The Exchanges

Q I spoke to Dara at Uber before the show. And he said, I just want to make sure I get this quote right because he might be a little bit upset if I don't get it right. Um, it's something along the lines of essentially you are often second or third in a market. How do you feel about that? And is that an okay strategy?

A Well, I'm a facts and numbers based guy. So let's look at that. Um, today we operate in about 50 countries around the world and Bolt is the number one most popular platform in more than 20 of them. So we're feeling pretty happy about where we are. And when we look at the trend of most of those other 30 where we operate, we are continuously taking share for the reasons I mentioned earlier. Uh, we offer a better value proposition to the customer. We offer better value proposition to the driver. And in a lot of those places, we're confident that over time, even though we're a second mover, uh, we've, we've done that before. I think we can catch up and actually become the most popular platform.

AI assessment note: “Bolt is the number one most popular platform in more than 20 of them.”

Q You've expanded well beyond the core category now. Talk to me about the decision to expand to other categories. And how you thought about that?

A We always had the ambition to build a replacement to your private car, and we knew that ride-hailing alone is not gonna do that. Ride-hailing is gonna be a huge business, but, uh, there needs to be other modes of transport we're gonna offer on the platform as well. Uh, but the first five years, we were just having no resources whatsoever. So ride-hailing consumed all of our attention, all of our money. And then probably in 2018, we for the first time realized that we actually have enough Budget that we can take on one new bet. And then we debated internally long. What is that going to be? And then we, um, took a gamble on micro mobility. So we decided we're going to be the first ride hailing company in the world to launch electric scooters on the platform as well. And that was actually quite a controversial decision, both from the employee point of view, but also from some of the investors, because they were thinking that first of all, it's a hardware business, so it's very difficult. Um, and second, you're going to be cannibalizing your own very profitable ride-hailing trips because there's a big overlap. Um, about 40% of the ride-hailing trips in a lot of these countries are, um, less than four kilometers long. So you're going to be cannibalizing that and pushing people instead of taking a ride-hailing car to taking a squatter instead, and you will have much worse margins the…

AI assessment note: “We always had the ambition to build a replacement to your private car”

Q Okay, so you have this realization. What do you do next? The majority of people fail, I always think, because they don't take the first step. What do you do post? I want to innovate in transportation in this shared new economy of not owning assets. What's the next step?

A So I started then as a nine-year-old kid, uh, by literally Googling, how do you start a startup? So I went to this Y Combinator and other websites. There was great content out there. Uh, and the top one advice was always that, uh, you gotta validate your interests of customers and the suppliers before you build The first line of code. So I did that. I, uh, set up this survey, uh, on social media, uh, in Google forms, uh, sent it to my school list, put it up everywhere I could, and hundreds of people replied. And they all said that, yeah, they agree. The taxi industry in, in Tallinn at the time was horrible. And it was so horrible that if you were to design a bad experience, that's what you would essentially design. So I can walk you through what, how it was. So you, uh, had 15 different companies. You had to call all of them. Uh, most of the time they didn't even pick up. Then even if they did pick up, the car almost never actually made it to you. Even if it did made it to you, then, uh, the car was in a horrible shape. It was, uh, dirty. The, the driver was rude. You could forget about paying in, uh, with a car. You always had to pay in cash. So it was just all around a really bad experience. And I was thinking that there must be a better way to, um, uh, to organize this. And so did hundreds of other people. So I was thinking after the survey, wow, this is great. This is the e…

AI assessment note: “by literally Googling, how do you start a startup?”

Q Did the message change to make it successful with those 50 or was it just persistence?

A I think that, uh, I was sort of optimizing the wording a bit as you would in sort of face to face sales because over time I saw what, what was getting me the better conversion. But at the end of the day, the pitch was simple. Uh, you're currently stuck under this legacy taxi company. That doesn't give you a lot of work. That's why you're sitting here in this taxi stand. Uh, you should sign up to this app. It's low risk. You're just gonna be paying a small commission every time you get a trip. So it should be a no brainer. And even then most of the drivers didn't want to sign up. So it was a big uphill climb to make it happen.

AI assessment note: “I was sort of optimizing the wording a bit as you would”

Q Wow. Okay. So what was the Series A?

