Good morning to everyone. Thank you for coming. Thank you for the interest. We'd like to get started with what will be our second Investor and Analyst Day. Thanks also to our friends at the Nasdaq for letting us use the venue, and being able to host the event out here in New York. Very quickly, you know the drill, but I've been asked to do this. A lot of what we'll cover today may contain forward-looking statements relating to such matters as continued growth prospects for the company, industry trends, and product and technology initiatives. The statements are based on currently available information and our current assumptions, expectations, and projections about these future events.
While we believe that the assumptions, expectations, and projections are reasonable in view of what we currently know and the available information we have, you are cautioned not to place undue reliance on these forward-looking statements, as our actual results may differ materially from those included in this presentation for a variety of reasons, so on and so forth. The idea of this Analyst Day that we carry out once every three years is to be able to stop and reflect a little bit big picture on the strategy and the long-term vision we have for the company. As those of you that follow us know, there's a lot of very exciting stuff going on. There are many moving pieces to the ecosystem we're building out, and we're actually quite excited as we see all these pieces coming together and generating quite a powerful platform.
Over the next two or three hours, we'd like to walk you through what these pieces are and the vision we have for them going forward. Hopefully then be able to engage in a constructive and interesting Q&A session with all of you and with the management team. Finally, lunch, to be able to interact with some of the key members of the MercadoLibre team. There are a lot more of us back in Buenos Aires, but it's great to be able to have four of the most senior execs here for an open and candid discussion. Having said that, we're going to start off with Marcos. Stelleo, our COO, will then cover the marketplace business. Daniel, our CTO, will go over our technology platform and what's going on in that front. We'll stop for a brief coffee break then. We'll reconvene.
Osvaldo will go over Mercado Pago, our payments business. At the end, I'll share some thoughts with you on how we think through the financial model. We'll break into Q&A. Having said that, let me hand it over to our CEO and founder, Marcos.
Thank you, Pedro. Good morning, everyone, and welcome to our second Analyst and Investor Day. Last time we were here seven years ago, we had revenues roughly of $50 million. Currently, we have annual revenues in the $500 million range. The idea today is to spend the morning talking with you about the initiatives that we have to try and grow this company, again, 10X, so that next time we're here, we are substantially larger than what we are today. When we started this company 15 years ago, we felt very strongly that the internet was going to change the world. That it was going to be a major thing, and that it was going to create huge opportunities. We felt every day that we needed to move very fast because the opportunity ahead of us was immense. 15 years later, we have traveled a lot.
We have built a great and sizable company that we'll be talking about today. What amazes me the most is that we still feel exactly the same way we felt 15 years ago. Every morning when we wake up, we feel that we need to execute very fast because the opportunity ahead of us is enormous. It's just as big, if not even bigger, than what it was 15 years ago. When you look, for example, at the penetration of e-commerce in Latin America, it's still less than 5% of retail. The relative size with respect to general retail is still quite small. It's growing a lot, but we think that this is going to be much, much bigger in the coming decades.
Euromonitor projects e-commerce to get to almost 50 billion by 2018, even then, the penetration of e-commerce is going to be less than 5% of retail. This compares to almost 12% in the U.S. Even in the more developed markets, e-commerce still has a long way to go, and it's in the very early stages. We're trying to put it into context. We're just beginning. This is day one for us. However, unlike 15 years ago when we were a couple of guys in a garage, today we have several advantages to capture this opportunity. The opportunity is just as big, if not bigger. I actually think it's bigger than what it was 15 years ago. We are much better positioned to capture that opportunity. Today, we have the leading e-commerce company in each and every one of the markets where we operate.
With a sizable advantage in terms of traffic. These are Comscore figures for August 2014. Desktop only, maybe in mobile, we have an even bigger advantage. We lead in each and every one of the markets where we operate. This is obviously a great starting point. We will be focusing Obsessively in these seven strategic initiatives. We have been focusing on these initiatives. Those of you who have been following our earnings calls will be familiar with them, but we will continue to focus on these initiatives in the coming years because we believe if we nail these initiatives, we can take this company to the next level. We can grow this company 10x from where we are today. Payments, let me start there, and Osvaldo is going to talk a lot about payments later today. Mercado Pago's opportunity is just enormous. Really large.
Even though today Mercado Pago is the largest online payment processor in the region, its growth and its opportunities ahead are really big. We're going to focus on growing Mercado Pago inside Mercado Libre, where it still has enormous room for growth. Outside of our platform, outside of Mercado Libre, where it is growing very fast, but also it has a lot of room for growth, and obviously on the mobile phones. Payment from the mobile phones, we believe is going to be an enormous opportunity. We have launched several initiatives in that front. Osvaldo will be talking about them, but I just want to highlight that in the next 3 to 5 years, we will invest and focus very heavily on payments because we think Mercado Pago can be even bigger than what Mercado Libre is today. Mobile is obviously another key initiative for us.
What mobile has done is our dream come true. When we started 15 years ago, our dream was everyone at some point in time will be able to access the internet and trade online in Latin America. Many people would say that will never happen in roughly a third or half of the population Latin America is poor or very poor. Those people will never have a broadband connection, a computer, et cetera. Today, everyone has a computer on their phones. Most of them are smartphones. Those that are not smartphones today will be smartphones in the coming years, and they will be able to access the internet and trade online very easily. It has made our market a mass market, and it's a fantastic possibility for us. It has enlarged our audience.
We have done a lot of progress in mobile, but there are still many opportunities ahead. Shipping, logistics, and fulfillment is a key area where we will be focusing in the coming years. We launched Mercado Envios less than two years ago. It has grown very rapidly. It's doing very well in the countries where it is operating, but it's just the first steps in an area where we will be focusing and investing heavily in the coming years. We will also be focused on attracting large brands, large retailers to our platforms. Historically, we had always tried to do that, but we hadn't been successful at attracting and retaining top retailers and top brands onto our platform. Occasionally they would come, but then they would leave. We had a lot of churn. That was because we didn't have the right product for them.
As we have increasingly become successful at solving payments and shipping and mobile, and providing for them a curated space, so that they can take care of their brands and position their brands the way they wanted to position their brands, we have become increasingly successful in this space. The brands and the retailers that are on our marketplace today selling very successfully is amazing. Every week we have a new brand coming over, and this has really gathered momentum. I feel that this is just the beginning, that the opportunities here are really big. The percentage of gross merchandise volume that is conducted by them is still relatively small, and we think it can be much bigger than what it is, and it's going to help us drive growth on the marketplace overall. This is going to be another key area of focus for us.
We will continue verticalizing our site. We did this first with motors, with cars several years ago, then we did it with real estate and services. More recently, we have done it with fashion and apparel, and every time we do that, we see a lot of success, we see a lot of growth, and we see how our network externalities become even stronger, how the power of MercadoLibre becomes even stronger and how we enhance our leadership position. We will continue to focus on verticalizing our marketplace, providing the best possible experience for buyers and sellers in the different categories. We will continue to focus obsessively on customer experience. The progress during the last two years in this front has been amazing. Stelleo will be showing how our Net Promoter Score has been trending.
I'm very happy with the progress we have made so far, but we still have a long way to go. We want to be the place where people have the best experience buying and selling online in Latin America. We're getting there, but we still have a long way to go. We have invested very strongly first in our product, but also in the tools and services we use to respond to our buyers and sellers and the people who browse our site. We're very happy the way we're going, but we will continue to invest heavily in this area during the next several years. Finally, our open platform initiative. When we opened our platform two years ago, we were the first and only company in Latin America to have an open platform.
Even developers in the region were not really sure what to do with it because developers in the region are more used to developing for software factories who do jobs for multinationals, and they are not used to being entrepreneurs, to seeing an opportunity in a platform and developing an application or a solution for people who are using that platform. We launched a $10 million fund to invest in developers who are doing interesting things on our platform. We have done 10 investments in some of the companies that we think are doing really exciting things on our platform. More importantly, we are seeing a lot of momentum. We really are seeing an ecosystem of developers building solutions for our sellers and for our buyers.
This is also a very strong competitive advantage because not only do we have a competitive advantage in the sense of the scale and the size and the traffic that we have, when we have an ecosystem of developers also focused on helping our users, it creates an even stronger moat. Those are the areas where we will continue to focus during the next several years. In addition to building these areas, we have created an addition to evolving MercadoLibre from a consumer-to-consumer auction site to the enormous marketplace it is today, where we service large retailers, branded retailers, small and medium-sized businesses, and we still continue to service consumer sellers, and we still have some people who like to buy auctions, and we still maintain those. We launched Mercado Pago to solve payments, Mercado Envios to solve shipping and eventually fulfillment.
We also launched Mercado Shops and our advertising solution, Mercado Libre Publicidad. Mercado Shops started being a storefront builder. It has roughly 120,000 sites that are using Mercado Shops. It is starting to become much more than that. It is also creating the place where people who want to have an omnichannel presence can manage all their e-commerce activity. They can build their site, but they can also, from Mercado Shops, manage what they do on their platform, on Mercado Libre, on their mobile phones, and on their store. That is the vision for Mercado Shops. It is becoming very successful in that front. It's a very powerful vision. We continue to be very excited with Mercado Shops. Mercado Libre Publicidad, that has over 13,000 active advertisers, started as text ads on our platform.
More recently, we launched product ads that are being incredibly successful and scaling very rapidly. These provide also a solution for people who not only want to list in Mercado Libre, but perhaps also want to buy our traffic and send that traffic either to products on our platform or to their own sites. Those sites might have been created through Mercado Shops or not. When you look at this entire ecosystem that we have created, basically provides a solution for everyone who wants to do e-commerce. We can build your site, we can build the tools that you need to manage your omnichannel sales. You can list on our platform. You can have your own site, but we can provide payments for your own site.
We can help you with shipping and logistics. We can help you buy traffic to send that traffic either to your site or to your products on our site. Basically, whatever you want to do in e-commerce, we will have a solution for that. That's the vision for our company. We want to be the operating system of e-commerce. I think we have made a lot of progress in that process, but we still have a long way to go. The opportunity is really enormous. This is the ecosystem that we're building. We will continue to make this ecosystem stronger. It's very interesting because we lead in each one of these businesses, but each one of these businesses helps the other one, and it becomes very synergistic.
Increasingly, we go and we talk to, for example, a branded retailer who wants to sell on our platform. It's very easy for us then to offer to them Mercado Pago on their own site. Then eventually to offer them our shipping and logistics solutions that right now only work for Mercado Pago. Maybe we will eventually open them up as we have done with Mercado Pago, that initially only solved for payments within our platform. Today, obviously, it's also very big outside of Mercado Libre. We have implemented all of these solutions only in a few of our countries. This is where we see the model really thriving, particularly Brazil and Argentina, that have all of the solutions I was mentioning before, Mercado Libre, Mercado Pago, Mercado Shops, Mercado Envios, have been launched and rolled out.
We see large retailers selling on the platform very successfully in large volumes and with a lot of scale. We see Mercado Pago growing aggressively. We see mobile commerce and mobile payments really taking off. We can roll out this, and we need to roll out this to all of our countries. That is an area where we'll be focusing very strongly in the next 12 to 18 months. We have spent a lot of time developing and adjusting each one of the solutions, and now we will be aggressively rolling out all of the solutions to all of the countries where we operate. That should be a very low-hanging fruit for us to continue growing. We will simply roll out all of our solutions to all the markets where we operate. Obviously, we will continue to iterate and improve on each and every one of them.
As I was saying before, we started 15 years ago with a consumer-to-consumer auction site, and we have been evolving it into an enhanced marketplace. Interestingly enough, even though offline it's very difficult to envision these two things next to each other, online, one can do it. We have been able to do it all under one brand, under one URL, and we have been building on top of each other. We maintain the leading position in the consumer-to-consumer trading platform. We still maintain people who do auctions. Roughly 2% of our volumes is auctions, 98% is fixed price. 90% of the items that we trade are new items, but still 10% are used items. Those 90% that are new increasingly are being traded by large retailers and by branded stores.
We still maintain our power sellers and our small and medium-sized sellers and our hobby sellers. It's very interesting because we are building on top of the strengths and making our moat even bigger, even stronger. Today, we already have scale. Even though we believe that 10, 20 years from now, we should be substantially bigger than what we are today, we already are a sizable business. That's important because this is a network externality business. The bigger you are, the stronger your moat. In the last 12 months, we had 90 million items sold on our platform, 37 million transactions processed through Mercado Pago, six million items shipped through Mercado Envios, 112,000 active stores, and 13,000 advertising using our Mercado Libre Publicidad solution. Already interesting scale, and each one of these businesses, again, creates synergies to the other.
