CMD 2020

Oct 13, 2020

Slides

Great. Well, thanks for having me. So what I'll try to do is to go through I have a pretty short presentation about 10 slides, pretty high level. I'll try to take like 10, 15, maybe 20 minutes on that and then we can, I guess, open to Q and A? So I think most people actually know BlahBlahCar, but we started as a long distance carpooling app. So essentially like allowing drivers to share their car when they drive along and passengers can book those empty seats in those cars just like they book your seat in a plane or a train or in a bus. And I'll jump to the next slide. Over the next 3 years or 2 years I should say since really 2018, we started to diversify into like a broader offering, which goes beyond this core use case and the unique use case of carpooling that we've created over the years and to tap into the larger mobility market or at least long distance mobility market, which made us go into buses, which will make us go eventually into offering trains as well on the platform. So most of the presentation is actually focused on that really like the last couple of years and the future. So I won't go back to sort of the root and the exciting story of BlaBlaCar in the early days, but really about the transformation we've done over the last really 2 years from becoming just a carpooling community to becoming more of a multimodal platform for travel. So if you look at where we are in that, I would say, transformation, we still obviously dominated by carpooling. So carpooling, as I described, is fundamentally a C2C marketplace. So it's drivers offering empty seats in their car, passengers booking the seats. So it's a fantastic business in a sense that it has 0 inventory risk. So essentially we don't pre purchase those cars. We don't obviously buy the cars. We don't employ the drivers, but just happen to be like math teachers or students driving across the country with empty seats in their cars. And interestingly, there is also no inventory risk for the drivers, right? So it's not just like you push the inventory risk elsewhere. There is just no inventory risk in this business. Essentially, you just like feeding up seats that are empty in cars. The business model is a transaction business model on carpooling. So essentially we have like a commission on every transaction and fundamentally it's a very high and gross margin business at 90%. Now the way we entered into becoming like a broader multimodal play was by offering buses. So we've done that in 2 different ways. And we've done that in 2 different ways because the European market is fundamentally different from the rest of the world when it comes to buses. So if you look at Europe, it's a market that actually liberalized or at least Continental Europe, I should say, is a market that actually liberalized quite recently in 2015 2016. And it was a market where people were not taking long distance coach. They were only taking either trains or carpooling to some extent. And that market get consolidated pretty quickly. So today essentially you end up with mostly 2, maybe 3 actors in Continental Europe. So we decided to launch through the acquisition of Rebus, what I would call the virtual operator, where essentially we contract bus companies that own fleet that they either deploy for school or local transport or events or corporate events or whatnot. And essentially, we contract them to run specific lines. So we do in that business take an indirect inventory risk in a sense that essentially you sort of prepay specific lines with bus operators, but you have a strategic contract, so you don't own the bus, you don't employ the bus driver and you end up with a very flexible network and ability to reshape your network. Yet, it is a fairly low gross margin business model because essentially you pre purchase to some extent the seats that you resell. So you do take some indirect inventory risk even though you don't own the buses. Outside of Europe, the way we want about buses is a slightly different game, where essentially those are very large bus markets, highly fragmented, very unsophisticated and mostly offline. So today, they're like 80%, 90% still offline. So it's your typical offline to online play essentially where you go in, you build technology. So in our case, we built like a GDS a la Amadeus for cyber for the online industry and an OTL layer. And essentially, we digitalize this sort of like old school bus industry in Russia, in Ukraine, in Brazil and in the future in every emerging market. So it's really an OTA play, which is very high gross margin as well. So it's 80% gross margin. You don't have inventory risk. Obviously, the carriers do, but essentially you sell, I would say, a few seats in plenty of buses, just like Booking.com in the hotel industry. Now the beauty of all of that, so I just highlighted the 3 legs, I would say, of the business we operate. For end user, all of that is aggregated within one app, right? So essentially, as a passenger, you just go and buy the car, you see all this inventory of car and buses will add train at some point and you just like one click away, one booking away from booking a seat in a car or a bus. So that's a brief summary. I'm assuming you see the slides by the way. I don't know if it works on this. Okay, great. I bet you won't go back into that, maybe just highlight a few things about the difference between those different business legs or business model. I think what's interesting is also to think about the market itself. If you look at the market of core pudding, it's a market where today I would say we created that market and we're very dominant. Like we don't have direct competition on core pudding. Whether it's in Brazil or in Russia or in France or in Spain. And it's purely about market creation. So if you ask me like what's the market size for carpooling, frankly, I don't know, because it's about like you're building a better experience and a better matching algorithm so that you can match drivers and