behaviors and habits that our employees, once they have, the whole organization have. This is our competitive edge. The second thing is a strong culture is a glue between the people inside of the organization and also enables the ecosystem, enables our companies to collaborate better from each other and create this ecosystem feeling. Last but not least, we're moving people's mindset from instead of increasing 10%, it's okay to increase 10%. We're looking for what are the things we're going to be doing right now that is going to bring us from 10% to 10x. Thank you very much for your time. Apologies for the technical issue here. I want to bring Fabricio, Euro, and Eoin here for a Q&A session where I don't think we're going to need the screen anymore. Thank you.
Hello everyone.
I told you I would be back. Good that we do not have slides now. We would have many slides now, but we said, let's make a quick Q&A because the screen is not working.
Yeah, I apologize about the screen, guys. It was my fault. On account of my paleness, I went too close to the screen and it knocked out. But it is always on its feet. It is due to be back. It is due to be back. All right. Now into the Q&A session, we are going to—
Euro is going to come to—
Euro, yeah, he is.
I think Euro said, "I'm over."
He is doing another demo back there.
Okay. Thank you, bro.
Euro, anytime you are like, "No, we are not just—"
Euro, take your time. We are here, but no problem. We can—it is fine. Fine.
Chairman is here and everything. Nobody is looking. We are going to pass around the boxes.
Please stand up, say who you are, speak close to the box, and we'll get to your questions. We have Fabricio, we will have Euro, and we have Vitti. Then we'll have a break, and then we'll go straight to the LATAM ecosystem where there'll be more time for Q&A there. Am I passing out the boxes? Who's going to ask the first question? Hey. Hey, without any further ado, it's Euro. Will, kick us off. Can we get the—Will?
We can't hear you. I think you are making us so difficult questions, so we cut them off. If you make a nice and sweet question, then we turn them off again. There we go.
Hi, it's Will Packer from BNP Paribas, and thanks for taking my question.
We have heard this morning about the creation of this lifestyle e-commerce offering, leveraging your data lake via AI to drive a differentiated consumer offering. Let's put China to one side with its very specific data rules. Is there an implication today that you need control of your assets to leverage the data? Do you need control over your assets to leverage the data? For example, you have some great assets in India, Meesho and Swiggy, but it would seem difficult to leverage the data from their assets with a minority position. Is the same true of OLX Brazil, where you need control to integrate it into your AI models? Or is that an incorrect perspective? If not, should you just be partnering with lots of different data providers? Thank you.
I will thank you for the question.
Many people make this question: can we do that if we don't control the company? The answer is no, unless you are doing something good that creates value. Then you can do everything, even if you have 10%, 20%, or 30% of a company. Actually, the way I like to see it is a little different. If the company doesn't want to do it, even if you control the company, they should have the power to say, "Get away here. You're holding that. Just create problems for me." If you are making new technologies that make the company run better, it doesn't matter who controls the company. We start here saying we have a vision. We are going to grow with that. We are going to make things run better.
If you have cutting-edge technology, if you have cutting-edge management model or culture, you saw Vitti showing here is Prosus strategy share. There were 50 companies there. In many of them, we had 10% or 20% because it includes the ventures investments. They left the meeting saying, "Why don't you invite us to more events like that?" We have 10%. The point here is, can we add value? What we are making is exceptionally good. If so, very good entrepreneurs should want to copy as soon as possible. If what we are doing is exceptional and the entrepreneur does not want to copy, then we should sell that company or change the entrepreneur. It does not matter if you control or not. Actually, I challenge you, if you are just doing things because you control the company, our bad things, we should not be doing that.
We are doing fine. Michu is an exceptional company. We copy things from Michu because they are exceptional. Michu copies things from Prosus. The point is not how much you own, it's how good you are. Euro, would you add anything to that? Pardon?
In terms of some of those flagship stakes like Swiggy, Michu, OLX Brazil, are they currently inputting their data into your e-commerce LLMs?
Are those companies putting their data into your LLMs?
Not yet, but as much as they see value, they will in the future. We started in Latin America because Latin America has so much data about everyone. That is the focus. The second one is India because also we have a very big ecosystem there and we have PayU that is core to our strategy. We are moving in that direction.
Just a request for the organization, since they have a mic, please turn the volume up.
Yeah, please speak into the mic. Please turn up the volume. We have a question over here.
Hi, Laura from Morgan Stanley. Thanks for taking the question. I just wanted to ask about the Large Commerce Model. I found it really impressive. Where are you with the rollout of this model at the moment? How many of your customers are using the agents for food delivery? If you can give us a bit more details on where you are and what the next steps are, that would be great. Thank you.
Thank you for your question. I just want to make one small comment. What you saw here is very fun because we are making questions, but it also happens automatically for hundreds of thousands of customers.
We just find a nice way to show h ow they think.
That's absolutely the case. What we have shown here is just, let's assume that you do this very slowly, just one question at a time, right? Everything we have seen, the models themselves, they are operating millions of times every minute in real time, and obviously no one touches them in the sense that it is the system that actually does that interaction with the model. Remind me again the question, please.
Where is it rolled out currently?
We started, yeah, we started in Brazil at this moment. We started in, that means for rolling out, that means in São Paulo. That's the first part where we go. From there, we are going to go to other cities gradually.
It is going to be initially for Latin America, initially for food, and then we start rolling with the other companies in Latin America. In parallel, we are initiating the same process with OLX, eMAG, and so on. That will come later on. I mean, the idea of the LCM is going to be, we believe it is going to be the operating system for e-commerce. That means operating system for e-commerce for all our e-commerce. We start in food for many reasons, but that is just the place where we start. It is not the place where we tend.
Just to give you one more number, if I am not wrong, the number is 1 million users in São Paulo using today. Our next intention is to move to 10 million in São Paulo anytime. I think it was last week. I am not sure if it already did.
The whole Brazil. That is how we are progressing that. This is so new. It started like seven months ago for the technology to be possible. I guess in six more months we will be running in most of the place.
Monique,
thanks. It is Monique Pollard here from Citi. I had a question for Gustavo, particularly around the incentives and the process way. You talked about how you have changed some of the incentives, particularly for all the employees, so it is more aligned. What are you doing to incentivize people within, I do not know, heads of the businesses to work together on the ecosystem? Are there specific incentives that are not just about their business, but how their business interacts with others or the benefits that they can bring to other businesses? How do you kind of codify that?
That is a beautiful question.
This is a phase two. We completed a phase one right now, which is we want to make sure there are incentives. We distribute value to our employees when value is distributed to our shareholders and investors as well. That was the first phase that we're doing right now. The second phase, which is our priority for the next 12 months, is to make sure that not only you see this beautiful culture in the corporate or in the top leadership of the company, but this flows throughout all organizations. Therefore you can connect them. Once we get there, the next step will be exactly creating the incentives where companies, they are incentivized to share value.
They say, "Look, I'm not only interested in my own company, but how can I generate value in A, B, or C companies that are part of our ecosystem?" This is exactly what we're discussing right now and we're implementing as well.
Just a suggestion, we have here the famous Christians. Orkun is here. Talk to them also to see. The way I try to do it is I don't start asking people to give me things. I start to say how the holding can help you, how the central management can help you. I try to give things to the company in terms of technology, culture, management model, investment. They should think the group is here to help. I think the big change to me that the culture does is that everyone is working together.
Our board has the same objectives and are discussing the same things. I can see my Chairman here. Thank you for coming. I have to say good things about the Chairman too. Our C level is very aligned. This company CEO, they meet all the time, really all the time. We create a way of working that is everyone together and is reflecting the incentives. There is another phase of the incentives that is going to be launched in one month that makes it also like a financial benefit for everyone to cooperate. Anyway, ask them and see what they think about that. If they say something bad, let me know. I'm just kidding.
Yes, hi. Thanks for taking my question. I'm Mandeep Singh from HSBC. My question is on your comment about becoming the largest consumer internet company in Europe.
If you could contextualize that with regard to the regulatory environment in Europe, do you think given the tough regulatory environment here, whether that is still okay for you to work with this or what challenges you think would pose this compared to your other markets to achieve that ambition here? The second part related to the question is also with your AI integration, how easy is it to use the datasets within Europe for you to develop your AI or are you totally dependent on other markets for that? Thank you.
There are two questions there, Mandeep. The first is, how have you found navigating the regulatory environment in Europe? The second is data sharing of users' information within Europe.
Good. I make a lot. I think the things are totally on to the other side.
That's why I'm not hearing you well or I'm getting old maybe, I don't know. But let me tell you a good thing. Six months ago, I made a lot of critics to Europe. I said the Europe mindset is a problem because Europe wants to block big tech at all costs. I think Europe is failing because they are blocking local big techs, but the big techs of other places of the world are selling here to Europe. What are they achieving if they just block the local companies to grow, but the other global companies are becoming global leaders? I criticized it a lot, this wrong mindset in my opinion. We want to make it different. We want to have a $200 billion European company.
The good news, and I'm happy about that, is that the last two months or one month and a half working with the European Commission authorities was very good in terms of speed. I told it many times, it's tech. The players are competing. We have a $100 billion player from the U.S., another one from China. If we take two years in this process, we are already lost. That's the reality. When we announced the JET deal, I said, "My biggest risk today is how much time it will take." Two years, it's too much. The last two months was very good. We filled the process this week before the expected time. I was happy because every time we send a 100 page, they read everything very fast and they make more 100 questions. Then we read very fast and we answer back.
You see, I see a momentum on Europe's thinking that we have to move faster to have also local leaders. I'm optimistic, but let's see because we filled like three days ago. Now we have to wait for the answer, but they said it's going to be quick again. We have to wait, I think, a few weeks, but it's more or less fast.
The second question was on data privacy and how we navigate that through Europe.
The old way of doing things is that if you want to do something like the LCM in the past, you had to aggregate all data in one place and then you train for that one. The beauty about these tools is that you don't have to do that in the sense that the model itself doesn't contain private data.
It contains the patterns that link certain behavior to certain outcomes. It does not have to have the identity of those behaviors. That means that once it has been trained, it can move, right? It can be incremented. It can be, let's say, trained more on other data. That is the key things. Of course, in the places where you can have data in one place, you can train probably deeper, but it does not mean you cannot do anything else elsewhere. The example that I have shown is that the model understands events of simpler, potentially of simpler in this case, from data that has been trained for iFood. That is because the technology enables that. It reads beyond its immediate boundary for what it has been trained. That also means that if you do not have the data, it still has value.
We have to think about this as something that is not predicated on sharing data. It is predicated on sharing learnings from data.
An example I like to give on this subject is a large language model. It can explain to you about a book, but it does not have the whole book there with all their data. We can learn that a user that likes salmon also likes jazz. I do not know if that was your example. We do not need to know that user and that is all. The important thing is the learning. That is what we are doing that is interesting. We are using reasoning technology. That is the same one that we discussed in the last language models to learn how people behave, but not necessarily that person. We are learning how everyone behaves. We can probably serve them better. We think we are going to comply.
We are a big company, so we have to comply with all local laws, obviously. A note on Europe. Europe overregulates. In local laws related to sustainability and to AI, the laws in Europe are more difficult than all around the world. I like very much privacy and sustainability. I talk a lot about that. If you put more work on us than in every country in the world, it's bad. That said, I believe we can, so I'm still saying Europe should be more flexible, but we can and we will work in a way to get the learnings complying with all local laws.
Silvia, you had a question there? Please speak into the box where I guess.
Thank you. Good morning, everyone. It's Silvia Cuneo from Deutsche Bank. I have a question on the user interface. Very insightful demo from Euro.
Just curious about how should we think about users potentially using like a WhatsApp interface rather than your own apps? How does this change the way you have insight into the consumers' engagements on your applications? Do you see this as a switch across countries or just in specific markets for specific ap ps? Thanks.
Did you catch that?
No.
I'm winning the hearing award for today. I'm a good hearer.
That's great.
She was talking about the different consumer interfaces, right, and whether you glean different information from a WhatsApp versus going directly to text or something like that.
There are tons of questions in that question, right? Assuming that you translate it correctly.
Anyway, first of all, the premise of using agents and large commerce models is because we assume that the way we are going to look at interfaces going forward is going to be different from the way we look at interfaces 20 years ago when we started all this. The main interfaces now are the browser, the applications, of course, and mobile, right? There is no reason for that to be the case, right? In the sense every surface, which is digital, is an interface. The first thing is that let's go to something which is pure communication, like WhatsApp. WhatsApp is so popular, so it should be, obviously, it should be one of the e-commerce platforms for there. Then a generation, an entire generation is now getting used to voice. They just continuously message and use voice.
The application that you have in your hands should give you an idea of where voice is at this moment in the sense that one year ago it was hard to make it work properly, and now it picks up accents, it understands, let's say, nuances, it is very reliable. If that's the case, then voice is going to be a viable one. Why shouldn't you actually buy things when you're in the car? It's the same thing. You have these glasses and you see something you want, why don't you do that? These kinds of things have been there as a possibility for a long time. You couldn't really act on them because you didn't really have the tools, you didn't have the technologies for that. We get to the point in which one by one these technologies are sufficiently good to enable all this.
I think that in the next one, two, three years, relatively short, you will have a transformation of the user interface.
Great. A quick note on that. The demo that Euro was supposed to show here today was a little bigger, but I think you reduced it because of time. You told me today morning. Our intention was to show the whole process. It can start in the iFood app. You can make an order, you know it's you. Then there is like a floating balloon, just like you have in WhatsApp. You can say, "Let's keep talking on WhatsApp." You can start talking on iFood and say, "Keep talking there." We talk on WhatsApp and we keep the conversation on WhatsApp.
You see it's a nice way to say, "I have the iFood, but I believe on that." We don't want to transform the whole world in ChatGPT because it's not the best interface. You can start on ChatGPT, not ChatGPT, on iFood because it is the best interface to order. Just keep talking on WhatsApp. If your order is late, not in the iFood case, your order is never late in iFood. Hypothetically, your order is too soon, but this is more possible. You just can get the WhatsApp and talk by voice, the way I think you use it voice. You can just say, "Quite arrived too early. I'm not ready to get the order." It interacts with you. That was the full interface he was doing. Just so you know, we test a lot.
We did like 10 or 15 tests of just dropping all the visual interface. Pure voice is not good enough. It's not exceptional. The thing is app, voice, and text also on WhatsApp. We are doing everything. As he said, the interfaces are going to evolve. We are already testing the next interface, for example, glasses or things like that, or other interface that hopefully we will be able to launch first and learn first.
All right. We're going to have to leave it there. It's now break time. Please stretch your legs. The bathrooms are over there. The refreshments are right here. The execs are here. They're going to be walking. You can talk to them. We'll come back at 11:20. We will be coming ba ck for the LATAM ecosystem with Diego, who's going to bring the flair.
He doesn't need a screen. Even if we don't get the screen back, he's going to go aca pella. Thank you very much.
Thank you.
Thank you.
T minus 3 minutes. T minus 2 minutes. T minus 1 minute. T minus 30 seconds. This is the rhythm of a system. Here we go. That's handheld one.
Hello, hello, hello. How are you, everybody? Nice to meet you. My name is Diego. I'm from Brazil, as you can see in my accent. It's a big pleasure to be here with you. It's the first time I'm here talking directly to the investors of Prosus, which is, for me, a big honor. I've been in the group for eight years.
I've been part of this story for eight years, starting in Movile with Fabricio eight years ago, then moving to iFood, and I've been also part of the Latin ecosystem that Prosus has in the continent. I'm heading this since the last three months. Thanks very much for joining. I hope everybody enjoys. The whole idea here is to talk about how we are envisioning everything in the ecosystem, then a double-click looking at the assets that we already have, the most prominent ones, double-click looking at iFood, and then going to Despegar. After this, we're going to have a nice conversation here, and Eoin will host a nice Q&A involving everybody here. Okay, as you can see, technology is working very well.
My whole challenge here is to make you dream about these lights that we made, but you will never be able to see it. Okay? The first idea here that I want to tell to you is, when we look to the past, Prosus already had good assets in the region. I'm talking here about Latin America. The whole point is that those assets, they were managed independently. When I say managed, don't think about like a very hierarchical thing. I mean, I'm not saying, "Oh, Diego just managed this because he's the CEO and the other guy." I'm talking about managing the entire portfolio, the entire ecosystem. The level of collaboration was low. In the end, we were not able to extract the value that we can do because we can collaborate. What we did during the last year is to start this movement.
We started to talk more. We started to understand more what everybody's doing. The more I understand about them, the more I can challenge what they're doing. Of course, I'm not the specialist, but I have good assumptions to understand and make questions about the performance, about the strategy, and so on and so forth. Also, during the last year, we were able to set up things that we believe are very important to disrupt. Two things that I'm going to also mention here during the presentation are, first, culture. Fabricio already talked about this, Vitti also, and management model. The way we communicate in our company, it's what makes things easy to come from one place to another and give agility and define pace. That's something that we work hard.
We came from assets that were far from each other, and now we have assets that start to speak the same language, assets that understand what the others are doing, and therefore collaboration comes in a very organic way. Now we are going to a next step, and this next step is what we call ecosystem. It's what Fabricio already mentioned here. This next step basically means, how do I make these assets that now collaborate quite well? I also have a certain level of integration that allows me, in the end, to extract more and more value. Make no mistake, when I say integration, I'm not talking about like the old concept of having an M&A and integrating the back office to save $10 million per year. That's not what I'm talking about. What I'm talking about is to integrate what really creates value.
Euro already comes here and said, "Okay, see these beautiful things that we are building regarding AI." Oh, integrating databases makes sense. Integrating the system of the procurement team does not make sense at this stage. That is the next step that we are doing now. We are getting the assets that were separated. We gave them the same rhythm, the same pace, and now we are bringing them together through the integration of things that make sense. That is what we are going to explore here. You could say, "Oh, Diego, but I mean, that is just a PowerPoint." Actually, that is not just a PowerPoint. That is more or less what we are already doing at iFood throughout the time, and I am going to go into details. Also in the region, we have another good example of this. Meli, Mercado Libre. Mercado Libre did exactly this.
Of course, they didn't do through separate companies, but in the end, when you take a look on the different business units that they have, what they are doing is that they're getting a lot of different business. They are integrating, for example, the database, and they're being able to leverage everything in order to create a company that has the same amount of customers that Prosus ecosystem has, the same amount. They are able to extract much more value. If you think that value is EBITDA, in terms of EBITDA, if you think that revenue is what is value, they are being able. If you think that valuation is what is value, they are being able. Why? Because this huge level of integration allows you to go further in value extraction. There is already an example, which is iFood.
There is another example, which is Mercado Libre. That is how exactly we are trying to do things now. Let me talk a little bit about iFood and how we have been doing that for all the time. At iFood, basically what we did is we built a vision that we could, at a certain point, be the main choice of a person that needs to solve a problem, that needs to look for a service on an ongoing basis, meaning every day, every week, every month. What we are talking here, it is about convenience. What we are talking here, it is about lifestyle. It is about your routine. We are not building a company that will not touch these aspects. How did we start that? Through food delivery. Why food delivery? It is simple. It is because it is the category with the highest level of frequency.
The more I touch the client on a monthly basis, the more I'm able to bring him for another experience. That's what we did. We started everything doing that. The second movement we did once we consolidated, once we reached everything that we believed in the food delivery, is to move to the natural adjacencies. We went to supermarkets, then we went to pharmacy, then we went to beverage and convenience, then we went to pet. The more we go, the more we are bringing frequency to the ecosystem. What's the beauty of this? The more I bring frequency to the ecosystem, the more I make my logistics, which is my main asset, more efficient. Why are you doing this, Diego? Is it just because the PowerPoint said that's the vision? No, no, no, no. It's because when you think about food delivery, that's the behavior.
Concentrated with the traditional banks. We are able today to have more accounts. We are able today to lend more money. We are able today to do more payment services than any traditional bank to the restaurants in Brazil. That is how you build an ecosystem. That is how you optimize certain aspects, such as logistics, and that is how you start to coordinate the value chain. What I mean by coordinate, the more you sell with me, the more I can provide credit. Therefore, the more I can manage the relationship of the restaurant in the ecosystem, not just because I am charging a lower or more expensive take rate, but because there are some levers that make sense in a more holistic view. That is what we did there. Now we are expanding this to the entire Prosus ecosystem.
We add in the same rationale other assets that are very important here. First one, Despegar. I'm not talking too much about Despegar right now because Gonzalo is my Argentinian friend, which should not be true because Brazilians and Argentinians are not friends, but I'm considering this as a new normal. Gonzalo, my Argentinian friend, will come here also to talk about Despegar. I'm not going into details. You already have too much information about Despegar when we announced the deal a couple of months ago. Basically what I can say in a nutshell is it's a company that is growing. It's a company that is profitable. It's a company with a very strong brand. It's a company, actually not a company. It's a business model that has something that is not beautiful. That's the reality. That's the transparency.
That's what about when you say face the brutal facts, that's what we mean. Tell the reality. And the reality is the OTA business has a low frequency, low retention. That's the reality. But then the ecosystem comes to play. Okay, if I, from iFood, see you ordering food in a city that's not where you live, and I know where you live, you traveled. Did you buy from Despegar? Checking the database, no, you didn't. I have an insight. Why not? And then you may say, what do you do with that information? And then you go back to iFood and you say, oh, iFood has the largest and most successful loyalty program in Brazil.
Why don't we integrate iFood and Despegar in the loyalty program and make the consumers of iFood, we are talking about 60 million a year, active consumers a year, also consider Despegar to buy? Take that person that traveled and did not buy from Despegar and understand better with the model that Euro showed here and try to understand why he did not buy. Maybe because he is not a rich person. He is a poor person. Oh, therefore maybe he did not travel by plane, but maybe by bus. Despegar does not give you the chance to buy tickets to travel by bus. Why not? If the database shows me there are 10 million people that buy iFood, that are in the loyalty program and that may consider. That is the rationale, for example, that we are bringing to Despegar.