We believe the opportunity is immense. In the last 12 months or last year, actually, we had 27 million people using our platform. 20 million buyers and seven million sellers. The region where we operate has 600 million people. It's not one country, it's 13 countries. If you want to do payments, you need to deal with 13 different financial regulators and different laws, et cetera. The opportunity is 600 million people, and we aspire to have 100 million people using our platform in a calendar year. The first time we said that, we had had four million people using our platform. Last year, we had 27 million. We have made a lot of progress, but we still are very far away from getting to 100 million. To be honest, there's no reason why we shouldn't get to 600 million people.
We aspire to grow substantially bigger from where we are today. Perhaps the key takeaway is we are just as hungry as we were 15 years ago. We believe the opportunity is just as big, and we are just as desperate to capture that opportunity as we were 15 years ago. But we are much better positioned to capture that opportunity than what we have ever been. Why? Because we already are present almost in all of Latin America, because we have built a brand that is synonymous with e-commerce in every country in Latin America, and it is a leading brand. It is the most recalled brand in the region. We have the leading site in terms of traffic. We have very strong network externalities already working on our favor. We have 24 million live listings. Whatever you look for in the region, you will find it on our platform.
That creates a lot of liquidity. We have unmatched prices, unmatched selection, and we are working really hard to have the best experience, and we have made amazing progress, and you will be seeing some of that today. We have an amazing technology team. We have 500 engineers and growing very rapidly, and they have built the best platform in the region. It is an open platform. We have an ecosystem of developers working on top of that platform. It is creating an amazing moat, and it is a team that has learned to adapt to the changes that the internet always brings. We are able to lead in this changing environment, and I believe in the coming years, we will increase our advantage with respect to our platform and our technology and our ability to adapt to the different changes that are constantly happen on our environment.
We have an experienced management team. I think almost everyone that is here today has been with the company for almost 15 years. We have solid financials. We are probably the only company that makes a profit in the region, but more importantly, we are growing very fast, faster than e-commerce. We have been doing that since we went public seven years ago, and we intend to do that for the coming 10 and 20 years. Again, just to reiterate, the opportunity is enormous. We are just as hungry as we were 15 years ago, but we are much better positioned to capture that opportunity because we are starting from a great leading position. Now I will hand over to Stelleo that will cover the marketplace. Thank you very much. Later on, we will have a coffee break and lunch, and we will also be entertaining Q&A. Thank you.
Thank you, Marcos. Good morning, everyone. MercadoLibre is the largest marketplace in Latin America by any measure. In the last 12 months, over 7 million sellers have published 24 million listings, thereby attracting more than 21 million buyers and generating over 90 million transactions for a combined merchandise volume of almost $8 billion. This level of scale generates the network externalities that continue to attract more buyers and sellers to our marketplace. It assures the liquidity that sellers need and guarantees that buyers will find what they are looking for. Since the last Investor Day three years ago, our marketplace has continued to grow at a fast pace, doubling volume and transactions. Growth has benefited sellers and buyers. For sellers, it provides the scale to operate more efficiently in an extremely competitive retail environment.
It allows them to capture the shift that is happening from offline to online commerce in markets that are still relatively under-penetrated. For buyers, growth has consolidated MercadoLibre's position as a major shopping destination, almost synonymous with buying online in the markets in the region. As our marketplace has grown, it has evolved from a pure marketplace where individual sellers connect with buyers, sometimes actually meeting in person to exchange payments and goods, into an increasingly more professional enhanced marketplace. MercadoLibre is no longer just a marketplace, but as we have worked on removing friction points in transactions, we have aggregated solutions that have transformed us into an e-commerce facilitator. In so doing, we have expanded the domain of our services even to third-party sites. Product selection has improved from an emphasis on consumer electronics to several other categories.
This kind of selection is unmatched by first-party retailers and has made us the leading e-commerce destination in all of our countries. Expanding into other categories require that we adapt the uniform targeted shopping experience we initially offered into more verticalized experiences, specifically addressing the needs of buyers and sellers in the different categories. These needs may include specific user experience features such as search tools or attributes for category, or even better ways of facilitating product discovery, such as curated content. Over time, entrepreneurial hobby sellers have seen the opportunity to grow with MercadoLibre. Many have graduated into very professional businesses with a few million dollars in sales per year. Today, we attract small and medium businesses that see our marketplace as a major sales channel in their multi-channel strategy. We have also adapted our organizational structure, hiring sales teams dedicated to attracting branded and large retail sellers.
Improving our interaction with our customers is an integral part of becoming an enhanced marketplace. In the past, this took the form of more reactive customer support as users contacted us to resolve issues. Today, we use information on payments and shipping to help manage our relationship with our customers proactively, anticipating problems and solving them to enhance customer satisfaction. Early on in our history as a company, we identified the need to work on reducing frictions that are inherent to any online transaction, particularly the payment component. This is how Mercado Pago was born more than 10 years ago. Osvaldo will present later on on how Mercado Pago itself has evolved in this time to become a leading payments facilitator both on and off MercadoLibre's marketplace.
Within the marketplace, integrating with several payments methods in different countries in the region has allowed for a simple and safe payment experience for buyers and sellers. Almost two years ago, we launched Mercado Envios in an effort to solve shipping, the other major friction point we saw in online transactions. Through Mercado Envios, we have integrated with the major carriers in Brazil and Argentina. Shipping products has become a more straightforward experience for sellers who have adopted our solution, and tracking and receiving them a more reassuring experience for buyers. Technology was necessary to achieve this integration with payments and shipping, but it has also enabled us to adapt to the shift in platform usage from the desktop to mobile as mobile devices have become ever more present.
MercadoLibre has led the way in mobile growth in the region, as we have identified the potential of e-commerce in this platform. We have actively encouraged downloads, and our apps are among the most popular both for iOS and Android users. At 16% of GMV in Q2, mobile is already a significant part of our business, and more importantly, it continues to grow very rapidly as this penetration was just 7% the same quarter a year ago. Growth in mobile is driven by the demand side of our business, as now almost a third of our new user registrations are coming from these devices. As cheaper smartphones and tablets have become more accessible in Latin America, entire user segments have leapfrogged the desktop, and an increasing number of our users navigate exclusively on mobile devices.
Of all platform innovations, Mercado Envios has experienced the fastest adoption over time, especially in Brazil, where it is already used in more than a quarter of transactions. This is an indicator that it has come to solve a real user demand. As we have gained visibility of all transactions shipped through our partner carriers, we have been able to negotiate volume discounts and improve service levels with them. Behind the scenes, technological implementation has allowed us to integrate with sellers' back ends, driving faster and more efficient shipping processes. Brazil is our most developed market, representing roughly half of all marketplace transactions. It is where we have deployed all of our strategic initiatives to facilitate commerce and seen significant traction in their adoption. Given these positive signs, we see the potential to continue launching in Brazil new initiatives such as a mall experience with brands and cross-border trade.
With the full suite of our initiatives deployed, Brazil has continued to grow rapidly and at a faster pace than smaller markets. Success of our complete model in Brazil has given us strong incentives to expand it to all of our markets, where we expect to accelerate growth. Consumers have become more sophisticated in the region and have gone beyond searching and buying primarily commodity products such as consumer electronics. This has paved way for growth in the lifestyle categories such as fashion, personal care, and home goods. As our selection has grown in general, it has also become less concentrated in consumer electronics and more evenly distributed among the different categories. As more categories have developed, we've seen the need to evolve from a one-size-fits-all approach to a tailored vertical approach per category. Our latest effort in developing a customized vertical has been in fashion.
This category has required us to adapt to very specific needs, such as preparing the site for seasonal promotions and product curation, giving more prominence to brands, and providing search capabilities in sizes and colors. Changing the experience of listing, navigating, and buying in this category has positioned MercadoLibre as an increasingly more relevant player in this space. As Marcos mentioned, vertical experiences are not entirely new for us. Our classified sites in motors, real estate, and services have had tailored experiences for years. This has gone as far as developing a different pricing model where sellers pay a fixed listing fee and obtain leads to complete transactions offline. As we have understood the supply chain in these categories, we have developed listing packages that cater to dealers with many cars and to real estate agents with many properties in their portfolios.
Similar to what we have done in fashion, we have created features that improve the demand side, such as search by make, model, and year in motors, map navigation in real estate, and geolocation in services. MercadoLibre is the leader in classifieds in the region, but we have identified interesting opportunities to acquire leading vertical properties in several markets. This has been the case with TuCarro in Venezuela and Colombia, with Autoplaza in Mexico, and Portal Inmobiliario in Chile and Mexico. In all of these examples, we have combined dedicated teams with deep understandings of their categories with our powerful ability to generate demand. As we have integrated databases into a single backend, we have obtained very positive results, allowing users to navigate more content from the different front ends. Traditionally, brands have been careful about controlling the entire retail experience and have been slow to take to marketplaces.
This resistance has given way to an increased desire to participate in the broad distribution that we offer. Our mall experience has accommodated the needs for brands to have more personalized experiences through official stores where they sell directly to consumers. Local brands in fashion, home goods, and consumer electronics, such as Ricardo Eletro in Brazil, have been the first adopters. Their movement has also attracted international brands such as Walmart, Diesel, Philips, Dell, and Sony. By the latest count, there were more than 350 official stores launched in Brazil, Argentina, and Mexico. We have been working on attracting brands and large retailers, but professional sellers, mostly comprised of small and medium businesses, are also a critical part of evolving to an enhanced marketplace. Professional sellers represent a small number of sellers, around 5%, and roughly half of our listings, but they generate 70% of our sales.
Their sales are significant since they bring high-quality products that buyers want. From our early days as an auction site for individual sellers selling used goods, MercadoLibre has consolidated its position as a major e-commerce channel for professional sellers selling new, fixed-item products. To complete our transition to an enhanced marketplace, we have invested resources to improve the experience we offer to our customers. As we have integrated payments and shipping, information is available for buyers and sellers, and the need to contact MercadoLibre has decreased. With fewer contacts, we are able to provide better service in an increased number of channels through which clients contact us. Customer satisfaction, as a consequence, has progressively increased over time. In summary, we are confident that the transition to an enhanced marketplace will continue to improve the user experience and to drive growth going forward.
Critical to this transition is our ability to execute on the technology front. Daniel Rabinovich will now present all of the interesting things we're doing in technology.
Hello. Thank you, Stelleo. Hello, everyone. I'm Daniel Rabinovich, CTO and Senior Vice President of Product. Today we're going to be discussing some of the technology foundations we have here in MercadoLibre and we've developed greatly in the last few years. A quick look into some of the things we've been doing product-wise in the last years, a word about our open platform. We're very proud of it. Technology foundations. For those who were here three years ago in 2011, we were in the middle of a deep technology overhaul. We had a great platform, very successful, but it was a decade old, it had two important flaws. First, it was monolithic, second, it was closed. It's very hard to grow a monolith.
We took a bold decision back then that was to complete rewrite our entire platform from scratch. We spent a year doing that without even releasing a single feature, and we released the new platform in early 2011. The most important thing we've done with this overhaul is the ability to split a large monolith into many little cells. What is a cell? It's like a little company. It's an autonomous unit. Its own code, has its own data, its own team, its own processes. Complexities among each cell is hidden from the outside, and cells communicate among each other using a well-established protocol called APIs. Are you familiar with APIs? That's great. This was really great for us.
One clear impact that this new way of doing things and organizing ourselves had in our lives, it implied a dramatic lift in the pace of execution and innovation. We used to deploy one huge version of our product into production a week, and now we're deploying 5,000 much smaller deploys into production a week. It's roughly one deploy every 10 minutes. There is no more central administration of anything. Cells, whenever they feel they're ready, they push code into production. This is very important for us, has been important in the last couple of years, and it's critical going forward as the ability to grow our team. Another side effect of organizing ourselves this way is the ability we have now to grow the team. When you add people to a centralized team, you're paying penalties. You're paying productivity penalties, diminishing productivity, marginal productivity.