passengers. But the amount of cars and empty seating cars is just humongous. So the challenge here is not like offline to online. It's not like looking at transport data. It's really about building a great product experience and it's about market creation. It's also very much driven by product and tech. So over time, we went from, I would say, a pretty simple sort of classified type product, where drivers would go on the platform and create a profile and let's say, I'm driving from Paris to Lyon and I'd like to get €25 per seat to something that's a lot more powered by machine learning, where today essentially we get what we are getting to a point where essentially the driver doesn't put a price, puts just a starting point and an endpoint and we match the driver in time, in price and in location and we define next stops along the way so that the driver can pick up and drop off passengers along the way and we optimize the price of every sub leg. So today it's 25%, 30% of the matching between passenger and driver is actually generated by machine learning. Just redefine the pickup and drop off point on the price of every sub segment. And long term, as you fast forward, maybe 80%, if not 100% of carpooling stops and prices will be defined by machine learning and not defined by a driver. That's important because as you do that, you clearly bring the concept of dynamic pricing into carpooling. You also create like a multitude of pick up and drop off stations or virtual stations, if you like, that just create a better network. So that's the challenge on carpooling. Blah Blah Blah Blah Blah is an interesting business because it's really a European thing so far. So it's really like the way the bus industry shaped in Europe with today essentially 2 actors. You have fleet bus out of Germany and you have us, we acquired a company called Webus to accelerate our presence in this market. So it's actually borne out of deregulation in Europe. It's a sizable market, but it's actually fundamentally capped by trains, because most of the markets in Europe have good and highly subsidized train networks, the like of SNCF and Deutsche Bank and so on. So it's there is like a pretty large market in the middle, but it's nowhere close to emerging markets. And then if you look at the what I call the bus marketplace segment, the last one on the right, again, it's a marketplace OTA type business model. It's a very large market. So very large market in countries like again Russia, Ukraine, all of LatAm, India, Mexico, and you're talking about tens of 1,000,000,000 of dollars actually of GMV generated by the bus industry in those countries. And it's interesting here because it's one of these rare, maybe one of the unique almost, I would say asset class in transport that has not gone online in those markets. So people, strangely enough, still buy 90% of the bus ticket in Brazil, in Russia, in Ukraine, in Mexico, offline at the station. Obviously, it's evolving fast. It's changing. The technology to crack that problem is actually not just building like a nice sort of, I would say, top user layer, but it's actually building all the technology stack, enabling actually those SMEs, so this sort of tiny Russian, Ukrainian, Brazilian bus business to actually digitalize the entire inventory and manage the entire inventory online. So the key, I would say, product and tech block we're building here is actually a GDS. So it's a global distribution system that allows us to do online distribution on buyback car. So you have like the hidden part of the iceberg is actually below Babacar and that's the GDS layer that partially came out of the acquisition of Bus 4 in Ukraine and Russia because we had built a fantastic technology to crack that problem in the region. And we combined that with our own technology and that's how we are entering the Brazilian market today. So starting from the supply side again, like we have these 3 legs. Obviously, the 2 new ones are the bus ones. If you look at the 3rd one, that's by far so the bus marketplace is by far the most scalable because the same technology we deploy around buses can be applied to trains, shuttle, essentially once you get like a coherent GDS, I mean IMS, GDS and OTL layer, you can start to connect like essentially thousands of operators on the platform. So today essentially our goal is to connect as many operators as possible in as many countries as possible and combine that with Carpooling to have like a highly differentiated offer. So now I talked a lot about the supply side, but I guess if we look at the passenger journey and why going multimodal for us is interesting, it first comes from the belief and a pretty simple one that most people would agree with that as you create or as you enrich your offer, you attract more demand, right? So we went from just being a carpooling app to combining carpooling and buses. And long term, the end vision is really to start combining all of that. And it's interesting when you own a very granular network with carpooling to combine that with other mean of transport because the car remains the universal connector. So if you think of most of the journeys you take by train, by plane, there is very often like a car at the beginning of car at the end. So the car remains the universal point to point connector for most of us. So the end vision is not just to put carpool and buses and train side by side, but it's really to start combining those mode of transport together. And by doing so, you end up having like very unique offer on your app, not just by carpooling, but by combining those journeys. So that's the I would say the unicity of USP of Delekard is to be combining coupling with this different mean of transport. Now, maybe for you to visualize what we mean when we talk about complementarity and granularity of the network, I think that slide is pretty interesting where essentially we show on one side the number of bus stations or cities we connect with blah blah bus in France and the number of pickup and