When you think about Simpla, which is our ticketing online platform, you can think about even Brighton United States as a good proxy. Simpla is the same. Simpla gives me. If I have to do something with you or not, changes everything. Let me tell you one thing. Push, which was something invented a few years ago, was like amazing until a person discovered one day that more and more people like us go there and block and say, "I don't want to get pushes anymore." Push strategy, and that's a very tactical example that I'm giving to you, gives you the ability, the ecosystem gives you the ability to enhance the efficiency of this strategy that changes the way you can grow. Simpla also gives me this. Let me give you another example of Simpla.
Transparency allows you to understand what can be the best strategy for it. I'm going to give you one example when we were talking about the Despegar and iFood integration. Finally, there is OLX as well. OLX gives me another perspective. OLX, for example, gives me the idea of how much a person is sensitive to price. The more you resell things, the more you have a certain behavior compared to people that just give away. This changes completely the way I approach the person regarding the offering, for example, at iFood or even Despegar, a person that prefers a flight or a bus.
The whole capacity to bring these assets together and, for the sake of Euro's presence here, integrate the database and allow this database to make the LCM stronger for me to understand which decision to take, the more I have this, the more I really can understand the right decision to make this company grow more. That is the entire idea that we have for the Prosus Latin ecosystem. What I am going to do right now is double-click and go into iFood, talk about five minutes to ten minutes, and then I am going to ask for Gonzalo to come here to talk a little bit about Despegar. Let me tell you a few things about iFood. First of all, iFood, it is a case where we see through all the time the capacity to grow and also to deliver profitability. How do we do that? It is very simple.
We do that through culture and management model. I agree with Fabricio. I mean, I've been down the road for many years. I worked with the capital markets for more than a decade, road shows, non-road shows, so on and so forth through all the world. Generally speaking, nobody asks about these things. The point is these are the things that make the change. Why? Because a company is not defined by one invention that you had. A company is defined by the capacity that you have to keep inventing yourself through all the time. Think about iFood. The way we started, it's not through how most of you think, like an app. Actually, we start setting up a call center with people getting calls and on the other hand, printing books that were a collection of menus.
We went to the condoms and delivered the menus to people. People, instead of calling straight to the restaurant and asking, "Oh, do you have Margherita pizza? Oh, is it with onion or without onion?" and having these weird conversations that we used to have like 10, 20 years ago, we solved this problem with printing books and also having the call centers. We evolved from this to the internet, then the internet to the app. When Android came, we said, "Okay, now I can go to logistics because the employees can have the chance to work from their hands." We went to logistics. We went to supermarket, which is tough. You know why? Because the capacity to integrate the brick-and-mortar operations of the traditional chains plus reinvent my logistics because it's, "Oh, now we have, oh, see? That's AI.
That's AI. AI was listening to me and now everything that you will see, let me go back here to this slide, everything that you will see, I didn't see these slides. Everything that you're going to see, actually AI did. Euro just put in his model and now I will have to look to these slides to understand what AI wants me to say. When you look to the story of iFood, what you will see is this company with the capacity to change. I mean, if you think about the supermarket delivery, it's a completely different logistics. I'm not talking about the same guy with a motorcycle delivery. It's a different bag. It's a different, actually, model as well because we also have, for example, the cars doing the delivery and so on and so forth. Totally different.
When you go to pharmacy, it's again a totally different thing. Just to give you an example, yesterday we released in Brazil for the first time the integration of our pharmacy business with digital prescription, which, by the way, is a company invested by Prosus Ventures and is part of our Latin ecosystem. Now people that have a prescription and can only buy delivering the prescription to the pharmacy can buy online for the first time. You go to iFood, you order, there's a way where you insert digitally your prescription and you can order. For the first time yesterday, medicine that demands a prescription was done digitally by iFood for the first time in Brazil. What is this? Again, it's your capacity to innovate.
When people ask me, "Oh, can you continue to grow or to deliver profitability?" I will say, "I will always have challenges. And one day or another, I will do like this because I'm not going to be able to do it." I completely have faith that we're going to be able to do that, as you can see here. Why? Because I have culture and management model. It's not because Diego is great or Fabricio was great. It's not because of one person. It's because the entire company moves towards a direction in a certain pace that it's ours. It doesn't come from one person or from another thing. It belongs to us. This story, it's what we will keep repeating. Another thing that we were able to do, and that's like for me, it's one of the things that I have more proud.
It's been recognized as the most loved brand. Maybe you may say, "Okay, that's just branding. You put money there, you bring great agencies and people will deliver great videos and everybody." It's not about this. First of all, I don't have the largest branding budget in Brazil. That's the first thing. Second thing, I dare you to find another, let me say, food delivery company that were able to be loved in their country in that standard. That's a business where people generally don't like you. There's a huge debate towards labor regulation. There's a huge debate about transport, about the impact in the cities. Most companies in the world are not loved because this business model is still being in development by the society, by the way you align all stakeholders around the concept.
Being here since 2022, it's something that is really, I mean, makes us very happy because that's a way for us to understand that we are building a sustainable business. It's not about the brand. It's about what is behind. If most of the people in the country are saying that's the most loved brand, it's because most of the people are understanding more and more all aspects of my business. That is tough. As I mentioned to you, when we started this business, the whole idea was to grow frequency. There is one dilemma in the beginning. Should I grow like hell to have all the customers in my ecosystem, or should I focus on frequency? The decision five, six, seven years ago was to focus on client acquisition.
As you can see, we tried to do as much as we could to acquire every single Brazilian. If you think about that, currently, we have 60 million active users in our ecosystem on an annual basis. If you think that Brazil has 210 million people, 40% of the population, unfortunately, is very poor. People basically ask in their homes, generally speaking, one person asks and the others do not. One person represents the family or the couple, and etc. Actually, we acquired already 80%-90% of the customers in Brazil. At that point in time, you may say, "Okay, the company will stop to grow." That is not what happened. Why? Because we are building the ecosystem that looks to frequency. We did not grow frequency here because we acquired new cohorts that start with a lower frequency.
Then we started to bring more and more people in that understand the business, understand the offering, understand how we can impact the convenience of their lives. Then they understand pharmacy, they understand supermarket, they understand pet, and they start to evolve throughout the ecosystem. An ecosystem that is not only looking to the consumer, but also is looking to the restaurant. At this point in time, only to the restaurant, not to the other merchants. Focus is another thing that we like a lot, even though today you may say focus, but remember, it's already like years down the road building this. We like focus a lot. At this point in time, the focus on the B2B side, it's the restaurant. When I say the restaurant, it means ERP systems. Think about toast.
It is the same, just a way for you to simplify the rationale here. Think about toast. We go from ERP to kiosks where you can go and order. All of this is linked with our fintech that processes the payment. Also, because we have this information not only on this side, but also on this side where I can see how much this restaurant sells, how profitable he is, how is his growth rate, how is the customer satisfaction, and so on and so forth, I can organize all this data and through an algorithm provide credit. Just to tell you one thing, the best delinquency rate in this segment in Brazil belongs to us. Why? Because there is an algorithm that basically uses machine learning to understand if a restaurant has potential to churn or not.
See, I'm not talking credit about looking at financial metrics. I'm talking about churn. If the restaurant churns, there is a chance of him not paying me anymore. If the restaurant doesn't churn, why would he screw with the main sales channel that he has? Doesn't make any sense. If he has two low ones, one with me, one with the other bank, he will default here and not here. That's the concept of ecosystem. That's the concept where everybody starts to understand the beauty of behaving well in the ecosystem and therefore making the ecosystem stronger. Moving forward, that's how everything that I told you translates into growth. What you can see here, it's basically the business that we built throughout the time growing.
There is the food delivery marketplace, there is the logistics business, there are the other categories, there is the dining, which is this ERP, kiosks, payments, offline payments strategy. That is how we built, and also the fintech, of course, that is how we built throughout the time this. All of them growing, all of them reaching good levels of profitability, all of them, more importantly, making merchants transact much more, consumers transact much more, and riders transact with us much more, make more money throughout the time and decide to stay in the platform. When we look to the marketplace business, let me call this all the adjacencies where the consumers interact and order for their convenience, what we can see as well is that there is a huge opportunity in the market. When we look, for example, to these ones here, they are still very, very underpenetrated.
We see a huge potential to grow in the pharmacy space, for example. The more we are now being able to shape the regulation in Brazil and allowing a journey that has less friction, the more we are seeing the retention rates going up and the more this business is becoming very interesting. When we look to the restaurant space, we are now going to a new phase, a phase where there is an integration between the online and offline space. This opens a huge opportunity for us that I'm going to show you how, showing the product in a few slides afterwards. Moving forward, we're still seeing the early days of iFood Pago. iFood Pago is our fintech. We basically already have the assets ready to allow us to keep growing, allow us to have good profitability, and also to manage risk very well.
We not only have the credit for supply, but also credit for growth for the restaurants. We have already the digital account. Eighty percent of the restaurants already have their digital account with us, and they are switching from the traditional banks to us. We are already increasing, penetrating more and more the payments. We already do 80% of the payments of all transactions that happen in our ecosystem. We have unique assets here, and that's something that is quite important. I'm not fighting here in the open market. I'm not fighting here with people that believe that their algorithms are better or worse than mine. It's not about this. It's about first having data that changes everything, the way you price, the way you understand risk. I have the sell-out data.
I know if the consumer likes you, if he will come back to you, if they are demanding more shifts from you, which means increasing your sales. I know exactly if the chance of you to succeed versus the other pizzeria in the close neighborhood. I know everything about this, and that changes everything compared to a bank that just gives credit. Second, I have the money in. The money is with me. I process the payment. I have the money. If I have a problem, part of my risk, I can reduce with the money that I have. Third, I have the principality. The example that I gave to you, if you have to default between two banks, I'm not your first choice. Why? Because I sell. I not only give you the credit, but I sell.
If you have to have a problem discussing your delinquency rate, it's better to discuss with somebody that doesn't have other touch points with you. Fourth, I have the sellers. I don't have to prospect them just to give the credit. I can just offer the credit because I already have them due to my relationship. Finally, these are the numbers. They are growing, but much more important here, they are growing in a very sustainable way. It's a positive cash flow business. There's a great delinquency rate. The returns, the way that we measure this, has an ROI that is very good, and we will keep investing here. The whole idea here is not to have the fintech as a sole asset, but it's to have a fintech as something that is part of something bigger.
We also have what I told you that will link the online to the offline. We call this Maquininha, which means the big machine in English. What is this? You may look at this and say, "Okay, this is a POS, right?" That is where you get your credit card and swipe your card and say, "Okay, I paid." That is wrong. That is what you generally see. What we have here is, yes, a traditional POS, but we embarked as software. This software, every time that you swipe your card or put your telephone number, this software goes back to my database and says, "I know who is this guy." This guy lives close to here, worked there, and the restaurant is here. Every time that he is at his home, he orders online in your restaurant.
Every time that he's working here, he never comes to your restaurant. There's a great opportunity here. What's the opportunity? The guy loves you. He already orders, but he never comes. Maybe he doesn't know that it's close. Maybe he doesn't know it's you. Maybe he just didn't change the behavior. Instead of keep sending promotions online, send offline. That's the rationale behind. The software just says, "Because you set up a rule, that's what we're going to do." You swipe your card, and instantly you get a cashback that says, "If you come back here next week, I give you a 10% discount." You may say, "Diego, but this is just a cashback." It's not just a cashback. It's a cashback to the person that needs that cashback. Same rationale.
You go to a restaurant, you swipe your card, and you say, and the software thinks, "Okay, so this person doesn't live in the city. Don't do anything with him. Don't give a cashback. He doesn't live here. He doesn't need that." There is another aspect. You go there 10 times. Don't give promotions to him. He likes you. He loves you. He likes the price. He likes the food. He has the behavior. Don't give a penny to him. That is now how we are generating traffic to the offline space. That is what restaurants are getting amazed by. We just finished all the tests back here in September. Since September, we did our go live. That is the result so far. We are already transacting BRL 100 million per month, and this number will keep going.
The problem today is that we don't have enough capacity here to deliver more POS. We'll keep doing, but of course, there's a physical limit here with the partners that we have. Moving forward here, we also have our kiosks that do the same. You may say it's just a kiosk. It's not just a kiosk. You swipe your card here, and it does the same. Even more importantly, I'm going to talk next, or actually in two slides, about our ads business. It's also ads business. I know exactly who it is. I swipe my card. I like Coca or Pepsi. I like Coca. Pepsi can say, "Okay, you have to offer something to this guy here." It's different. It's not only a kiosk. I have the data to make retail business. That's not what we are envisioning. That's a reality for us.
That is why in the offline space, we are being able to close the entire demands of the restaurants. You can see the financial service, iFood Shop, which is our B2B platform that sells for the restaurant inventory, the CRM and POS that I just mentioned to you, and also the logistics. We are growing very fast in numbers in terms of orders and the dining. As you can see here, from one year to another, it was a 15-times growth. Of course, we're not going to grow 15 times every single year, but this business will keep growing. Every time that we grow here, we fuel especially the fintech business. As I mentioned, we also have a great business here that is growing very fast and being very relevant. That is the ads business. I just broke here in three pieces.
The first one is the food delivery ads top placement. In this business, we already have 1% penetration in terms of the GMV. When you look to the grocery business, we already penetrated 2.2%. When we look to the investments in the platform that we manage for the restaurants, their ads for the restaurants, we already have 5%. This sums and gives you 6.4% penetration. What is that? It's basically capacity of ads that I have. As soon as I have this ecosystem working, the more I understand the consumer, the more I can make the ROI work for the restaurant, and not only restaurant, but also to FMCG companies, to drugstore chains, supermarket chains, and so on and so forth. The more I'm able to do that, to deliver this ROI, the more I'm able to make these guys keep investing in the platform.
I could also talk here about AI. I will not take too much time about this, but what is important for you is first, we're not talking about AI here because now it's the frenzy. We do that since 2018. We've been working with Yuru and his team here since 2018 in iFood. We already have 174 proprietary AI models that are live and run the operation of iFood. Currently, we do 14 billion predictions per month. We already have wrappings, which means integrating the capacity of an algorithm with a third party such as OpenAI in more than 30 examples. The next frontier is the agentic bots as Yuru showed here. The first phase of ourselves was to do two things: reduce costs, as you can see here. That's the example of the requisition cost that we decreased by 20%-30%, and also to increase the experience.
Currently, 90% of the tickets in the CX, in the customer experience department, is managed by an AI algorithm with NPS higher than when a human does. It has been like that was the first phase. The second phase, it is exactly what Yuru said before, and we are going to invest heavily on this. I talked briefly about our loyalty program. It is important to say here how we are moving from a transactional loyalty program to a loyalty program that really delivers loyalty. There is a difference between just giving discounts and really making somebody say, "I'm going there because it makes sense," like Amazon did in the U.S.. Here is a good example. Currently, 13 million people already pay a monthly subscription to be part of Club. Club is the name of the loyalty program. Club, it is basically a program that is started with food delivery.
Since last September, we started to evolve to other categories. Since September, the growth of pharmacy, the growth of beverages, and growth of groceries are the biggest in our lifetime. Why? Because Club has been able to make them think about iFood when they decide to go to shop. That is the beauty of this. Club will keep evolving. We will keep doing things to maintain this capacity to increase the orders. That is the number of orders that we do from Club, from clients that use the loyalty program, order at iFood. Also, as I mentioned to you, I did not remember that this was here. As I mentioned to you, we launched it here in September. See the impact that Club is already having in these verticals. More and more, we are going to give the capacity. If you ask me, "How does it work?
Does it make money? It does make money. Does it grow? Yes, does it grow? How does it work? I have to show you three algorithms of AI to explain how does it work. It's not about a linear relationship that you do in a software that makes everything work, grow, create loyalty, and at the same time gives you profitability. Also, we announced it two months ago, the partnership with Uber. That is also the idea of building an ecosystem. An ecosystem does not mean that we're going to acquire every company. An ecosystem means that we're going to acquire companies. We're going to develop business, but also we're going to have real partnerships. I'm not talking about this like piece of paper commercial partnerships where you just sign something that lasts for three, six, seven months. It's not about this.
The tech teams are working hard to make an integration that will be the best integration in the world. You will not need to sign in to use Uber at iFood, and you will not need to sign in at Uber when using Uber at—I'm sorry, let me repeat this. You don't need to sign in at Uber while you're using inside iFood and vice versa. It's going to be something magical where you just enter and you can order and they know you. They know your payment preference. They know your cashback in the order platform and so on and so forth. This opens a lot of possibilities to us. We are going to discuss the integration of the loyalty programs. That's the next step. We are going to discuss partnerships around the logistics because there are many, many synergies between both companies.
Don't forget that Uber has its fastest business in Brazil, not in the ride-hailing, but on the moto-hailing, which means the same person, the same vehicle that does most of our deliveries. There's a huge opportunity for us to look for synergies, especially in terms of synergy when I talk about logistics. That's another thing that we're going to do. The whole idea here is we got our information. We sent to a third party. They got their information, sent to the same third party. They cross-checked the data, and the results were astonishing. There's a minimum overlap between our database. We didn't imagine that. Not myself, not Dara. When we talked for the first time about this, both of us said, "Okay, we know it's not going to work, right? Yeah, but let's try." That's how we do in tech. Let's try.
I can be wrong. I was wrong. The point in the end is there's a huge chunk of people for me to acquire, and there's a huge chunk of people for Uber to acquire as well. This gives us the ability to not only have what we have on our side, the Club, but also extend that to something much more powerful for the consumer. When we bring Uber there on the mobile space, we bring Simpla, Despegar, OLX, and we bring many other assets that enhance the value proposition of not being a transactional program, but being a loyalty program. Now I want to call here. Where, where, where? Oh, where's Gonzalo? Gonzalo, please come here and join me for us to start to talk a little bit about the exciting things we are doing in Despegar. He's not here? Oh, Gonzalo disappeared.
Oh, Gonzalo is here. Welcome, Gonzalo.
Thank you, Diego.
For you that love soccer, Gonzalo supports River Plate in Argentina and was in the U.S. cheering for River Plate in the World Cup. Unfortunately, Brazil, we won again, but Gonzalo. Anyway.
It's okay. It's okay. We're just world champions, but it's fine. Thank you, Diego. Thank you. Hello, everyone. It's great to be here today and tell you all how excited everyone at Despegar is of joining Prosus and probably more importantly, to partner with iFood in developing and growing the Latin American ecosystem. What is Despegar? Despegar actually has a story that is quite similar to iFood in the fact that Despegar has been leveraging technology to transform commerce, in our case, the travel industry.
Despegar was founded in 1999 in Argentina, and it had a first phase of growth that was expanding geographically across the region. When that finished, the next stage came by introducing all possible travel verticals. Basically, we had hotels, packages, car rentals, vacation rentals, insurance, everything. We had everything. You name whatever travel product out there, we put it on the platform. That led us eight years ago to an IPO in the New York Stock Exchange that triggered yet another phase of growth that came by consolidating the industry through several acquisitions across the region, in Brazil, in Mexico, Andean countries. We went out and started to consolidate the industry. That led us now to the merger with Prosus, which for us is another opportunity for growth by leveraging the ecosystem and with this partnership with iFood. What is Despegar today?
Despegar is the number one travel company in the region. We have over $5.5 billion of gross bookings every year transacting through our platforms, transactions that generate almost $800 million of revenues every year, $175 million of EBITDA we had last year. We have a very nicely diversified geographical footprint, especially important in a region that is sometimes relatively volatile in individual markets. Probably more importantly, we have a very healthy mix of non-AR products that in the travel industry means much higher revenue per order. What does it mean and why is it that we are so happy and so excited to be joining Prosus now? It is because we believe that we are in a very strong starting point to join Prosus, to partner with iFood, and to leverage the ecosystem for further growth.
I can tell you, Gonzalo, a few things here to all of you on how we are doing that. This is something important to tell you how these things happen in our kitchen. Of course, when you are doing an M&A, an investment, you have your Excel spreadsheet where you put a lot of assumptions. You try to understand what can work, what cannot. As soon as you close this, because that's what you decided, the whole point is not to look to the Excel spreadsheet anymore. The whole point is to get everything that you want and say, "I'm going to test like hell." 99% of the things will fail, right, Gonzalo? 99% will fail. The point is every 1% that fails, you get knowledge.
Oh, the consumer prefers this, prefers that." "Oh, the hotel doesn't like this, doesn't like that." "Oh, it doesn't work for Despegar, but works for iFood." "Oh, it doesn't work for iFood, but works for Despegar." You keep building knowledge, knowledge, knowledge, knowledge. Suddenly, boom, you reach the 1% that works. That's what changed your life. That's how life works. Generally, people think that we are like smart people that have a great idea and it works in the first place. It doesn't work like this. What I'm going to show you here is the first 99%, things that we are testing and we are learning. Fabricio mentioned here, he said, "Oh, people are already doing three to five tests in one month, and that's it." I mean, we are already doing several tests in five hypotheses that are already live.
We are learning a lot. I'm going to give you a few examples here. The first thing I want to show you is when we, for the first time, were able to cross our database, we discovered that the database that covers both companies is tiny. That's a huge opportunity. How many clients I can bring from Despegar in Brazil, of course, to iFood, and how many clients from iFood I can work hard to transport this to Despegar. That's the first big opportunity that we saw that materialized to us. The database could be different, and we could also work considering this other scenario. The point is the main opportunity in terms of database, in terms of client database, is the fact that there is low overlap. Can you change for the next one, Gonzalo?