When we have these cells, since the complexity among these cells is hidden from the outside, we're more comfortable doing so right now than we were a couple of years ago. As you can see there, we basically doubled our team in the last three years, and we're going to continue to grow our team aggressively going forward. Switching gears to product and technology, but features and things we've launched in the last couple of years. First, I mentioned that we did a technology overhaul in the back end, we also did that in the front end. We knew we had to have a beautiful product, beautiful and usable product. I think we didn't have that in our first decade.
We assembled a nice group of designers, not only graphic designers, human interaction designers, ability designers. We started, as Stelleo mentioned, relying on curation. We think we redesigned our platform in a way we're proud of. The challenging part was not to design the beautiful or usable thing. The challenge for us was to do that in a consistent way. We are a large team. We have many properties, and the hard thing is that if you land on every corner of our platform, you must feel that you're in the same place. It's not that easy when you have a distributed and large team. We developed many tools, style libraries. I don't want to bore you with it, but tools.
We've assembled tools for every one of those in the internal team, most interestingly, in the outside teams that are using our open platform, they can develop on top of our platform in a consistent way to present a consistent product to our customers. Another area we've focused on heavily is the use of data. We store everything. Every single click on the system, we store it. For a long period of time, at least a year. I'm not talking about analytics here, I'm talking about the ability to use data to make our product better. Use that data, process that data in some automated way, and give feedback to the product so the product can change automatically based on data or based on what we're learning from our users. I picked one example. Search results. Here. Search results.
What you're seeing here in the horizontal axis are the rows. Row number 1 to row number 50. On web, we have 50 results. On the vertical axis is the CTR, the click-through rate, how much each one of those products are clicked. This is data for just one query on the system. We have hundreds of millions of those, just one. If you aggregated all these queries, you will see a perfect curve, like a hyperbola curve. In this case, when you zoom into just one query on the system, people who search for, in this case, an iPhone, you will find outliers. Search can calculate and predict how many clicks will the particular listing get, but it's a prediction. You have to process the feedback on that prediction.
As you can see, this is a nice example because the now old-fashioned iPhone 5s, it's not being quite heavily clicked, and it's enjoying a very relevant position in our search results, as opposed to the new iPhone 6, which is buried in, I don't know, the position number 18, I guess. It's quite clicked for that position. Search can understand that and correct that in real time, and that's for one of the queries. This is just for one query, and we have many of those. I wanted with this example to give you a glimpse of how in the back end information is processed. This is a lot of information. This is just one query out of millions of them, out of several operations, and we can chew and crunch all these numbers and give feedback to the product in real time.
We're doing many more of those kind of experiments and features, we will continue to do so in the future. Another thing we've been doing, as Stelleo mentioned that, is the ability to verticalize our platform. By that we mean that we were born as a horizontal product, the typical one size fits all. You know that one size does not fit all in many cases. We started working on specific verticals. We say verticals, it crosses all the tiers of a platform. We've been working on real estate, for example. The challenge here is to how to adapt, for example, an open search to a particular vertical like real estate. If you're familiar with real estate platforms, you will see that almost no one of them have an open search box.
They typically offer a map or a very well-structured form. It's not easy to process open queries for real estate. We haven't got this choice. When a user go into a homepage, she may write a real estate related query. First, we need to understand that this is a real estate related query. Second, we have to understand it. We cannot say, "We're not processing that query. Go to the real estate category homepage and perform that query again." It's not going to work like that. In this case, I've searched for apartments in Belgrano, which is both a neighborhood and a street in Buenos Aires. As you can see there, search tries to manage that and show a high concentration in the Belgrano neighborhood and also apartments in the Belgrano street, in this case.
Another vertical we've launched recently is the fashion vertical, as Stelleo mentioned before. A very important one. Each vertical has different challenges. In this case, there were two critical features. First, the ability to show variations. By variations, I mean maintaining inventory by combinations of size and color. Also the ability to show beautiful pictures and to attach different sets of pictures to specific colors. You may think that this is pretty easy for a particular retailer, and it is. If you're just one retailer, this is basic stuff. When we have a marketplace with millions of sellers and hundreds of thousands of professional sellers, understanding that one picture is beautiful is not a very hard concept. You cannot ask a computer if a picture is beautiful or not. Sadly, you need machine learning for that, for example.
For sizes and colors, to standardize as you may know, different retailers have very different way of describing sizes and colors. Very different. In order to present a consistent product to our customers, we need to standardize many different way of expressing sizes and colors, and to present that in a consistent way to our buyers. That is a challenge for a marketplace. Another challenge is the uploading of new listing, the creations of new listings. If you're a singular retailer, you have all the time in the world to describe your inventory. As a marketplace, we needed to adapt our APIs for the professional sellers, and also we needed to adapt all our selling flows for a C2C and a small B2C. You can list a pair of shoes from your mobile application, and you will somehow have to describe size and color.
If you're many of them, you have to describe all these variations. That was the challenge for us in the fashion vertical. Fashion is a nice segue for this neat new features that is called official stores. We've been really successful in the first part of our history delivering a great product for C2C and the small and medium businesses. It was hard for us to onboard brands, big box retailers, or maybe manufacturers. There were many aspects of this problem, commercial aspects. Product-wise, they needed differentiation. A brand requires to be treated differently as a small B2C or even particular user. We created the concept of a store. We differentiated them among our marketplace. We created this verified icon as an endorsement, so we start hammering with this icon everywhere in the platform.
We gave them the opportunity to customize their background, which is unique and particular for these official stores. In this case, the Wrangler brand has this beautiful background there. You can also search from within a store. If you're searching using the regular search box, you can check this checkbox there, and you're going to search from within the store. You can search within all the official stores in the system. As you can see, we're getting closer and closer to delivering a mall experience within our marketplace. Instead of just launching a new thing, we are progressively delivering a mall experience into a regular marketplace. We believe that is possible and it's very synergistic with the rest of the platform. Shipping. Stelleo covered that, our business model there. Product-wise, shipping was very important for a couple reasons.
First, since now we own the relationship with the carriers. Remember, we don't touch inventory yet, but we own the relationship with the carriers, and that is critical for several reasons. First, we can negotiate the best fees with them, and that levels the playing field with all sellers in our platform. We have scale, we can negotiate the best fees with the carriers, and we can pass on those discounts to every seller in the platform. If you're an individual, you're now paying the very same shipping fees as a large retailer, and that's a game changer. Second, a tiny detail, we now know where packages are. Before shipping, we had to rely on the feedback system to know if the transaction was completed or not, and that was far from perfect.
Right now, we know if a package is shipped and when the package is delivered, and this perfectly closes the loop with the payment system. We can now know exactly when to release the fund, what to do with the funds. If there's a problem, we can know that in advance. There are many things that we can do to prevent problems from happening and to deliver a nice experience. We can send a notification saying, "Our packages have been shipped," and, "Your package has been delivered." We also developed this free shipping option in which sellers absorb the shipping costs in return for a higher conversion. As you can see there, it's hard to see, but it's green there. You can purchase something. Also, in this screenshot in particular, this describes a new functionality among shipping, which is available in select cities.
If you purchase something in the morning, if you're in Buenos Aires with some sellers, you can receive that in the afternoon. If you purchase that in the afternoon, you can receive that the next morning. We also launched an initiative that is very similar to the lockers initiative you're probably familiar with. You can deliver that instead of a particular home or a particular address, you deliver the package to a carrier branch in this case. Saving the last mile is faster and cheaper, it's getting traction, particularly in Argentina. Again, switching gears, another great project in the last couple of years was going mobile. I remember that we launched the first native mobile application just weeks before the last Investor Day on October 5, 2011. Right now, we're proud of the progress we reached.
We are now able to deliver a very specific experience for the four main platforms, I mean, desktop, the web mobile, Android, and iOS. Again, we're not reusing stuff. We're developing for native on iOS and developed a completely different product for Android. Of course, we're reusing the backend part. We're relying on the same API, but we're delivering a specific experience, and that is a challenge because we're basically doing four times the same thing. Actually, the desktop and web mobile have many more synergies, so 3.5 times the same thing. As the product got better and better, we saw clearly impacts in the penetration of mobile. You can see there, this curve speaks for itself. We're now at 16% of our volume coming from mobile, but it's the progress what I want you to pay attention to.
We don't know where this is going to end, clearly, it's gaining momentum and momentum. We see this trend only going up. Since we have all this infrastructure for developing mobile, we started developing payments, leveraging that and developing payments off-platform, in this case, both an mPOS, a mobile SDK, and a wallet. I'm not getting here much deeper because Osvaldo is going to cover that in great detail in a few moments. I wanted to dedicate a moment on our open platform because it's something that we definitely launched in the last couple of years after we said we were going to do that in the last Investor Day, but we launched a couple of months after that. The most important thing for you to understand here is that it was not a product designed for the community. We created this platform for ourselves.
We created this huge service layer from scratch. We mounted all our web properties on top of this platform. Someone's going to kill me for trashing our logo this way. These four colors represent our four properties. Please, this is not a consumer-facing logo. All of our web properties are mounted on top of the same API. Remember, this is a decoupled environment. A front-end developer on MercadoLibre is working as if she were a third-party developer, there's no difference. We're treating ourselves as clients. All the mobile applications you've just seen are developed on top of the same API and also the backend tools. We had this progress in NPS, in the Net Promoter Score, because we're delivering better services. We implemented salesforce.com, for example, as our main backend tools for handling our relationship with our customers.
salesforce.com, guess what? Is talking to the same APIs. After we've done all that, we were like eating our own dog food. We were the largest possible client for API in Latin America. We're the largest ones. It was like a no-brainer and a logical step and a minor step to open that platform to the rest of the development community. That is the most important thing for you to understand. For us, the open platform is not just a feature, it's sharing the things we are using for ourselves. It's not that easy. When we released, we thought it was like if you build it, they will come. We opened our platform and started waiting, and of course, nothing happened. We needed to push the developer community a little bit. We issued a developer conference.
I'm wearing a T-shirt that we used in that event. It says, "Our API, your ideas," it's like a logo with a puzzle. We've done three developer conferences. We issued hackathons. As Marcos mentioned before, we also created a fund to invest in companies that help our ecosystem. It took some time, but it took off. Right now, we're seeing an enormous amount of innovation from our development community, and it comes in many flavors. Sellers integrating directly. Sellers in our platform, professional ones, have technical departments, they integrate directly and link their system with our systems. Real estate agencies, they have a huge amount of properties, they also integrated with real estate agencies and real estate networks in Latin America. We went up from 70,000 properties to more than 1 million, because now inventory is synchronized in an electronic way.
Also e-commerce companies in Brazil, we call them e-builders. They facilitate e-commerce in the region, integrate MercadoLibre, and offer our platform as one of their channels. Many flavors. External innovation was really important. To put that in just one number, we're proud to say today that 25% of our new paid listings are coming through our open platform. That is huge. That was 0 two years ago. The beauty of that number is that each one of those 25% listings, excluding real estate, those are not individuals. Sellers that list their inventory through our open platform are necessarily professional sellers, either integrating directly or using third-party tools. Wrapping up, we think we've developed a very strong technology foundations.
We performed this deep technology overhaul, also performed a design overhaul. We use big data to allow our product to learn from our users and give feedback. We verticalized some categories. The latest one was fashion, many more to follow. We launched this official stores product to enhance the ability to deal with brands, retailers, and manufacturers. We made strong progress in shipping, developed this strong drop ship solution, we're probably getting a lot more focus in this area in the following years. Strong progress on mobile from virtually 0 in the last investor conference to 16% of our GMV right now. Last, we've opened our platform, as we said we would, reaching 25% of new paid listing coming through our new platform. That was it. Thank you very much. I think we have a break here. We have a break?
Yeah.
Now with coffee
brevity is the soul of wit, and as avowed Anglophiles we are, we read our Shakespeare. We're running a little bit early, so why don't we do the coffee break now, and then we'll reconvene and go through the other presentations, and hopefully that'll give us a little bit more time for Q&A at the end.
Okay. Let's move on to payment. In our prior Investor Day three years ago
We said that we were very excited about the opportunities for Mercado Pago, that if we executed well, we would be able to grow at a faster rate than e-commerce in Latin America in the coming years. That is, fortunately, exactly what happened. In the last three years, we have nearly tripled our TPV and quadrupled the amount of transactions done with Mercado Pago. At the same time, we are also seeing, when we look at the future today, we also believe there are lots of opportunities going forward that will enable us to continue growing at a fast pace. Therefore, what I'd like to do today is spend most of the time during the presentation talking about which are these opportunities for growth we see for the coming years.