drop off points mostly generated by machine learning on carpooling. And I think when you see that, you realize that actually you're building 2 very complete different network, like the topology of the carpooling network is highly granular. So in other way, carpooling is not very good at doing center Paris to center of Brussels. Buses and trains agree to that. Corp putting is good to go from some like suburb of Paris to your small town in Brittany. That's really good at that because the alternative is either you drive your own car or you take a train on a bus and another train on a taxi to get there. So by combining those 2 essentially, we tend to respond to 2 slightly different use cases on a very similar audience, which is a pretty young price sensitive audience. Now the other thing that's interesting here and going more into financial or unit economics topics. The beauty of carpooling is because we have this unique inventory, we have like a very strong captive audience and very high quality user profile. So when you use carpooling, you need to sign up, you need to give us a mobile number, you need obviously to pay online, you verify your ID, you give us your license plate of your car and so on and so on. So we created because you need to create trust in carpooling, we created like a very rich community. And essentially, the consequence of that is we don't buy traffic, right. The main thing on carpooling is carpooling brings free traffic for free. So if we look at most of the markets, we actually don't spend much in marketing. And that's what you see with the bar chart on the left actually, 87% of the traffic on the Black Car and the passengers booking the corporate rep on Black Car are not paid, right? So it's quite unique to a normal OTA, like normal OTA do depend a lot on paid marketing and SCA and essentially if you end up spending a lot of money on Google to attract traffic, we don't. So the beauty is as we start to add actually other mean of transport, starting with buses as I just described, actually you don't change your CACCO that much because you benefit from this very wide audience that historically was only booking or putting and then you start to increase your share of wallet within this audience and they tend to come back even more because they see that you have more than carpooling. So essentially you have like a very interesting phenomenon where you have TV is improving because you increase your share of wallet per user. And in fact, even your CAG and that's what we show here over time, we don't show the exact number because we want to keep that confidential, but you see the relative drop of the cost of acquisition per new passenger is dropping because essentially you're better at responding at any demand because you start to aggregate the entire market, which for us is the case in countries like France, Russia increasingly and Ukraine as well increasingly in Brazil. I talked about a bit about that on core pruning specifically. As you can tell, I mean, the old story sort of to some extent rely on I think carpooling as an element in the equation. So people often ask us like, why don't you launch the pure buzz business in some countries where you don't have carpooling. I think all the magic in terms of whether it's CAC to LTV or having all these granular trips comes from carpooling and the fact that we created like a very strong community around carpooling. Again, the future of carpooling is also very much around technology. So over the last 2 years, and I mentioned that a bit earlier, we've been building great technology to move from, I would say, a classified business on which you bolt on a transaction based model, which was by the car like 3, 4 years ago to move a machine learning play where essentially you create another barrier to entry beyond liquidity of the marketplace around like being smart at essentially like finding pickup and drop off along the way for drivers and being smart about optimizing the price point at which you can find a match between the driver and the passenger. And I believe that's pretty fundamental because as we do that, essentially it becomes extremely hard to replicate a business like that. Extremely hard because you need millions of people with a trusted profile, which is like the first element of matching. You need to know who's in the car and feel comfortable with who you're going to share the car with. And what's something else that's powered by volume and data, which is like what are the good pick up and drop of the area in a country and what's the right price to match supply and demand. And again, today, we slowly transformed and it's really transparent to users, that's the beauty of it, the complexity is really integrated. But today, like 25%, it's even more now than 25% of carpooling matches are purely generated by machine learning. So meaning like drivers did not plan to do this detour or pick up or drop off people in a specific location and they did not choose the price. It's been purely computed as a by a matching algorithm. So we believe that's the future of carpooling. And as we do that, we increase usage of carpooling. And as we do that, essentially you get this fundamental advantage over the oven enough transports such as buses initially and probably more as we go along. So I guess these days no presentation is complete without a COVID update. So obviously, the current crisis has been bad for transport, bad for tourism, Like it's been a pretty rough ride in Q2 2020. Having said that, we've seen like a pretty remarkable rebound starting in June and still today, by the way, the curve is still going up. So we are right now at like minus 15%, minus 20% versus 2019 in term of passengers transported on by the car. And it keeps on slowly going up. And to compare that to Othemenoft Transport actually buses are still and we know that because we operate by bus in Europe, buses tend to be 50% to 60% down on 2019. Train seemed to be 30% plus to 40% down depending on countries. And obviously air travel is like 70%, 80% down. So surprisingly or not surprisingly, carpooling has rebounded extremely well out of this crisis and still I mean, it's still essentially sustaining those level right now in October for several reasons. 