I'm going to give you a few examples of what we are testing. What we are already doing is there are like 10,000 people in Brazil that already have access to this category at iFood. You may say, "Oh, this is very simple." It's not simple at all. It's not only like a banner where you click and that's it. There's already a very interesting integration here that we are now seeing how people react. How many times do people click there? If I change this to the bottle, how many times do they click? How many times does a person travel during the year? This, I don't know, four times. Okay. How many times is he clicking here to plan his next trip? We are learning with this. We're learning how people interact.
As soon as you click here, you come to this space inside the iFood app, and then you start to understand, and it's already integrated. Most of the accounts are already integrated. People just flow, and they know exactly, and Despegar knows exactly who is myself. Another example that we are doing here is what happens if I offer a cashback for you if you download the Despegar app? You may say, "Oh, Diego, you are giving money to somebody to download the Despegar app." No, no, no. Think differently. There is a price that Despegar pays to Google or to Instagram to make you click and download his app. Am I able to acquire the same client to make the same download for a lower price at iFood? That's the cashback. The cashback is not an extra expense.
Actually, it's a more efficient expense because Despegar already has this expense. That's another thing that we are testing. The way we are testing, it makes the client, once he is inside of the two companies. Here we say, "Make four points at the Despegar loyalty program if you click here." Then the person comes and also goes to—I'm sorry. He clicks here, and he gets a cashback at iFood to use in future purchases paid by Despegar because he's already paying. At the same time, he gets the points at Despegar, which for Despegar is almost for free. That's another example of a test that we are doing here to learn how clients react. Inside each of these hypotheses, there are 20 different tests to understand how clients react. Can you go to the next one, Gonzalo? That's the last one.
I'm going to show you some good examples here. Despegar has a very interesting asset, a fintech called Coin, like coin. What they do basically is it's a buy now, pay later business that has a great performance in terms of the liquidity that is already used by Amazon. It's already used by Magalu, it's one of the main retailers of Brazil, among other very important marketplaces. Now we connected iFood there. What's the hypothesis? The hypothesis is maybe there are people, low-income segment, that by the end of the month, they have a birthday of their kids, and they don't have money to buy a very fancy meal. You may say, "But Diego, if the person doesn't have money to buy a meal," no, the person has money, but at that point in time, he doesn't have enough money to do that special thing.
You may say, "But it's only one transaction for a low-income segment." That is what you need because those people do not transact a lot as the high-income segment. We are bringing Coin to solve this. The great thing is the results are amazing. First, the AOV gets higher, the frequency gets higher, and the DAO, the daily active activity of that user, increases by three-four times. What we are learning is doing buy now, pay later for low-income segments allows me to unlock a certain behavior that was not able to happen because there is no credit available to that person. Again, you may say, "But credit for food?" Do not forget, it is only like $5. The chances of the person not paying $5 in 15 days, in 30 days, is very low. The first result here is amazing. Of course, this is a credit business.
We're not going to scale from one day to another, but it's something that is making us very, very, very proud. At that point in time, we have iFood at Coin app, and you do everything from there. So iFood is already integrated there. The person goes to Coin and does everything there. The next step, we bring Coin to our checkout, and then we make it available if the person is from the low-income segment or we don't make it if the person is not from the low-income segment. Ok ay. Thank you very much.
Thank you.
Now we have the superstar.
Wow. What another mistake of mine to put two Latin Americans on the stage and tell them to stay on time. Wow. All right. Let's try and do Q&A correctly this time. Where are the boxes? There's one. Who's got a question? Marcus. Right over here.
You can throw it. Stand up there, Marcus. You can catch it. You can do it. Go on. Should we just throw them? They're meant to be thrown. There you go.
Yeah. Hi, Marcus Diebel from J.P. Morgan. I have two questions for Diego. The first one is on AI. Obviously, we heard a lot about it in the last presentation. If we just stick to the core food segment, how do you think iFood at this point in time compares on AI, user application, customer experience, and everything that comes under it to other leading food delivery players in the world? Yeah, maybe that's the first question.
I visit the U.S. and China every single year for 10 years. We are at the same level. No problem at all. Actually, if you compare to Europe, we are much better, much better, much better. It's incomparable.
With China and U.S., I think we are at the same level. When I say I visit, it does not mean I go there to see corporate presentations. We go there to work there, to see the models, to understand how things work. I am comfortable with that.
Perfect. Perfect. The second question is on competition. Obviously, a lot of noise around Meesho. I am sure everybody is impressed by the synergies that you have and particularly what you plan in the future using these synergies. If anyone's strategy is to come into your market with heavy vouchering, aggressive marketing, is it your view that you can just withstand just by being better executors, having the network, or is there also a function of probably outspending or at least increasing spending on top of good execution as well?
Yeah. There are a few aspects here.
The first one is some people ask me, "Oh, they are big." I say, "What's the difference about Amazon and Mercado Libre when Amazon went to Brazil?" It's the same. What was the difference between eBay and Alibaba in China when they compete? It was the same. I mean, they were like a small company versus a big company. Right? The question here is not size. That's the first thing. The second thing is what's the definition of strategy? Coming to say there's $1 billion in vouchering is not a strategy. It's go-to-market. Somebody can come and spend $1 billion, but in the end, a consumer only will stay with you if it makes sense. Every time that somebody asks, "I mean, how are you preparing for this?
Can you compete with this?" The question that I do to myself is, "Do I already have competitive advantage in my business model? Am I building new competitive advantages?" The answer is yes and yes. You can lose, yes, if you have a company with better competitive advantage. You can get hurt because somebody spent $1 billion but does not have competitive advantage. Yeah, you can get hurt. That is part of life. I mean, every time that you are competing, you can get hurt. The point is who survives and who gets stronger. I am confident that first, we have been able, with everything that you saw, we have been able to realize that we have competitive advantage. Second, I could show you more tennis lights that, unfortunately, I cannot, where I am building new competitive advantages that are making my business stronger and stronger. I am confident.
I mean, it's going to be a good fight if they come. It's going to be funny. As it will always be. I was there when we fought Uber. It was the same discussion. "Oh, there's coming an American company here, $100 billion valuation, a billion dollars to spend. You were screwed." And then came Rappi with a billion dollars from SoftBank, and then come Glovo. It's part of the life. The question is also, do you have competitive advantage or do you have $1 billion? I'd rather prefer to have competitive advantage.
All right. Where's it going? Let's go over there to you. Robert. Nice. Nice relay.
Perfect. Yeah. Thank you. Yeah. Robert [Fink] from Kepler Cheuvreux . My first question is about, yeah, also a follow-up question about Meituan. I believe that Meituan, when it enters markets, oftentimes focuses on the lower tier of the market.
I believe that you're actually launching also a lower-tier product tailored also maybe to fend off competition. Is there a risk of cannibalization of your more higher-end product? Yeah, can that cannibalize? The second question is maybe about B2B, more longer-term question. Yeah. How large do you see the opportunity in B2B? I think right now you're still predominantly a consumer internet company. Of course, yeah, in B2B, there's also a lot of opportunity with restaurants, with, yeah, people that own accommodations, hotels. How large is that opportunity? Could that be over time just as large as the B2C side of the business? Thank you.
First question, low-income is a big opportunity in Brazil. We are working on this since 2018. Learning, learning, learning, learning, learning, learning, learning, learning. We thought we found the product.
I mean, we just found that like a year ago, more or less. Now we are expanding this to Brazil. We're still in just two big cities trying to finish the learnings, and then we will roll out. Why is it so difficult? Because it's low-income. You get a very small AOV, you apply a take rate, and you have to pay for everything else, including the delivery rider, right? Some people say, "No, we are going to Brazil to do low-income." Don't forget that Brazil is not the same as other places, right? In Brazil, you have to pay for social security. In Brazil, you have to pay for social benefits. In Brazil, you cannot just tell somebody, "Work 18 hours a day." It doesn't work like this. It's different. It's easier to do low-income when you have lower expenses.
It's more difficult to do low-income when you have more expenses. Yes, it's an opportunity. It's not easy to crack the code. Regarding the B2B, B2C, B2B is not a priority. We have just the seeds. We do, we learn, we grow, but it's not a priority. It's not where I'm putting money. The opportunity is unbelievably big, but the challenge is also very big. It's not easy to make the integration. It's a very low-margin business, so scale counts. Therefore, the level of investments should be big. It's not the right time for us to do that. We should wait for a few more years to have this as a priority.
Maybe you can quickly elaborate a bit more on the cannibalization risk of a lower-tier product.
Yes. There is risk. You are totally right that there is risk. How do you solve this?
You solve this through UX. We had the initial results where when the product starts to work well, it's called hits, this product in Brazil. The initial results show a high level of cannibalization. How do you do that? You come back to Euro's platform and you say, "This guy doesn't deserve hits. I mean, he has enough money. This guy deserves." It's on the UX that you solve cannibalization. How do you do that? You test 100 hypotheses, 99 don't work, one works, and that becomes your framework.
Perfect. Thank you.
Great. Will? We'll get you. We'll see you there one more.
Hi. It's Will Packer from BNP Paribas Rouxanne. My understanding is that the government is considering reforms to the meal voucher ecosystem in Brazil, particularly at fee levels.
Could you remind us how big the meal voucher segment is, what the opportunity for iFood for meal reform would be, and whether you think reform is likely? Thank you.
Yes. So the meal voucher, for you that do not know this in Brazil, which is similar to Argentina and to France, there is a regulation that says if you as a company provide a meal voucher, a credit specific for food, you can use the same level of spend as a tax reduction. There is an incentive for companies to give food, to provide food for money to people that work for a company. That is the rationale. The size of this market in dollars is around $25 billion. It is not like huge, but it is very qualified. It is for food, and the person only can eat food. That is the size.
What happened is that we started this movement in Brazil back in 2019 when we saw that the incumbent players of meal voucher companies, they did not integrate with iFood. They said, "I do not want to integrate. I do not care about the consumer." I would love to say a word here, but it is very unpolite. That is basically what they said. I do not give to the consumers. They rather prefer that the consumers had only to go offline to use their card. We went to Brasília, we did the lobby to say, "We have to change that. It does not make sense for the employee." The government changed the law. The law basically now says you have to interoperate and you have to make it portable.
If you have a benefit in one company because your company decided to, you can make the portability to other companies. We set up a meal voucher company. This company is growing fast. We are reaching this month 1 million users, 1 million clients in this, which already makes us a very important player in the market. Now the government is saying, "I mean, why does iFood charge 3% MDR, the fees that you mentioned, and the other companies charge 7%? It doesn't make any sense." There is a discussion regarding cap fees, not in a sense that the government is doing something that is not liberal. The point is it's unfair. It's really unfair. There is a market power in some companies that makes this kind of fee structure work. That is what the government is working. It probably will be reduced.
It doesn't affect us because we have already a good level of fees.
Great. All right. Cesar, I'll get you in the next one. We have to go. We're running out of battery on the screen. We will go to the India ecosystem now. Thanks, guys. Fantastic.
Thank you.
Thank you
for the India ecosystem.
India isn't a portfolio. It's a platform. We are building a lifestyle ecosystem in motion. Here, growth doesn't happen in silos. It flows. Food to payment, payment to loyalty, loyalty to credit, credit to repeat. Every click trains the system. Every transaction makes the next one smarter. Every service improves the others because this isn't user acquisition. It's user circulation. People don't just visit. They stay. They loop. They compound. They do it daily because food drives frequency, and frequency drives everything else. This isn't just scale. It's ecosystem scale. Live. Local. Accelerating.
Powered by Prosus.
Hello everyone. I hope you're having a good time, learning a lot about Prosus, our different businesses. I'm here to talk about a topic that's close to my heart, and that's India. Over the next hour or so, we will be joined by a couple of guests, and then we'll have enough time for Q&A. India is in a unique situation right now: progressive policies, favorable demographics, stable government, and generally favorable macroeconomic situation. Therefore, it's quite a destination for almost all investors and corporates across the world. That's general knowledge. I think that you know that, so I'm not going to talk about that. What I'm going to talk about is Prosus, our history, our significant milestones, and our vision for the future. Before we jump into the future, let me take you into the past a little bit.
Let's go to year 1300. This was a time when almost all trade routes across the world went through India. The Emirs of the Iberian Peninsula, the ancient Romans, the Chinese, the Egyptians, almost everyone was in India. A country so prosperous that it contributed 25% at the time to world GDP. No surprise that it was called the golden bird. Why was that? That was because of the spirit of innovation and enterprise. Innovation was not new to India. It was not new in 1300. Zero was discovered and so on and so forth. Even after that, that innovation and that enterprise has continued. From the grand mines of Golkonda to lab-grown diamonds of today, from the rocket designs that were developed in the Deccan Plateau in India to high-tech space missions launched by the Indian Space Research Organization.
That spirit has been hastened because of the liberalization that the country saw. That spirit has been hastened because of the ever-expanding digital public infrastructure. That spirit we see in our portfolio as well. This is a picture from one of the many local markets in India where micro-sellers come and sell their wares at a specific time on a specific day to a limited set of people. Meesho empowers not one, not ten, not one hundred, hundreds of thousands of micro-sellers to come online and sell their wares at a time of their choice, on a day of their choice, at a price they want and to whoever they want. Food delivery is not a new concept in India. Dabbawalas of Mumbai have been there for ages, for decades.
Swiggy saw that opportunity, made food delivery more efficient and more scalable, and took it to beyond Mumbai, to hundreds of towns and cities across India. At Prosus, we saw that spirit. At Prosus, we saw that spirit of innovation and enterprise, and we backed India when there were questions about the addressable market, when there were questions about exits. We did not know at the time that India was going to grow from number nine in GDP to number four in GDP. We believed, and we believed with conviction in that spirit of innovation and enterprise. Because of that, from our very humble beginnings, almost a decade ago, today we manage a portfolio of $6.5 billion in India, and I'm not including the $4 billion of cash that we have already returned to shareholders.
More than the cash and more than the gains, I think it's to notice the impact that we have created on the society. Through our portfolio, today we address 200 million annual transacting users. We deliver 100 orders per second, and we transact $15 billion in GMV, and we process $80 billion in payments. This is not a disparate set of investments. This is not a few random assets that we have tried to invest in over time. This, to me, is already a loosely coupled ecosystem where companies are coming together, collaborating, finding synergies, and in process, growing faster, becoming more profitable. Let me share a couple of examples with you. There are many, many examples that I can share of this. Swiggy and FarmEasy came together last year to launch India's first medicine delivery on Quick Commerce platform.
Today, in Bangalore, you can get medicines at home, prescription and non-prescription, in 10 minutes. Swiggy owns the customer and the last mile delivery. PharmEasy, another one of our portfolio companies, owns the relationship with pharma companies in the first mile, and they co-own the middle mile. This product, since launching last year, has seen 9x growth in a very short period of time. Another example, Meesho and PayU, two of our portfolio companies. Meesho has commerce data. They see a lot of transactions every day. PayU has payments data. When you marry both those, you can underwrite a merchant much better than you individually can. Therefore, when they come together in a format where they can offer sellers or merchants on Meesho credit, it becomes a much more successful, a much more interesting proposition.
You can see there is a situation where you launch a first-of-its-kind product coming together, and there's another situation where you make the underlying business much more profitable. So far, it's suffice to say that, look, I think it's already looking like a loosely coupled ecosystem. As we see going ahead or going forward, and as we see the world in that ecosystem flywheel that Fabricio shared with you earlier in the day, which had four key segments: food delivery, e-commerce, payments, and experiences. I just wanted to share with you a couple of examples of companies that represent one or two or three of those segments. Let's start with our oldest investment in India, Swiggy. It's been eight years since we invested, and I'm super impressed in how in those eight years, Swiggy has changed from a food delivery platform to a customer convenience-focused multi-product business.
It is not as if these are very different business lines. They have no interaction with each other. No, they are very extremely interconnected. They share the same customers. In some situations, they share the same delivery infrastructure. Of course, they share pieces of the modular tech stack that Swiggy has built. What that means is that Swiggy can now optimize cost, drive synergies, and launch products one after the other in a very seamless manner. Swiggy started as food delivery, added Quick Commerce, which is a substantially large business today, bought DineOut to offer in-restaurant services, launched innovations like 10-minute food delivery, snack, minis, malls, a co-branded credit card. All of that comes together under one roof through the loyalty program. The numbers most of you know. I will not go, at least for now, in sort of very detail on these.
Suffice to say that the food delivery business is growing, growing reasonably fast and turned profitable. We're very happy about that. Quick Commerce, in my history of investing in India for the last 15 years, it's hard for me to find another product with a better product-market fit. It's magic. It's dopamine, the way it works in India. Therefore, even at a very large base, Swiggy continues to compound the Quick Commerce business. Let me give you another example.
[Foreign language].
Meesho falls in the e-commerce segment of the ecosystem flywheel that Fabricio Bloisi presented this morning.
It's amazing how, despite having two large horizontal e-commerce incumbent players, Meesho came from behind and became the largest horizontal e-commerce player. That's because of one thing: because of first principles thinking. India is not one. India is many. It's an extremely heterogeneous country. The way the top 50 million Indians behave is not the way the next 100 and the next 200 behave. They behave very differently. They buy very differently. They interact digitally very differently. Therefore, Meesho is the product for the next 100 and next 200 million Indians where price is the hook and less convenience. Therefore, you will see that in a short period of time, it has now grown to 1+ million sellers, 120+ million listings, and almost 4 million orders per day.
Another example I wanted to give from the experiences segment of the ecosystem flywheel, and that's Urban Company. Another example of first principles thinking.
[Foreign language].
[Foreign language].
There's many nice metrics to share about Urban Company.
There's many good news that I can share. They just recently filed for an IPO, hopefully list this year. Of course, as a company, they're doing very well, profitable, and so on and so forth. I think two things I'm extremely proud of. One, the impact that they're creating in the Indian society that you saw on the video. The second is this number: average consumer rating of 4.8. Just imagine, just think for a moment how non-standard this service is. When you pick a product from a warehouse and drop it to your home, in my view, it's an extremely standard service. Maybe you expect that the product is left at the stairs. Someone expects it's left in the garden, but that's it. It's a pretty standard product. Look at this: woman salon and spa at home.
I know your expectation of what a good spa experience is versus yours or versus someone else's is completely different. Home cleaning, what's a clean bathroom for me versus you, Christian, or versus you, Yuro, completely different. Of course, I like cleaner and probably Yuro does not. It is an extremely non-standard service. For that kind of service, this kind of average consumer rating, growing fast and extremely profitable. Those were the three examples of companies that I wanted to introduce to you in one of the three, each of the three segments. With this, I would like to invite Anirban to talk about PayU, which lies in the fourth segment of our ecosystem flywheel.
In the past, payments and credit meant managing multiple disjointed systems. Today, PayU is bringing it all together in India.
One platform, unified and powerful, trusted by over 450,000 merchants, processing a quarter of e-commerce transactions, authenticating three-quarters of card transactions, and processing half of real-time payments, UPI in India. We process $90 billion in total payment volume. We are the preferred choice of leading businesses powered by a robust tech platform, handling 10,000 transactions per second and 99.99% uptime. PayU combines the power of three very powerful platforms: payments, credit, and banking SaaS. Each strengthens the other, creating a seamless embedded financial experience that will accelerate the needs of India's digital commerce story. Innovation at PayU means impact at scale. Take FlashPay, where biometric authentication replaces OTPs, improving success rates by 200 basis points and reducing card latency by 80%. Or WhatsApp Commerce. Merchants can now sell and collect payments directly inside chat, boosting conversions by 7%-10%. What we're building today is more than tech.
It's the backbone of tomorrow's digital economy. That's where our credit offering ties up with our merchant ecosystem at checkout for our consumers or their partner app for merchant lending. PayU, powering India, powering you.
Hi. Glad to be here. Glad to see all of you. I'll talk a little bit about who we are and what we do and how we are positioned for the next set of, next phase of growth. Okay. As you know, PayU serves merchants with two things. We provide them the most seamless payments, and we provide them solutions to grow their business. PayU is what I call a universal payments platform, meaning we serve all kinds of merchants. We serve commerce, travel, food, all of the digital commerce segments. We also are one of the largest players in bill payments and tax payments and every kind of merchants.
We have a special place with the top 500 merchants in the country. We are their preferred provider. We also serve mid-market and SMBs in the country. PayU's strategy is twofold. We want to be the best in class in payments. Second, payments is a platform for broader cross-sell of solutions, including credit for our merchants and their consumers. I'll just tease apart this a little bit. If you see on the top, you will see two sides on the payments. Like Stripe, like Adyen, we are a merchant payments company, which is this side. We are also an issuer processor. We provide most of the issuing banks' processing services for processing payments. Three in four card transactions in India are processed through PayU's authentication platform. Recently, we invested through a strategic partnership with a company called Mindgate, which processes one-in-two UPI transactions.
If you know, UPI is the platform that puts India on the digital, global digital map, which is real-time mobile money in an interoperable way. Mindgate does one-in-two transactions in India. Why are we on this side? What is this two-sided play? Look, digital payments is not a solved problem. It is not a solved problem in any part of the world, especially in India, where the median smartphone is a $100 phone. There are many, there are lots of friction, lots of hops. How do you simplify those hops? How do you create the best experience and the highest conversion? Is how we ended up being on both sides. We have the best infrastructure, the best pipes. We are also able to launch solutions the fastest.
As a result, you know, we have the best conversion, which leads to most of the top merchants in the country working with PayU as a leading provider. That is about best-in-class payments. The second is payments provide an opportunity to cross-sell many solutions into merchants to grow their sales, things like WhatsApp Commerce, things like offer engine and couponing engine. Payments also lead to credit for us. We lend to the suppliers and sellers of merchants. Ashu gave the Meesho example. We also do BNPL and checkout finance for the consumers because India is very fundamentally credit constrained. This platform of best-in-class payments with embedded credit is how we are very uniquely positioned in the digital commerce ecosystem. I'll talk about payments a little bit, then talk about credit. In payments, we have continued to grow very nicely over the last few years.