It's usually credit cards, bank accounts, and also cash alternatives in each of these countries. Two things are relevant, speaking of Latin America, in terms of funding options. The first one is related to credit cards, and it's called installments. Installments are very popular in each of Argentina, Brazil, and Mexico. By installments, I mean if somebody is buying a laptop that costs $1,000, they can, rather than pay it up front, decide to use their credit card to pay in up to 12 installments. Probably they will pay $100 per month and in total, $1,200. Many times consumers are, in Latin America, most concerned about whether they are able to pay the monthly installment rate than what is the total amount they will end up paying.
It's usually credit cards, bank accounts, and also cash alternatives in each of these countries. Two things are relevant, speaking of Latin America, in terms of funding options. The first one is related to credit cards, and it's called installments. Installments are very popular in each of Argentina, Brazil, and Mexico. By installments, I mean if somebody is buying a laptop that costs $1,000, they can, rather than pay it up front, decide to use their credit card to pay in up to 12 installments. Probably they will pay $100 per month and in total, $1,200. Many times consumers are, in Latin America, most concerned about whether they are able to pay the monthly installment rate than what is the total amount they will end up paying.
The other thing worth mentioning is that bank penetration in Latin America is significantly lower than it is in the U.S., it's roughly 40% across the region. Sometimes it's 35%, sometimes it's nearly 50%, but it's roughly 40% across the region. Therefore, we do need to offer alternatives for people who don't have a bank account. What we do varies on a country-by-country basis, but the mechanism is more or less the same. For example, in Brazil, there's an instrument called Boleto Bancário, where buyers can print an invoice with a barcode in it and then go to any branch of any bank or even to a kiosk or to a drugstore, and they get scanned that invoice, and the money is in their Mercado Pago account within 24 hours.
The money is sent to the seller, to the Mercado Pago account of the seller, and he or she is able to withdraw it to their bank account. We offer a buyer protection program and a seller protection program to ensure both parties against fraud. Mercado Pago today is available in six countries, in Brazil, Argentina, Mexico, Chile, Colombia, and Venezuela. The value propositions for buyers and sellers is, for buyers, it's about security, about convenience, about ability to pay with any means of payment they have, and also the ability to pay in installments. On the seller front, it's mostly related to, it's very easy to integrate and to set up. If you were to integrate with each of these 56 or 50 or 10 or whatever means of payment, it would take a lot of time.
To start using Mercado Pago takes just a couple of hours, and you can include Mercado Pago in your website or in your mobile app in a variety of ways, which are very easy to do. It also has a very high conversion rate because we have a huge base of users and of stored credit cards. We have tens of millions of stored credit cards. If somebody's coming to your site and you have Mercado Pago installed, it's very easy for them to pay with a card that was already uploaded into the Mercado Pago system. In terms of business models, we have two revenue sources. The first one is financing, because when somebody pays in installments, we charge them an extra to the buyer to be able to do so.
What we charge is less than what we pay them to the bank or to the financial institution to discount those receivables. We make a profit, a financial profit, for the revenue from financing operations. The second one is a processing fee, which we charge merchants when they process the transactions off Mercado Libre. This is in the order of between 4% and 5%, although we do give discounts to some larger sellers. When transactions occur within the Mercado Libre platform, we don't charge a processing fee. We used to in the past, but what we did several years ago was to bundle the Mercado Pago processing fee with Mercado Libre's final value fee into a single fee. In this way, we make it sort of free, if you want, to use Mercado Pago.
We also made the use of Mercado Pago, the adoption of Mercado Pago, compulsory among buyers in these countries. Yes, among sellers in these countries. How have we done? In the last three years, we have grown transactions at a CAGR of 54%, going from 10 to 37 million, and total payment volume from nearly $1 billion to $2.8 billion, growing at a 43% CAGR. We have grown very strongly, and we believe that we can continue to grow at a fast pace for the coming years. What are the main opportunities we see going forward? We're trying to align them across what we call five avenues for growth. Those are on-platform growth, off-platform or merchant services growth, financing, mobile payments, and then buyer and seller credit.
What I'd like to do for the rest of the presentation is go through each of these opportunities and explain where we stand and what do we see as the opportunities going forward. First, let's talk about on-platform or transactions that occur within Mercado Libre. We believe that we have cracked the code for driving Mercado Pago penetration of GMV. Probably the most successful case to date is Brazil, where nearly three-quarters of the volume is paid with Mercado Pago. We have, a few years ago already, bundled the, what we said, the Mercado Pago within Mercado Libre. What we did in the last, I'll say, 18 months were, first, significantly improve our fraud prevention algorithms.
Fraud has been low for a long time, but we worked a lot to improve our approval rates so that most of the transactions, the vast majority of the transactions that go through the system are processed right away, and there is no delay in those payments. That has helped us drive up the penetration of Mercado Pago. Also, as Marcos and Stelleo mentioned, during the last 18 months, shipping has grown a lot in Brazil. Mercado Envios has grown a lot. We have found that Mercado Pago and Mercado Envios are extremely synergistic, because in order to be able to use the shipping we provide, you need to pay in advance with Mercado Pago. Therefore, both in Brazil and to some degree, too, in Argentina, Mercado Envios has driven up the penetration of Mercado Pago.
Finally, during the third quarter this year, the quarter that just finished, we started to make what we call integrated payments for some items in Brazil. We started with those items or high-ticket items, those that are above BRL 500. In order to buy those items, you need to pay them right away using Mercado Pago. It's sort of a departure for the traditional model, where you could first buy an item and then pay for it using Mercado Pago or arranging with a seller. Now you have to pay for those items right away. The initial experience has been very good, and therefore, we are already reducing the limit from BRL 500 to BRL 400 in Brazil, analyzing the possibility of launching this in our markets.
We believe that as we implement these 4 points in the other markets, we'll be able to accelerate the adoption of Mercado Pago in those countries, which has grown a lot in Argentina over the last 12 months, has grown in Mexico, and it has really grown, too, in Venezuela and Colombia. With regards to Chile, we are about to relaunch Mercado Pago in Chile. We haven't put much effort in the past in Chile because of the local regulations and interactions with the processors, but that has been solved now, has been taken care of, and we plan to relaunch Chile before the end of the year. In terms of merchant services, merchant services has been growing, I would say, at an even faster rate than Mercado Pago, gaining share from Mercado Pago, but it's still a fraction of the total volume of Mercado Pago.
So far, the volume has been more concentrated in Argentina and Brazil, mostly, and to some degree, Mexico. We have really focused in these top three industries, which are travel, coupons, and retail. First, we believe there is a huge opportunity to continue growing in these countries and in these industries. Also, we believe that there are many opportunities along, probably four other areas. The first one is expanding to new verticals. There are some verticals we have not yet focused on. One of them, for example, is recurring payments. We have just very recently launched a recurring payment solution for Argentina and Brazil. This will enable us to target, for example, subscription models or industries which have a very low ticket that are bought frequently, such as digital downloads or apps or games or music.
Therefore, we believe there are many categories where we can really continue deploying Mercado Pago. As we said, so far, we focus mostly in Argentina and in Brazil. We're starting to gain traction in Mexico, and we have plans to focus more, first in Colombia and then in Chile as we relaunch the product before the end of the year. The third one is cross-border payments. We are seeing many companies, mostly from the U.S. and Asia, selling into Latin America, and many times they need to process payments domestically because large proportions of the credit cards in Latin America, mostly in Brazil, work only for domestic transactions. Therefore, Mercado Pago is a great way for them to collect money domestically and then have a bank wire them funds cross-border. The fourth one is what we call open platform.
It goes in the same lines as what Danny discussed. Basically, in the past, in order to use Mercado Pago, the seller needed to have the buyer go to the Mercado Pago website and complete there the transaction. What we've been doing in the past, first, we launched APIs so that sellers could integrate us, we built around those APIs, software development kits, SDKs. These SDKs, what enables sellers or developers to do is to have It's more flexible and easier for sellers to integrate Mercado Pago into their checkout in a way that probably the buyer isn't even aware that Mercado Pago is processing the transaction. The transition in the checkout is really smooth. It all occurs within the website for the seller, and the buyer, really, at no point has any doubt that the transaction is occurring at that same site.
In a way, it works in a very similar fashion to what, for example, Stripe or Braintree do in the U.S. Moving on to financing. Nearly half of the TPV we process is paid in installments. If we were to take the total TPV, 80% is paid with credit cards, and of that 80%, 60% is financed. In total, it's 50% and it gets financed. The column to the right is the number of installments users decide to pay with. As you will see, 12, 6, and 3 are the most popular ones. 12 is the most popular one. When buying mostly high-ticket items, consumers opt to pay for it in 12 installments. I think it's worth mentioning how the business model works here, because it's rather particular. Let's talk about Brazil, which is the highest market.
It varies on a country-by-country basis, Brazil is the largest market for us, they are probably the one country where the industry works in a different fashion. When somebody buys in installments in Brazil, the merchant gets paid in installments. We get paid as a merchant over 12 installments when somebody buys in 12 installments. However, first, we take no credit risk. The credit risk stays with the issuing bank. If somebody with a Bradesco card is paying us, if after three installments, for some reason he stops paying his credit card, that is a risk Bradesco assumes and not us.
Second, even though it do require a lot of working capital to work in this way, given that we have all of these receivables, we are able to discount those receivables in the market and sell them at a convenient rate because these are very low-risk receivables. These eventually have the risk of the issuing bank, they are guaranteed by Visa and Mastercard. They have very low-risk receivables. We are able to discount them at a lower price than the one we charge buyers for paying us in receivables. That's where we make a profit, which we call financial revenues in our P&L. This is how financing has worked so far. What we started testing a couple of months ago during Q3 is what we called, in Portuguese, parcelado sem juros, or zero-interest financing.
Of course, there is no zero interest, what we do here is instead of charging the buyer for the privilege of paying in installments, we charge the seller. In the past, we used to charge sellers a final value fee. MercadoLibre used to charge sellers a final value fee of between, say, 6% or 7% and 11%. What we did was we maintained that, we also introduced a new listing type in Brazil, which we call Diamond. We charge this new listing type 15%, and this listings type has the ability that whenever a buyer pays for it, he can pay in any number of installments that he or she wants, and we will charge him the same.
There is no extra cost for the buyer for paying in installments, the sellers do have to pay 15% instead of 7% or 11% for offering this. We started testing this in Brazil three months ago, only with power sellers, the initial results have been very good. What we have seen is, first, more buyers paying installments than before. Second, there is a shift toward longer installment periods than before, as we expected. With the pricing we use, we have been able to maintain the margin for the financing part of the business, therefore it has enabled us to increase the monetizations of the items that are paid in installments in this fashion.
We started with power sellers three months ago, and now we are starting expanding this beyond power sellers to some other sellers in Brazil and planning to roll this out to Mexico in the near future. Moving on to payments. Until a couple of weeks ago, the only way we could get payments through Mercado Pago was when somebody was using MercadoLibre's app and paying for an item using the MercadoLibre app. We were working on three channels and three initiatives for payments, and I'd like to cover them today. We'll go into each of them. Each of them will get a slide. Let me just introduce them. The first one is SDK for in-app payments. The second is what we call a new wallet to a buyer's app, and the third one is a mobile POS. Let's talk about each one.
First, the software development kits. This works in a very similar fashion as we described for desktop applications and for websites. The one difference is that we see e-commerce migrating from desktops to mobile, and within mobile from browsers to apps. Therefore, what these SDKs allow is for developers or large retailers who have their own app to include payments within their app without the user noticing and using the Mercado Pago technology. It's exactly the same as what happens when you use Uber, and you don't even notice that they are using Braintree in order to process your payment. We just rolled out these SDKs a couple of weeks ago, and we are seeing lots of interest from the community of developers to include them into their apps. The second one is Mercado Pago's own mobile app.
As we said, MercadoLibre's app allows you to process payments within MercadoLibre, but if you want to send money to a friend, you are not able to do that. If you wanted to pay for a transaction, which you closed in MercadoLibre, say, a couple of days ago, but did not pay at the time, you were not able to do that. That's why we launched the Mercado Pago app two weeks ago. It also allows, beyond those functionalities, it lets you, for example, if you're paying on a website that uses Mercado Pago with your mobile browser to conclude that payment using this app, which is more convenient since it has already stored in it all your financial information. We launched this two weeks ago, and it's already top ranked among the top five financial apps in Argentina.