1, you don't need to think about the network. So as I said earlier, it's community based. So in June or July, when things reopened, when every company struggled thinking like, well, should we put the bus line or train line or should we fly planes here and there, and you don't know because the demand is unpredictable. Essentially, we had a lazy job. Like we did nothing. Drivers came back, passengers came back and they started to book again. Like within a few weeks, we were like, it was 60%, 70% of the network was back up actually within a very short period of time. We've also seen that on the passenger side, there was a clear preference in many markets for carpooling within this audience, because you minimize the number of contacts. So you had lots of passengers thinking that well, if I need to choose between a bus or a train or a carpool, a driver be with 3 people in a car than 40 people or 100 of people in a bus or a train. So funnily enough, actually the biggest challenge we have right now is actually making sure we can get the drivers back and get like new drivers on board, but we have too much demand actually. So the last to give you a sense in same well, in stable markets, but I'll take France as an example. In France actually, we are at 15% higher demand year on year versus 2019. If you are in France right now, it seems surreal. It seems surreal because trains are completely empty. The countries have shut down, but that's the point. Essentially the fact that the other mean of transport are being depleted, actually carpool becomes like the alternative. So we see like a model shift toward carpooling from buses. So we see like bus network are like quasi shutdown, trains networks are operating but at lower scale. So you see a transfer toward carpooling. So it's interesting for us because, A, it's very sustainable. We don't have fixed cost in the business. So we do operate at lower volume than expected obviously, which is not great, but it's highly resilient. And as it's going to bounce back, I mean, we all hope it's going to bounce back, I don't know when, maybe spring 2021, maybe earlier, essentially we gained market share and we gained essentially more customers in the customer mix. So it also enables us to think a bit more like mid term and long term to 2021 and beyond and not burn too much cash along the way while you're being relevant for users. So today we're one of the most relevant on this like young low cost segment I would say, one of the most relevant transport network in countries like France and Spain and Russia and Ukraine and Brazil. So to some extent, I would say the length of the crisis is not such a bad thing for us because it's really, I mean, the differential pain, I would say, between what we suffer and what competition is suffering is going in the right direction for us. So that's COVID. Maybe to give you a sense of where we come from and the impact of COVID. So back I was saying that earlier in the presentation, but the run rate of the last 3 months actually pre COVID was around 85,000,000 passengers, growing roughly at 55% year on year. So late 2019 2020 was actually an early 2020 was actually an acceleration for us. Obviously, that acceleration went down with COVID. We anticipate we'll do around 50,000,000 passengers in 2020, obviously down on 2019. We expect and then frankly it's crystal ball in term of demand, but that model assumes partial recovery in H1 90% recovery of overall demand in H2. And remember, like the fact we grow in 2021 is not a function of market growing. It's a function of like offline to online migration accelerating emerging market, which we see now actually during the COVID phase and the fact that inherently we've built actually larger market share during 2020. So the drop of 71 to 50 is not the drop you see in for bus companies or even train companies and never mind airlines. So we believe actually because we've built a bigger relative market share during the crisis as it comes back actually we should be able to partially maintain in most markets that market share. So last but not least and getting out of the COVID topic for a minute, it's always been important for us to also do something that makes sense, I mean, for us, for the employees, for the users. So as a reminder, actually, the activity in 2019 saved 1,600,000 tons of CO2. And that's something that's not been like a back on the envelope calculation. We've done that with like a consulting firm actually that came and analyzed all the model shift actually that carpooling is creating and it has a massive positive impact, which is good in many ways. I mean, it's obviously good in the current environment. It's also like a very strong sense of purpose for many of our drivers. So if you look at the carpooling drivers, they're not professional. They're just people sharing their private space. And of course they do that for financial reasons, but not just right. They also do that for a deeper sense of purpose around the social benefit of meeting people and the environmental benefit of not driving alone. And that enables, I would say, all the economics I talked about are also powered by the pretty strong brand image we created. And maybe a good proxy of that is the NPS, net promoter score of Bebblacar in every market is anywhere between 4570 actually, which for transport is extremely high and higher than any of them in North Transport. And again, that translates to that sense of purpose, the fact that people feel they're doing something that makes sense for the future and the fact they have a fantastic experience does translate into those economics and these better sort of unit economics idea I described on the free traffic. So I'll end on that note with slightly outdated numbers, but actually not quite, quite