A So we took a very unconventional approach. So first of all, uh, we realized that we had to raise some money. We saw that all the sophisticated VCs didn't want to invest. So, uh, we had to swallow our pride and be like, okay, we need to raise money. I was the company's going bust and we need the money to continue this fantastic expansion. So I contact anybody I knew who had any sort of money. So I remember this one meeting was with a local real estate company in Estonia. So I go there probably back then, like a 22 year old kid and they'd never even heard of a startup. They'd never heard of a tech company. And then I'm pitching them about how our metrics are doing, what the story is going to be. And then I remember at the end of the meeting, they were like, we've never invested in anything like this. We've never even done any investment outside of real estate, but you seem like a good guy. So we're going to invest half a million dollars in the company. And looking back, it was complete insane that they did that. But I mean, they invested at the valuation of about fifteen million. So by now they made like easily more than a hundred X return on that investment. And the only reason they did was because they'd never even heard of any of our competitors. So they just look at it on the merits of the business. And that turned out to be much smarter than all the sophisticated VCs who tur…

AI assessment note: “they invested at the valuation of about fifteen million.”

Q What do you do at this time, dude? That is, like, it's unprecedented. The last time was Spanish fucking influenza in 1917. What do you do?

A So COVID was this fascinating episode for us. So we were doing probably two hundred million of ARR. It dropped 85%. Every other company in this industry started laying off people. They laid off 30%, 50%, whatever they thought was, uh, needed to, for the company to survive. So I remember we had this meeting over a call with, with the top management, and we decided we're not gonna do any layoffs. We're gonna be the only company in the industry that's gonna retain all of our people. And it's gonna be a huge gamble because we lost 85% of our revenue. And instead what we're gonna do is that, um, we will, uh, do a salary reduction to everybody for 20%. And then we're just gonna bet the company that in six months this is gonna go by. And because we're the only company that's keeping our team intact, we're gonna accelerate out of this faster than anybody. And we gambled the company on that. And that turned out to be fantastic. So actually the team morale we got from doing zero layoffs while everybody else did. Was just such a massive morale boost that the people like were even opting in that some people did for 30, 40% salary reductions to get us through those next six months.

AI assessment note: “we decided we're not gonna do any layoffs”

Q You've expanded well beyond the core category now. Talk to me about the decision to expand to other categories. And how you thought about that?

A We always had the ambition to build a replacement to your private car, and we knew that ride-hailing alone is not gonna do that. Ride-hailing is gonna be a huge business, but, uh, there needs to be other modes of transport we're gonna offer on the platform as well. Uh, but the first five years, we were just having no resources whatsoever. So ride-hailing consumed all of our attention, all of our money. And then probably in 2018, we for the first time realized that we actually have enough Budget that we can take on one new bet. And then we debated internally long. What is that going to be? And then we, um, took a gamble on micro mobility. So we decided we're going to be the first ride hailing company in the world to launch electric scooters on the platform as well. And that was actually quite a controversial decision, both from the employee point of view, but also from some of the investors, because they were thinking that first of all, it's a hardware business, so it's very difficult. Um, and second, you're going to be cannibalizing your own very profitable ride-hailing trips because there's a big overlap. Um, about 40% of the ride-hailing trips in a lot of these countries are, um, less than four kilometers long. So you're going to be cannibalizing that and pushing people instead of taking a ride-hailing car to taking a squatter instead, and you will have much worse margins the…

AI assessment note: “We always had the ambition to build a replacement to your private car”

Q Okay, so you have this realization. What do you do next? The majority of people fail, I always think, because they don't take the first step. What do you do post? I want to innovate in transportation in this shared new economy of not owning assets. What's the next step?

A So I started then as a nine-year-old kid, uh, by literally Googling, how do you start a startup? So I went to this Y Combinator and other websites. There was great content out there. Uh, and the top one advice was always that, uh, you gotta validate your interests of customers and the suppliers before you build The first line of code. So I did that. I, uh, set up this survey, uh, on social media, uh, in Google forms, uh, sent it to my school list, put it up everywhere I could, and hundreds of people replied. And they all said that, yeah, they agree. The taxi industry in, in Tallinn at the time was horrible. And it was so horrible that if you were to design a bad experience, that's what you would essentially design. So I can walk you through what, how it was. So you, uh, had 15 different companies. You had to call all of them. Uh, most of the time they didn't even pick up. Then even if they did pick up, the car almost never actually made it to you. Even if it did made it to you, then, uh, the car was in a horrible shape. It was, uh, dirty. The, the driver was rude. You could forget about paying in, uh, with a car. You always had to pay in cash. So it was just all around a really bad experience. And I was thinking that there must be a better way to, um, uh, to organize this. And so did hundreds of other people. So I was thinking after the survey, wow, this is great. This is the e…

AI assessment note: “I, uh, set up this survey, uh, on social media, uh, in Google forms”

Q Okay. So then we launch and it's difficult. You're managing demand supply. How does that go?