We just launched this in Brazil a week ago, but we don't have yet numbers for the ranking, but we see that the number of downloads is scaling fast. The third one, this one we are just starting to pilot test with some users in Brazil and some users in Mexico, but the idea is to roll it out next year, is a mobile POS. POS penetration in Latin America is lower than it is in the U.S. For example, in the U.S., there is one POS every 30 people, whereas in Mexico, there is one POS every 180 people. So it's one-sixth the penetration of POS in Mexico compared to the U.S. We believe we do have some interesting advantages.
We have relations already with millions of sellers, and hundreds and thousands of those are small and medium businesses, which many times they also have a physical presence, but many times they don't have a POS. We could enable them to charge credit cards when somebody shows up at their store. Also we have a strong brand and we have big scale, so we are in a good position to negotiate with the processors in each country and to bundle it with the volume we already have. The image here is for a POS that will work in Argentina eventually. In Brazil and Mexico, though we will test the solution, both countries are strong on chip and PIN, so the dongle we'll have to use will read a PIN, and we'll have a keypad for you to enter the chip information, the PIN information.
Finally, we are expanding into cards and credit. We launched first a co-branded credit card in Brazil last year, it's ramping up the number of cards that are issued every month. We are working with Mastercard and with the local bank, with Cetelem. We have a huge database of users, it enables us to target those users who we believe could use more credit. For example, they are trying to buy something, they get a reject on their approval for the transaction. We send those clients to the bank who check their credit rating and see if they are willing to offer more credit to them. We are running the co-branded credit card in Brazil together with the bank, on a revenue-sharing agreement. Therefore, we are not taking the great risk here either.
In Mexico, we are announcing this week that we are launching a prepaid card. The reason for launching a prepaid card in Mexico is, first, that bank penetration is low. Second, that there are very few prepaid cards in the market. Third, that in Mexico, there's a, I would say, beyond having many people who want to do e-commerce but don't have a card to use it on the web, there's also a growing concern about online fraud. Many people who do have a bank account would rather use a prepaid card to buy something online because they feel it's more secure. The third one is seller credit, we are still in the planning stages for this one. It's something we're working during next year. Here, the idea is we have lots of financial information for our sellers.
We know how many payments they have received every day for several years, many times. Therefore, we are able to evaluate the credit risk probably better than most banks. We are willing to give them credit, for example, for working capital when there was a special occasion, say, holidays or Thanksgiving or whatever, and they need to use their capital to buy X inventory to increase their sales. We are willing to finance that. Because also we'll be able then to collect money from them as the flows of capital get in, as the flows of money get in when they start selling that inventory. In summary, we believe we have huge opportunity in front of us. We believe we've cracked the code for MercadoLibre in Brazil, we are replicating that across countries.
Merchant services is growing very strongly, growing at a faster pace than average for Mercado Pago, and we believe that we can continue expanding into new categories, new countries, and new segments. In financing, it's doing fine, but we believe that with parcelado sem juros with low interest rate, we'll be able to increase the amount of volume we finance, first in Brazil and then in Mexico. We are just getting into mobile payments, but the opportunity there is really big. Also we are starting to launch cards, co-branded cards in Brazil, prepaid cards in Mexico, and seller credit next year, probably in Argentina. With that, I'd like to pass it to Pedro, who will cover the financial implications of the business initiatives we've been seeing throughout the day. Thank you.
Thank you. Great. The idea now is to be able to get into the financial model and to walk you through the implications of what we're seeing. Before I get started, just one clarification. Our objective and our goal is to continue to build out the preeminent e-commerce platform throughout Latin America. We don't view our financial model as the end in itself, but we view it as a critical tool to be able to carry out what my colleagues have just walked you through. A big part of what I hope to be able to transmit to you today is the philosophy that guides our thinking through that financial model and the conversations that we're having internally around what that financial model should look like in an industry that's extremely changing and dynamic, and where financial flexibility is key.
Before we do that, let's just recap what we told you 3 years ago and what has happened since. Back in October of 2011, we essentially highlighted the following as the most relevant aspects of the model that we wanted to deliver. Continue to drive strong and diversified top-line growth. Sustain what we view as a very healthy gross margin profile. However, one that is being impacted by mix shift as we see very strong growth and mix gains from our payments business, which is a lower margin business. Focus on efficiently managing OpEx so as to be able to drive scale that can partially offset some of those margin compressions. Deliver earnings in an e-commerce space, as Marcos said, where many other players are running very tight margins or no profit at all. Maintain our focus on cash flow generation and cash flow positive business units.
Finally, make sure that our balance sheet and our capital structure is efficient enough for us to be able to carry out everything we need to over the next five, 10, and 15 years to really capture the enormous opportunity in front of us. I think we've delivered quite consistently on all of these fronts. In terms of revenue growth, if we look at four-year growth, the business has been delivering a 30% CAGR in U.S. dollars, nearly 40% CAGR in local currencies, with what has been up until now a relatively flat pricing and take rate structure. In terms of diversification of the business, the business continues to be well diversified geographically. Brazil represents slightly more than half of our business. It's by far the largest market. We've seen improvements in our exposure to some of the more complex and complicated markets such as Venezuela.
As we've began to highlight today, enormous opportunity in some of these newer markets where we will begin to pay more attention and more focus. When we look at the other countries, which is still less than 10% of the business, there obviously are some very interesting economies there. Chile and Colombia, the two largest ones within that segment, should be drivers for future growth. Mexico continues to under-index, then there are some other interesting countries there. When we think of the business by revenue source, it also has significant diversification. Our payments business is growing extremely well. Advertising, shipping, and our classified offering that most other e-commerce retailers don't even offer.
On the right, what you see there is that because of the way we account for our marketplace business, where we do a lot of bundled pricing, a lot of the revenues that are being captured by the marketplace business actually have a very important component of services being offered by the other BUs that get monetized in the marketplace revenues. If we look at our Brazilian business that has the most deployed service and pricing model, what you see is that nearly two-thirds of all marketplace transactions already have one component of the services being offered by the other business units tied in. Whether that be simply a payment processing, a credit offering through financing, the usage of Mercado Envios, or more than one of those services.
In essence, the very strong growth of all these other BUs is what is allowing us to increase and improve the monetization on the marketplace, which in our financials comes across as marketplace revenues. In reality, is very largely driven by the growth of the other BUs that don't directly monetize because of the bundled pricing strategy. Diversified by geography, diversified by revenue streams, check. Gross margins. The story has played out much as we anticipated. We've seen about 400 basis points of gross margin compression, driven practically entirely by a payments business that has been growing more rapidly than the marketplace business, and consequently, because of its lower gross margin profile, has been depressing gross margins.
Mercado Pago penetration in terms of TPV as a percentage of GMV has more than doubled since 2010. We've been able to offset some of the compression that you would have seen by also improving the relative gross margins of each of the different BUs. In terms of what we could have anticipated the gross margin structure depression to have been back in 2010, it's actually played off better than we thought because the margin structure of the standalone payments business has actually improved. In terms of operating expenditures, the other area where we can offset some of the gross margin compression, we're quite pleased with the way we've been managing the model, it's what we're calling scale the right way.
If you look at the two areas where we do not want to manage the business to scale, technology and product development is the first one. That's the lifeblood of this company, and it's what will allow us to be successful over the next 10 years. We've actually increased that in terms of percentage of revenues by about 200 basis points. Customer acquisition and customer retention, given how early stage we still are in the e-commerce opportunity, is another area where we're not looking to manage the business to scale. We'll get into more detail on all of this in a while. That hasn't. It's remained relatively flat since the last Analyst Day, at about 9% of revenues.
Where we have aggressively pursued margin improvements that we can reinvest back into the right areas of the business has been G&A overhead, legal and compliance, bad debt, and fraud loss provisions, where we've driven about 300 basis points of margin improvement. Compensation and headcount-related costs, where we've driven about 200 basis points of margin improvements. Generate margin where we can to free up margin for the areas of the business where we need to continue to invest in aggressively. Check. Earnings. We do like earnings. We have delivered consistent earning growth, a four-year CAGR of nearly 30%. Slight decline in our net income margin, but still extremely healthy net income margin at roughly 25%. We remain committed to running our businesses in a cash flow positive manner and generating free cash flow.
If you look at our business with some adjustments to take into consideration certain capital controls in Venezuela, primarily where we've invested in real estate as a measure of preserving the value of our assets, the financial model actually converts net income to cash quite consistently. That should continue to be an area of focus for us going forward. Finally, to maintain an efficient capital structure and a robust balance sheet. Probably the most relevant thing going forward is, as you're all aware, we have recently, on a balance sheet that had never carried any leverage, placed some leverage on it. We issued a $330 million convertible bond in June of this year, and we believe that that gives us even greater flexibility to be able to carry out the financial plan and the business plan that we have over the next five years.
We think that even more than what our balance sheet looked like three years ago, we really have the right capital structure to be able to capture the opportunity we have in front of us. Now getting into sort of where we are today, the conversations we're having, and how we're thinking through from a financial model perspective, how we can accomplish the vision that my colleagues have outlined. Clearly, there is a commitment to invest efficiently and to be able to capture all this opportunity that we've outlined. One way in which we think this is, first of all, we need to make sure that we continue to capture the growing user base that's out there as more consumers move online, as more online consumers start consuming online.
It's critical for us to remain focused on the long-term goal Marcos set out of the 100 million active users. We've worked really hard to subliminally plant this concept of an enhanced marketplace, we're going to coin the term officially, which is transitioning from the old marketplace model to a marketplace that has a significant number of service overlays and of value-added services that make buying on a third-party marketplace as efficient and as compelling to consumers, both buyers and sellers, as it is from a first-party marketplace, yet at the same time, maintaining all the financial benefits of the third-party business.
As we accomplish that and we complete this transition, and buying and selling becomes more compelling and easier on MercadoLibre, that will allow us to increase user engagement to drive more share of wallet from our users, which in turn should drive more scale. Allow us to accomplish the goal of continuing to sustain a profitable business model and generate more capital for us to allocate in continuous investment and innovation, which is really what allows us to then improve the technology, improve the experience, which gets us back to capturing more users and starting once again this virtuous cycle. The balancing act, and this is very similar to what we said three years ago, is how do we find the right balance between the investment in growth that all this will require and continuing to deliver the solid financials that we are committed to?
Here we go. First of all, just in terms of context and the region, we've consistently said that we do not believe that macro is a determinant factor of our business. That how we execute our technology and what we do are really the most critical catalyst for our growth going forward. Having said that, macro backdrop is relevant for currency, which it does affect our reported numbers, and as we grow and we become a larger portion of overall commerce, it does indirectly affect us. In those terms, I think we're quite optimistic when we take a midterm look for the region. Most forecasts for the next three years show acceleration in GDP throughout most of the economies in Latin America. That improved GDP should lead to greater currency stability than what we've witnessed recently, less headwinds from a very, very strong dollar.
Also remember that our business does allow us significant natural hedging to currency oscillations in that we have both revenues and costs fairly well distributed across the different geographies. Because of the asset-light nature of our business, we are able to move costs around efficiently to take some advantages of depreciating currencies. If you look at Argentina, for example, where we've been in a cycle of a depreciating Argentine peso, we've been able to position costs there efficiently. When you look at OpEx margin gains that we've been able to drive from loading costs in Argentina over the last two half years, they've been significant. For the first half of 2014, we've been able to drive over 200 basis points of OpEx margin improvements by the way that we've allocated and move operations and costs around into countries that are seeing stronger depreciation.
From a more secular perspective, when we look at the industry-specific context, we also see, and we've walked through this, very favorable trends. Mobility will be an engine for sustained user growth. Large portions of the population in emerging markets will primarily access broadband through their mobile devices. Latin America, as you all know, has more mobile devices than people. As the conversion from feature phones to smartphones continues, all those users will become potential repeat users of MercadoLibre. That, in turn, should continue to drive the trend of commerce moving offline to online as retail goes from being still less than 5% of overall commerce in the region to more than that.
We're also seeing consolidation within the sector, where the largest players that are able to reinvest more aggressively in the user experience and in customer acquisition are concentrating a larger amount of the e-commerce pie. When we look at the secular trends, we think that all three of these bode very well for us. Let's look at each one of these, how we think through from a financial perspective, how we accomplish maximizing capturing the growing user base. First of all, the pursuit of mobile growth will be critical for us. When you look at the percentage of overall investment in mobile marketing, it's obviously started from a very small base and is accelerating and should continue to accelerate. That, in turn, is driving what has been a very successful shift away from desktop commerce to mobile commerce.