good. So 22 countries with close to 100,000,000 members signed up and the run rate pre COVID was 85,000,000 passengers on the platform. So maybe I'll stop there on the presentation if that's okay. So I chose to talk a bit about the business model initially, which is which I think is interesting to understand like what's under the hood. But I have to take questions. Thanks. Yes. Thank you, Nico. We so everyone listening in, you can type your questions in the chat or the Q and A sort of products on Zoom there. But I think Bjorn has a question to kick off with. Yes. Maybe we could start with a different mix of modalities. As of now, of course, as Spas is fairly new to the business, carpooling is the vast majority in that mix. How do you see that more long term? Will it be more of an even mix between the different modalities? Or, yes, what do you think? No, no, it's I mean, it's clearly going to go with over time like the share of buses is going to increase. So if you look at the pie chart I was showing at the beginning, that's been built in the last 12, 18 months essentially, it was 0 18 months ago on the platform. The blah blah bus business in Europe has been growing pretty fast. It's more capital intensive to grow, more operationally intensive to grow because you need to contract bus operators and set up lines and so on. Again, even though it's highly flexible and you don't pay for buses, it's still a more incremental job. The bus marketplace, business was growing like over 100% year on year pre COVID and even now it's positive growth actually on a year on year basis even during COVID. So that segment, I would say the non European bus play has the potential to be as big or bigger than carpooling in a pretty short time horizon of, I don't know, 2, 3 years, if you need to say. It's a giant market. And the thing to understand actually in, if you take markets like again, like Russia, Ukraine, Brazil, and then you can extend that to other markets where we operate like Mexico, India to some extent. There is not much online distribution actually. So in a strange way by creating our own asset with carpooling, we were the only player having like 100 percent online distribution of trips. So today, in some of these markets, we, I mean in most of the market, all of the market, in fact, I named Ukraine, Russia and Brazil specifically, BlaBlaCar is a bigger online portal than any other bus online OTA essentially. So when we start plugging buses, we it becomes, I mean, I would say much easier for us to fill up those buses with demand. And we have much better economics in terms of CAC to LTV to a simple bus OTA. So that's the play. I think in those countries, we can actually grab a disproportionate market share of the online distribution of buses, because it's non existent. Like if you look at competition, it's not. I mean, every time you have like a few local companies, but it's not that big. And that's part of the story by the way with BuzzFoort. BuzzFoort had done a phenomenal job on the supply side that built a great technology connecting, you're building that GDS layer, this IMS layer connecting to these very fragmented world operators. What we really bring in is we accelerate that because we were also working on that. But we also brought in demand. In 2017, if we inject the black card demand on those thousands of operators that have been connected online, you just like you tap into a market that's like you well, in Russia, it's like maybe $4,000,000,000 to $5,000,000,000 of in Russia and Ukraine, it's maybe $4,000,000,000 to $5,000,000,000 of bus market, that's still 90% offline. And last but not least, the part of the I would say the bad story of COVID is less people travel right now. The good story of COVID in that sense is more and more people book online. So today, the reason we see it's still like positive growth in markets like Russia and Ukraine, it's not because people travel more, it's people because people book more online. And essentially that at least from what we see on the platform, the offline to online effect is greater than the reduction in the temporary reduction of travel. So I'm very bullish actually on these markets because it's pretty rare that you have like a large asset class, like buses in those markets that have not migrated offline to online and clearly it's bound to happen. And we ended up with carpooling to build like your free demand traffic for executive those trips, except people do it with carpooling on by the car. And as you add over enough transport, of course, they take buses like in Brazil, like that's the only way you can travel essentially, like there are no trains. So when you start plugging buses on 2 Belga Cargo, those are the same, there's the same audience essentially. You're just increasing your share of wallet on that audience. So long story short is yes, I would anticipate that and it's hard to predict, but I would anticipate that within a few years you see as many, if not more bus bookings than carpool bookings. The only slight caveat on that is as we integrate more supply onto the platform, it's interesting what's happening because we start to attract what I mean, what I would call like initially bus only demand or if you had trained, it could be trained only demand. And when they discover carpooling actually they switch and they use carpooling. So in fact actually they tend to feed each other. So the fact that you have more and more supply and you attract more demand onto the platform has the tendency to grow carpooling as well actually. So in terms of you have sort of like future cycle going on. Thanks. Great answer. And so we have a question from the audience on the monetization and sort of monetizing new countries. And you obviously generate new revenue without raising cost much. And can you cover which countries now are strong monetizing now and which one are beginning to monetize or about to monetize? And then what's the threshold monetization in terms of scale or? Yes. So there is no, I