A So the odd thing about this business from, uh, for us for day one was that it was very asymmetric. So the consumer side was from day one actually very strong. So we got the tone of free media. Um, it was a great story. Everybody hated the tax industry. There was this young kid who tried to fix it. So we got hundreds of initial signups coming into the app. And that was great. But the problem was almost every single one of them got a really bad experience because we barely had any drivers online. So they opened the app. The car similarly didn't show up or the experience wasn't great. And, um, I got really into a panic mode because I was thinking that we can't just waste this first initial burst of demand that came in because if all these people get a bad experience, we're going to get a bad rep and we're never going to be getting out of this, this problem. So I, I did everything I could. So I went on the streets of Tallinn, Signing up taxi drivers one by one again, this time at least I had an app so I could sign them up on the spot. And that turned out to be a very effective sales tactic. What I recommend to more founders nowadays, which is that, uh, you just, I just got in the car and then I just wouldn't go out of the car before the driver had signed up. So I just was like, Hey, like, give me your details. I'll just do the account for you, download the app, set it up. And I was…

AI assessment note: “the odd thing about this business from, uh, for us for day one was that it was very asymmetric.”

Q Wow. Okay. So what was the Series A?

A So we took a very unconventional approach. So first of all, uh, we realized that we had to raise some money. We saw that all the sophisticated VCs didn't want to invest. So, uh, we had to swallow our pride and be like, okay, we need to raise money. I was the company's going bust and we need the money to continue this fantastic expansion. So I contact anybody I knew who had any sort of money. So I remember this one meeting was with a local real estate company in Estonia. So I go there probably back then, like a 22 year old kid and they'd never even heard of a startup. They'd never heard of a tech company. And then I'm pitching them about how our metrics are doing, what the story is going to be. And then I remember at the end of the meeting, they were like, we've never invested in anything like this. We've never even done any investment outside of real estate, but you seem like a good guy. So we're going to invest half a million dollars in the company. And looking back, it was complete insane that they did that. But I mean, they invested at the valuation of about fifteen million. So by now they made like easily more than a hundred X return on that investment. And the only reason they did was because they'd never even heard of any of our competitors. So they just look at it on the merits of the business. And that turned out to be much smarter than all the sophisticated VCs who tur…

AI assessment note: “we're going to invest half a million dollars in the company”

Q Have you seen that play out in reality? And if so, how?

A So, um, we have a couple of these extremely, extremely intelligent people at Bolt who really defined the company. I'd say Oliver, our lead software engineer has been like that, uh, like he was able to do alone, but It took probably 10 average engineers to do the same work, and it was fantastic. The company would have otherwise never, never made it this far. Um, and the other one is our head of growth, uh, Pavel, who's been with the company for about eight years now, who runs all of our global, um, incentive, uh, logic of how exactly do we figure out which is the small subset of customers who's worth targeting, which are the ones that you don't target. Uh, and I think we have one of the most sophisticated frameworks for how do we do marketing of any company I've ever seen.

AI assessment note: “I'd say Oliver, our lead software engineer has been like that”

Q So then we had 50. What did we do then?

A So once I had validated the consumer interest and I had got the first drivers on board, I started building the app. So I was back then 19. Uh, everybody else, uh, in my high school was studying for exams. But, uh, I had gotten lucky. I had completed one of these, um, national, um, computer science Olympiads. So I had got to a good grade there and they had given me a free pass to university. So that gave me a free pass effective. I was like, I don't care about the exams. I'm going to get into uni anyway. So that gives me six months to work on this product. Uh, so they spent the, all the other time I had aside from being in the taxi stands by actually coding. So I was building the consumer app, the driver app, trying to build a backend. And I realized very quickly that this is going fine, but I'm not that capable of programming that I can build all of these systems as quickly as I need. So then I went to my older brother who was, um, 15 years older and, uh, had a great career in tech and then used to be one of the early employees at Skype back in 2004. So he had a great network of engineers. So I went to him and I was like, we need to get some engineer on board to help him with this. I need a co-founder. And, uh, I think we talked to, it must've been at least 30 people. And every single one of them turned me down because why would you join? It's a 19 year old kid. They have no ex…

AI assessment note: “I started building the app. So I was back then 19.”