The opportunity to accelerate that mobile growth is probably even larger because much like everybody else in the industry, when we look at mobile conversions, they still are smaller than desktop conversions. Primarily a consequence of a desktop experience that's been optimized over 15 years and a mobile conversion that has still only been optimized over two, three years. As we continue to narrow that spread between conversions on mobile and desktop and accelerate our pace of investment in mobile marketing, that should drive even increasing accretive growth from our mobile initiatives. That's the case not only in mobility, but across the board. We are committed to investing more aggressively in customer acquisition as we feel increasingly more comfortable with the quality of the service and the platform that we're delivering.
If you look at our actual spending on customer acquisition, it has been increasing consistently over the last three years, even at lower ROIs. In other words, we're no longer maximizing for the ROI of the incremental usage, but more and more we're maximizing absolute numbers of new user acquisition. When you look at the three-year CAGR for new registered users, it's actually been quite solid at 20%. Which is larger than the rate at which the addressable user base is growing. Finally, you've seen throughout a lot of the previous presentations an increased commitment to the markets that go beyond Argentina, Brazil, which are the two largest e-commerce markets in the region, an increased focus in capturing the opportunity that's available in Mexico, in Colombia, in Chile, in Peru, in Uruguay. These markets, we believe we under-index them.
When we look at the number of internet users that exist in those markets and the number of registered users we have, our reach is lesser in those markets than in the bigger markets. We are increasing aggressively our levels of investment in those markets over the next few years, and we're quite confident that that should increase the rate at which they return revenues and business to us. This will be an area of growth. The transition to an enhanced marketplace also is a key area of investment focus for us. In terms of payments, we're seeing very solid growth. Osvaldo walked you through that. The important message here is that there are important financial offsets to the gross margin compression that payments generates.
The first one is that when we look at the impact on OpEx through improvements in bad debt, improvements in collections, improvements that we've been able to carry out in the fraud loss rates of our payments business as we get better and better with our fraud modeling capabilities. The payments business does drive OpEx improvements that partially offset some of the gross margin compression. Second is that the payments business is also a critical driver of improved take rate, in part because it generates additional revenues through financing and processing, but also because as the payments business makes selling more efficient and easier for our merchants because they no longer have to worry about settlement, it allows us to carry out pricing more aggressively. The more payments is penetrated, the more leeway we have on pricing.
Shipping is another area that we're getting more aggressive with, that we're investing more in, we've done so in a very financial efficient manner so far. When you look at the overall penetration of units where Mercado Envios is present, it's gone from nearly nothing a year ago to almost 20% on a consolidated level, and as we showed you earlier, in some markets like Brazil, rapidly trending towards 30%. Now, the way that we manage that from a financial perspective is essentially we book the revenues on the shipping business net of cost. Because right now our focus is on adoption and penetration, we're not trying to drive profit from that shipping business, but essentially to charge our users what we are being charged by the carriers and logistics operators that run those businesses for us.
When we look at the middle slide, the gross revenues from shipping, what we're charging our users, and the gross costs, what we're getting charged, and the net of this is what you see flow through the P&L. That number has been managed to get to as close as zero as possible. We've made tremendous strides in terms of offering more and more shipping and fulfillment solutions with barely no P&L impact. That's the vision going forward, at least as we continue to run the current model. Shipping, much like Pago, has the same positive impact on our take rate.
Not directly, because it's not driving incremental revenues, but the existence of a much better service offering, much lower costs for our merchants as they're able to both pay less through shipping and have an integrated shipping solution which makes selling on our platform much easier. That also gives us greater flexibility on the pricing front as we generate significantly more value for our merchants to be able to extract some of that value, were we to choose to do so. Limited P&L impact directly, very positive in terms of the flexibility it gives us to capture some of the value that's being generated if we decide to do so. Finally, we've talked a lot about category expansion.
Moving both away from consumer electronics in terms of getting stronger and having greater share of market in a growing number of product categories, and also changing the merchant base or adding to our existing merchant base of consumers to consumers and small and mid-size businesses to enterprise merchants, brands, and retailers. That also strengthens our financial model. What we see is that as we gradually have less mix from consumer electronics and the number of large retailers and official stores and large retailers grows at a consistently faster pace than the overall number of sellers, we have take rates that are both higher than consumer electronics in many of these new categories and take rates that are higher, the one data point there, on official stores and larger merchants.
These newer categories and merchants actually strengthen the financial model as they penetrate the marketplace more deeply and broadly. An additional benefit of the transition to the enhanced marketplace is that we already begin to see, as all the pieces of the puzzle come together, improvements in our user engagement metrics. If we look at units purchased per buyer, units sold per seller, if we look at total payments carried out by merchants or receivers, if we look at the % of overall units that we are able to cross-sell financing, all of those consistently continue to trend upwards as we gain more share of wallet of buyers, of merchants, of payments receivers, and of people financing. Those operational improvements in user engagement metrics translate quite nicely into improved monetization ratios of our user base.
In terms of revenue per listing, revenue per receiver, or perhaps the ultimate goal, which is the average revenue per user, as we have more services that we're able to offer and monetize. As the experience of buying and selling gets better, and therefore users' frequencies goes up, the ARPUs improve over time. As the enhanced marketplace continues to grow and penetrate, we hope to see this consistent trend occurring going forward. As we then begin to think through, how do I continue to generate the scale-driven OpEx additional margin to reinvest in capturing more users and reinvesting in continuing to roll out the enhanced marketplace, there are a few key areas of focus for us. Headcount or headcount-related costs are the largest line item we have. We've really focused very much on tight headcount control.
If you look at our CAGR for revenue growth since the last Investor Day, it's been roughly 26%. When we look at the headcount CAGR, it's been less than half of that. It's been 12%. Really trying to drive all these improvements through technology and through efficiency and not throwing headcount at the problem. Any of the headcount ratios that you were to measure, whether that be revenue per headcount, successful item per headcount, TPV, GMV, have shown extremely good evolution over the last few years, and that's one of the guiding principles for us going forward. However, just one caveat here. When we look at salaries and wages and compensation, although we do see improvements and we see important scale there, it's not as marked as on the headcount front because it's also critical for us to continue to attract and retain talent.
In other words, what we're trying to do is to hire the best and the brightest and to compensate accordingly, but be very careful to not get too carried away on headcount per se. Salaries and wages should be one of the key areas of control in OpEx going forward. When we look at G&A beyond just salaries and wages, very much focused on continuing to scale that. If you look at G&A as a percentage of revenue, it's been consistently coming down. That's not only driven by what I just showed you above, salaries and wages, but it's driven by overall overhead, which has also been consistently trending down. When we look at G&A ratio to sales and marketing and product development, which are the two areas of our P&L where we think we need to continue investing, it's been consistently coming down.
This is the shift in terms of investing in the right areas, away from G&A and overhead and towards sales and marketing and product development. From a CapEx perspective, when we look at what we're calling operational CapEx, the CapEx that's more of a normalized run rate, this continues to be a fairly asset-light and CapEx-efficient business model as a technology company. Our CapEx ratios adjusted for operational CapEx continue to be less than 5% of revenues. That's freed up a lot of capital for what I'm calling here strategic investments. Whether that be M&A activity, to acquire either businesses or technology, or also property, plant, and equipment, which allows us to efficiently manage certain capital controls that exist in some countries and in other countries to invest in buying our own offices, which eventually will have a payout going forward.
When you look at the reported CapEx, I think it's important to note that there are certain distortions that are driven by capital controls, and there are certain distortions that are more investment cycles where either we're getting a little bit more acquisitive or we've invested in our own property, plant and equipment, primarily offices, which will then yield better returns going forward. As we're able to drive these efficiencies, the key area where the investment will continue to occur will primarily be in technology and product development. When we look at product development as percentage of revenues since the last Investor Day, that number has been trending upwards. There have been about nearly 200 basis points of increased percentage of revenue in P&D. A significant amount of that, as Daniel showed you before, has been in engineering headcount and engineering talent.
When we look at headcount number, it's grown from about 300 to nearly 500 engineers, hopefully will continue to grow at a very solid pace. Here's another way to look at the improvements in throughput, which is how much product and technology develop we're able to CapEx to capitalize, which is a financial indication of how much more deployment and how much more technology development we've been able to carry out over the last few years as a consequence of investing more and more in the platform and in product development. This is the part of our IT that gets seen in D&A and not directly in product development.
A lot of this investment in technology and the platform has been not just in number of engineers and engineering itself, but also in other areas such as site ops, which appear in COGS, connectivity costs, hardware costs, data center redundancies. If we look at product development CapEx, that drives most of what I had called the operational CapEx before, it has been trending upwards, although it's still efficient. It's nearly 5% of CapEx when we look at H1 2014, up from roughly 3% last year. The final metric indicative of the way we're thinking through how we need to accomplish this is the number of engineers that we've onboarded as a consequence of M&A activity, so of acquisitions.
As you've seen, it's picked up significantly over the last few years, that's one way we think that we can accelerate our pace of innovation, is to continue to be very selective, yet more acquisitive in terms of either what I guess is now being coined as acquihires, or in many cases, buying companies not just for their engineering talent, but for actual intellectual property and technology that they have that's interesting for us. In summary, what are some of the guiding principles that we think through when managing our financial model? Then we can get into greater detail in Q&A. Prioritize top-line growth and capturing the opportunity. Right? Again, the financial model is a tool to be able to capture as much market share going forward as possible.
Continuing to invest, but in what we consider the strategic areas of the business: product development and technology, user acquisition, and user retention. To be very focused on driving as much scale as possible on the less strategic expenditures, mainly G&A, where headcount and compensation are the largest line item, so very focused there, and also on the other general corporate overhead areas. Freeing up capital to deploy and also capital that we've raised to invest in regions we think we underindex and that have a lot of opportunity, new businesses that we've outlined many of these for you. Then there's a clear opportunity to selectively become more acquisitive as the ecosystem of technology companies in the region has become more interesting, and there are a growing number of assets out there that could be potential targets for us.
Wrapping up and finally being able to get to Q&A, we really do encourage you to ask us as many questions as you want. I think that's probably the more interesting part of today's proceedings. I'm going to take us full circle and wrap up where Marcos started in many aspects, and it's really the way we feel about the business. When we rang the bell yesterday here at the Nasdaq, it was sort of one of the few moments that we've allowed to pat ourselves on the back and say, "Wow, we've accomplished a lot over the last 15 years. We've accomplished a lot since we took the company public 7 years ago." We've built the preeminent e-commerce platform in Latin America. We've created a brand that really is synonymous in the minds of Latin American consumers with buying online.
We've been very successful in this transition from a Marketplace 1.0 or Marketplace 2.0 or Marketplace 3.0, that was the other alternative of how to call this, to the enhanced marketplace model with integrated payments, integrated shipping, integrated credits, fulfillment solutions. We believe that that places us in a much better position for the next 15 years than the one we were when we started the business out of a garage in Buenos Aires. We have a better financial position. We have access to capitalization and to capital if necessary. We have a technology platform that is truly cutting-edge today and are committed to continuing to invest to maintain that cutting edge.
We have a robust ecosystem of services that are no longer just ideas on a PowerPoint that we can share with you, but actual working operations in shipping, increasingly so in logistics, in payments, in financing, which allow for a business that's more diversified and that generate much more significant buyer and seller lock-in and stronger network effects. When we piece everything that we've shared with you today, we truly feel a lot more optimistic about the long-term prospects for MercadoLibre and how we go from $5 billion of market cap to $10 billion to $15 billion to whatever the number is over the next many years. We really think we have a great shot to accomplish that. Thank you so much for having sat through the presentations.
We're now going to move ourselves onto center stage, we'd like to take as many questions as you have. We've left plenty of time to be able to do that. We also encourage you to stay for lunch and to be able to talk to us in smaller breakout tables. I was just going to say, if someone would rather just scream out their question and not wait for the mic, that's great too.