would say, good timing to start monetizing. In that sense, it's very much like a classified type play, where you spend several years essentially building your C2C marketplace, making sure you essentially you are dominant as a C2C marketplace and then over time you start monetizing. Today, we are monetizing most of Europe. So most of the key markets, I would say, if you relate that to volume rather than countries actually, 70% of Europe is monetized. And as of a few months ago, 0% outside of Europe was monetized, even though we have more than half of the GMV outside of Europe. Now the reason we haven't done that actually was the growth and the scale actually we were reaching without monetizing in most of these market was great. And it felt like essentially a better model to spend less marketing and not monetizing than monetizing early and spending more marketing. So essentially think of that as your no monetization is just a marketing to build to some extent to build dominance in those markets. We started to monetize Russia just literally like a couple of months ago as I mean it was the plan was to actually do that in April, but when it was technically ready in April, the country was shut down for transport. So we are starting to monetize those markets outside of Europe and that's going to be the story of the next 12, 18 months essentially of between now and end of 2021, we will start monetizing most of that GMV. And again, progressively in those countries. And it gets pretty interesting because as you do that, you have like a mechanical sort of EBIT uplift because essentially in these markets, all the cost is already there. Like essentially like all the teams, the marketing, the platform, I mean, all of that is already there. So essentially as we stop monetizing, we have like a massive EBIT uplift in those markets that we can easily reinvest into either other markets or interestingly, we are defining that playbook between how fast you grow bus versus how quickly you monetize carpooling. And there is like a nice equation to multimodal equation to find between carpool monetization, how aggressive you want to be in buses and how, I would say, how you want to monetize the bus traffic. So we're playing with that in Russia. And there is a nice equilibrium where you that you can find where it's all about monetizing everything as fast as you can. It's about finding the right equilibrium on monetizing carpool, monetizing buses, which are monetized from day 1, by the way, in all the markets. And then you can tune like the right equation of growth and monetization in those markets. So that's what we're getting into right now as such as we speak. We are deploying those models in Russia and after that we deploy that in Ukraine and Brazil. Thanks. I think on the slide follow-up on that revenue, I think we at VMV, we have sort of we've put out there like a long term target that Dabla should be able to generate like €5 per passenger or member per year. And there's a question here on that and then that you think that's realistic? Yes, I think it is. I mean, it's I mean, today, it's higher than that in markets like France, right? So if you look at the revenue gross margin per passenger in France, it's much higher than 5 already, right? So I mean, if you think of it in a single booking in France, we tend to make €3,000,000 on average in carpooling, even more actually. So it's between €3,000,000 and €4,000,000 Now the ticket size in emerging market is lower. So you tend to make more because the trips tend to be around $10 You take like even as you monetize, you tend to take like $1, $1.5 per booking of gross margin, which is more or less the same as revenue. It's 80%, 90% gross margin in this model. Now as you add buses and you get more frequency, you should be able, I think in all these markets to get around the €5 market. So I would say globally, yes, it's very achievable, I would say almost without multimodality, but even more so as you get multimodal to get like above EUR 5 average revenue per user per year on your active user base. Great. There's many questions here. But so one is about the economics of connecting trains to the Blah Blah platform or app? Yes. So it's a very good question. The essentially the commission you get or the tech rate you get on train is pretty bad, right? So as the thing to think about, like you tend to take few percent points that you get on trains. So if you look at that like standalone, you're aggregating trains and getting like 2%, 3%, 4% tech rate on train. You need to sell a lot of trains to create a business. When on carpool, we take 22%, 23% actually tech rate on something that's on average €20. The reason it's interesting is, a, it comes for free because if you take like European markets, essentially like everyone takes the train. So essentially it's something we're not offering the audience that they're booking anyway. And B, where it gets very interesting for us and the way we think of train is combining journeys. So it's creating train is combining journeys. So it's creating these unique experiences where you combine a train and a carpool. And people do that, like when we survey, I mean, people do that because, hey, when we survey the user, they tell us, but we don't even need to survey user. Actually, we know that 15% to 20% of our meeting points in falls are train stations, which means people naturally combine those strips or they try to combine those strips. As you do so essentially, we start selling I mean essentially we will attract more demand on the platform because we have trends, but mostly because they know they can find some unique combo of your train plus carpool and then you end up selling more carpooling on the back end or the front end of a train journey. So you think about as if you go from Paris to I don't know, Caius in the French Riviera or some like that. The best journey might be actually a high speed rail going to Marseille and then the car approved for the last 100 or so kilometer. You'll never