Q Johannesburg and Lagos. You are still respectfully a kid from Estonia, from Tallinn, and you've never been there. How do you launch in Johannesburg and Lagos?

A So, um, That was the thing that also really differentiated us, uh, from most companies was that we really always look, uh, extremely pragmatically at everything. And we always try to calculate the rise of everything. And we sort of go from first principles. So the logic for us then was that, okay, what do you really need to get this service off the ground? So effectively you need hundreds of drivers in the ecosystem, but okay, what do you need for that? So we just started running these online ads for drivers and we saw that hundreds of them were signing up because unemployment rates in some of these countries were huge, like 30%. So you had a lot of people who wanted to make extra income. And then we thought, okay, what's next? We, uh, just put up an online ad, um, for, um, hiring the first employee and some young kid from a university signed up as a sort of part-time job. Uh, and then we had this Skype interview with him back in like, 2016 when nobody was hiring remotely. And we told this kid over a video call, Hey, we're gonna figure out a way how to send you a card so you can pay for utilities. So you go and you find an office space and then you start calling all these drivers and start training them. And that's what we did. So this young kid just trusted us for a video call, caught the first office space. We sent him a list of drivers. He called them and got them into the o…

AI assessment note: “He called them and got them into the office and trained them up.”

Q Okay, so we have Johannesburg now, and we have Lagos now, and we're looking at this going, huh, Maybe we underestimated the potential of this business in different parts of the world. Do we have a strategic discussion now in management and say something's changed?

A Absolutely. So we were just monitoring these dashboards every few hours. We couldn't believe what was going on because just the consumers signing up were by the thousands every day. We had these thousands of drivers signing up who wanted to all get on the platform. So very quickly, um, had to update our sort of priors and be like, okay, like this is actually going to be fantastic. Let this This might be the entire company six months from now. And that's exactly what happened. So we, uh, we created the SWOT team of people in Tallinn who identified, okay, what are the next 15 countries in the world where we could replicate the same thing? So we just continued with this original exercise of then taking the next cities from the list, running ads in those places, figuring out which ones have the best traction and just launching there. And we went very quickly from having no presence outside Europe into suddenly being live in about 15 countries. So we launched all the major African economies. We even launched in places like Mexico without having no presence there, not flying anybody there, just full remote launches. And, and almost all of these launches turned out to be fantastically successful.

AI assessment note: “created the SWOT team of people in Tallinn who identified, okay, what are the next 15 countries”

Q What do you do at this time, dude? That is, like, it's unprecedented. The last time was Spanish fucking influenza in 1917. What do you do?

A So COVID was this fascinating episode for us. So we were doing probably two hundred million of ARR. It dropped 85%. Every other company in this industry started laying off people. They laid off 30%, 50%, whatever they thought was, uh, needed to, for the company to survive. So I remember we had this meeting over a call with, with the top management, and we decided we're not gonna do any layoffs. We're gonna be the only company in the industry that's gonna retain all of our people. And it's gonna be a huge gamble because we lost 85% of our revenue. And instead what we're gonna do is that, um, we will, uh, do a salary reduction to everybody for 20%. And then we're just gonna bet the company that in six months this is gonna go by. And because we're the only company that's keeping our team intact, we're gonna accelerate out of this faster than anybody. And we gambled the company on that. And that turned out to be fantastic. So actually the team morale we got from doing zero layoffs while everybody else did. Was just such a massive morale boost that the people like were even opting in that some people did for 30, 40% salary reductions to get us through those next six months.

AI assessment note: “we decided we're not gonna do any layoffs.”

Q What do you do then? You vertically integrate and buy them? You partner?

A My view is that most of these companies Like Waymo over the longterm don't want to operate a car network. It's just too much hassle. So I think that these companies can build what they're fantastic at, which is building the software. And then there's going to be companies like us that actually do the real world operations. Because I think most people underestimate how difficult it is to operate a million or five million or ten million vehicles in hundreds of cities, comply with all the local regulations. Collect the payments, deal with customer support, clean the cars, charge the cars, et cetera. I think it's, it's sort of just a scale of complexity. Most people can't even fathom, and they just hand wave and think it's going to be easy. No, it's not. Not to mention the insurance, the financing, the procurement of the cars, et cetera, et cetera. Like, I think that ride-hailing companies like us are absolutely going to be pivotal for these self-driving car companies to actually go to market.

