It's hard to project, isn't it. It's always the same when you create a project. At the beginning, you have a lot of skepticism because if it doesn't exist, people think there's a good reason. And actually, sometimes there isn't. That's precisely the principle of entrepreneurs to say no, it doesn't exist and in fact there's no good reason. So we're going to do it and now it will exist. What makes BlaBlaCar work is that it's well-known enough that everyone finds each other. It's the marketplace logic. It means there need to be many offers and many demands meeting. The two biggest countries for BlaBlaCar are Brazil and India.
You create a company that brings together 100 million happy users worldwide and it's not enough for you. You created BlaBlaCar and it's not enough. You think how can I be even more useful today and you create a platform dedicated to solidarity, excuse me, that engages companies and individuals for associations. In short, what's going on in Fred Mazella's head, one of the founders of BlaBlaCar? What's going on in Fred Mazella's head at the time of AI? On the founders of DIFT and also president of France digitale, the multifaceted and new founder of the prize of citations.
Yes. So that was unexpected, huh. You weren't expecting that one.
No, but I loved it. And I just wanted to launch the jingle and then we'll talk. We're off [laughs]
[music] white [music] 40 shades of next
multi-entrepreneur. So that's it de facto. We are here at Vivatech. We question many entrepreneurs about their report to AI, whether it has a positive, negative, or completely neutral impact on their business. I'm curious to hear the double entrepreneur that you are with a company that was already very established and growing worldwide and then with a company like Comnif which is younger with perhaps more latitude to make rapid changes with these tools. So there and colivierie we wanted because BlaBlaCar is 2007 or 2000 2006 2006 makes 20 years. Happy birthday. Vivatech makes 10 years. BlaBlaCar the 20 years.
Yeah. So well no, that opens even more doors, you know. That is to say it goes even faster. The time between the idea and the realization can be even shorter. That's what I see with AI, that between the moment you have an idea and the moment you manage to develop, even just a first prototype, now it takes days where it used to take months. So it changes a lot of things. It's much more dynamic. That is to say you can test things much faster. It costs you less to test, so it costs you less to innovate.
So it's true for this prototyping part on the early stages, but when you talk about a launch, you love to launch, we know you're a project initiator, and when you have something much bigger, so is it more of a facilitator or a threat? If you take BlaBlaCar, carpooling has inherently algorithmic optimization for routes, transport, etc., matching, and does AI come to shake things up in a positive or negative way?
It's positive because it allows development teams to go much faster, even just to optimize the code. You know, you code three or four times faster today than a year ago.
But you know, we've heard this word SaaS apocalypse with SaaS companies seeing their stock collapse, and not everyone is a SaaS. But still, is it the same issue for an entrepreneur like you to think tomorrow anyone in their bedroom could code what is BlaBlaCar? And what is my difference if not my community?
Well, there's the community, the brand, and then there's all the customer service behind it. That's not easy to set up. After with AI, you can optimize that, but there's still know-how that isn't necessarily in machines. And then, obviously, what makes BlaBlaCar work is that it's well-known enough that everyone finds each other. It's the marketplace logic. It means there need to be many offers and many demands meeting.
I think there's also something else you said before and I remember, which is that you're in the business of humans and therefore trust. And this community associated with the brand, size, and many things, efficiency, service quality, at some point, people are asked to trust each other for one to get in another's car. It's sure and it's not just tech; people know behind it there's a human company, reactive, innovative, with people handling service, and it's not just an app. There's a real company behind it, obviously.
And it's not just French today, it's 15% of our volume in France only. Yes, 15% of our carpooling volume is in France and 85% outside France. We're in about thirty countries, we just announced a few days ago the opening of new countries in South America. And yeah, it's completely international. The two biggest countries for BlaBlaCar are Brazil and India.
How did you do it? We have little time, but you bought companies that did the same thing. You decided to install everything. We had three methods for international expansion. There's simply the method of recruiting on-site. That's very, very hard. It's the hardest of the three methods.
Because you need to find someone on-site with entrepreneurial spirit who thinks this service doesn't exist in my country but it will exist. It's hard to project; it's always the same when you create a project. At the beginning, you have a lot of skepticism because if it doesn't exist, people think there's a good reason and actually sometimes there isn't. That's precisely the principle of entrepreneurs to say no, it doesn't exist and in fact there's no good reason. So we're going to do it and now it will exist. And so when you go to find such people in a new country to launch, they need to have a strong entrepreneurial spirit. Obviously, we help because we bring the product, customer service, know-how, but it's still difficult to find the right people first. There's another option where within your teams you have people from the country where you want to launch, and you tell them we're going to open that country. And then they have all the know-how, culture, knowledge, and they move back to their country and launch the service there. That's magical. We did that when we launched in Germany. Otherwise, there's the third method which is indeed buying small companies. We bought, well, we bought a company of different size. It ranges from two collaborators to 400. So we bought up to 400 collaborators, it was in Turkey. And you buy what? You buy the community, the tech, you what? It depends on the size. When you buy a team with say 2 to 8 collaborators, you really buy an entrepreneurial team that will have the energy to launch the service in the country and you give them a boost because you give them a working product with funds to get known and know-how from customer service to the playbook of how you deploy a working service. So you give them a big boost. After when you buy much bigger companies, up to 400, there you buy know-how too. And that's completely different because typically in the case of Obilet in Turkey, you buy people with a very well-functioning structure and they typically deal with bus ticket sales. So it's even a related activity to what we do and today completely integrated obviously in our multimodal playbook which includes carpooling, bus, and train. But there they are specialists in Turkey primarily for bus ticket sales.
And before talking maybe of DIFT, we talk a lot of BlaBlaCar but still a success, an international success, there aren't that many in France. You said there were these three methods, but remember you also have an algorithm, finally you have a culture of this. When you say we don't know if it was a good idea, why it wouldn't work in a country, I think you've designed this experience, you often said for example that countries with a train culture, long distance, it's rather a favorable factor in memory and then there are also maybe other cultural things that mean we trust each other in certain societies and less in others.
We broaden that because typically one of the countries that works best is Brazil, and there's no train. So indeed, there's that, so actually we broadened. It's true that when we had to discuss it, we created this kind of filter. But actually, in the end, we broadened. I think everywhere there are roads and mobile phones, BlaBlaCar can exist now. So if you ask AI to give you all your countries and say what are the key factors that mean in which other country I should go, you see, to find a sort of sociological-economic algorithm, but then you also need to have the means to launch. It's not just identifying, it's good to launch there, but then you need human resources, financial resources to launch properly because it's not just opening the service and saying it's open, that's enough for a marketplace. You know, it's like sand. So for example, if you take two grains of sand in your left hand and two in your right and try to make them meet, you have no transaction. It needs many. You need a big handful of sand in your left, a big handful in your right, throw them at each other. Hop, you'll have more touches, you'll have collisions that schematically give at BlaBlaCar. No, but it's transactions. No, but so it's the probability of encounters, satisfaction. You really need many offers and many demands for a marketplace to work. And that's the big difficulty. In fact, at the beginning when we launched BlaBlaCar, the first remarks were always a bit the same. It was 'I'll never reach the critical mass for it to work. How do you want to find a black sedan when you need one?' Well, it turns out today you type 'black sedan' and you find one. It's a matter of critical mass.