think of that on your own. So the platform is going to find that for you with the smart stops and the machine learning on carpooling. And by doing that, actually that last segment is something where a driver could easily get like €15,000,000 or whatever, like €20,000,000 on the last leg and we take 25%. So most of the value is going to be created by what we call intermodal trips where we combine with other mean of transport. But train standalone, like in the post European market is tricky. Do you know like what proportion roughly today is intermodal where someone uses first the bus and then the carpool or train or carpool? Yes. As I said, it's about like, I mean, we see that, I have to check exactly the data, but it's around like 15% to 20% of BlaBlaCar journey actually originate or end at the train station. So assuming they don't live in the train station, they probably take a train or came from a train. And then like what type of train journey? Is it just like a short leg, a long leg, like we don't have all this information. But again, today people naturally do that, but we're not creating even the opportunity to think of that, right? So whereas if you start combining those means of transport and again, it just I guess if you take a step back, it just makes sense like what do you see when you arrive at the train station? Cars. I mean, this is full of cars. It's just like that's how people do the last part of the journey essentially. So combining the 2 makes a lot of sense. Combining bus and carpool would make sense. It's less logical because a train is a bus is essentially slower than a car. So you don't have like that speed element in the bus. So I would say we do people I mean, we see people do that, but it's a lot, lot less interesting than Trane. So again, like going back to Trane, like it's really about combining trips at the end. And at a very basic level, well, you grab that share of wallet from passengers on your platform and you still make like 3%, 4%, 5% depending on the operators on that GMV, which you will not take over ways. Are we are you monetizing data lines? And has that sort of product line just to take a step back, it's the commuting app. So it's carpooling for your regular daily journeys. So going from work to your home. So Blabla Line was taking off like extremely well actually pre COVID. Now it's restarting. So it's coming back on. We don't do any marketing. So it's like something it's community based, so people just come back. It's pretty new, so it's hard to look at year on year. If we look at year on year, they look good, but it was very tiny actually a year ago. So it's restarting. It's a bit slower to restart than Blackhawk right now because I think you still have a lot of people working from home right now. So it's still like the main trend is not to go to work. In fact, I'm at the office today and I'm alone. It's like the empty office for me actually right now. So today we're focusing on France and it turns out that because of and I won't get into all the details, but because of some regulations that passed actually last year in France, actually you get a lot of subsidies from regions around short distance carpooling. So for us, the activity today is monetized by that. So in France, it's actually a profitable standalone. It's pretty much breakevenprofitable as it grows. So that's kind of the beauty of it. It's not necessarily the model we deploy in every country, but it makes like a fantastic playground where essentially we're being subsidized to create this great high frequency product. And now we need to decide before next year to start launching that in another market and then like we need to define what's the monetization model of Blaber Lines. But before even talking about the monetization of lines, the beauty of it, if you take a step back is to say, if you get like the tens of millions of users we have on Blackbaud car to actually use a product, not like once a quarter or 5, 6 times a year, which is roughly the frequency we see today for active users on buyback car, but you get them to use a product like 4 times a week. So the frequency we see on buyback car is like 4 times a week. We do like roughly between 15 20 trips per month actually on lines. You have fantastic engagement of the community. So even as a product, even if you were to only lightly monetize that usage, Essentially you create like a captive, an incredible captive audience. And then you can, you should not monetize those sort of like repeat trips from home to work, you can actually monetize anything else. So today what we see is that we have 2 type of audience onlines. One is like the true car dealers that will repeat the same pattern over and over again. And then we see passengers booking what we call a one shot or a one off trip actually on the blah, blah line journey. So essentially like your home to work or a leg of your or a portion I should say, of your home to work might become a better Uber for someone else. And that's super interesting because we realize that those sub segment of blah, blah line journeys, you create like a massive network of commuters. And by doing so actually you just created like a phenomenal network and some subset of the network becomes highly valuable as a short distance journey that you can highly monetize. So we're playing with that in France and we realize that part of these journeys are highly monetizable. Some other journeys that you repeat all the time are less monetizable, but they create the massive frequency. So there is like a follow-up question on that. That is the long term idea to create like a blah, blah super app with where you add on other services. You've done a venture with Voy testing that and payments and there's a lot, yes, like a super app phenomenon? Yes, I would say, yes, and the super app is actually black card. So we've been in a sense that today that's where we aggregate buses. And even though we created an independent app for commuting so that we have the technical agility and