We have retained our market leadership in terms of revenue market share. We are also one of the three profitable players in the industry. By the way, do not assume number two and number three are the profitable players. Profitability is something that is an intentional strategic thing for us. It comes from two things. It comes from serving the largest merchants and having more share there because we then enjoy a little bit better economics. It also comes from the ability to cross-sell many different solutions. Cross-sell makes up about a third of our revenues across banks and merchants. Those two things power profitability for us. The third dimension is also very important. Increasingly, with scale, we see a lot of operating leverage. Our cost per transactions have gone down significantly over the last three years.
Here, there is much more scope to grow. Like with Fabricio, we call it being the lowest cost producer. We see a lot of scope here, especially with GenAI. We are experimenting with many different parts of GenAI. What's already true for us is fraud risk management, customer service, onboarding, stuff where you actually have a lot of people in payments or in financial services touching the stuff. That's where we see a lot of scope for improvement. Of course, in tech, both in testing, in security reviews, in a lot of this. This is work, again, like Diego said, some of this we started years before this all became in fashion and very strong partnership with Euro and team to take this forward. There's also a lot more to be done in terms of making the salesforce much more intelligent here. We are scratching.
This is chapter one of what GenAI can do. If we continue with GenAI, I think we can keep driving this cost per transaction down very significantly. With massive scale and profitability and lower and lower costs, payments also will produce great economic returns over time. Before I talk credit, I wanted to talk about something that the whole industry has gone through over the last three years. Over the last three years, the most important thing that happened in fintech was we went from being unregulated fintech to regulated digital financial services. We are fully regulated, both payments and credit. Credit also saw a moderate cycle over the last 18 months. Most of those changes are baked into our baseline now. It did impact short-term growth and profitability for us and for the industry. I see a very clear path forward from here.
We are now fully licensed on both sides of our business. Going forward in the longer term, I actually think regulations are a moat. They increase the entry barrier from anybody starting up to now we have 10 players of note and really three players with real serious scale in this business. You know, I think most of this is now well behind and the future looks really exciting. I'll talk about credit. Our credit is very integrated, synergistic with the lifestyle e-commerce business. Partners like Swiggy and Meesho and lending to the sellers and restaurants there using the cash flow data on both sides is how we do supplier and SMB lending. For consumers, we provide them buy now, pay later, and checkout finance. We are one of the leading providers of this in the country. I think this gives us an edge.
We can do integrated economics across payments and credit and value-added solutions with merchants. Overall, it creates much more stickiness and helps merchants drive their sales up. Our credit focus, we have tried many things. This is where we will be focused in credit. Coming to our numbers, you know, over the last three years, we have grown well in spite of the speed bumps. We have grown about 30% CAGR. We have also diversified significantly, even as payments have grown. Both SaaS as well as credit, both of which are structurally higher margin than payments, have also grown significantly. You see both growth and diversification in significant amounts over the last few years. In summary, I'll say three things. Look, this digital India and the Indian digital economy is a rocket ship. It's going to grow 5x over the next five to seven years.
Two large macro forces. One is the digitization itself. Second is consumption driving credit. And PayU is well poised to drive both of that and benefit from both of that. PayU is the only platform truly, deeply ingrained into the very fabric of digital commerce in India. The largest merchants and ecosystems, the largest banks, we are already their leading partners. This position, you know, will enable us to grow very strongly and continue to deliver great results. With that, you know, thank you. I'll call on Ashu for the next piece.
Thank you. Thank you, Anirban. Very exciting. I would have loved that welcome video back, actually, Eoin. It's very interesting. Just to share where we are today with three companies that have hit liquidity: MakeMyTrip, Flipkart, and Swiggy. We have now unlocked $4 billion of cash gains. This is not the end.
We have many companies who are mature and ready to list, and therefore will add to this gain number in the short term. That is also not the end. There are many in the pipe which are at a stage which probably a couple of years from now or three years from now should provide such gain unlock at the time. Of course, for the long term, we are sowing the seeds as we speak. Are we happy with that? I think all these companies on the slide are all structurally good companies. They will continue to grow. They will perform well. They will independently do very well. How can we improve their journey? How can we increase their pace of growth? How can we increase their profitability? I think that is where the process becomes unique in all of these situations.
We will, from now on, be much more intentional. I showed you a couple of examples of enabling interaction between companies, but we will now be much more intentional in building this ecosystem, in putting companies together, driving synergies, and helping them grow fast. We will be more intentional now in the type of knowledge that we share, which includes AI, but beyond AI as well. We have food delivery across continents. We will ensure that knowledge sharing happens at that level also. At the same time, we will continue to invest in areas which will make our ecosystem much stronger and much richer. If we are able to do that, and if we are able to change the trajectory of these companies substantially, what we will be able to unlock will be far bigger than what we have today.
What we will be able to unlock is a truly dynamic, a truly humming, and a large India tech ecosystem. Thank you so much. Now I have a surprise guest for you, Harsha, the CEO of Swiggy. We talked a lot about Swiggy during our talk previously. Harsha has been kind enough, flown in all the way from India. Thank you, Harsha.
Hi, everyone. Thanks, Ashu. Thanks for having me here.
Let's start with a few questions from my side, and then we'll, of course, open it up for the audience. Harsha, what is Swiggy? We are a food delivery company. You have quick commerce. You have dine-out. You have minis, malls. What is the soul? What is the vision? Where are you going? How does all of this come together?
Cool. It is a deeply personal topic. Everything we do at Swiggy is united by our mission.
It's all about offering unlimited joy through unparalleled convenience. I think why it is deeply personal and super exciting is also all to do about the Indian context. I think India is going through its own moment, you know, strong economic growth overall, GDP per capita going well. I think we're on our way in general with lots of exciting things to look forward to economically. I do think some of that is also coming at a cost, at the cost of maybe some basic quality of life indicators, like you just don't get like maybe in some of the developed countries, like on a good day, my commute to work takes an hour. You just can't help but wonder, saying, "Okay, we're getting all of this economic growth.
Can we do a little bit more to let Indians enjoy their economic growth and not just get it? That is at the center of like what we get excited about at Swiggy and everything I wake up to do is all about that. Everything is just about that. I think what's also deeply fulfilling for us is that, you know, offline, India has it quite bad in convenience. Online, I think we would probably be the world's best, if not amongst the world's best in terms of the range of offerings. You have so many offerings. Like quick commerce doesn't work in any other part of the world. Like what you talked about, Urban Company services like that don't exist in other parts of the world.
Even if you take like businesses like food delivery that exist in many, many countries, I'm biased, but I do think that India food delivery in terms of like service parameters and you know how the overall experience works would be in the top three globally. For us, as you know, a developing country, so-called like to be a very developed force in online convenience, I think fills my h eart. That is what we are here for at Swiggy.
We're very happy to be associated. I have an anecdote to share. When my mother comes to Bangalore and visits me, she, and this is an example of how much product love people have for Swiggy. Her biggest quip is, "I see the Swiggy delivery guy more at home than I see you." This is the number of times the guy comes home.
Every day, there's three or four or five deliveries. Thank you. Harsha, let's go back a little bit. When you started in food delivery, I still remember I reached out to you in 2015. 2015-ish, yes. And we were doing 300 orders per day. There's probably like 20 food delivery companies in India at that time. What do you attribute this success and this growth and where you are today? What's in the Swiggy org that makes this possible?
I think in 2014, 2015, the thing that really worked for us is that I guess we just stayed true to the concept of unparalleled convenience. We didn't think that unparalleled convenience was going to come when the restaurant was going to do the deliveries. It was going to be unparalleled inconvenience in India.
For us, it was quite obvious that this is the only way you can actually make a customer happy. I think we were in the significant minority amongst a bunch of companies that just chose to do the same thing, but we played it differently at the time. At that time, it was quite unconventional. Just that one decision and executing on that and figuring out how to make that on-demand model work in India was the first. On-demand mobility had existed, but on-demand mobile delivery was a whole new thing. To be able to figure out how to let a service like that exist and scale it in 10 cities, just that allowed us the opening because our cohorts looked very different from our peers, and that just allowed us another day to fight.
It's amazing to see the growth from 300 million, 300 orders per day to millions of orders per day. Was the evolution of quick commerce similar in terms of sort of how you thought about it and then how you executed on it?
No, I think a lot has changed. The quick commerce that we see, I mean, if you look at food delivery today, largely 95%, it looks like food delivery that we did in August 2014. I think in quick commerce, what defines quick commerce is changing like every three months, six months, I think, because it's such a large, it's effectively retail at some level, and there could be so many user segments, so many categories. A lot has changed since we first kind of began our journey with quick commerce.
When we started, we assumed that we were going to build a 7-Eleven equivalent for India that comes in 30-45 minutes. We just thought of quick commerce as probably just a distress, top-up, indulgence, impulse purchase mission destination. I think a lot has happened since then. I think consumers were surprised to find like batteries in year one. Consumers were not surprised to find bedsheets on the platform last year. I think consumer imagination also expands quite crazily as platforms get the salient. I mean, it is still kind of a blurry line. I think it is almost like all the players in the category are figuring out what the boundaries of quick commerce are. You talked about pharmacy. I do not think we thought about all of this when we started, but we are finding out as we speak.
Yeah, truly, I think quick commerce in India is quickly becoming the go-to format for almost all online retail purchases. Food delivery has evolved in India to a stage where you're profitable, you're growing fast, reasonably fast, et cetera. In your mind, is there a question on profitability in quick commerce? Is this TBD, or do you have comfort based on what you see today?
No, I think to the doubters, skeptics, I can see why it looks like that. We've also gone through this journey in food delivery. We look and understand the levers of the business quite well. We feel pretty good about the size of the category is staying in line with expectations. I think we continue to see strong growth. It's already larger than food delivery in India, the quick commerce category, and it's not showing any sign s of slowing down.
I think the levers exist. It's up to us now to make sure that we navigate our own business towards that place.
What has changed for you, Harsha, personally after the IPO? Has life changed considerably?
If you ask any of my leadership, management team, all of us, I think the one word that we'll probably say is that we're feeling a lot more liberated.
Oh, sure.
After going public, because when you're not public and you're about to go public, everyone has their own version of what being public means like. That just creates all sorts of fear, uncertainty, doubts, saying, "Am I going to be able to do this? Will I not watch the stock price? Will I be able to be short-term, long-term, whatever?" I think our experience has been quite good.
It is an unknown, but then I also realized that as a founder, I've gone through a million unknowns in 10 years. What's a couple more? I just didn't think about it like that before. When you think about it, I think our market, I mean, even investors, analysts, the overall feedback has been constructive, whether it's good or bad. I like that it's constructive. It's not like myopic or anything. As long as you're able to explain what you're trying to do, I think there is some comfort around, "Okay, this is what happens," and then you do well. You get a pat. You do badly, you get a knock. I think that's all right.
Yeah. The good news is the feedback comes every day, and the bad news is the same. Probably one last question. It has three parts, Harsha.
One is, where is Swiggy going? What's the end game, if at all, for Swiggy, for food delivery, and for quick commerce? Right. What is the big solve? What is the big unlock?
Look, I think from here, everything we're doing is going off the rails of the same thing that I started my conversation with, India's economic growth. I think if I had to oversimplify overall Swiggy's growth strategy, everything we do serves maybe the top 150 million consumers in the country. That's the limit of maybe what we go after, 150, maybe 200 with a little bit of a stretch. And our deep belief is that every passing year for the next 10, 15, 20 years, more and more users are going to come into the part of the pyramid that we play in. We just need to keep solving for the same problems.
We need three, four large businesses that will just compound based on the users flowing into the part of the pyramid we play in to build an ultra-large outcome. That's on the business side. If I had to go in each category, it has very, very different nuances. I know you mentioned like the food delivery business is growing reasonably fast, but I think it has the potential to grow reasonably fast for a very, very, very long time. I think if you were to just put that in perspective, Beijing has more restaurants than India as of today. That's how early we are in like India's eating-out history. That is changing. As every country goes through its own economic growth, then discretionary income goes fast into the restaurant industry. The rails that we're building on top of are definitely going to grow faster than the economy.
For the last year, for example, 13.5% was the rate that the organized restaurant industry grew at. What we love are the rails. If I have to use a crystal ball here, I mean, if the industry is going to grow at like 10%-15%, and we can continue growing at 15%-20%, what you will see will be very unusual for most countries that have food delivery today. We are already like 25%-30% of the overall organized restaurant industry, Swiggy and Zomato. I think that number is going to start hitting 40, 45, 50% of the overall restaurant industry. At some level, the industry is going to evolve very, very differently because of how early online food delivery happened in the country's history of eating out. I think that is something, it is not going to happen by itself.
I think we have to do so much more because our job is not just a channel transfer, saying people are already eating enough offline, so we just need to convert that to online. I mean, every category, in quick commerce, it's about, of course, like another super large category, I think, never-ending penetration that will continue over the next couple of decades, I guess. Overall, if I have to think of what we will be doing, what the soul of the company will be 10 years later, it'll be the same. Hopefully, between the things that we're doing today and maybe one or two more big game openings that we have, I would love it if we're able to serve 100 million users 15 times a month.
That's amazing. Last eight, nine years, association with Swiggy has been amazing. Thank you so much.
Thank you so much.
We look forward to an even longer one. With that, ladies and gentlemen, we'll have the Q&A. I'm looking for Eoin and Anirban to join us. For all the... As we...
Very suspenseful music as we put the chairs up. I realized it was an incredibly cruel thing to do minutes before lunch to talk about food delivery. We are almost there. Do we sit or do we stand? We might as well use them.
Yeah, let's stand.
Oh, Jesus, now it's awkward.
Yeah,
I'm tired. Okay, let's get straight to questions. There's one right here because we haven't had any questions from over there. This is, I'll go back to you, Cesar, I owe you, Monique, then next.
Okay. Perfect. Thank you. It's Michael de Nobrega from Avior Capital Markets.
The first one is that in India, Prosus has quite small interests into all the different companies. As to Tash, could you elaborate on the influence that you see Prosus has over these companies to collaborate, share information, cross-sell? We have seen different examples of, for instance, Rapido entering food delivery. How do you deal with the individual interests of the companies?
Sure. I think on the first one, win-win for everyone is the only way we can influence in this situation. I gave you a few examples, Swiggy and PharmEasy coming together, PayU and Meesho coming together. There are many other examples where we can either through enabling these connections at the time or bringing super specialized expertise from our AI team or from iFood. We are able to kind of, for example, in Swiggy's case, Swiggy and iFood work closely together.
Those combinations are already happening. We will, as I said, we're going to go much faster on that. As for your other question, I think Harsha is a better place to answer because Harsha owns 13% of Rapido.
How much?
Whatever.
Not that much, but we are also an investor in Rapido, a small investor. Look, I think each company has their own strategy, and sometimes they may come after new openings that they're interested in. I'll just maybe take you a few steps back into food delivery history itself in India. I think Ashu talked about how there were 20 players in 2015. In 2017, Uber and Ola threw their hat in the ring. In 2019, Amazon threw their hat in the ring. In 2021, there was an entry of like a more decentralized network called the ONDC that was about to come in India.
We, and credit to us and Zomato for having seen these threats off, I genuinely think we do a pretty good job of serving the consumer. It is not easy to just get an opening that you can go take a home run with. I think it will be interesting to see if there is an alternate take to food delivery that can grow the category some more because we are waiting for some more growth as well. We are definitely super, super agile and paranoid. I think these are reactions that we are going to make. If we see a new opening that is going to be exciting, we are going to be all over it. Asha talked about the modular tech stack, et cetera. There was an emerging potential opening in food delivery around, let's say, cafe kind of food.
There was a player, Zepto, that launched a Zepto cafe, et cetera. The moment we decided to do something of our own as a pilot called Snack, it took us 16 days to put the application out there. We are going to come, if there is an interesting opening, we will be out there in weeks trying our own luck with the customer to grow the category. We are not going to wait and watch.
Great. Thank you.
Okay, Cesar, are you ready? Do you have a question? Okay. Go long. Just give it to him. You better catch this, Cesar. A lot of pressure. Well done.
Cesar from America, I have two questions, please. The first one is on quick commerce. Do you expect some consolidation to happen? Do you think you are going to be part of it?
Second one, still on quick commerce, do you think this could evolve into distributing more products than just groceries and pharmacy? Do you think it's moving closer to actually e-commerce and probably distributing a wider range of SKUs? Superb. Thanks.
For the first question, I think the real answer for all of these questions about market structure and consolidation is market structure is a function of market size. I think in 2019, we found out that food delivery has its own limits. It's going to be like a $4 billion-$5 billion category growing at a certain pace. That was a moment of clarity for the category that consolidation was imminent and Uber sold to Zomato, et cetera. At this point, it looks like we're headed for a $30 billion-$40 billion-$50 billion kind of category in quick commerce in three to five years.
I would assume that that kind of size can support more than two players. Can it support like five, six? Unlikely. We should expect some more consolidation amongst, like if you look at the structure of the category today, there are three major players and four fringe players. There would be some consolidation. In some cases, there may not even be a need for consolidation because, you know, let's say the legacy e-commerce may want to continue the offering in a bit to stay relevant for their consumers. It's going to continue, but it's not going to really expand the pie as we're seeing already today. As for the second question, honestly, I think that's already changed the breadth of the selection that is on offer. We would now have like the widest coverage of the widest selection at the country level.
I think we've come our own long way. We started quick commerce with 2,500-3,000 SKUs in 2020. You would now see between 30,000-40,000 SKUs on quick commerce already. The kind of needs and the categories that we're going after are already changing. However, I wouldn't expect that every single category that can be done in traditional commerce can be done in quick commerce. Like you can't do fashion, probably. I still don't know how that would work because fashion is a very selection-oriented game. You need a lot of models sometimes. White goods is probably going to be hard. There are some reasons why maybe a little bit more expertise will be needed in stuff like mobile phones, et cetera.
There are a whole host of other categories in home and kitchen, toys and stationery, beauty, personal care that are all going to be, our sense is that they're going to be QCOM dominant versus ECOM dominant today.
Anecdotally, the behavior is changing as it flows. The first port of call is inherently becoming the quick commerce platforms. If I find it there, then I go to something else.
I think a lot of this is also the expanding imagination of the consumer. It's almost like our null searches drive our next selection edition. Nobody was searching for bed sheet in year one. As users got more and more selections, they were like, let me try my luck. When enough try their luck, they're like, okay, let's see where we go next from here.
I mean, Jeff Bezos said that the users have like divine dissatisfaction and they express it on the search box every day.
Thanks. Monique Pollard from Citi. I had a question for you, Ashutosh, in terms of how you think about these listings that might come up in the Indian ecosystem. And do you see them basically as an opportunity for value crystallization or capital returns? Or do you also see them to some extent as an opportunity to potentially buy more of those assets, own more of those assets, particularly if there are assets that you see as incredibly critical to the ecosystem? I'm just thinking you're going one way, and then we've seen in Latin America and Europe some purchases of public market assets.
Absolutely.
Look, I think anything that enriches our ecosystem and makes it more dynamic and more vibrant and strengthens it, we will be interested in it. Now, the question to buy or not buy is beyond that. There is a question of valuation and a hundred other things. The focus at this time is those four segments that Fabricio talked about. Anything that falls in that is, of course, of interest. Other than that, I think we are long-term believers in almost all of these businesses. More often than not, we will continue to keep our trust in them.
All right, Cesa r. Robert .
Yeah, thank you. Robert from Kepler Cheuvreux . I have a question about PayU. I saw that you made a couple of acquisitions or investments in companies that were innovative in certain fields of the payments landscape.
Why did you decide to make those investments instead of developing those technologies, those capabilities internally in your own company?
Sure. Look, we are mostly builders. We build, but once in a while, you find an opportunity that accelerates your journey. That is when we do an M&A or a strategic partnership. In this specific case, the one that we did last year, we took a 43% stake in a company called Mindgate. Mindgate is already deeply entrenched, providing UPI platforms to all the banks. We are very strong inside the banks. We are very strong on cards. On the merchant side, we are strong in all the areas. UPI, they have one in two transactions in the country today. It is hard to replicate that. We thought it was the opportunities in creating much more newer UPI solutions. Credit on UPI is just happening.
UPI is going from being a payment method to a full payment system. The opportunity to do many more new things is much bigger instead of trying to just beat them at their game inside the banks. That's where we thought it's better to partner strategically.
Perfect. All right, we'll do Warik and then go long to Mandeep.
Thanks. Warwick Bam from RMB Morgan Stanley Just another one for you, Anirban. Just in terms of the unit economics, I mean, you've got significant scale. Certainly, the transaction volume is very impressive. Just give us a sense of how your positive profitability, what sort of margin you expect in the payments business. You're getting - 2% at the moment. How long should we wait before you get to sort of sustainable profitability? What does that look like?
Yeah, fair question. Second half of the year, we broke even.
Look, long term, till UPI pricing comes in a bigger way, it's not in our control, so we don't account for it. I think near term we'll be in the 5% range soon enough. From there, I expect it to expand. That's just core payments. The SaaS and the vast layer on top adds another few percentage points to that business. With more scale, hopefully it keeps compounding for a very, very long time. We'll see, hopefully, much more profitability.
Thanks.
All right, we're going over here to Warik. Go long there. See? Use your rugby skills. Uh-oh, Jesus. Well done. That could have gotten awkward. Tough catch. Nice. Very good.
