A Product Market Fit Show | Startup Podcast for Founders
00:06:47 · Pablo Srugo, Hermann Frank
[00:06:47] S0: I mean, it was like the railroads, then telecom, then then AI. You know? But it was the AI of twenty years ago. [00:06:51] S1: Yes. And actually when Tim Altman came up with a statement of like, we need to invest $7,000,000,000,000 into AI infrastructure, people thought he's crazy. But if you look at telecom, just in the last ten years, and it's a mature industry, they have invested over $3,000,000,000,000 in infrastructure. So the overall amount invested in telecoms, like in wireless, since since the nineties, is way higher than what Sam said we need to build global AI infrastructure. So for us, it didn't seem that crazy at all. [00:07:22] S0: And so... And just tell me a little bit about the original business, and maybe just the timeline in terms of when you start thinking about Yes. [00:07:29] S1: So we kind of settled on telecom, and then we decided to look into, okay, what might be reoccurring problems in that industry? [00:07:37] S0: And when was this? Just to kind of get... Have that context. [00:07:39] S1: Around 2000, like, 1819. [00:07:42] S0: Okay.
00:07:43 · Hermann Frank
[00:07:43] S1: We kind of looked at everything from like how are networks built, like how are they monetized, like how does distribution work. And the first thing that we saw is, if you look at the world, like only about half of it was really covered with high speed wireless, and the rest was not. It was just like massive wireless gaps, entire continents, parts of countries, even in countries like Germany and The US, you drove like ten minutes outside of a big city, you did not have high speed wireless. And so we like looked at, okay, why is this the case? And it's just like, it's a density problem. Like the the less dense the population is, the higher the costs are. And there isn't this oligopoly of Nokia, Ericsson, and Huawei at the time, and most of the world that still is... They're still running, that if you understood anything about hardware, you saw that they're overpricing these like closed loop systems that they're selling to telcos by about a factor of a 100. Because if you put enterprise wireless hardware next to it, like it was about a 100 times cheaper because they were like open standard white label wireless hardware. And we thought, okay, if we can figure out how to build a cellular grade network with this standardized hardware, not this black box closed loop system, we could make deployment much cheaper, and therefore make high speed wireless much more ubiquitous than this. And that's what we did. Like we we set ourselves a rule, like you need to be able to order this on Amazon or Alibaba, like no other hardware, nothing custom, and then we developed our own software hardware stack. So we put a real telco back end that usually lives on a big data center into a tiny like Intel Mac computer, and we built in our apartment two cells. And then we walk from one to another, and it worked great. Okay. We need a vessel to test this at, like, higher scale at, like, the apartment won't do it.
00:09:39 · Pablo Srugo, Hermann Frank
[00:09:39] S0: And who... When you say we, by way, who... Who's involved at that point? [00:09:42] S1: So it was my cofounder Dennis and I, like... And Dennis and I are now cofounders and gigs. And we had one intern, like a friend that was like helping us. [00:09:51] S0: And are both of you like technical types? [00:09:54] S1: I'd say I'm technical of of the both of us. Like Dennis is... Without Dennis, like, we would not have a a company. Right? Like, I I... We always joke, we could have a 100,000,000 in the bank, and like, nobody wouldn't get paid because I'm just terrible at, like, anything that is administrative and operational. So we complement each other incredibly well. So we then went out and built a WiFi network in Berlin. So we walked from cafe to cafe, and then we asked them if we can deploy this this WiFi inside of their inside of their cafe that basically, like, put together resembled a cellular grade network. And then you could walk through an entire neighborhood and get stay connected on this network. And it worked worked amazingly well. We had like over a million customers in the first year, mostly tourists that were just like roaming around the city. So we basically became the number one public WiFi network in the city. [00:10:50] S0: Did you fundraise for that? [00:10:51] S1: I financed it myself at the point. I'd like... I I was like in it for like to find the one big problem, and then we once we kind of figure it out, and we wanna scale, then the idea was to go out and race. [00:11:02] S0: How many users do you have?
00:11:02 · Pablo Srugo, Hermann Frank
[00:11:02] S0: How many users do you have? [00:11:03] S1: We had over a million users after a year. [00:11:05] S0: Wow. I mean, a million is is actually... It's it's a big number of paid consumer users. Right? Like, did you... [00:11:10] S1: Not paid. Just the... Just users. That's how it came to gigs when we tried to figure out the paid part. [00:11:15] S0: Oh, so you're offering, like, you're on Google Ads or wherever you are, and you're marketing, like, free roaming for tourists sort of thing? [00:11:21] S1: So you can run any type of spectrum on the on the antenna. Right? So we just ran it on WiFi spectrum when we made it into a WiFi network. So when you were somewhere in the vicinity of one of those... [00:11:34] S0: Oh, you would just see an open... You'd see an open WiFi? [00:11:37] S1: Yeah. We'll see a free WiFi, and once you're in one of the nodes, you're you're everywhere. And we would have, like, either a flyer at, like, the cafes that we had these nodes installed in, so we had, like, free Internet basically to backhaul whatever you were consuming, and that was how it spreads. [00:11:55] S0: What's the cost behind that? Like, who's... This is a naive question of somebody who knows very little about telco, like, you're serving this WiFi, what's what's the
00:11:55 · Pablo Srugo, Hermann Frank
[00:11:55] S0: What's the cost behind that? Like, who's... This is a naive question of somebody who knows very little about telco, like, you're serving this WiFi, what's what's the [00:12:02] S1: cost for you? So the hardware, the total cost like to... Per node came out to like a couple $100. And the other kind of hack that we had was that we used the y... Like, we went to a cafe and said like, hey, your your WiFi is not secure. Let us install this very secure WiFi for those and those and those reasons, and you can offer better WiFi experience to your customers, and their WiFi will work in the entire neighborhood. And what we got in return was free Internet access. So we had to pay zero for the actual Internet. Right? [00:12:35] S0: Right. They're the ones paying, like you're piggybacking off of their plan.
00:12:35 · Pablo Srugo, Hermann Frank
[00:12:35] S0: Right. They're the ones paying, like you're piggybacking off of their plan. [00:12:38] S1: Exactly. So we got gigabits upon gigabits of like free backhaul capacity out of it. And then about a year in, we're like, hey, this is clearly working, like, what are we gonna do with it? Like, we had plans to scale this like 10 other markets, and we like tried to figure out how to monetize the best. And the idea was we're only in two neighborhoods, we don't have a desire to like cover the entire city, that that's where like tourists are mostly here. And what we wanted to achieve is that you could buy an eSIM, and when you're within reach of our network, you would roam on our on our network, and we would not pay anything for the traffic. But when you're not, you could roam on a national LTE network. And we were very much expecting to find an API to allow us to provision a device on the network, and then tie in ourselves. So basically, like a Stripe like experience in telecoms, if you will. And it didn't exist. And we were very surprised as to why that doesn't exist. So like that kind of led us down a rabbit hole, and we understood that the industry just like was not able to produce this like developer experience, like this horizontal developer experience that we've kind of came to be used in in pretty much every other industry, like hosting payments. Right? Like, you you just have an API for pretty much everything. And once we understood why, and like, what would happen if we would solve it, we understood, okay, this is the biggest problem we've ever seen.