the product agility redefine a slightly different product paradigm for commuting, actually everything is built so that everything can be connected through API and all the blah, blah line trips can and will and we're already playing with that will be connected to BlahBlahCar. So essentially like BlahBlahLine is really going to focus itself on the repeat use case of commuting, which is mostly monetized today through all kind of government schemes. There are some very strong one in France, but not only there are many markets actually subsidizing some of that. We're helping some of that because we see that as public transport. I mean, it's really like a substitute to public transport in a way. But then essentially if you connect all these trips to buyback car, you have more of, I would say, one off demand that we'd see those trips as, hey, it's a fantastic like 40 kilometer journey or whatever to go from X suburb to the city center or whatnot. And the willingness to pay for this one off journey is extremely high. So essentially like Grab a Car is built to receive every 3rd party, I mean today like trains and buses from all these emerging markets, including Baba lines. Lines. So we see that at some point, it's going to become like another supply feeder onto BladderCar. And yes, we've been experimenting with Voi, but that's the same idea. It's like fundamentally and that's very experimental with VoIP because the booking model of e scooter is very different from what we do traditionally. But whenever it's something you prebook slightly in advance, you can put that on the Belair Car platform. So we're rebuilding that's part of the introduction, I guess. I mean, we went from a carpooling community to building around that community like a much larger GES and OTL layer when you can connect actually all type of supply. Great. I know we started a little late, but hopefully, I will have time for maybe one more question. And I think there are several questions around geographical expansion beyond the countries you are today and especially with an emphasis on China and why we're not there and then maybe the U. S. Also. And then most importantly, why you're not in Sweden? Maybe, I mean, like broad answer on that. The, we, I mean, we had like a big phase of geo expansion where we went from like essentially 1 to 22 countries between 2012 and essentially 2016. As you can tell from the presentation and the discussion, the focus right now is mostly on product expansion and building that multimodal play. So I would say we want to demonstrate that at scale in several markets. Some in Europe, I mean France is obviously a great playground for us, but also Russia, Ukraine, Brazil. I think once we have that, it would make tons of sense actually to go back. So I don't know if it's like a year down the road or 2, but it's not I mean hopefully, it's not like 5 or 10. Then you have this new playbook of, okay, you can enter a market with CorproLing Plus buses. And then yes, you would think of Asia, you would think about the rest of LATAM, you would think about some countries in Africa. After that specifically on the 2 giants there, the 3 giants there, Sweden, China and the U. S. But in terms of big markets, I mean, China, we never really looked at China. I I'd love to get some intelligence on that, but it feels pretty hard to create a business like that in China when you looked at how the like of Uber and many other pure C2C or B2C marketplaces have struggled existing in China. To some extent Didi is doing or was trying to do more or less what we do. So it feels the competitive landscape in China is a bit scary, right? And the amount of effort and capital you need to put to exist in China. It feels like today at our scale, probably an unreasonable risk. It doesn't mean there is no market. I believe there is probably a market. It feels pretty hard for us to crack. And I think we have like offer I mean, we demonstrated there are lots of other larger markets where competition. US is different in a sense that the model of core pooling as we do it has never taken off. I mean, for all kind of reasons ranging from cost of motoring is very low. The car insurance framework is complex and essentially that if you take a passenger that pays for a journey, even if you don't make a profit, you could have some issue if you have an insurance with you a lot less protected than in most European countries. And last but not least, there is no good public transport. So it means like finding the right or I should say, the level of liquidity you would need in the marketplace so that your pickup and drop off point are relevant for users because they cannot take like a tube or tram or anything to get to the meeting point would be enormous. So it's been tried. A company like Lift actually started as Zimmeride, which was the blood alcohol of the U. S. I've always talked to John and Logan, actually the founders of Lyft and again before that Zimmeride. And that was the analysis as well that it was really tricky to crack the liquidity problem in a country like the U. S. Where you have no first mile and last mile solution without a car because you ask the drivers to become taxi driver and it's really tricky in that model. So today, I guess we've been proven right to some extent not to go there. I mean, because no one has done it. And I think it's hard to crack still today as a market. So there might be an opportunity in the Northeast area where I think the density and maybe makes more sense. Maybe we do that one day. I would say right now it's not top of the list in terms of things to do for the next couple of years. Great. Well, thank you. I think we have lots of other questions, but I think we've taken so much of your time. And thank you for being so generous with it and answer candidly on all these questions. And with the questions we haven't answered, we'll try to follow-up ourselves with that. But everyone, thank you so much. And Niko, especially a big thanks to you. Thank you. Thanks. Bye. Bye.