AI assessment note: “ride-hailing companies like us are absolutely going to be pivotal for these self-driving car companies”

Q Why gross bookings? It doesn't necessarily determine user happiness. So like frequency of trips would suggest actually that I love it and I just can't get enough of it, but they may not be very much. It may be just quickly around the corner or a mile. Why is gross bookings the focus?

A So my view is that it's the only thing that matters. And if you ask me what's our retention, what's our activation rate, what's our CAC, I couldn't tell you really any of those things because I don't think they're real, like they're completely irrelevant. So the only thing that at the end of the day matters is what's your GMB or gross bookings. And, uh, I think it encapsulates everything that is either going good or going bad in the business. And the reason for it is that if you have great retention, then consumers will keep on returning to the platform. They will keep on doing GMB. If they like it, they're gonna be increasing their frequency over time, which means they will do even more GMV, and they will also get their friends to their platform, which will bring even more GMV. So again, that's the only North Star metric we're focused on. And of course, there are some teams who look at CAC, who look at the acquisition rates, you know, who look at the retention rates, but that's a clear secondary priority for us. That's never been the North Star of the business.

AI assessment note: “I think it encapsulates everything that is either going good or going bad”

Q Okay. So then we launch and it's difficult. You're managing demand supply. How does that go?

A So the odd thing about this business from, uh, for us for day one was that it was very asymmetric. So the consumer side was from day one actually very strong. So we got the tone of free media. Um, it was a great story. Everybody hated the tax industry. There was this young kid who tried to fix it. So we got hundreds of initial signups coming into the app. And that was great. But the problem was almost every single one of them got a really bad experience because we barely had any drivers online. So they opened the app. The car similarly didn't show up or the experience wasn't great. And, um, I got really into a panic mode because I was thinking that we can't just waste this first initial burst of demand that came in because if all these people get a bad experience, we're going to get a bad rep and we're never going to be getting out of this, this problem. So I, I did everything I could. So I went on the streets of Tallinn, Signing up taxi drivers one by one again, this time at least I had an app so I could sign them up on the spot. And that turned out to be a very effective sales tactic. What I recommend to more founders nowadays, which is that, uh, you just, I just got in the car and then I just wouldn't go out of the car before the driver had signed up. So I just was like, Hey, like, give me your details. I'll just do the account for you, download the app, set it up. And I was…

AI assessment note: “I went on the streets of Tallinn, Signing up taxi drivers one by one”

Q So what is the future of these marketplaces?

A I think that most of the world has already stabilized. So if you look at this industry, barriers to entry now are so high. The technology getting places like London going requires such a massive need of capital that nobody's willing to fund in this new environment. So, so I don't really think that there's going to be new competitors entering the market in a traditional sense. I think where this innovation is going to come from is new modes of transport. Either it's going to be something new like electric scooters, or it's going to be self-driving cars, but I think it's going to be a completely different thing that's going to shake up the market, not the existing company.

AI assessment note: “where this innovation is going to come from is new modes of transport”

Q So then we had 50. What did we do then?

A So once I had validated the consumer interest and I had got the first drivers on board, I started building the app. So I was back then 19. Uh, everybody else, uh, in my high school was studying for exams. But, uh, I had gotten lucky. I had completed one of these, um, national, um, computer science Olympiads. So I had got to a good grade there and they had given me a free pass to university. So that gave me a free pass effective. I was like, I don't care about the exams. I'm going to get into uni anyway. So that gives me six months to work on this product. Uh, so they spent the, all the other time I had aside from being in the taxi stands by actually coding. So I was building the consumer app, the driver app, trying to build a backend. And I realized very quickly that this is going fine, but I'm not that capable of programming that I can build all of these systems as quickly as I need. So then I went to my older brother who was, um, 15 years older and, uh, had a great career in tech and then used to be one of the early employees at Skype back in 2004. So he had a great network of engineers. So I went to him and I was like, we need to get some engineer on board to help him with this. I need a co-founder. And, uh, I think we talked to, it must've been at least 30 people. And every single one of them turned me down because why would you join? It's a 19 year old kid. They have no ex…

AI assessment note: “I started building the app. So I was back then 19.”

Q What have been your biggest lessons on what works for driver supply and what doesn't?