Got it. Here. Great presentation. Thanks for doing it. Scott Devitt with Stifel. My question ultimately relates to logistics. Different markets globally have succeeded with e-commerce, with one P models, hybrid models, or what I guess suppose you're now calling enhanced marketplace, then pure three P models. If you look at Brazil, which is your most developed market, some of the owned inventory players have begun to become more aggressive with third-party services. Marcos, I think earlier in the presentation, you had mentioned not yet touching inventory, but alluded to the possibility of touching it. Pedro, in the financial presentation, talked about the asset light model and retaining that approach.
I'm just trying to connect the two in terms of ultimately how far you think you can go beyond where you are today in retaining asset light, and what are some of the incremental things that you can do. Ultimately, and specifically around Brazil, do you think that you ultimately need to become capital-heavy and be more involved in logistics aspects? Thanks.
Okay. We will experiment with everything. I think the end game will also be a combination of many things. Very efficient drop ship for certain sellers, consumer sellers or hobby sellers, very likely fulfillment for professional sellers, brands, large retailers. We will experiment with everything and whether we own the fulfillment centers, we partner with people who own them, or we rent them and run them ourselves, remains to be seen. We will, in the next 10 years, experiment with everything and whatever we see works better in those strategies. I wouldn't be surprised that the end game is going to be a combination of all those things, plus lockers, maybe drones. It's going to be a combination of many things.
Just to complement that very quickly, I think there's two questions. There's, what's the service model? As a three P, do you need to offer warehousing solutions for your merchants? Do you need to have a sorting center where the packages pass through your sorting center and you can repackage them and then more efficiently route them? That's a service question, and I think there, as Marcos is saying, we'll get involved in all of those. We'll try sorting centers. We'll try lockers. We'll try just direct merchant to buyer delivery. The second part of the question is a capital allocation and an operational efficiency question, which is, in order to offer that service model, do you need to own the asset on your own balance sheet, or can you partner with logistics companies that are specialized in that?
I think we've always said we're agnostic long term. We need to do what allows for the best fulfillment and most efficient operation. Right now, given the starting point, most of what we're doing on all these fronts is through 3P. There's none of this on our balance sheet. Where we end up, I think we'll get it right, and I don't know if that's on the balance sheet or off the balance sheet. It's not on the balance sheet right now.
Scott.
Jim.
My question is on take rate. In the past, it's been you don't really manage take rate. There's different forces between the different businesses that impact the take rate. Pedro, I sense that there is a little bit more room that you may be impacting that. You talked about shipping and payments and adding value to sellers and maybe finding ways to extract that value. Can you talk a little bit about how we should think about take rate and how you will or will not be influencing it in the future?
Sure. I think we continue to say that we don't manage the business to take rate. That's exactly how it is. Pricing is not something that we believe is the way to drive financial results. What we're trying to show is that as the different pieces of the ecosystem come together and they begin to grow their penetration, there are natural take rate advantages of that. If you have larger non-marketplace businesses, they're accretive to take rate. Also, it was one of the slides I showed, as more and more of the marketplace transactions have one of these services overlaid, and because we have this bundled pricing model, there are benefits that are being generated, and we're willing to extract some of that value. Take rate continues to not be a metric that we have targets for or that we manage aggressively for.
We actually think that this is very early stages, and it's not the time to maximize that. The natural consequence of the very positive traction that we're having with shipping, with payments, with credit, with these different services, should, over time, be accretive to take rates.
While positive impact on take rates, it's not something that you're going to be proactively raising the prices for.
Yeah. We don't envision ourselves raising final value fees. We haven't done that, I think, in a while. We have brought up caps in certain countries. As our services become better, as we monetize the transactions that are occurring as we monetize GMV into revenue more efficiently through Mercado Pago, through better completion rates, we naturally see improved take rates.
Thank you.
Hi, Mark Miller from William Blair. Can you expand on the increased spending for customer acquisition? I'm hearing you say that scale and network effects are ever more important. As you move more aggressively in this direction, how should we think about that playing through the financial model? Specifically, as you step that up, what's the immediacy of the return? Should we think about this potentially being somewhat dilutive to short-term results but creating a longer-term payback? What's been your experience so far? Then, just anything more you can share beyond just stepping up the investments in mobile, what channels you're going after more aggressively?
Yeah. What we're trying to convey here without getting into specific guidance, which we don't offer, is that when we survey the e-commerce landscape, what we see is tremendous opportunity, a large user base to be either acquired or to drive more volume from. We see a financial model that today delivers very healthy margins. What we don't want to get caught in is the trap of let's worry about our financial model and our margins over the next two or three quarters and miss the enormous long-term opportunity that exists. What we were trying to convey here is that there are a lot of users out there to be acquired. We have all these great new services that we want to make consumers aware of, and the right decision is probably to make sure that we're investing in customer acquisition, in marketing, and communicating.
If that's at a faster clip than what we had been doing in the past, we are very comfortable doing that. What percentage of margin, what basis points it's going to be? Again, we don't offer guidance, what we're trying to convey here is we are definitely not managing the business to stay. If I've been investing X percent in sales and marketing, customer acquisition, I need to be very cautious to not invest more than that, because then what's my quarterly number going to look like? Not at all. The platform is more robust today than it's ever been. The user experience is better than it's ever been. This is the right time to bring more and more users onto MercadoLibre so that they can experience what we're offering.
That's great. Just to follow up to that, I think we would agree that's the right strategy of adding more value as the company is and the larger you become, the more benefits to the whole ecosystem. Just as we see that play out, does this take a couple of quarters when you activate a customer? Should we think about it across years? Just that payback period would be helpful to understand. Thanks.
The metric that I showed is more of a short-term ROI metric. That ROI tendency I was showing you essentially measures the gross merchandise volume generated from that consumer within the quarter, that's the one that we've been willing to drive down. When you look at our customer lifetime volume models and other things, you might get different curves. I do think that, again, as everything we're managing for the long term, and when we're thinking, bring in a new consumer, introduce him to MELI, the true question is, what's the long-term value of this consumer? We're not managing it, again, to get a short-term return. Therefore, we're willing, by definition, to give up margin in more customer acquisition. There's probably 30 people back in Buenos Aires in the marketing team hearing me and saying, "Here we go. Budget 2015.
I don't think they're going to get what they want.
I had a question in terms of trying to get an update on competition. There's different types. There's offline, online, cross-border, would hope to get a sense for how that may be evolving over the past year or two in some of your largest markets, how that directly impacts your business in terms of things like new customer acquisition costs or market share. Is it more that it's growing the market than impacting you guys in any direct way?
Okay. It's a little bit open question in terms of how competition has been evolving. Let me start by saying what hasn't changed is that we have a lot of competition from many players that has been a constant during the last 15 years. I think what has also been a constant is that everyone except us has been losing quite a lot of money. I would say we continue to look at all of our competitors, take from them what we like, learn from them as much as we can. We continue to focus on ourselves, on our strengths, which are, as we mentioned before, a lot of traffic, unmatched breadth and depth of inventory and prices, get better in the things that we want to get better, which are the ones we talked about today, shipping, payments, mobile, verticalizing experience, bringing large retailers.
I think we continue to do very well. We think that going forward that we will continue to do very well. The question is very open-ended, there's not much more that I can say about that. I don't know if anybody has anything.
I think what you said also in your presentation is, of course, the opportunity is enormous, and it's natural that competition will continue.
Yeah. We expect to have a lot of competition in the next 15 years as well. We feel comfortable with competition. It's one of the things that make us go to the office every day, 15 years later. We like competing.
Hi, guys. Over the years, we've talked a lot about which metric portrays the health of the business, most importantly. We've talked about local currency GMV growth, items sold, local currency Brazil growth. What metric do you think gives the most accurate view of how the business is doing now?
In terms of our marketplace business, I focus a lot on successful items, and then local currency revenues as well. GMV is useful. Revenues, perhaps as we start to overlay all these services, is going to give you a better indication of the marketplace business. In terms of payments, obviously, total payment volume and the number of transactions processed is very useful as well.
How do you guys feel about the gap between items sold and revenue growth in Brazil? We talked a lot about the other initiatives that are happening in a lot of the non-marketplace businesses. Should we see those growth rates converge over time or continue to diverge?
I think we've always been pretty candid about this. When we look at unit growth in Brazil in the low 20s in the last quarter, getting very specific, it's not a metric that we necessarily are excited about. We think it's a metric that can improve. It's probably somewhere around where the market is growing. If you look at longer periods of time, Brazil has consistently delivered higher growth than that, no reason to believe that if you execute well, we execute well, we can't return to that. I think what you're seeing now is you're seeing tremendous strength in our Brazilian operations in payments. The shipping business has been growing extremely well, which has some improvements in monetization. You capture that in revenue, you see the marketplace still growing +20%. Longer term, we think that can certainly accelerate.
Again, we're not looking at this quarter by quarter periods. No, 21 is not something that we're saying
What we would like to do, that's the one thing that we have always said since we were last time here seven years ago, is grow at or above e-commerce. We have consistently been doing that. We think we continue to do that, we hope, we expect to continue to do that for the next 5-10 years. If we can do that, I think we will be in a very good position. Occasionally, some competitor that is investing very heavily can grow faster than us for a quarter or a year. I think if we can continue to grow at or above e-commerce for the next 5-10 years, we will be in a great position because we are starting from a very good position as well.
Thanks.
Hi, good morning. It's Jeffrey Walkenhorst from Copeland Capital. I have a few questions. The first is on capital allocation. I know that you've paid a growing dividend for the last several years, although it's not something that I saw talked about today or mentioned. What is your specific philosophy around the dividend?
I personally, as a shareholder, enjoy collecting dividends. I recognize that we have a huge opportunity. Honestly, I believe the opportunity today is bigger than what it was 15 years ago, and it's much easier for us to capture it. We obviously understand that fulfillment is an area where we will have to focus and invest heavily. Right now we're paying dividends. I don't expect that to change. I think the policy has been 25% of prior year's net income. We're not committing ourselves to anything other than I think what I just said.
I think we like consistency. We've always said we're very committed to the dividend policy we have. Our earnings should continue to grow. That overall guiding principle of a quarter of the previous year's net income seems right for this stage in our development. We can reassess that a few years down the line and let's say let's go up. Again, I think there's so much opportunity for capital deployment within the business that that's probably a better return to shareholders than augmenting what percentage of earnings we're paying through our dividend policy.
Okay. Just to summarize, I guess, if earnings grow, you might expect the dividend to grow, but you're clearly committing to grow in other areas as you need to invest in the business. Okay.
Correct.
Thank you. I'm not as familiar with the model as some may be here today, can you talk about the relationship? We know eBay is a large investor, and you've had a long-standing relationship with them, they're also a competitor in many of your markets. Talk about the dynamic between the relationship and how it's evolved.
The relationship is still very good. Although we compete with them in payments, they have a free classified site in the region. We have a good relationship, they are a substantial shareholder, we compete in the marketplace with them as we compete with other players as well. We get along well, we interact occasionally and exchange views on where the industry is going and what things need to be done.
Okay, thanks. One last one. The countries where you're under-indexing are your smaller regions or countries. How, in terms of the investments you're likely to launch this year or next year to hopefully accelerate the growth, what might those entail in terms of OpEx and margin impact? Thank you.
Yep. I can probably give you some more clarity on what they look like, not guidance on margin impact, right? I think if you listen to both Stelleo, Osvaldo, and Marcos, we laid out a product roadmap where clearly there is a significant number of services that we offer that have made the experience in Brazil or Argentina that has that more complete product roadmap very successful. To roll those out to Mexico, to Chile, to Colombia, first and foremost, that entails engineers and engineers focusing on the rollout of that technology. That's one piece. It's product development. The second piece is, as we get more comfortable with the experience we're offering, we get greater, I don't want to say urgency, we think it makes more sense to invest in customer acquisition.
That means more marketing budget for those peripheral countries that have had very small marketing investments historically. You should see some of that incremental investment come up in sales and marketing in the customer acquisition lines. I would say short term, it's probably those two things. It's product development to deploy more quickly in those markets, and then marketing and communication investment to bring consumers on board to those new improved platforms.
Thank you. My question is a follow-up of the last question that you just answered on Mexico and Colombia. What will it take to grow faster or to make these two countries more relevant for MercadoLibre? It's a higher focus, as you just mentioned, or is there anything in these two markets that prevents money to be a larger part of your earnings?
Yeah, I think I addressed some of that in my presentation. Brazil is, of course, the country where we have our complete model of solutions deployed. As we saw, it's the country where we're seeing higher growth even compared to some of the markets you mentioned. We believe with that strategy deployed in Mexico and Colombia and some of the smaller countries we've been mentioning, we can accelerate growth in them as well. You should expect to see the same kind of developments we've seen in Brazil in those.