Thanks for taking my question again. Just a question probably for Ashutosh. India is already a $4 trillion economy, and I think the ambition is to get to $10 trillion.
From your ecosystem perspective, which verticals do you see benefiting the most from it? Do you need to get into any new vertical which you do not have right now to capture the additional growth? Just a question for PayU as well. I think payment is a very exciting segment. I wonder what is the overall market opportunity you see for your space? Is there any further verticals you are not operating in right now which can be of interest for you? How big can PayU actually become? Thank you.
I think I can start with, by the way, great chat this morning. Thank you for that. Thanks. I think, look, India is a very interesting market at this time. You look at cement companies, like forget tech for a moment. Cement companies in India are growing 15%-16%. Therefore, you are right.
As the tide rises, it will raise all boats. I think the technology companies will disproportionately enjoy the benefits of that raise. Now, as to your question on what will benefit the most, I think consumption economy in general will benefit the most. Because what is happening is India has reached $2,500 in GDP per capita. When that number was reached in China last, any incremental GDP on top of that was spent more towards discretionary spends. Therefore, discretionary spends like food, fashion, retail, commerce, travel, et cetera, et cetera, will benefit disproportionately. Thankfully, in almost all of the segments in that ecosystem flywheel that you saw, we have representation. I think what we are looking for is how do we go beyond the top 100 million-150 million consumers that Ashutosh talked about? Is there a new product? Is there a new Meesho, for example, for mobility?
That's where we found Rapido. Or is there a new Rapido, for example, I don't know, travel? I'm just making this up. The next demographic cross-section of the society will be our continued focus. Anirban.
Okay. Look, overall, as I said, digital commerce will grow 5x. If you break it down, payments will grow about four- to five-times over the next five to seven years. Credit will actually grow seven times. Because as India becomes middle income, as Ashut says, like there is the India is very fundamentally credit card constrained. SMBs also are very, very underpenetrated. Massive opportunities. We say India will, we call it 247 inside the company. India will double, payments will at least become 4x and credit will become 7x. We are deeply entrenched in the ecosystem. Hopefully we can meet or beat those numb ers.
All right. Will.
Hey, guys are getting good. Hi, it's William Henry Packer from BNP Paribas. Examine. You surprise me every time. We heard about the e-commerce-led LLM growth focus for Prosus Group. We also heard that Swiggy currently doesn't share its data with Prosus. Could you talk us through how you think about the risk and opportunity associated with any such data shar ing? T hanks.
Fabrício planted that question. Does he?
I don't know which specific kind of data.
I think, look,
we're a public company, so I'm not even sure how much w e do.
To the broader point, actually, there's enough connects between Eoin's team and Madhu at Swiggy so that we are probably sharing almost all of our findings, our best practices, which we implement elsewhere in the world with Swiggy. Specifically on data, we may, I don't know, there's a DPDP question involved or what? Data privacy question, I mean.
In terms of absorbing all those trillions of data points that are developed from Swiggy as you interact with these consumers, currently you do not share them with Prosus Group. That is going to be the case for the foreseeable future.
We w ill have to come back to you on that. I think as for the best practices, as I said, in food delivery, using AI and without AI, we are completely, completely connected.
All right. Let's go one more. We got eight seconds before lu nch.
Hello. Hello. Thank you. It's Michael again. Just a quick follow-up on the last question on PayU. You mentioned that you are focusing still on the merchant credit. Where do you see growth on the consumer side? It seems that could be a much larger potential. What is preventing you from going into that?
No, we also do consumer credit.
Consumer credit is actually 60% of our book, and merchant credit is 40%. We do, as I shared in my slide, we do BNPL on partners like Swiggy. We also do installment loans because fundamentally India has 50 million credit cards only. In real terms, there's the demand for and probably the opportunity to serve 150 million people. Big opportunity, agree. We just are doing it very synergistic with our ecosystem. Once we start a relationship with BNPL or checkout finance, as we call it, then we cross-sell them personal loans and consumer loans. It is a big opportunity, yes.
All right, let's leave it there. It's halftime. It's lunchtime. Thank you for your attention for the first half. Lunch is upstairs. Please proceed at an orderly pace. I know you're hungry.
Check out the QR code that's over there by the hologram to interact with the AI agent. Enjoy lunch. We will be back here in just over an hour. We will be dealing with the European ecosystem.
Minus 30 seconds. Aí foi que o barco desabou. Nessa que meu barco se perdeu. Nele está gravado só você e eu. Só você e eu. Yes, okay. Okay. You're right. You have to invite the people into the chairs and do your thing. Here we go.
Good afternoon, everyone. I'm not going to stay here too long. What I want to do now, while you take back place in your chairs, I want to have a very short conversation with the hologram, you know, the hologram which is down there. That hologram there has been learning from your interactions with the CMD app all morning. This is the QR code for the app, in the same way that the LCM learns from millions and millions of interactions with our platforms. Let me ask the first question to the hologram.
What are the three most common questions that you have asked to the application this morning?
Oh, sure. So the most common questions people have been asking are, what's the strategy and future direction of Prosus, specifically around lifestyle e-commerce ecosystems? What's the schedule and who are the speakers? And what's on the lunch menu? Those have been the top three so far. Want me to dig into any of those?
No, no, that's—thank you very much. That's very interesting, of course. Another question that we got in the last, let's say, 20 minutes of our lunch is that, but how does the large commerce model work?
So the large commerce model, or LCM, is basically an AI model that leverages a ton of cross-platform commerce data. It's all about making super smart AI-driven decisions to optimize everything from pricing to customer experience.
It's a big deal for their e-commerce ecosystems. Want to hear more specifics?
Okay, that's not too bad. By far, however, the most common question we had is, how do you pronounce the name of our Head of Investor Relations? So the guy that looks like Ed Sheeran.
Haha, you must be talking about Eoin Ryan. That's pronounced like Eoin Ryan. And just for the record, I'm not confirming any lookalikes.
Anyway, we're set. Thank you. You do look like Ed Sheeran, right? It's a matter of fact. You have the application here. Please interact with this. The voice that you heard is the voice that responds to you. And it has been learning, and there's a lot that he can share about Prosus. With that, I'll welcome you to the second part of the program today. I'll see you later. Thank you so much.
Europe, a region rich with deep potential, ready to ignite its next chapter. Like we've done in LATAM, what we are doing in scaling India, Europe is next. We're here to ignite it, restore it. This isn't a collection of companies. This is a system in the making. OLX connects scale with community. eMAG, a one-stop engine for digital retail. Jet, poised to deliver the velocity and the frequency that scale ecosystems. Iyzico, running seamlessly underneath. Secure, integrated. Separate brands, shared intelligence. This isn't a single transaction. It's a multi-brand, multi-platform system. Each user insight improves every service. Each click trains the system. Because in Europe, synergy isn't soft. It's scalable. It's strategic. Our brands don't compete. They compound. Faster, smarter, together. This is how platforms become a system. How a region becomes a growth engine once more. Powered by Prosus.
Hello everyone. You thought I was gone?
No, I'm back. I'm in charge of Europe. I'm in charge of everything that, if something has a problem, I'm in charge of that too. So I'm back. Hope you enjoyed our lunch with brigadeiro. Many people asked, what is this black thing? It's brigadeiro. If you are in a party in Brazil, a children's party, that's the best sweet that we eat in Brazil. I ate many brigadeiros. If someone on your table is, they are not going to be here after the event. I'm going to eat them all. Hope you also tasted the brigadeiro. Let me talk about Europe. In the last, in the morning, we had Diego talking about Latin America, and we have Ashu talking about India. Who is in charge of Europe is not me, actually. It is Roberto Gandolfo. Roberto Gandolfo was in charge of food delivery in Brazil until one month ago.
There was Diego, and in charge of all food delivery operations in iFood is Roberto Gandolfo. He left Brazil. He was fired from iFood, and he started working in Prosus on July 1. Next time we talk, today I think he's unemployed, but July 1, he's going to be in charge of making sure the ecosystem works in Europe. He's out, he's on vacation because actually he's unemployed today. He arrives in Amsterdam on July 1, and I'm going to make the Europe presentation. You have to deal with me. Next time, Gandolfo will be here. Roberto will be here to talk to you about the results on what I'm going to show you today. First, I am extremely bullish about Europe. Extremely bullish. We have so many talents, very good in technology, in development, very good in some specific areas.
For example, I was in Cambridge and Oxford the last week, seeing many deep tech companies, great universities. We have many startups here and many unicorns. The size of this market, as I showed in the morning, is more or less the same as the U.S. The reality is, even though we have so many good ingredients, our six biggest companies value half a trillion. The six big companies of America value $15 trillion. Oh my God, for sure there are opportunities here. In food delivery, our biggest company values just the European part, EUR 3 billion-EUR 4 billion. The biggest one in the U.S. is EUR 80 billion. In China, $100 billion. There is a big potential if we can explore and succeed on that. What is our plan? Build on top of a strong foundation. OLX, I put here the two biggest companies.
OLX is quite big, quite successful, quite good. If you are thinking, what is the secret? Christian is going to come here and show you everything. Prepare for the best part of the day. Together with OLX, we have Just Eat, as soon as the European Commission approves. It is a very good base. On top of that, lots of technology. We started to have an ecosystem in Europe. I do not know if you have it today, but we started to have, putting together Just Eat, OLX, but also eMAG and Iyzico. A quick overview on that. First, actually, we already start. It is a good start. That is how it looks like, this ecosystem last year. In 2025, the numbers are going to be better than that.
Last year, we were talking about almost a billion transactions, $25 billion dollars or euro, I do not know. Hope it is euro because euro is much more valuable now. Let's say EUR 25 billion here, 100 million customers, $7.3 billion. It is dollars. Okay, but let's increase this goal soon to be EUR 7.3 billion, but today $7.3 billion and $0.7 billion EBITDA. Just to start, we are going to do much more. That is our starting point. This starting point is based on first OLX. As I told you, we are very bullish on OLX. OLX is doing good. It is growing, expanding its margin, is competing well, good product, good leadership, good culture. You are going to hear just after me the main program of the day, Christian talking about OLX. Now you have to deliver. Now, no pressure.
OLX, it's a very good foundation for a strong European ecosystem. We have more than that. We don't talk that much, but actually, I like very much what we are doing in Romania. We have eMAG. eMAG is an ecosystem. We have Genius in the center. We have not only the eMAG. eMAG is like the Amazon of, or even better, much better than the Amazon of Romania. Very high market share. Together with eMAG, we have many other businesses, including credit, same-day delivery. You saw all the lockers here, so you can deliver for millions of customers. This is the same day, fresh food in groceries, fashion days. We have a strong player in Romania. In my opinion, one of the reasons we didn't talk a lot about that, or you didn't ask, is because this business was losing money for some time.
Tudor, that is here, Tudor, raise your hand, made an amazing turnaround the last six months. What I challenged Tudor was we want to have a best-in-class in the world company doing retail, and our benchmark is Mercado Libre. The bottom line of eMAG grew dramatically. We overdelivered our budgets. Our expectation for this year is to grow many times, I think around five times growth in their EBITDA. I think in their EBIT, five times growth. The big challenge for them is same level of efficiency in terms of profitability than Mercado Libre. That is what I believe they are, that is what they are doing, that is what I believe they are going to do. Because I have more information than you that is not only seeing the numbers, but seeing the grit, innovation capacity, speeds, communication of the company.
I have lots of expectations we are going to keep overdeliver in eMAG. Tudor is not making a presentation today here, but he is going to come on stage for Q&A. Please make good questions about eMAG because next capital markets day, this is going to be the star. Besides eMAG, we have Iyzico. Usually, we do not report in detail about Iyzico, but Iyzico is growing a lot, 87% in local currency. It is focusing in Turkey. We have also not only payments, pure payments, but their wallets with a very good product, growing well, leader. I believe keeping this growth rate, it is going to be also an important part about our ecosystem in Europe. As I told you, food and payments and fintech should work together to create a strong ecosystem. We have Orkun here. Orkun, raise your hand.
Please make questions about Iyzico in Turkey when we have the Q&A in a few minutes. This is the three businesses we usually talk about. A few quick numbers on Europe. Today is around $3.5 billion. I believe it is going to grow more than two and a half times. This number does not include today, obviously, Just Eat Takeaway. We do not know how we are going to start including them, but you can expect a lot of growth. The results in this area are good and growing, mostly OLX this year. However, we expect a lot of growth here from OLX. Iyzico is quite profitable too, but we expect a lot of growth for Iyzico. EMAG this year is going to contribute a lot with this EBITDA too.
We have a quite good growth trajectory that is going to grow at least three times in the next few years. It's going to be much more, I'm sure. These are the big numbers. What Eoin told me was just give a quick interview, a quick overview. What they want to ask you is, and Just Eat Takeaway. Talk a little about that. That's why I am the volunteer to talk a little about Europe. I'm going to show you some of our perceptions, a few numbers, a few of them you already know. I'm going to show you a little how we see Just Eat Takeaway compared to iFood. I, as I told you, cannot answer on questions on when you are going to do that, exactly what you are going to do, because we now have to wait for the European Commission.
What's our story on Just Eat Takeaway? First, it's a very interesting company. We start, our starting point is a company that is leader in the U.K., Netherlands, and Germany. This is very strong, very valuable, and we are very excited to have the leader company in these three markets. We are also in another 14 markets. Look, to be leader in the U.K., Netherlands, and Germany is very valuable. Hope you can dream with me what we could build starting from leadership in these three countries and having 14 more markets. 650 million orders, close to $400 million EBITDA, 360-380 is the guidance they gave. Remember, we cannot talk about their number. This is their public numbers. 60 million customers, 350,000 restaurants. What can we add? Prosus has a strong tech record on innovation, growth, profitability. You know a lot about that.
We showed you Euro and Diego talking about exactly that in the last presentations. We have Jet has strong positions, and we have a clear plan to accelerate Jet's performance. We're not going to share the plan in details, but we're going to share what we can and how we see the future of Jet here. Moving on. In the first slides of the day at 9:00 A.M., I told you stock day paradox. We believe we are going to succeed and have the best food delivery player in the world, but we have to face the brutal facts. The brutal fact that you are thinking is, first, we have some business performance issues, let's say it this way. Many people think the business could deliver more. I agree. I have to be a little, how can I say? We are still finishing the transaction.
We have not approved yet, the approval yet, but I agree with people that think that Just Eat should be delivering much more. Much more. I agree with you. Business performance should increase. Second, it is a 10,000 people company. We have to integrate culture, align people, make the whole company communicate better. It is difficult. It takes time. Third, we need world-class tech and innovation. This is a requirement to just succeed. A few comments about that. Many people, not you here in this room, not you looking in the streaming, but many people said we cannot change the culture in Prosus in a few months or in one year because it is a too big company. It is old. What can we do? Actually, the event today is about that. We did it.
I'm very, very confident we will quickly move Just Eat Takeaway to be much more oriented to results, growth, and innovation. I think the presentation today reinforced our confidence on that. The presentation today reinforced our confidence in world-class tech and innovation. We showed a lot of things today that we think can help Just Eat Takeaway to grow faster. I'm going to focus more on the business performance. What kind of idea we have that we can really improve the performance besides those two? First, that is iFood know-how that we believe can help to accelerate growth at Jet. I'm going to have like four or five slides detailing more about that. Briefly, we believe tech and product can learn. We can share from Just Eat, but share from iFood on tech and product. A few clear examples about that.
People asking me, can you give clear examples about that? There is a very strong culture of tests in iFood. We don't care about the idea of Fabricio. We have 100 hypotheses, and we test all of them. Some of them work, and that's what we scale. It doesn't matter if it's not a top-down thing. We create hypotheses, we test fast, we scale what works. The iFood culture on that is very aggressive, generates a lot of results. We believe Just Eat Takeaway can learn from iFood on that. For example, you can also on AI, there is a 10% increase on conversion. I'm going to show more details about that because of AI. Our loyalty program is very successful. There is lots of learnings here.
Our offer quality helping the restaurants through information is a competitive advantage and also the reducing delivery cost and using AI to deliver. All of that are things that we clearly can see that we can share important things to Just Eat Takeaway. This was just a summary slide. I don't want to stay in the summary. Let's go to the other more detailed slides. First area is tech and product. A few good things to share how we see the tech and product culture. We run a lot, you can see here, hundreds and hundreds of tests with restaurants, 600 meeting hours, listen to the restaurants, creating hypotheses, testing hundreds of them, scaling what worked. You could say fine, so what? Sorry. 20% improvement in restaurant churn over, I think, one or two years because of the improvements that we are doing in restaurants.
On clients, customers, the same thing. Forty-two sessions in one test in one session. Doesn't matter who has the idea. Doesn't matter what Fabricio thinks or Diego. What matters is create hypotheses, test in one week. If it works, scale. 64% conversion, much better than the industry standards. The same thing with drivers, lots of meetings, improvements, very high app approval and score. During the lunch today, someone asked me the following question, how many months I need to put this product in Just Eat Takeaway? That's not how I see it. This is not like you just take a product and you put it in another country. If you do that, I think it would be replicating a mistake that many other companies did, that is just replicating products. What we can do is share learnings and culture and sharing technologies, not like a ready product.
I think Just Eat Takeaway has to reposition itself to say, I'm ready to win again. I'm here to win, to be the number one in Europe, then the number one in the world. We can invest for that. We can help on that. We can share the knowledge. We have to make Just Eat Takeaway iterate, learn, and innovate as the best in class in the world. There will be things we just move from Brazil to here. That's not the most important question. The most important question is, are we moving fast enough to create a global leader company? That's the challenge I have to do in this tech product. If I can do that on tech and product, Just Eat Takeaway will turn around and start to grow again quite fast. That's our biggest challenge. Second area is customer journey.
You have a few examples here too. Discovery, 50% of the traffic is originated for a fully custom personalized home. We also, during the due diligence, we do not operate Just Eat Takeaway. I want to remind you all how Just Eat Takeaway does that. This is substantially better than what we see in Just Eat Takeaway today. Fully personalized using LCM or using AI to define everything and improve conversions. Same thing with merchants, same thing with support. Support is an example that I think Diego showed. 65% of tickets are resolved by AI chat with bigger customer satisfaction than done by real person. We have substantially smaller costs and better customer satisfaction at 65% of the tickets. There is a lot to share and hopefully to work together with Just Eat Takeaway to improve within these areas.
Third area, I think it's the most important for this presentation, loyalty program. The loyalty program from iFood, it's not written here. I think Diego talked about it briefly. Today we have 36, 37, it was March. I think today is more than 42 million-43 million orders per month and millions of customers, customers are not here, using our loyalty program. This is profitable. Customers are happy. It incentivizes cross-sell aggressively. It starts in food delivery, then it goes to pharmacy, to grocery. Now we are sending trips and events, all of that in a profitable loyalty program. How we got that there? Maybe you can think, yeah, Amazon did it with Prime, so we did it in iFood. That was the most absurd thing I said in the whole day here. Loyalty program is extremely complex.
There is a presentation we have internally that we did 1,200 tests on loyalty program before we get to dynamics where the customer is happy and it's profitable. Because look, if you think loyalty program, free delivery, I just saw Amazon Prime, you charge $9 and you get free delivery. It's like a machine to destroy money in an unbelievable speed. Doesn't pay off. So loyalty program, we have to understand what do you offer, what is the benefit, what is the profile, how the customer enters, how it evolves, how they perceive the offer. Use a lot of data on that. You have to make money on the average, so you know there are outliers. You have to control these outliers. We worked on that for three years until we said, I was in iFood already yet by that time.
Prosus always said, loyalty program is fine, but if you scale this thing, it's a bomb. Because you cannot scale that before you have the right economics. Now we have the same margins per customer in the loyalty program as outside the loyalty program, but customers are using much more here. This experience of running that for five years with these numbers for sure can make Just Eat Takeaway move faster. 100% sure. There's a lot of value here. Obviously, we use a lot of AI to do that. Also, delivery costs, not only costs, but quality, delivery time. We deliver in less than 30 minutes on average. Driver productivity, which means drivers are making more money, delivering more things with a lower cost. We are very good on that. There is an area that we see lots of upsides on Just Eat Takeaway.
I did not want to make this presentation. I do not want to show any Just Eat Takeaway number. We have not completed the transaction. We have to wait a little more. I want you just to see that there are areas that we are global class. We really believe Just Eat Takeaway can benefit from that. I will be back here to Q&A in a few, I think in one hour. Please write down your questions and I can answer many questions to Just Eat Takeaway. I cannot answer questions related to the regulatory approval. We have this nice slide made by our legal team. The legal team said many regulatory approvals were already obtained, like U.K. and Canada and a few other countries. We filed this week, I think Monday or Saturday, the document with the formal approval request.
It is ahead of the time that we planned with the lawyers and everyone else. I am very happy that we are moving faster and ahead of the time. That said, European Commission will take the time they want and we will talk a lot about that as soon as they answer our official feeling over the next few months. The offer period, we always said we expected to close inside this year. Obviously, we keep this expectation. As a good Brazilian tech founder, I am optimistic we can do more than what we originally promised. I will not comment more on that today. My biggest priority is to move fast. The faster I can move here, the more I can create competitive advantage and I can execute well on Just Eat Takeaway. We are confident we are doing well. We have lots of knowledge.
Now let's wait for the European Commission, but they have been fast lately. I am optimistic. I think this is the last one. As I told you, this is our plan. Someone made the question during the results day on Monday, do you have plans to do more here? Probably yes. Let's be honest. Can I comment on that today? No. I cannot say who I want to invest or buy or integrate or merge because my big priority is Just Eat Takeaway. For two reasons. One, I want the European Commission to talk first. It's not my time to talk. It's time to listen from them. There is a second reason. Just Eat Takeaway is the foundation.