A At the end of the day, About 80% of volume is, uh, driven organically from word of mouth. So nothing can beat that even today. Uh, and all the paid channels make up about 20% in a mature phase. However, once you start in a market, of course it's different because nobody's heard of your product. So the question really is how do you get the first couple of hundred drivers and the first couple of hundred customers on board? And generally we see the most effective for both sides is a combination of PR. So you always got to get some first launch media and you get some exposure from that. And then the other thing is just paid online ads. And for us, generally, the most effective turned out to be Instagram and Facebook ads. So that's how you get the first couple of hundred people excited. And then after that, your product needs to be great, because if it's not, then all your cohorts will fall to zero. If you actually have a fantastic value proposition, then the other way around, your key factor is going to be positive. You're going to be exponentially growing.

AI assessment note: “the most effective turned out to be Instagram and Facebook ads.”

Q Johannesburg and Lagos. You are still respectfully a kid from Estonia, from Tallinn, and you've never been there. How do you launch in Johannesburg and Lagos?

A So, um, That was the thing that also really differentiated us, uh, from most companies was that we really always look, uh, extremely pragmatically at everything. And we always try to calculate the rise of everything. And we sort of go from first principles. So the logic for us then was that, okay, what do you really need to get this service off the ground? So effectively you need hundreds of drivers in the ecosystem, but okay, what do you need for that? So we just started running these online ads for drivers and we saw that hundreds of them were signing up because unemployment rates in some of these countries were huge, like 30%. So you had a lot of people who wanted to make extra income. And then we thought, okay, what's next? We, uh, just put up an online ad, um, for, um, hiring the first employee and some young kid from a university signed up as a sort of part-time job. Uh, and then we had this Skype interview with him back in like, 2016 when nobody was hiring remotely. And we told this kid over a video call, Hey, we're gonna figure out a way how to send you a card so you can pay for utilities. So you go and you find an office space and then you start calling all these drivers and start training them. And that's what we did. So this young kid just trusted us for a video call, caught the first office space. We sent him a list of drivers. He called them and got them into the o…

AI assessment note: “So we just started running these online ads for drivers and we saw that hundreds”

Q I spoke to Dara at Uber before the show. And he said, I just want to make sure I get this quote right because he might be a little bit upset if I don't get it right. Um, it's something along the lines of essentially you are often second or third in a market. How do you feel about that? And is that an okay strategy?

A Well, I'm a facts and numbers based guy. So let's look at that. Um, today we operate in about 50 countries around the world and Bolt is the number one most popular platform in more than 20 of them. So we're feeling pretty happy about where we are. And when we look at the trend of most of those other 30 where we operate, we are continuously taking share for the reasons I mentioned earlier. Uh, we offer a better value proposition to the customer. We offer better value proposition to the driver. And in a lot of those places, we're confident that over time, even though we're a second mover, uh, we've, we've done that before. I think we can catch up and actually become the most popular platform.

AI assessment note: “So we're feeling pretty happy about where we are.”

Q Okay, so we have Johannesburg now, and we have Lagos now, and we're looking at this going, huh, Maybe we underestimated the potential of this business in different parts of the world. Do we have a strategic discussion now in management and say something's changed?

A Absolutely. So we were just monitoring these dashboards every few hours. We couldn't believe what was going on because just the consumers signing up were by the thousands every day. We had these thousands of drivers signing up who wanted to all get on the platform. So very quickly, um, had to update our sort of priors and be like, okay, like this is actually going to be fantastic. Let this This might be the entire company six months from now. And that's exactly what happened. So we, uh, we created the SWOT team of people in Tallinn who identified, okay, what are the next 15 countries in the world where we could replicate the same thing? So we just continued with this original exercise of then taking the next cities from the list, running ads in those places, figuring out which ones have the best traction and just launching there. And we went very quickly from having no presence outside Europe into suddenly being live in about 15 countries. So we launched all the major African economies. We even launched in places like Mexico without having no presence there, not flying anybody there, just full remote launches. And, and almost all of these launches turned out to be fantastically successful.

AI assessment note: “had to update our sort of priors and be like, okay, like this”

Q Have you seen that play out in reality? And if so, how?