Two questions from me. The first is on mobile GMV. You'd mentioned the 2Q that was 16%. Just wondering if you'd give some granularity on how that looks in Brazil and Argentina and then some of the smaller markets. The second question is on competition, specifically in Brazil. It seems like you've got a few competitors, B2W is one of them, that are getting more aggressive there. Just, let's say, looking two or three years down the road, would you expect to have higher or lower, about the same market share as what you have today?
In terms of opening up mobile GMV per country, I don't think that we're doing that. Actually, I don't think anybody opens up their mobile GMV at all. It's something that we did until Q2, but it's great. I would love to know what our competitor's mobile GMV is actually on a per country basis. In terms of market share, we aspire to have more. We aspire to have more market share than what we have today in Brazil and elsewhere.
You've heard me say this before, just one thing, right? It's very important to understand that the market share gains that really make this a 10, 15, $20 billion company, you shouldn't be thinking those as what's the share that they're stealing in e-commerce from B2W or from player B. It's what are the share gains that as an online industry, we are stealing from offline retail. 95% of consumption in the region, more, 97%, is occurring offline. If we focus on what our online competitors are doing, we're really missing the big picture. This is not a zero-sum game. This is very much a positive-sum game going forward. It's not a coincidence, you've also heard me say this 100 times, that the market where we've consistently grown the most is Brazil, which is the more competitive market.
It's because as an industry, both ourselves and other online players are really putting out there compelling online experiences for consumers to stop shopping offline and start shopping online. It's also not a coincidence that the less competitive e-commerce markets in the region have had more anemic growth rates because consumers there are faced with a much more limited amount of options to move from offline to online. This is not spin. We think that better online offerings benefit all e-commerce players to the detriment of bricks and mortars.
Michelle Moran from Morgan Stanley. On M&A, I was wondering what you can share with us in terms of what you might be looking for, just given that you do have a lot more resources to work with now post the convert.
Where's the pipeline? Let me see. I think what we said through the convert, and we kind of shared the thought process, right? We said there are the natural adjacencies which are similar to the M&A we've done in the past. If you look at some of our M&A history, there are marketplaces out there typically that have taken a vertical approach in a specific category, and that now have reached a certain scale and that we've known them for a while, we've monitored them, we've had conversations throughout the years, and now could or could not be the right time, depending on multiple factors, to carry out a transaction. The latest one we did in May, which was Portal Inmobiliario, is a good example. It was a real estate-focused property with absolute leadership in Chile and a very strong position in Mexico.
Think a version of a Trulia or a Move for those markets. There are other players similar to those. There you're tucking in a vertical marketplace. In the payment space, there could be opportunities either to acquire share, so companies that have a good overlap or don't have too much overlap in terms of merchant base as the payments business becomes more scale-dependent and the regulatory hurdles become more difficult. Players like ourselves are 100% capable of complying. Perhaps smaller players aren't. There could be opportunities there to buy market share or interesting IP, whether it be in the payment space, some sort of interesting technology for retail, recommendation algorithms, shipping efficiencies. Logistics and fulfillment is another area where we said that as we get more involved, there are companies that have that DNA from their beginning and that could be interesting for us.
Not necessarily warehouses or vans, but WMS technology, TMS technology. All the technology that powers logistics operations, which are quite complex. I think then we also said that we begin to see through the open platform, through our Mercado Shops initiative, that increasingly there's a natural evolution in our business towards getting more and more involved with the ERPs of our customers. Customers on the merchant side want to less and less have to access our marketplace user tools and sell through there, and more and more want to be able to use their own tools that simply integrate seamlessly. There could also be opportunities in the whole ERP space that allow us to better and more deeply integrate with the kind of merchants that we're beginning to go after in the whole enterprise segment. I think those are the areas that typically we're looking at.
I think a common thread in all those areas is acquiring talent, which you mentioned in your presentation, of course.
Could you talk a little bit more about your customer acquisition strategy for brands? You've made some really nice progress over the last, whatever, 12 months or so, and even less, acquiring quite a few very high-profile brands. What have you found has worked well to accelerate that, and what has been less successful?
I think I can speak to that. First, I mentioned in my presentation that we have changed our org structure and actually hiring sales teams that are dedicated to brands and larger retailers. That has been important in bringing them on board. Of course, many of them had not understood the opportunity in MercadoLibre because of our closed platform in the past. That has also been very important for us as we have partners, e-builders that provide integration for these larger retailers. It's become possible, whereas in the past it wasn't, for them to, for example, load thousands of SKUs onto our platform. I think I also mentioned that their expectation in terms of the experience and how they can manage that experience was something that wasn't possible before considering we had sort of this horizontal, one-size-fits-all approach.
As we have made our platform more prepared to receive those sellers with more customization, it's also become easier to attract them.
Okay, Stelleo. I think Mexico, ever since I can remember, has been under-punching for you guys for a while now. What's changed? What's the strategy going forward? I think, Marcos, I know you guys pay a lot of attention to what other companies are doing globally. You've seen another marketplace's platform internationally make the strategic decision to take themselves off the dominant search engine and indexing by that search engine because inevitably you're going to be competing with them. As you think about your long-term future, do you have the luxury to be taken off Google over the longer term?
I believe you're referring to, for example, Alibaba, that is not indexed in the search engines in China and whether or not we're going to take that route. I think right now we are comfortable where we are with being client partners and competitors of Google. I believe in the next 10 years, that question is going to become less and less relevant where we're going into a mobile world, into an apps world where search for e-commerce is
It's been less relevant, frankly, and it's going to, I think, become significantly less relevant in the next 10 years. It's not one of our highest priorities to focus. Obviously, we want to actually improve our search engine optimization as much as we can for desktop. I believe in the next 10 years, it's not going to be that relevant of an issue. With respect to Mexico, we're also frustrated about its under-indexing. What has changed is Mercado Pago's conversion ratios are really improving in Mexico. Mercado Pago is a great tool to help the marketplace grow faster. It brings down overall fraud ratios in the marketplace. Mexico continues to have higher fraud rates than our other countries.
As we are getting Mercado Pago approval rates higher and chargeback rates lower, we're starting to feel much more comfortable that we can be much more aggressive pushing Mercado Pago inside our marketplace, and hopefully making it grow faster and in a healthier fashion. I don't know, Stelleo, you want to add something?
I think I would add that financing and the installments without interest we saw in Brazil. Osvaldo mentioned we will launch them in Mexico, and we expect to see financing grow in Mexico as well. Shipping is something that we expect to launch in Mexico soon. We hope to see the same kind of growth rates we've seen in Brazil and in Argentina. Those are the dimensions that we can control within MercadoLibre, of course. Pedro just mentioned before that what we can control, which is sort of this move from offline to online. In Mexico, we have seen that happening slower than we've seen in other markets. We expect that to change over time.
We're on to round 2 here. Can you address the long-run importance of cross-border trade within your markets, both in terms of opportunity for MercadoLibre, and I know you have a number of initiatives underway. You didn't really discuss it here, but how important can that be over a longer term? Also from a competitive standpoint, of course, Alibaba's talking about very fast growth with AliExpress, and especially in Brazil. Can you address the extent to which that's a changing competitive dynamic for you?
I think the opportunity is really big longer term. I do not see that as a competitive offering today in the sense that if you want to buy something, it's a very different offering. If you buy cross-border, you typically have to wait 60 days to get the product. Oftentimes, the product gets delayed in customs, and you have to go and personally pick it up in customs. It has lots of complexities that need to be solved. Clearly, some international players have solved it and are seeing a lot of traction there and a lot of growth. We see it as an opportunity. We are focused today in most of our businesses inside the country, not even cross-border regionally. We are the leading marketplace for intra-country trading. eBay has historically had, I think, a quarter of their business coming from cross-border trade. For us, it's negligible.
It's a huge opportunity, and it's obviously a market that is growing, and growing fast.
I think also just compliment, there's probably different models that will emerge as to how to facilitate brands' entry into Latin America. Just from informal conversations, for example, with some of our own mid-sized power sellers, what they'll say, and it's pretty obvious if you think about it, is, "Wow, so that means that I don't necessarily have to ship item by item from Argentina to Mexico or from Mexico to Brazil. I theoretically could, in the future, use your entire ecosystem and maybe send a pallet, have you fulfill that pallet, and then use the marketplace to sell straight from inventory that I already have in the country. That would be fantastic for me because I wouldn't have to set up any physical assets in these new markets I'm getting into." That's why I think when Stelleo showed the slide, the opportunity is there.
I think there will be different models to accomplish that. One is the direct-from-China individual parcel model that AliExpress is pursuing. Osvaldo talked a little about also trying to monetize cross-border globally through our payment solutions and to offer payments and processing solutions for some of these global cross-border trade companies or marketplaces. There's the whole notion of helping brands that want to disembark in one of our markets to do so with some of our local assets and capabilities that we're building out.
Just wondering, philosophically, how you think about the payments business, in terms of being an integral part of the marketplace company.
Yes.
Both short-term and long-term, for yourselves and others in the industry.
Sorry, just one thing. I always do over-unders with them. The over-under on that question was 10.
Yeah.
I think it was over.
10 what?
10 questions before someone brought it up.
It's a great question. It's interesting because a few years ago, we were hearing a lot of people complaining about a major player taking away their payments unit out of the marketplace without full consent, apparently from all of their shareholders, and that was a huge issue and a huge complaint from shareholders. Now we are seeing the opposite, right? We're seeing shareholders complaining about, "Hey, we need to take this out." Obviously, we are very happy the way we are managing our business. I think you've seen today that the synergies are enormous. We couldn't imagine ourselves separating payments from the marketplace at this point in time. We just couldn't. We don't know how it would operate, one company separate from the other. We do fraud prevention. We don't charge for payments on our marketplace.
We have a single fee which covers payments and financing. The final value fee covers everything, right? We're not discussing between Osvaldo and Stelleo, how do we allocate that revenue. Basically, Pedro makes a Solomonic decision. We're very happy the way we're operating. I think as long as you are the leading platform, there are many, many ways to create value having these two businesses together.
I think there's something to be said also about life cycle. I am absolutely convinced, and I think even everyone involved in the whole eBay, PayPal debate would agree that given the stage of development where we are, whereas Osvaldo said the payments business is still a fraction in terms of TPV to GMV, that it makes all the sense in the world to keep those combined. We can revisit the theory when the payments business is multiples of GMV. I'm not at all convinced we reach the same conclusion. We still see tremendous synergies there. Even using the China example, synergies are there again. For right now, I would say the debate for us is interesting to watch from afar. It doesn't really resonate at all because, again, it's still a small portion of TPV, is still a fraction of GMV.
Maybe it won't be in fashion a few years down the road.
I think you mentioned the official stores. You have a higher take rate on that relative to the marketplace. Are there offsetting costs or other services that you provide to the brands that offset the benefit of that take rate? Also, is there a certain point where you think that you get to a tipping point as far as the number of brands that you have on the official stores, that basically every brand has to be on there or they won't be competitive in e-commerce?
Tipping point. I can do the financials.
Okay.
From a take rate perspective, I think your intuition is correct. The official stores are typically brands that are using a greater number of the services. They're all selling practically all their inventory on the integrated payment model, you have to pay through Pago to buy. Many of them are using shipping solutions. Many of them, if not all of them, are using the bundled credit solution. By the nature of those businesses, both because of their sophistication but also the fact that what they want is for us to handle the whole set of e-commerce complexities for them, they do have consumption of a much more complete set of the different services. That's what drives the higher take rate.
In terms of thinking of it as a tipping point, I don't think that's how I would think of it. I think for brands, there is an enormous opportunity to participate in our marketplace and its distribution. It's something that many brands don't think about. It doesn't cross their mind because they really haven't considered the possibility. At one point, it wasn't possible technologically. Now it is. Some of them are just not aware that this is possible. Part of our work is getting the word out there and convincing them to join. I think the ones that have joined have seen early results that are very good, the tendency is to see more and more brands come.
Lunch?
Lunch is on us. I was going to say one more thing, lunch is on us, because I always wanted to do that. Lunch is on us. We'd encourage you to stay, stick around, and we can chat some more.
Thanks for your time and your questions.
Thank you.
Thank you.
Yep.
Yeah.