If I have a company with 70 million customers that are used many times for very good technology, I can expand this ecosystem, not only investing in someone, but creating value faster. That is my big priority. I am going to do that. Next time we meet, I am going to talk more. Hopefully, we will be able to talk more about that. That is what I could share about Just Eat Takeaway today. Open for questions. However, we have a star today here that we hope you enjoy. I would like to ask here, let me see. Yes. OLX. Lots of our growth in terms of EBIT came from OLX. We have Christian here today. Hope you enjoy his presentation. Christian, please come here.
Some say move fast and break things. At OLX, we say move fast and make things. We move at the pace of technology.
In this race, we're way ahead. That's why we're the number one destination for motors, real estate, and jobs, serving sellers, dealers, agents, buyers, and users in nine key markets across Europe and South Africa. How can a company that's already growing faster than all its peers grow even faster? We'll prototype, wireframe, optimize, and make every beta better and better and better. We're outpacing our competitors with world-class technology, AI solutions, tools, and utilities to optimize user experience and make our platforms more efficient, easier, smarter. AI is our turbo to keep us ahead and drive the key levers for growth as we accelerate towards 20%+ annual revenue growth and 50%+ our EBIT margin. Every day we're moving faster. Buckle up.
Thank you. Hello, everybody. Thank you for having me here and obviously Fabricio introduced me. Thank you very much, Fabricio.
My name is Christian. I joined OLX about 15 months ago. I have the pleasure of presenting you OLX today. OLX is a leading business with nine markets which we are in. Most of them, we are market leader. We have nine brands that we are entertaining in different categories. We have been able to grow across the last couple of years, as you can see, from about $485 million to this year, $777 million. We have been able to do this also by scaling our profitability to now 35% EBIT margin, which means that our business is not only a business that grows fast, but it's also a business that obviously with its scalability is able to continue to drive the margin.
On the back of our very clear strategy, and I'll come to that in a second, we have also raised our ambition level. The ambition level, as you've seen in the video, means that we are expecting ourselves to grow in the near future and also in the longer-term future by 20% + per year in terms of revenue growth. We're expecting also to reach EBITDA margins of 50% + across the next couple of years on the back of the scalability. Now let me give you a better understanding of what the group looks like and also what our focus is. We are focusing on three categories, three main categories, which are jobs, motors, and real estate. Obviously, verticals that are very well known to most of you who have been following classifieds for a while.
We are doing so by now also pivoting away from what we call the C2C, which used to be general classifieds, towards more a B2C-driven business because it allows a far better monetization. If you look into the slides, you will see that the core categories have been growing much stronger than the non-core categories, namely 22% for the core and 10% for the non-core. What you can also see is that the core categories are delivering far higher EBIT margins than the non-core are delivering. This is clearly depicted from the fact that the monetization of the B2C customer is something that helps very much not only the growth, but obviously also the margins. The second thing that we are entertaining, which you can't see on the slide, is that within the clear strategy and within the move of B2C, we have also started to align our platforms.
We are running in certain countries, or in most of our countries, we are running a horizontal distribution channel, as I call it, and we're also running a vertical distribution channel. What we're doing now is we are aligning the horizontal with the vertical to make sure that the content, the inventory on both sides, that the customers on both sides are having the best and beneficial experience, which on the user side helps obviously for lead generation and traffic, and on the customer side, obviously the lead delivery and obviously the faster and the better sale of cars, houses, and obviously also finding new jobs. Now let me go one deeper into the different categories, starting with motors. Motors is our largest and obviously most successful category that we're having.
We have been growing, as you can see, motors across the last couple of years on a very successful level in the general around mid-20s, which is something that we want to continue going forward. It is a leading destination on the verticals as well as on the horizontal in most of our countries. This is something that we are not only keeping, but that's exactly where we have this end-to-end accountability where we make sure that the person who runs motors has clear accountability for the overall product and also in tech stacks that we're having on the verticals as well as on the horizontals. The way we drive it is obviously we are very much focused towards business customers, but also we are also very much focused on our user experience.
The business customers, just to give you a bit of an example, obviously we are a product and we are a customer-led company. The products that we're introducing into the market is something that we help either the user to better choose and transact, and obviously the dealer in this case to turn his inventory faster. Because inventory fast turn means that they have less on their lot, and obviously less on their lot means they have less working capital, which allows them to also show greater profit margins, which obviously we then can monetize. On the one hand, you have a dealer tool, and as you can see, the product that we're delivering allows for higher weekly adoptions.
It also creates a multi-use account for large dealers, which is for them a big help because it allows them to monitor their overall inventory, which is something that we have just created. As you can see, this has been rising by 10 times just in one year, given the fact that the large dealers like to adopt this sort of products. On the buyer side, on the user side, the aid to AI posting is something that has become very popular. That means that basically the user can from the ads create an own AI video, which obviously gives them another way of looking at the car. Those are just two products that we have launched. There are car history reports for users. We are getting far deeper into the CRM tools also for dealers.
We are working a lot on product to basically make sure that our customers on the one hand, but also our users, are not only more engaged, but also are returning to us more often. The second largest category with us is real estate. Real estate, again, we are in most of our countries the market leader in terms of the horizontal, but also in terms of the vertical. We are doing this with a numerous number of brands. As you can see, real estate has been growing across the last couple of years with somewhere averaging high teens. This is a category where we believe we can do more and where we are expecting growth to exceed 20% again across the next couple of years. This is something where we are on the path of growth acceleration.
We have since about six to nine months started with monetization initiatives. I'd like to call out that one of the monetization initiatives, as you can see, has led to an ERP increase of 44% from March to May 2025. That means that basically the way we have been offering agents, the way they can present themselves, and the way that we have monetized led to a 44% increase in the average revenue that we are deriving from insertions. On the other side, we have also started to work on Matchbox revamp. I think the most important elements there is the 40% seller reply rate that is up. On the other side, also becoming far more efficient, namely the reply time is down. This again is something that we are working constantly on.
As you can see, there are a couple more initiatives that are coming, like the geo-based offering, the new value-added services product suite. We are at the moment revamping the whole real estate with new product, with new offerings that are fitting the users and are fitting our business customers. Therefore, we are very hopeful of increasing our revenues across the next couple of years by more than 20%. Jobs is the youngest category that we're having. Jobs is something that we are aware of the cyclicality of the category. This is something where the AI, and you've heard a lot of AI that Euro has presented, and we are working very closely with Euro and the iFood team on this overall AI topic. AI is something that we will be very much using in the jobs category going forward.
The jobs category is something, as you can see, which on the right-hand side, we are very much reshaping the jobs market. It is about matchmaking. It is about creating far more efficiency on the employer side. It is helping our blue-collar worker because the blue-collar is a space where we are to do CV passing. We are helping them to basically present themselves to the employer in a much better way. That is where the AI is helping quite a lot. We have now started also to extend ourselves more into other industries, which are more getting us into gray and white, which, as you can see, is banking and insurances. This is something which is very promising and where we also see as a category us growing very strongly going forward.
If I wrap this all up, we have significant room to continue to grow our revenues and also of the profitability. The three main pillars on which we are basing this, which is nothing unknown for the classified, is clearly pricing. The second is volume. Volume means that our agent, we are acquiring more agents, more dealers, and more employers. They will put also more intervention on us. This is the second pillar. The third pillar is the product enhancement. With all the products that we are referring to, we will allow to also monetize and create new packages, which again will help us to continue to drive revenues by 20% + across the next couple of years, reaching an EBIT margin of 50% + across the next couple of years.
One of the things that I think needs to be mentioned because it's something that will also continue to support us is what we believe we can also extract from the ecosystem synergies, which are certainly lying very much on the general classifieds in the goods and the services space. We will on the panel talk about it, but we are already entertaining with eMAG certain ideas of what we can exchange to try to make this ecosystem work much better because we are having used goods, they're having new goods. The question is, can we inject goods into each other and then see how the consumers react into it? Those are a lot of elements that we're going to test.
We have already started the electronics refurbishment with Flip, which is a daughter of eMAG, where we are already introducing ourselves and introducing this to our customers with a little test in Romania, which is going very nicely. Obviously, together with eMAG, we're very much working on the large commerce model and seeing what, especially from the horizontal, we can deliver to make the large commerce model also in that space far more vibrant. Thank you. Now we have a panel, and let me welcome very much Monique, who is going to moderate the panel. Thank you very much for being with us. Obviously, my fellow colleagues, Tudor and Orkun, the CEO of eMAG and the CEO of Iyzico.
Perfe ct.
Over to yours. Excellent.
Thank you, Christian.
Thank you.
Thank you, Fabricio, for the bigger overview on building the European ecosystem.
I think given we've just heard from you, Christian, and that we haven't heard as much in the past on both the Iyzico business and the eMAG business, maybe we can kick off with a few questions for Tudor and Orkun, and we'll then hand it over to questions from the floor. So Tudor, maybe if I can start with you, the eMAG business had a really exciting year, 2025 inflected to profitability with strong growth. Maybe you can talk a bit about some of the key drivers and what's in store for 2026.
Yes, FY 2025 was a very good year for us. Basically, what we did is looked at our portfolio because we've made a lot of investments post-pandemic, and we decided to accelerate what really works and divest from the areas that we figured they will not produce long-term profitability and growth.
We accelerated a lot on the marketplace model, on the original marketplace, and here especially we accelerated the fulfilled marketplace where the business doubled year over year. We accelerated a lot our loyalty program called Genius. Here, the number of paid subscribers grew by 50%. Last but not least, we accelerated a lot the out-of-home network. We built an out-of-home locker network, which we grew by 40% last year. We grew from 5,000 points to 7,000 out-of-home points. Looking to the future, we feel that next year in FY 2026, to these three pillars, we will also add the fintech arm, which should boost a lot the marketplace GMV.
Excellent. As you mentioned, next year you're going to be working on the fintech arm.
Given we're talking about the European ecosystem, will there be an opportunity for you to work with Orkun? Are there learnings you can take from the Iyzico business that will help you with your fintech journey?
We've been working a lot with Orkun and also with Anirban from PayU. What we've learned from them until now is how to set up a business in a regulated environment like payments and credit. We also took from them a lot of insights related to how to structure the technology for such a different product. Last but not least, in terms of user experience, how to present a BNPL product or a payments, a wallet product to the consumers.
Excellent.
Orkun, maybe if we turn to you in the Iyzico business, given the sort of wealth of opportunity I would imagine that you see available in your home market of Turkey, whether that's the size of the population, the online penetration in a market like Turkey, how do you balance sort of growth in the home market against thinking about growth opportunities within the broader European ecosystem, potentially starting with Central and Eastern Europe?
Thank you, Monique. First of all, we would like to continue, of course, our growth pace in our home market, Turkey. Last year, we could grow our business revenues 87% in local currency. And how will we continue this pace? We will continue to nurture our ecosystem, e-commerce ecosystem by serving our merchants, but also our consumers.
Right now, our consumer base who's using Iyzico digital wallet in Turkey and Iyzico as a payment method reached to 7 million, is reaching nowadays to 7 million. Right now, we are not only proposing our merchants accepting payments, but managing their financial services via Iyzico, also doing campaigns, cashback campaigns, cashback campaigns like Diogo explained this morning in a very smart way, also helping our consumers not only to pay with credit card, but with different other methods in a secure way, in a protected way, or also using some other alternative payment methods or open banking technologies. Also this year, last year, we could test our in-store solutions. We won't be only online, we will also follow our merchants who have offline presence to help them to increase their sales in their stores.
I think that it will be a scale-up year for especially in-store for Turkey. On the other hand, thanks to the Prosus European ecosystem, we are also always discussing, but we will continue to discuss how to maybe use our existing Iyzico products like issuing white label wallet, digital wallet, e-money capabilities outside of Turkey on different group companies or group ecosystems to better serve our ecosystem.
Great. And when you think about that and you make the parallel to some of the things Diego was talking about earlier with iFood, do you see there's a big potential if the Just Eat Takeaway transaction closes to work with them as well?
I see as if it's after the regulatory approvals. I think that as they have a very strong ecosystem, I think that we can learn from each other and we can share some of the best practices that we are using at Iyzico, even some products that we have already built at Iyzico in order to serve their ecosystem. They have a three-party consumer, restaurants, riders ecosystem, and I think that we have the different product sets, financial product sets in order to help them to nurture th is ecosystem. The discussions are ongoing and the discovery phase.
Understood. And Tudor, maybe if we come back to that loyalty scheme that you have, the Genius loyalty scheme. I think you've got now 1 million paid subscribers in Romania and that you're trialing it, doing free trials in some other markets.
Just wondering what you're finding the Genius loyalty customers, what they like, what they don't like, if there have been learnings from the iFood Club program that you've been able to bring across.
Yeah, we started Genius, I think, four years ago. It started as basically a prepayment for free delivery. In time, it grew to an ecosystem loyalty scheme like Fabricio showed earlier on stage. Basically now a user receives delivery returns, priority in customer care. He receives dedicated offers from all the businesses within eMAG Group. Looking to the collaboration with iFood, what we learned from them were some very good tactics in customer acquisition. From the learnings we've had, we will grow our targets for new customers in FY 2026 compared with what we deliver in FY 2025. Also now we are working on some insights related to personalization.
Also like Fabricio showed earlier, the personalization which is done for a loyalty user, it's very interesting what was built in iFood and we will adopt this within eMAG through the LCM.
Excellent. Maybe one for you, Christian. Earlier in the year when we had the deep dive on OLX, you talked or you hinted at the opportunities of some learnings you could take from the eMAG business when it came to your pay and ship and sort of capitalizing on the goods opportunity. I just wondered if there had been more collaboration since then between OLX and eMAG or if there was more sort of slated for 2026.
We started and what we're doing at the moment is we are aligning on the same day because the lockers is something that is of high interest to us.
We had to postpone a bit because on both sides there are other things that we were prioritizing, but this is one of the things that we're working on. I think the more important thing as well is I think is getting more understanding around the listings. How can we sort of share inventory? Because again, we are C2C and he's more B2C. The question is, can we inject stuff into each other? We need to understand what is the customer overlap. This is something that we need to work on and that we have decided to work on to make sure that we're really capturing what is really important in the ecosystem because I think this is going to drive really revenue. This will help us in the margin.
The logistic is something that basically to me sits behind, which will help us to basically be a better deliverer to our customers.
Excellent. Okay. I think we're happy to take some questions from the floor if people have questions for. Oh, there we go. Okay.
Hi, Nadim Mohamed from Standard Bank Securities. Just a question for Christian. It seems like you've deprioritized pay and ship and it looks like deprioritized horizontal platforms in general. Just would like to understand what the future is of that business and how you're thinking about it going forward.
Yeah. I think to come back, we have not deprioritized, but we have started to focus on what we are really good at. I think pay and ship is nothing we're good at.
That's the reason why I think cooperation with eMAG and others, Iyzico and others around payment makes a lot of sense. When it gets to the horizontal, there's no deprioritization. What we're using now, the horizontal is obviously we have very strong elements being real estate, motors, and jobs, which we want to keep, and we want to use the traffic in a much better way than we did. That's the reason why we're aligning the whole product and tech stack across the vertical and the horizontal. There's no deprioritization on that at all. It's just making more use and better use of the traffic because that's a lead generator that w e can monetize.
Marcus, yeah.
Thank you. Yeah, Marcus Diebel from J.P. Morgan. Two questions for Christian on classifieds. The first one is on M&A.
Clearly, I mean, there's a lot of talk about food delivery and the synergies that you have. Clearly not asking sort of like what targets are, but it feels like there's at least big opportunity in Europe. A lot of assets are up for sale, different size, different holders. What would be sort of like your idea? Is it really that you can bring a high degree of synergies even in new markets, or is it that you're maybe a bit more cautious given the evaluations? I just want to get a bit more feeling what Prosus currently thinks about M&A and classifieds, and then have one more question.
The first, what we've seen is just built on organic. Our whole plan is built on organic growth. Obviously, as always, if M&A comes up, it's something that we'll look at.
What I always say is big is not a quality in itself. It only makes sense if there are top-line synergies that can derive. This morning, it was already said that Diego said, "We don't think about cost synergies," which is right. Our businesses are growth businesses, and you only do transactions if you see top-line synergies. If there would be some and valuation would also be, let's say, adequate, maybe our shareholder would allow us to do a transaction.
Perfect. The second question is ju st on investments, organic, on the assets that you have. Where are we in terms of investments? I mean, in classifieds, it's always that there's a period of investments, and then hopefully you harvest a bit later when margins go up. Where are we in the Prosus portfolio?
Do you feel that there are still meaningful investments coming, or is it now the time that we should also get even more excited about the classifieds margins going forward?
The way I was presenting it is that there's a lot of scalability. If there's scalability, it means that basically we have investments, but we are able to monetize those investments at a higher rate. Therefore, I would suggest that the investment that we have done are some that we are catching up in terms of value, and others where we are investing right now will allow us to continue to monetize. That's the reason why you see that margin expansion from now, roughly 35% towards 50% and more.
Hi, it's Will Packer from BNP Paribas.
Just following up on the pay and ship comments, is it right to think that pay and ship penetration of your horizontal businesses declined and stalled? Has that been an important driver of the margin expansion? Could a bit more color there be helpful?
It did not stall. It was a conscious decision of ourselves to deprioritize, to make sure that we are driving profits. Because pay and ship, in the end of the day, is a subsidizing business, at least in our case. We came very clearly to the conclusion that it is better to keep it, and we are doing it, but at a much lower level because we do not need to subsidize at that price. That is the reason why you are probably hinting towards the growth on the non-core categories, which has to do with the fact that we have deprioritized pay and ship.
In that context, how's traffic holding up versus your c ompetitors? Is it having an impact?
It doesn't have any impact because those are the things that we're monitoring. Before we did it, we obviously did certain tests because I had certain, or we had certain ideas around it, and we saw very clearly that there is no traffic loss that we are incurring in the space. The traffic to us is not on the pay and ship. The traffic to us is the 64 million daily listings, the inventory, the freshness of the inventory. That is what really matters on the general classifieds if you think about the horizontal.
Th anks.
Pleasure, Will.
Cheers. Do I need to press anything? No, I'm on. Giles Thorne from Jefferies. A question for Tudor. If the power of the ecosystem is so strong, why did you sell Taz?
Yes, good question.
Taz, for us, it was subscale. It was, from a technology standpoint, the investment that needed to be done there were huge. Important for us was the ecosystem, and we sold it in a manner in which the current owner is still part of the ecosystem for the foreseeable future. The genuine customers still benefit from the Taz now vault offering, and we do not get to, let's sa y, pay the bill for it.
Thanks. It's Andrew Ross from Barclays. One more question. Just to put some context around the 20% revenue growth you're talking about going forwards, can you just remind us what the take rate is in your markets, I guess, if we think about it as a percentage of gross profits or commission pool? Are we kind of talking about a monetization gap compared to Western Europe, but it's easy to close?
Or just kind of put some context around what drives about 20% growth? Thanks.
Andrew, it is a take rate conversation, as always. I mean, if you look into the markets where we are, those are markets that are developing. The take rate today compared to what you would see in, let's say, Western part of Europe is still low. The margins and the money the dealers and agents are making in those countries is still large. On the one hand, it is a catch-up. It is a pricing. Obviously the volume, as I was saying. We do not have all the dealers and all the agents, albeit we are the number one. There is a lot of inventory that we can still get on, which we will be monetizing.
The third one is obviously the product enhancement, which will allow us to tap more into the commission pool and all the cross-profit margin of the dealer.
Yeah, thank you. Robert [Fink] from Capital Chevreux. First, I have a question for Iyzico. Could you give us a bit more color on the merchant mix, on the types of payments that you're processing, so like large tech platforms or digital versus in-person? What are kind of the key growth drivers in the growth areas in your company? If you could give some more color on that. A question for eMAG. Of course, a strong position in Romania, but what is kind of the international expansion strategy? I think you're expanding into Bulgaria, into Hungary. Do you have maybe plans in the future to expand into more markets? What is kind of the long-term game there?
Thank you.
Thank you. First of all, for the merchant mix, just to give you the magnitude, we processed last year more than TRY 7 billion volume in our platform. Our merchant mix is like all the global brands who are operating e-commerce in Turkey, large marketplaces from East and West, from China and the U.S. operating in Turkey. They are all working with us. If you are buying a T-shirt in e-commerce in Turkey, there is a 90% chance that you are paying with Iyzico. If you are buying a sneaker in Turkey, there is a 100% chance that you are paying with Iyzico, all the sneaker brands that you all know. They are working with us. Not only global or large Turkish brands.
We have more than 120,000 merchants in our platform, which means that 99% of our base is small, medium enterprises in Turkey who are doing online commerce. The key pillars today, it was growing e-commerce in Turkey and also our growing merchant base, also more global brands coming to Turkey. For example, last three years, all international car dealers in Turkey, they are accepting in-advance payments when you reserve a car, and almost all of them are working with Iyzico, electrical cars or classical cars. For the future, what we are looking, we are now, as I said, we are discovering, we discovered especially, we are following our merchants. Most of our merchants, they have offline presences too. Last year, we tested lots of offline payment products. We are partnership, or some of them are on.
We have now presence in their stores, and we will scale this presence to help our merchants on this one.
Thank you.
Thank you. We have a strong presence in Romania, where within our ecosystem, there are, let's say, six businesses offering value to consumers. While in Bulgaria and in Hungary, we have two businesses. We have eMAG and Same Day, the courier business offering value to the ecosystem. Soon we will start with Wolt in food delivery in Bulgaria and in Hungary. Related to expansion, we see that the ecosystem, all six businesses and adding more, will create value for the consumer. On the very, let's say, short to medium term, the priorities for us is to enhance the value that the ecosystem brings in Bulgaria and Hungary.