A So, um, we have a couple of these extremely, extremely intelligent people at Bolt who really defined the company. I'd say Oliver, our lead software engineer has been like that, uh, like he was able to do alone, but It took probably 10 average engineers to do the same work, and it was fantastic. The company would have otherwise never, never made it this far. Um, and the other one is our head of growth, uh, Pavel, who's been with the company for about eight years now, who runs all of our global, um, incentive, uh, logic of how exactly do we figure out which is the small subset of customers who's worth targeting, which are the ones that you don't target. Uh, and I think we have one of the most sophisticated frameworks for how do we do marketing of any company I've ever seen.

AI assessment note: “we have a couple of these extremely, extremely intelligent people at Bolt”

Q What did you do in those six months? I mean, we all remember nothing happened.

A So we did a couple of things. So first of all, we optimized every single thing we could in the business. We're already frugal going into COVID, but that really made us like question every single line of the P and L and we, we squeezed everything we could to, to make the business more efficient. The other thing we did was we really started preparing. How do we come out of this much stronger than going into it? So we actually launched the number of markets. So we were setting them up and it was great timing because all these drivers were low utilized. They obviously didn't get trips. There was no traffic going on. So it was very easy for us to get into these countries and sign up to drivers. They, they, they had no other alternative. And then what we did was that these markets.

AI assessment note: “first of all, we optimized every single thing we could in the business.”

Q So what are you nodding? So if you were to add a sixth, what would you like to add as a sixth?

A We have this, uh, wish list of probably 50 ideas that we're looking at every year and thinking about what are we going to launch next year? So the one I'm currently bullish about is, um, uh, dining payments. So in many parts of Europe and the world, it's actually not a common thing that you can scan the QR code, you see the menu, you can order and pay from that. And you can integrate it with everything else we already do in the bold food ecosystem. So you can have a loyalty program there. That's all attached to that. So whatever you do for delivery or you do, uh, uh, the dining experience or any of these sort of foods related services, like they can all tie into a really nice holistic ecosystem. So that's what we want to build next.

AI assessment note: “the one I'm currently bullish about is, um, uh, dining payments.”

Q Why gross bookings? It doesn't necessarily determine user happiness. So like frequency of trips would suggest actually that I love it and I just can't get enough of it, but they may not be very much. It may be just quickly around the corner or a mile. Why is gross bookings the focus?

A So my view is that it's the only thing that matters. And if you ask me what's our retention, what's our activation rate, what's our CAC, I couldn't tell you really any of those things because I don't think they're real, like they're completely irrelevant. So the only thing that at the end of the day matters is what's your GMB or gross bookings. And, uh, I think it encapsulates everything that is either going good or going bad in the business. And the reason for it is that if you have great retention, then consumers will keep on returning to the platform. They will keep on doing GMB. If they like it, they're gonna be increasing their frequency over time, which means they will do even more GMV, and they will also get their friends to their platform, which will bring even more GMV. So again, that's the only North Star metric we're focused on. And of course, there are some teams who look at CAC, who look at the acquisition rates, you know, who look at the retention rates, but that's a clear secondary priority for us. That's never been the North Star of the business.

AI assessment note: “I think it encapsulates everything that is either going good or going bad”

Q Why were you skeptical? Why did you change your mind?

A I started using ChatGPT when it came out more than a year ago, and there was a lot of buzz about it. And all these CEOs were coming out saying like how it's going to transform their companies. And I'm on one hand, a big techno optimist, but on the other hand, I tried it out. I asked a bunch of people in the company, try it out, see how it can optimize the work. And it came back very negatively. Like nobody really thought it's going to move the needle with the exception of one part of customer support. And that was generally the one part that I was thinking, optimistic is going to move. But now I think in a year, my thinking has switched the significantly, both on a personal level and on a company level. So on a personal level, uh, I now use these LLMs basically every day.

AI assessment note: “I asked a bunch of people in the company... And it came back very negatively.”

Q Did the message change to make it successful with those 50 or was it just persistence?

A I think that, uh, I was sort of optimizing the wording a bit as you would in sort of face to face sales because over time I saw what, what was getting me the better conversion. But at the end of the day, the pitch was simple. Uh, you're currently stuck under this legacy taxi company. That doesn't give you a lot of work. That's why you're sitting here in this taxi stand. Uh, you should sign up to this app. It's low risk. You're just gonna be paying a small commission every time you get a trip. So it should be a no brainer. And even then most of the drivers didn't want to sign up. So it was a big uphill climb to make it happen.

AI assessment note: “I was sort of optimizing the wording a bit as you would in sort of face to face sales”