Related to other markets, we are constantly looking, and if opportunities will arise, yes, we will take them. Perfect. I think in the interest of time, we're going to have to leave it there.
Thank you very much, Christian.
Thank you very much,
and Tudor, for that overview of the European ecosystem.
Thank you, Monique.
Thank you very much.
Thank you.
Finance is a system logic, a logic that moves with discipline. We take bold bets, but only when return outweighs risk. Disciplined, aligned, always. We don't just track units. We track compounding motion. Every transaction, every signal adds force to the whole. Finance doesn't follow the business. It empowers it. We're not just scorekeepers. We're architects of momentum, enablers of growth, fuel for the ecosystem. We build for frequency. We model for momentum. We scale by design. This isn't just a snapshot.
It's a system, a flywheel, always learning, always accelerating. This is finance, not at the edge, at the core, powered by Prosus.
Good afternoon, everybody. I really trust that you've had an enjoyable day so far, and that it's been most insightful. I've been with the group for more than 26 years, and during which time we've transformed into the global e-commerce technology company that we are today. The pace of change that we've seen has never been faster than in the last year. Fabricio, Diego, Ashu, they all talked about our strategy to build out ecosystems in the key geographies that we're covering. My goal today is to tie it all together with the financials and to show you how we expect to continue to scale the business profitably, and not just incrementally, but to accelerate.
We've moved from deep losses just two years ago to more than $400 million in e-commerce profitability today. As Fabricio outlined, we will be in the billions in just a couple of years. As you've just heard from the intro video, me and my finance team, we're not just here to count the beans. We are here to enable growth of the group, but we want to do that with discipline. In doing so, we need to create real value for you as shareholders. I know I'm talking to an audience of numbers people, so I hope you'll be as excited as I am. Let's get started. I think a good place to start is, what are my priorities? Where will I be focusing my attention and energy and effort? As our results have shown, we are fully committed to sustained and profitable growth.
We will look to invest and to fund our growth, and we will do that with the necessary discipline required. That's what we've done over the last couple of years, and you've seen some of that result. That continues. I will ensure that we maintain an optimal capital structure, and we will have to sometimes make the necessary trade-offs to do that, but I can assure you we'll do that with discipline. We're here to deliver value to you as shareholders, so increasing our free cash flow and sharing more with you over time is going to be very important. In discussion with many of you, you've highlighted the difficulty that you sometimes have in valuing us. My goal is to achieve the valuation that we deserve, and to do that, to ensure that our financials and the disclosures are aligned with this goal.
I will come back to that in more detail. Please be in no doubt, we are committed to this peer-leading profitable growth. You have heard from the team today about all the various excellent opportunities that our businesses and our investments have, and by connecting that together in the regional ecosystems and how that can drive value for us all. Our vision here is not just to take part, but actually lead and create that. The good news is we are well on our way. Today we are growing faster or double the pace than most of our peers, and that is primarily driven by the excellent growth coming out of OLX, as well as iFood and our payments platforms. You have seen today that the runway for growth at iFood specifically is substantial.
Diego showed that excellent slide showing how they've layered new opportunities on top of others, and so on building the business. You have seen iFood Pago, Macro Nonna, and its dining offering, all of those things driving incremental value for the business. You have seen some of that coming through strongly in our FY 2025 financials. This will continue to add more and more value to that ecosystem. Likewise, Christian has just shown with his team at OLX how they're rolling out new products and offerings to our partners across the key verticals of real estate, autos, and jobs. These offerings have been a major contributor not only to the revenue growth, but a substantial margin enhancement that you have seen at OLX.
PayU has now fulfilled all the complex regulatory requirements of India that had to go through, and the team is now fully focused and aligned and can concentrate on growth and profitability. It's a real priority for us in an important region, and there's much potential, as Ashu outlined. I am confident we will continue to grow strongly and reach the ambition of doubling our revenue by 2028, as Fabricio outlined earlier. That includes the Jet acquisition from FY 2027 onwards. With revenue nicely on track, we've established now a three-year track record of improving our profitability, and we have the full intention to continue on that curve. In three years, we've improved our e-commerce profitability by more than $800 million, bringing us to $655 million on an Adjusted EBITDA basis for 2025. This speaks really to the operating leverage and efficiencies that drove the product expansion across the portfolio.
We are not resting at this point, and we will do more as the teams have shown throughout the day to expand margins further from here. On the next slide, you can see that we have already improved margins by 15 percentage points over the last three years. While our revenue growth is outpacing our peers, margins remain lower. That is the opportunity that we now see in the next three years ahead of us to really step that up. Let me give you a couple of examples. The core food delivery margin for iFood is 28% on an adjusted EBIT basis. If you look at the iFood business on its consolidated level, it is about 17%. Why? We are still investing in the fintech and iFood Pago initiatives, as well as meal vouchers and a number of other things. That can clearly only grow from this point forward.
Similarly, with OLX, it's gone from 25% on an Adjusted EBITDA basis to 35%. Christian showed you that his ambition is 50% in the next few years. I will take you through how we see that unfolding in a few minutes. The teams have taken you through some of the individual plans throughout the day that will help to improve and drive the growth and margin expansion I'm talking about. Rolling it all up from a financial standpoint, I see three areas for us to focus and hit our goals. First, we need to continue to invest in the businesses that are front runners and winners to date. iFood, OLX, and Iyzico all have funded plans and have the ambition to do a lot more. Secondly, we will continue to support those businesses that have unrealized potential or have experienced some challenges of late.
eMAG and PayU are very much on the right track at the moment. They have strong and solid plans in place to improve from here, and I expect meaningful results in the next three years. Finally, we will continue to invest in building the ecosystems. You have seen that we have closed a Despegar transaction, and we have showcased how we already are working fast to integrate and just make that business better. Of course, we are seeking to bring Just Eat Takeaway into the family. Just Eat has leading and profitable assets in significant markets, and you have heard from Fabricio about our plans to reignite some of the growth in there. Moving on, very importantly, we have improved profitability and growth and margins. We need to deliver cash flow. That is a real true driver for me of value.
We've increased our free cash flow over the last three years by more than $1 billion. We have reached an important milestone that Prosus reached at Tencent, its first positive free cash flow in the past financial year. There is another example of the process of today. It is quite fundamentally a different company from just a few years ago. As we continue to scale, the cash flow should grow with the profitability. Ecosystems will drive that, and it will also obviously enable us to share more of that with you as shareholders. I mentioned upfront that we look to invest and fund our future growth, and we will need to do that with the necessary discipline. That is an easy word to just say, but it is a core principle for me, and I do believe we can do both.
As a follow-on, the group has a strong foundation in the current capital structure that we have, and we'll make certain trade-offs. I think, as you can see from the next slide, we have a long-dated profile in our debt, and our cash position remains very healthy despite the funds that have already been reserved for the Jet transaction. Turning to the next slide, and I know there's probably some debt investors in the audience, you can see that firstly, we remain committed to our investment grade rating, and our operating businesses continue to perform well. Improving the quality of our debt profile as our dividend stream from our businesses up to the center will improve will just further enhance our debt capacity and debt profile. We have some upcoming maturities.
These are, as you can see on the previous slide, very manageable, and we have options at our disposal to sort of refinance and enhance our profile from here on. Let's now look at the area that receives rightly a lot of attention from the investment community, capital allocation. We have been clear that our priority is to deploy our capital in the ecosystems, focusing on generating strong returns. Our capital allocation matrix looks as follows. Firstly, we will invest behind our ecosystems, making both big and small checks to optimize returns. If we can find assets in the regions that will further strengthen these ecosystems, offering scale, improving technologies, and obviously, it needs to come at a reasonable valuation, as has been the case with both Despegar and Just Eat Takeaway.
Similarly, we will invest in frontier technologies and innovative companies with products, services, or technologies that we can deploy across our regional ecosystems. That will help us run faster and better. These actually ought to be smaller check sizes, and importantly, we will continue to invest in ourselves through the open-ended share buyback. That has meaningfully also impacted and helped narrowing our discount. Just to remind you, this locks in value actually on a daily basis as we increase our exposure, not just to Tencent, but also to the rest of the underlying portfolio. In the next slide, we show our capital allocation at work, and we have deployed almost $8 billion of cash in the last 12 months, investing mainly in enhancing the ecosystem through the Just Eat Takeaway and Despegar transactions. We have also backed innovation that can actually accelerate some of these aspects.
Two examples of that is one, Anirban spoke about the Mindgate acquisition in India, and we also did Paynet in Iyzico in Turkey. We have actually done quite a bit here and have done it in a disciplined and thoughtful manner, as I have just outlined. While we have added to our portfolio through the acquisitions that we have made, I think we have also been quite disciplined in the last 12 months. We have reduced our exposure in various other areas. We have taken some decisive actions to actually dispose of businesses that either do not fit within the portfolio or do not meet actually a returns profile. In total, we have crystallized $2.6 billion of value. On this slide, I am showing you that whilst the group have made significant investments, we actually believe that returning capital to our shareholders is as important. This is the trend over the last seven years.
Over this timeframe, our open-ended share buyback, as well as the cash buybacks that we've done before that, returned more than $50 billion of capital to shareholders and reduced the Prosus free float by 36%. To put the $50 billion in context, just a few weeks ago, that was almost half our market cap. This is a significant sum, and our current program remains one of the largest in the tech space as a percentage of shares outstanding. Similarly, as part of our ongoing journey of capital returns, we obviously need to look at dividends. Given our strengthening free cash flow position, we are able to share more with shareholders.
We've doubled our dividend this year, coming from a historically low base, but we clearly have the ambition, as our free cash flow grow, to share more with shareholders and enhance your returns through the dividends that we will declare. Now let's turn to the discount, and I will start by saying that reducing the discount has been and remains a key priority. We seek to address that really in three ways. The first one, which I think is very important, probably the most important one, is that we need to continue to scale these businesses within the ecosystems that we're backing, growing revenue ahead of our peers, expanding the margins, as I've spoken about, and generating a free cash flow. That is what will get us to a position where we're really into a Prosus plus situation, like Fabricio outlined this morning in his slides.
Secondly, we will continue to highlight the value in our portfolio through increased disclosures and telling and sharing more with the investment community, like we have started recently with some of the deep dives that our IR team has done and through days like this. Thirdly, we will continue to optimize the value and create NIV for shareholders through the share repurchase program and other forms of capital return as they present themselves. Let me dig a little bit deeper into these buckets. Let's talk very specifically about the first one. You have heard Fabricio talking about his desire to double revenue by 2028. Throughout the day, you have heard from the teams talking about the drivers of growth. I am actually just recapping it all here.
This implies that we will more than double our revenue by 2028, and it essentially then accounts to an organic growth rate of between 15% and 20%. Just to be clear, this includes Despegar from June 2025. We have assumed, just an assumption at this stage, that Just Eat Takeaway, we have only included it from FY 2027 onwards. Here you can see the revenue contribution from the various regions as we see it. All the regions, with LATAM and Europe expecting to grow more than two times. On India and PayU, we have been appropriately more conservative as the business turns itself around. You have heard from Diego. LATAM growth for the next three years will really be driven by all the various growth initiatives that he has highlighted. In Europe, a lot of it will come through OLX and hopefully Just Eat Takeaway from 2027.
Lastly, I think we're optimistic about the measures and changes that have been put in place at PayU India. As Anirban described, we expect a significant improvement and growth. Looking then at the profitability, yeah, our intention is actually to triple our EBITDA in the same period. That will again require iFood, OLX, and other businesses like eMAG and PayU to start making meaningful contributions. All right, Tudor?
Yeah.
That profitability is again just the regional expectations. You can see that Latin America, Europe being by far the leading contributions, and India, we forecast it more modestly. We've given ourselves ambitious, but what we think are achievable targets. Fabricio will say, "We ought to aim for a lot more," and he's right. We're doing this so that I think you can hold us accountable as we aim to deliver on our commitment of profitable growth.
We're not just doing that, but I think to give you the understanding that Prosus and our business will evolve quite substantially over the next few years and that there's more than investment in Tencent and that the rest of our e-commerce portfolio can create real value for you. Talking about valuations, moving then to the second bucket, I think we are committed to also make the business simpler and easier to understand and value, and sharing information like today is part of that process. It's a request that I often hear from investors as well, and I'm committed to actually help you on that journey. To help shed more light on the valuations of our businesses, on this slide, I've shared with you what the analyst ranges are of the various businesses.
I've shown iFood and OLX as the two largest ones, and then the rest of our non-listed e-commerce businesses in the third bucket. What you can see from this slide is that the valuations of our two largest owned investments, they're actually valued far below where many of the listed peers trade. In some cases, like iFood, this is despite having industry-leading revenue growth as well as margins. OLX margins are actually showing they're growing ahead of most peers, and their margins are catching up rapidly, as you've seen from the presentation that Christian gave. In our view, these seem to be very conservatively valued by the market. I hope that by sharing more information, we can have constructive conversations about our businesses and assist you to value them better and for us to make it simpler and easier for you to understand the value.
Lastly, the buyback. The share buyback has had a significant impact over the last three years. I've shared with you how it has enhanced NIV by 15%. It's put more than $15 billion back in the hands of our shareholders, and we will continue the buyback. That is our intention. I started the day with my priorities, and hopefully you can see how those priorities hopefully will translate into us achieving many of our objectives. I hope you feel that you have a clear view of how we shall achieve the ambitious targets and accelerating revenue growth and improved profitability that we are aiming for. Thank you for coming, and thank you for listening, and I look forward to answering your questions in a Q&A session.
Hey, Nico, Nico, over here. Hey. Hey, it's Eoin. Listen.
I was up making this all night, which was even more awkward because I was in a hotel room. I think putting a capital markets day on his birthday is pretty miserable. At least what we got to do is give him a cake. Don't burn yourself. Happy birthday. We'll share it all at drinks, and we can stare at the Q&A, but well done. All right, we're going to switch now to the Q&A section. Could I have all of our leaders up, and let's get this done. It's the last of the day, and the thing in between the cocktails and us.
You see, Nico, we made a capital markets day because of your birthday. Congratulations. Hope you're enjoying the gift.
This time we are going to sit on the seats. All right.
Happy birthday. Okay.
If it goes well, we get a piece of cake.
Yeah, if you answer well, you get a piece of cake. If not, you'll be asked to leave. Okay, Joe, you had your hand up nice and quickly, so let's get this started. Is everybody ready? Last part of the day.
Only difficult questions. We only want the difficult questions.
Okay. Euro knows the questions already because he's been talking to Tokan. Joe.
Hi, it's Joe Barnet-Lamb from UBS. I've got a couple. My first one, you spoke about food and payments as the foundational layer of the ecosystem strategy due to frequency of purchase. In LATAM, you obviously have incredibly strong assets in those areas in Brazil, but less so outside. How important is food or payments to that foundational layer? Can it work without them? If not, what are your ambitions in LATAM beyond Brazil?
We have very good assets in Brazil, as you said. In India, we have Swiggy that is a very good asset, and we have PayU, that is the clear leader, as you saw today. In Europe, we expect to close a very good asset that is Just Eat Takeaway that is a very good foundation. We have some payment business in India, but in very specific locations, for example, Iyzico in Turkey only. My point, I do not think we are in Brazil too strong. I think we will have a stronger position all around over the next one year too.
Sorry, just to clarify, sorry. I mean, in LATAM, outside of Brazil, i.e., you have got food and payments in Brazil. You have now got a whole load of other ecosystem stuff outside of Brazil, but that foundational layer, you do not have outside of Brazil.
Yeah, yeah.
Now I got your question. I preferred when I didn't want before.
You asked for a hard question for Brazil. I'm here for it.
You're right. Our first goal, I think, is to execute that exceptionally well in Brazil. To remind you, Despegar has a very strong position outside Brazil too. We don't have food delivery outside Brazil. You're right. Goo d question. Yeah, good question.
That's it. Nothing to add. I mean, priorities. The pri ority now is Brazil. If we crack Brazil, we have the playbook to go to the other countries.
Got it. Thank you. Second question. You spoke about ramping dividends and obviously having increased them 100% this year. Now e-commerce obviously is not a drain. What's dictating the rate of dividend ramp?
Is there any specific ambition to pass through the Tencent dividend, or do you just look at it each year based on the priorities of that year? I'm just thinking, yeah, why 100% this year? Why not more? Why not less?
Why do you think so small? Yeah, just the Tencent. We are going to be much more than Tencent. Prosus plus Tencent. What dictates the rhythm is our strategic priorities, is the opportunities that are available for us, and the board should say that's the way we want to keep expanding our dividends. I don't think many of you suggest that we should have a playbook on how we do capital allocation for the next five years. I think this would reduce our strategic flexibility substantially. I think the real thing behind that is a lack of trust that we can take the decisions with discipline.
We are not going to have a playbook. That's how we do. We distribute 100% of Tencent, or 40%, or 50%, or 60%. We are going to keep looking to the opportunities. We postponed, for example, some decisions for six months so we could analyze better the environment. It was quite good. Many of you suggested we should distribute the cash we had six months ago. I think we used it in two transactions that I'm quite confident are going to have a very good IRR. We have to keep assessing that. I think you should give us a little credit because now it's 100%, so it's good. Let's celebrate for at least one week, and then hopefully we will increase our EBITDA a lot, as you saw. Tencent is going to keep growing a lot.
We will reduce our costs, and we will check if we should invest it somehow growing. We would. If you can distribute more in terms of dividends or buybacks, we would. I think, and that's the way the board sees that. If you take a decision like that one, two, three years in advance, it's not smart. In a normal world, in our current world, it's impossible. No one can predict what's going to happen next week. We were talking about the news this week. No one can predict what is happening right now. If you refresh your news, you don't know what's happening in the world right now. It doesn't make sense to us to say this is our capital allocation for three years and lack this capability. I understand there is a lack of trust for now. We are going to keep showing results.
I think if you look to the last one year, we took good decisions in terms of distributing results, increasing dividends, reducing costs. If you keep doing that for some time, thinking only about Tencent is going to be small.
Thank you.
You want to respond?
Hi, there. George Wood from United First Partners. I was hoping you could answer this quick question about the EC's recent developments in terms of, well, I'll get straight to the question. If the EC requires you to divest your Delivery Hero stake, will you be happy to do that? And have you thought about how you will do that if that is the requirement that they make to complete the Just Eat Takeaway?
Your lawyers, they are going to run away, take you out, and throw you away from here. So be aware of this kind of question.
That's why he wore a tie. He's trying to sneak in.
Prepare it for everything. We can't comment on that at all. We feel that everything ahead of time that you're going to answer in a few weeks. We are confident we are going to create a big European leader. I know that's what's written. I should keep on the written thing until the European Commission says something. Sorry, next time I will talk more.
Willie. Sorry, who are you?
Hi, it's Will Packham from BNP Paribas . So a couple of quick ones. Tencent's your most important asset, and the stake's now below 25%. You're committed to the discount, which is NAV per share accretive. On the other hand, as your stake shrinks, your influence at Tencent could fall, and it's an overhang for Tencent and its shareholders. You're committed to the buyback while the discount is elevated.
Could you remind us what that means? The second question is, you've got $11 billion to spend. Can you give us color on the priority areas by geography and segment, having given us so much detail today? Thank you.
This is an easy question. I could answer, but because it's Nico Marais, maybe Nico, you can answer this question. No, you are with the mic.
Yeah, yeah. I'll start with your second question. I think, as I've covered, on a capital allocation perspective, the focus for larger checks will be within the ecosystems that we discussed. Now that can make that easier, just better. Smaller checks, frontier technologies, and that. That hopefully is, so that's the focus. Longer term, but for now, we need to settle Despegar, and we actually need to complete the Just Eat Takeaway transaction as well. That's a more short-term focus.
In terms of the discount element, look, we think the discount is still too high. There's more we can do. I think I've outlined the three things around that we want to focus on to do that. It starts with building the ecosystems of profits and making the e-commerce side of our business a lot more valuable. I think the market ought to then give us credit for that if we deliver on that. I've spoken about valuation aspects and sharing more with the market, and we can enhance that with the buyback to continue with that. It requires all three of those things. I don't see it just as, oh, it's a buyback as a solution to improving the discount.
I think part of his question on the regions was, do you have a region that needs to have more of the capital allocated to it, or is it just pure opportunity-based?
I think Diego think s it is Latin America.
For sure. We can have a good debate. We will decide. We will let you know.
Yeah. We really do not have it. I think this is part of the strategic flexibility we have. What I think is good. Some people ask it like, what do you have to invest? What is your budget per region? I think we can be flexible. I think this is good.
I think we had one over there, and then we have Tim here. Actually, one behind you, and then Tim. Go ahead.
This is Simon Kahl, Kahl Investments. Just I have two questions.
First, the discount, is it a problem or is it an opportunity? That's my first question.
It's both because I think we want it to be lower. It ought to be lower. Whilst it's there, obviously that's what allows us to put in a share buyback program that, as we execute it, we increase the per share value of Tencent as well as our other assets. Longer term, I think we would like to get to a position where we get to the Tencent Plus position, where our e-commerce businesses generate the profits and the growth that we're aiming for. At that point, yes, there probably will still be a discount, as there are many other investment holding companies that have that. You do not then need to use that to actually create underlying NAV. You can do that by operating and building great businesses.
My second question is on, basically, it's very simple. What do you think is the probability that those demands in a year from building ecosystems don't materialize? In light of that, how do you justify these bid premiums for buying control of assets that you're going to build this ecosystem on? Because in my view, it seems like, at least I'm not aware of super apps that have been developed through M&A. I'm just like, in your eyes, what is the probability that this will fail and then you're still paying these big bid premiums in building them out?
It's 8.7%.
No, no, but qualitatively, just hearing your thoughts around that.
Can I just jump in to say one thing? I don't think we're talking about super apps here. Definitely. It's an ecosystem strategy, which means that each asset has to generate its own value.
On top of this, there are integrations that will generate an alpha on top of this. This is not a super app. I do not expect necessarily to have all the assets in Latin America at iFood App. Which ones will be there? The ones that make sense. I show you an example of Despegar at iFood App. If the results are not good, I turn off. That is not a problem at all. Anyway, Gonzalo needs to perform the Despegar and make that asset great.
Also, if you think about the user experience, there are other ways to integrate users between apps. Even the conversational interface can create eventually some interactions that make sense. For example, the super app, big play, it is in China. Also, they have a conversational interface through WeChat that was important to create their big ecosystem there.
It is not like we have to develop a super app. It is that we have synergies between our offerings. We are not saying that we are going to invest. You said something. How to justify? It is 8.7. You can update your model. How to justify the big premiums? Despegar is a company generating $180 million in EBITDA. We are talking more or less about 10 times EBITDA, less than that, right? Yes, 10 times EBITDA. Just Eat Takeaway is $400 million this year, $380 million. We are talking a little more than 10 times EBITDA too. I know in the past we did some transactions that you made critics. I understand the critics. We lost some money in some transactions. We were talking about 10-20 times revenues. Now we are talking about 10 times EBITDA. The level of discipline is bigger.
The connection to our core is much bigger. We operate a similar company with bigger volumes. The synergies we are talking about are not like magic. What you demonstrated is magic. The synergy we said, it's not magic. It's like we know how to do it. In Brazil, in Despegar, iFood has 70 million customers, 15 times more than the 4 million customers that Despegar has. You said, how to justify this overprice? Look, I don't think there was that overprice, and I don't think that justifying that is like a magical thing. I think if we do poorly, it has a good ROI. We expect it to do exceptional. Come back to the next CMD, and we are going to look for you to show the exact probability. It was zero. It was not 8.7.
I will. I will. Good luck.
Thanks.
Sunil, did you have one? No? No question. Oh, you did.
Following on from all the questions on the discount to NAV, yes, the CFO, I think, agreed or conceded that buying back shares is not necessarily the only route to narrow the discount to NAV because we've started this buying back program, say, two, three years ago, and I don't think the discount has narrowed materially after spending $50 million or is it euros, right? Now that you are very confident in your e-commerce strategy and we can see the energy that Fabricio and team are bringing, why don't you focus on investing more into the business or returning money back to the shareholders? Let us decide what to do with that money than you buying back your own shares. That's just an observation from my side. Thank you.
Just in terms of the I agree with you. There are other things that clearly will move the discount as well. I've highlighted those. Just to remind you, though, that when we started the open-ended share buyback program, the Prosus discount was in the mid-50%. It's now 28%. I do think it has had a significant impact, the buyback. Your points about investing in the business, that's what today was about. That's what Just Eat and Despegar is about. I think that we're not misaligned.
If I may add one more observation from my side, I think when I look at the discount to NAV, whenever Tencent runs away, the discount to NAV widens, and when it's flat or going down, it narrows. Just an observation. Clearly, I agree with you.
There's a market element of capital that maybe move out of China towards China, and we're impacted by that. I don't think that's the only thing. I think, hopefully, the things of peer-leading growth, the profitability that we're creating, the fact that we now have free cash flow positive, extensive, the clear strategy that we've outlined here today in terms of pursuing the various ecosystems, I think that at least an element of the discount reduction that we've seen should be attributa ble to our efforts.
Yeah, I think I agree 100% with Nico. I think the direction is positive of the discount is narrowing, not because of the buyback, but because I think the company is making progress. I think it's a good direction. You just said you spend $50 billion. I don't think we spent $50 billion because we are buying back shares.
What makes the shares of everyone that is here as our shareholders more valuable? What is good? It is good because it is an opportunity. So we are just investing ourselves, what makes sense now. I think for lots of time, for some time at least, you think this is the best way for Prosus to generate value. I think we are showing that we have other ways to generate value, and we are very confident on that. Good news. Let's just generate more value. That is why we are here, and we are going to keep returning capital, not with a formal capital allocation like table for the next three years, but really executing well, increasing the numbers, and finding ways to give the money back.
I will. We will do it on Andrew and then over here.
All right. Just two quick ones from me.
If I look at just on Just Eat for Fabricio, if I look at the development of the U.K. and the German consumer internet market, it looks a lot more like the U.S. than China and Latin America in that you have more established verticals and not sort of big ecosystems. I mean, are you confident that you can sort of break the norm and build Just Eat into a broader ecosystem in these markets?
Just Eat, it's Prosus. Prosus is going to be to use Just Eat and other Prosus companies plus technology plus to have a stronger B2C company. I think so. Look, I think we can do amazing things not in one month. Sometimes I talk to investors, they say, "Why don't you fix all the problems next month?" Because then you create all the problems, and probably it's just like a fake solution.
We have to first complete the Just Eat deal. We have to make Just Eat operate exceptionally well. We can do what you are doing. If you look today at a photography, you say, "I don't believe in that." You were right today. If you look many months ahead, I don't know how many months, but if you had, look, you wouldn't do this question on iFood. You say, "But iFood is different." Today, it was not four years, five years ago. We built it. Lots of work. Diego worked a lot. Lots of people worked a lot. We built it. Prosus is here to build. We will build into this position over the next, I don't know, one year, one year and a half.
Fabrício, if I can jump here, I think we should think this question from an innovation perspective.
You were right. When you look to the past, that's how you define the market. The question is, if there are still people eating at home or going much more to the physical space, to the dining, how do I change this behavior? It's very difficult. That's why you have to test 100 times, make errors in 99, find one, and then scale. You can change. That's why, again, we come back to the discussion of culture.
Yeah. Since you raised this point, we are ahead of so much transformation. There's so much transformation ahead of us. Like the investment companies, your ones, are going to change how they work in the next two, three years. The food companies, the grocery companies, the travel companies, the internet companies, the search companies, the ads companies.
That's why I started talking about the Prosus way. Look, our competitive advantage is to have passionate people with lots of grit, moving fast, and capacity to innovate. Because the change is going to be massive. If I am in Germany operating with the assets I have, let's say Just Eat if we close the deal, plus other things around during this transformation, and we are an aggressive, innovative company, we are going to win, I believe. If we are not an aggressive, innovative company, we are going to lose anyway. We, you, everyone. That's why I start with culture. Can we play to win in this world that changes so fast? Yes. Then Prosus is amazing. We have 2 billion customers, innovation capacity, 40,000 people in cash. If we can't play to win, then we are going to lose.
I think we can win if we deliver what I started with.
Let me materialize with two examples that we did at iFood, just for you to understand the concept. Back in 2018, we said at a certain point, we're going to acquire everybody. We need to increase frequency. How do we do that? Oh, we need to do a loyalty program. You saw the numbers today. Beautiful, right? We started in 2018. Failed, 2019 failed, 2020 failed, 2021 failed, 2022 failed, 2023 found the first sign. It is as important as you are. Let me give a second one. When we were in 2023 and we were going out of COVID, people were going back home and orders were going down, Fabricio came to us and said, "We need to grow. Doesn't matter how. We need to grow.
Don't come with COVID excuse." You know what happened? We found a business called WhatsApp that currently represents 25% of our orders two years and a half after this. Why? Because somebody came and said, "We need to keep growing. Don't come with the COVID answer to me." The culture of innovation has to come, test a lot of things, and find a solution.
In 2017, 2018, 2019, what we heard in Brazil again was, "How do you think you can build a big food delivery business if you're competing against a global leader in America and a very capitalized Japan, capitalized player? It's impossible to win on these markets." We are doing well. The point is, okay, the picture, there are problems. Our job is to overcome it. We will fail f or sure.
If you have the right culture and right people, we will also win.
Got it. If I may just ask a quick question, Nico, happy birthday, by the way. Just on the given the strategic focus on ecosystems and so on, how are you thinking about the other listed e-commerce assets? I mean, quite a disparate group of investments there. Are you looking to maybe crystallize value or monetize some of those to fund some of your other ambitions?
Yeah, like I said in my presentation, discipline in terms of the portfolio is important. We did sell $2.6 billion, and there are internal targets to continue on that path, and we will do more in the year ahead.
Andrew.
Thanks. This is Andrew from Barclays.
The slide you put up earlier where you highlighted that the analyst valuations are conservative in your opinion, and then combined with the slides you've given, the point of e-commerce portfolio generating a lot of EBITDA and cash flow kind of three years out, which, if you do it, should create a lot of fundamental value. I guess if we get to that time and that fundamental value is not being reflected in sell-side valuations or, I guess, more importantly, the share price, how does that kind of change your thinking about then the future structure of the group? I guess what I'm asking, does it make sense for Tencent to be part of the holdco when everything else can stand on its own two feet if you want everything else to get valued properly?
Firstly, I'm quite confident that if we actually achieve that, the markets may be wrong in a point in time, but in the long run, they tend to get it right. My expectation would firstly be that if we deliver cash flow, cash flow can't be ignored. I don't think so. In the unlikely situation that we get into a position where, because essentially what you're describing is that we will be successful of executing this, market will understand, describe value to it, the discount therefore will then remain high. That would be quite a disappointing outcome at that point. That's not what we're aiming for.
You know, there are two ways to look at the world. There is the way the glass is half empty. I know, I know. You like sometimes to make these questions. There is the other way.
The glass is half full. We are creating lots of value. We are moving faster. We are innovating. We are winning. We promised lots of things. You criticized the promise. We delivered everything we promised. I will tell you the half full. People are trusting more in what we are delivering. You saw today why I am trusting more in what we are delivering. Maybe in one year, people are going to say, "Shit, I should have bought this company when the discount was 28," because now it's a 9% premium, just like Berkshire Hathaway. I don't know. I don't know. Let's keep the hope on the half full glass. If you just look today, everything is going to explode. You are going to be wrong because we do things right, and we are going to keep doing a few things right.
All right.
Tim, if you ask a question about the discount, you're in trouble.
Hi. Tim Acker from Allan Gray. The first one's just a comment. There was an earlier question on, is the buyback the right idea? Just my humble opinion, and I think many shareholders would agree, we think it's a great use of capital, and it enhances NAV per share. That's just my opinion. The question is on Tencent. Currently, you've only been selling Tencent shares to fund the buyback. Is that still the intention to continue that way, or would you consider selling Tencent to fund other M&A?
That is the current intention to continue the way we are.
All right. Quickly to Cesar and to Silvia, then to Robert.
Yes. Hey, Cesar from Bank of America. I have two questions.
The first one is, do you think it's always possible to fix culture in companies that you think about acquiring? The second one linked to the buyback question. Do you think it could be a good idea to slow down the funding of the buyback from selling Tencent shares and maybe, as you increase your free cash flow, use your own free cash flow to fund the buyback?
The culture and the answer to the culture.
I think you can do whatever you want with culture. It depends on leadership. If you have the right person with the right culture that you want, this person will make the changes that will be necessary. Sometimes in a hard way, sometimes in a soft way, sometimes in a month, sometimes in a year. It depends a lot on the context. You can fix. How you fix that?
You set the right vision. You make the contract with the people saying, "That's how we're going to behave." You never negotiate your culture. You say, "If you're not behaving as we agree, if you're not following the vision that we are setting, you should not be here. If you are doing, if you are changing, if you are starting this, you will stay with us." The more people start to change for all the time, the more the company will start to materialize that culture. It is possible. Time depends on context.
A funny story. I arrived here one year ago, and people said, "You know, you have this funny English. Looks like Brazilian. This Brazilian thing." I do not know if the Dutch people will talk about this culture, etc. It was funny. The first meeting, I said, "I want to get feedback from everyone.
Please talk. Give me your feedback." There was absolute silence. I said, "Please say something." I talked here for one hour. No one gave any feedback. One year later, we are much more aggressive. We are communicating much more. You build culture. I learned more about Dutch, about our culture in Prosus. We changed the way we behave. It is possible.
Yeah. Your question was on, as the free cash flow improves, whether we can allocate some of that also to a buyback. That can be a possibility. I think, obviously, we want to share operational cash flow through dividends and enhancing that. If buybacks is a more efficient way to actually return capital to shareholders, then we can look at that.
All right. Quickly, Sylvie.
Yeah. Thank you. I have a high-level question.
When I think about the move to ecosystems, it reminds me of how in food delivery, there was a move from marketplace to delivery to other categories. The question is, is this where the industry is going, you think, or do you see still a place for single vertical players? Given the network effects from ecosystems, do you also see in food delivery or in classified other players potentially in the competitive landscape moving to an ecosystem model? Is this the direction for the industry as a whole, in your view? Thinking about the disclosure, as Nico mentioned, we're going to get some data. How do you think we should assess the success of the ecosystems? Should we look for frequency of usage of existing users, or is it going to be in terms of cost efficiencies, for example?
Wait.
Was Sylvie that wore the Brazilian shirt last time?
Yes.
Sylvie, you got my Brazilian shirt last time, and I gave a prediction last time. Brazil is going to win the World Cup. Oh my God, we lost the next game. My predictions now are much better. This time is going to work. I said the silly thing. I forgot what share.
The first one was about the ecosystem, the verticals. I can take you if you want. Let me tell you one thing. The PowerPoint will tell that you can just build a big business, and then it's easy to roll out to other verticals. That's the worst thing that you can hear from somebody. It's very difficult to do this. Very difficult. Remember how many times we discuss it for all the time? It's vertical.
The main point here is to understand what are the great assets that you built that really impact the next vertical that you have. Let me give you an example. I told you that we are scaling fast pharmacy in Brazil. Why? We had pharmacy for like three, four years, and we did not grow like this, but now we are doing. Why? Among other things, we understood that delivering in 10-15 minutes changes everything. Changes everything. Why can I do that? Because I have a logistic asset that gives me this capability. See, it is not about just having the vertical. We had the vertical four years ago, and we were not able to grow fast.
What I'm trying to say to you is you need to be very sure that you have the right asset for what changes the game when you go to the other vertical. How you discover that, it's tough. You need to test a lot of things and understand because you think you know which is the asset, but you don't know which is the asset. In the end, my main point here is everybody can become an ecosystem or a super app. Yes, but executing it, it's very, very diffic ult.
The easiest way to destroy a billion dollars. You create a super app, and you put all the verticals, and you just subsidize everything. The pharmacy thing, we are talking about that for four years. Much before I left iFood, we were trying every month new hypotheses. Now we are growing a lot profitably.
It's very difficult to do it profitably. That's why we have to test a hundred things. I was reading Twitter before I came here, and someone, David, I think you are there in the streaming, and he said, "I think Euro just took all the data about the users, sent it to LLM, and the LLM wrote something." I said, "No, my God. No, we are working six months to fine-tune a model, 60 models different, test it a hundred times." Sometimes you see a slide, and you say, "No, you just go there and do it." Yes, then you lose a lot of money. Why are we being profitable and growing? Because you have to do the technology right until you create a competitive advantage, then you can scale. That's what we are trying to do.
Your question, other part of the question was, is everyone going to do it? It is very difficult to do it because sometimes there are different companies. I think Prosus is well-positioned to do that. What we built in Latin America is very difficult to replicate. It is very difficult to replicate the Chinese ecosystem where, for example, in WeChat, you have so many players. Maybe because it is very difficult to do, we will have a competitive advantage, and we will play better than other companies. That is our bet. I think we have much more than 8.7% chance of succeeding.
You also asked how to measure. I forgot to answer this. It is very difficult to tell you an average KPI. Why? Because an ecosystem is very different. How do you measure the impact that the food delivery business has on my fintech business? You cannot get that through frequency.
You have to get a very specific thing. The best thing for me, it's difficult for you that are in the public market, is to look at the impact on cohorts. If you take the frequency of iFood, it did not change because of pharmacy. Why? Because food delivery is this size, and pharmacy is this size. If you get this cohort, you are going to see a bump in terms of increase. It is difficult for you to see from the outside, but I will always look at the cohort and how this cohort can be replicable.
We will share data. The one data for me is growth. We should grow faster. Otherwise, it is only PowerPoint. I hope to come back here in the future and say, "We are growing faster.
Let me show you the data that proves this growth because of what we are doing. If you're not growing faster, it's not working. We'll come here and say, "This is not working. This part is working. This is not working." That's how life is.
Okay. We are very close to time. Robert, do you want to take us home? Can I do a question? Can you do it for the last quote? It's going to be a good question if it's the last one.
Yeah, yeah, yeah, yeah, yeah. I have a question for Nico about capital returns. Of course, over the last few years, Prosus has accelerated its capital returns through buybacks and dividends. If I look at the buybacks, of course, you're limited because of the daily volume on the Euronext stock exchange in the amount of buybacks that you can do.
On the dividend policy, if you look at the buyback tax exemption that requires a consistent dividend policy. Yeah, these limitations, how does that impact kind of your long-term capital return program? Do you, for example, foresee consistent dividend policy? Yeah. How does that kind of work? Maybe a quick question about the India ecosystem. We already had a couple of questions before about the India ecosystem, which is, of course, a bit different instead of the European ecosystem and the LATAM ecosystem because it is a minority state, small states, where, of course, Fabricio in his strategy is really focused on being an operator. For me, it is not fully clear. How does that work in India?
Maybe I can start, Nico, and then Nico start. Yeah.
Ashu said he was going to start, and then you—
look, I completely concede that the starting point in India is different from what it is in LATAM. If building this ecosystem, realizing synergies was illogical, then I need to own people. I say, "Hey, I own you. You do this. I'm telling you." If building an ecosystem, connecting companies, realizing synergies is logical, then I do not need to have that influence to make that happen. That is the approach we are taking in India. Over time, of course, as we see opportunities to strengthen it further, we will keep on investing, acquiring.
Let me add to Ashu's point. You are right on your comments and on your concern. Again, try to see that as a movie, not as a picture. We changed the whole strategy that we are seeing today, I think, internally, maybe three months ago.
We are going in that direction. The starting point is what you are seeing because that was the previous strategy. Have 10% in many companies. We did it. Over time, we will have a stronger local ecosystem. Why? I know that because we have exceptional people there like Ashu that has the goal to have a very strong ecosystem. We are going to keep saying, "Ashu, deliver that to us." We are going to get there. The starting point is a little different. We are going to change over time.
You clearly know your Dutch tax regulation quite well. It is a little bit of a technical thing, but I'll try and be as clear as possible with it because the two things are quite different. It is a very important element and advantage that we as Prosus have.
When we created the Prosus listed vehicle, we actually, through the Tencent stake, were able to create a very sizable, call it taxable capital base of more than $120 billion. Now, that base helps us in two ways. Firstly, from a dividend perspective, we can distribute dividends essentially. Dividends, we do it through capital reductions, and effectively do that very efficiently then for shareholders that elect that way. We have done that over the last few years.
$120 billion.
$120 billion of—
a few years at least. Three, four, five. Okay. Okay. Let's keep working on that.
The second part relates to specifically the buyback elements. Now, the Dutch tax fiscal environment is actually quite punitive for companies that actually make share buybacks. We have actually been well shielded by that because of this high tax base that we have.
Our sort of call it capital element on a per-share basis that we buy back that effectively does not create any profits is actually very high. It's into the $40s per share. We further then qualify for certain exemptions based on the size of the dividend. I think that's what you were alluding to that comes on top of that. Clearly, as we grow our dividend elements, that's just going to be an answer. My answer to you is, despite the Dutch environment that's a little bit anti-share buyback and distribution, because of how we've set up the company and the big $120 billion shield that we have, we're actually distributing the cash and doing the buyback in a highly tax-efficient way for our shareholders.
Okay. We just have time. There is one more.
My version for that is, wow, we have a lot of tax benefits. Why there is this discount? Update your model, lots of tax shields here.
Thanks. Mandeep from HSBC again. You're building a great AI for your own usage, it seems. Do you have any plans to monetize that outside of Prosus at all? If yes, in what ways can you do that?
That was the initial idea that we decided, that the biggest impact we can have is internally. That's what we decided to, first of all, enable all our companies to do the best we can and try to do it as fast as possible in the most impactful way as possible. Having said that, this is new technology, technology that we own.
We can do what we want with that, and we are very confident we can use it to create the next wave of AI employment, which means not hiring people, but hiring agents to do the work of people. This might have opportunities outside. We have never made, after the decision to use internally, we have not made the call whether this is going to be commercial or not. For us, for now, we just focus internally.
Because our scale is big. We should first use internally, have a lot of benefits, then also sell externally. I think that we might get there. It is just not the focus today. Maybe next capital market day, next Nico birthday, we talk about that.
All right. We are going to leave it there.
A few final words. I lost my mobile phone. Very strange when you lose your mobile phone.
Final words. First, I'd like to say thank you all for coming, for staying a full day with us. We tried to share a lot. Some people gave me the feedback that it's nice because we are very open. That's the way we operate. Being open and sharing everything, bad news, good news, sharing what we are learning. We hope we increase the trust. We hope you decided to be a shareholder for the long term for us. Thank you for coming. I would like also, before we finish, to say the people from Prosus there, Roxanne, everyone else, please stand up. You saw that we spend a lot of time. You too, you're from Prosus. Please stand up. You too. I ask you to say thank you for our CEOs, everyone else. Thank you for organizing.
I think they spend a lot of time creating a very nice day to share with you. Thank you, the CEOs, the heads, and everyone else from Prosus. One more step to increase transparency. I enjoyed a lot. Hope you enjoyed and hope we are going to see again soon to keep sharing about our stories. My last note, we have a video to say goodbye. Goodbye not so long, but we will have drinks for the next one hour. I invite you all, if you do not have to go home now, get there. Let's get some drinks together and let's celebrate the last one year, but even better, dream about the next one year. Thank you. See you next time.