Bản chép lời đầy đủ của Segment: Peter Reinhardt & His College Roommate And Cofounder

to grow up, actually. 00:02:06.400 --> 00:02:12.159 Like not a lot of cars, just go run around in the fields, catch worms and grasshoppers and whatever else. 00:02:12.560 --> 00:02:16.479 You know, I know it's kind of hard to bring yourself back, but like just in general, what kind of kid were you say you were? 00:02:16.479 --> 00:02:20.000 I was a very, very introverted, very quiet math nerd. 00:02:20.240 --> 00:02:25.199 So, you know, math team, chess team, very into all the nerdary activities there. 00:02:25.199 --> 00:02:34.639 Like in high school, I took calculus freshman year and then was off to the University of Washington, took honors calculus, and then sort of applied math on up to complex analysis and stuff like that. 00:02:34.639 --> 00:02:39.599 So I was like, I was the scrawny little kid in the back of the university classroom trying not to be noticed. 00:02:39.599 --> 00:02:43.759 And so at some point in, I think at some point in middle school, I decided that I wanted to go to MIT. 00:02:43.759 --> 00:02:47.520 And I literally wore an MIT shirt almost every day of middle school and high school. 00:02:47.840 --> 00:02:50.719 Was there any, like, were you exposed to startups at all as a kid? 00:02:50.719 --> 00:02:53.520 Was there any stories of entrepreneurs or anything like that? 00:02:53.759 --> 00:03:00.560 No, I didn't really get startup exposure until maybe freshman year of college with Calvin and Ilya and as roommates. 00:03:00.560 --> 00:03:04.400 And Calvin introduced us to Hacker News and started reading all Paul Graham's essays. 00:03:04.400 --> 00:03:06.000 And that was really my first exposure. 00:03:06.240 --> 00:03:10.479 And at what point did you meet Calvin and Ia, the guys who would end up becoming your co-founders? 00:03:10.800 --> 00:03:13.520 They were my roommates, uh, freshman year. 00:03:13.520 --> 00:03:17.919 So Calvin and I had met each other in math competitions in Washington. 00:03:18.159 --> 00:03:18.960 I'm Calvin. 00:03:18.960 --> 00:03:25.840 I was one of the co-founders and CTO at Segment, and I've known Peter for a long time at this point. 00:03:25.840 --> 00:03:32.479 So Peter and I met actually way back in the day, back in junior high, I want to say. 00:03:32.479 --> 00:03:39.199 We kind of always known each other through the like elementary junior high math circuit, kind of always at a distance. 00:03:39.199 --> 00:03:43.120 Like you'd see the other person's name in the top, I don't know, 10 or 20 people. 00:03:43.360 --> 00:03:48.000 So was it a little bit of like kind of like friendly competition from a distance in the early days? 00:03:48.319 --> 00:03:49.680 Yeah, I'd say so. 00:03:49.680 --> 00:03:54.240 Uh I frankly, I don't think Peter and I were particularly close back then. 00:03:54.240 --> 00:03:59.199 But where we really got a lot closer was when both of us ended up getting into MIT. 00:03:59.680 --> 00:04:02.159 So we kind of met up at pre-frosh weekend. 00:04:02.159 --> 00:04:07.919 Calvin and I were wandering around campus, just like exploring it, checking out dorms that we thought might be interesting. 00:04:07.919 --> 00:04:09.599 So we walked to a dorm that we liked. 00:04:09.599 --> 00:04:13.520 We went to a random floor and we heard some voices, and we walked into a random room. 00:04:13.520 --> 00:04:16.399 And Ilya was there hanging out with someone else from his school. 00:04:16.399 --> 00:04:18.319 And so that was how we met Ilya. 00:04:18.319 --> 00:04:20.079 And we hit it off. 00:04:20.079 --> 00:04:24.399 And we spent the rest of that weekend together, and then we decided that we wanted to be roommates in the fall. 00:04:24.399 --> 00:04:27.439 So it was a super, super kind of lucky happenstance. 00:04:27.439 --> 00:04:29.920 So he was like the most normal person that we'd met. 00:04:29.920 --> 00:04:32.319 He was like a very Bebley friendly nerd. 00:04:32.319 --> 00:04:33.680 Everyone else was so weird. 00:04:33.680 --> 00:04:37.279 Like the freshman, like MIT is a great normalizing function, actually. 00:04:37.279 --> 00:04:41.120 Like people come in very weird and they leave like only somewhat weird. 00:04:41.120 --> 00:04:50.319 And you know, everyone else was like LARPing or like, I don't know, like obsessed with some obscure thing, which is very cool, but uh was hard to like, hard to make friends. 00:04:50.319 --> 00:04:55.920 And he struck us as by far the most normal person that we met that day, which if you know it'll he is maybe saying something. 00:04:56.319 --> 00:04:57.920 What would you say college Peter was like? 00:04:57.920 --> 00:05:02.079 Kind of different from maybe growing up or high school Peter, was it, you know, mostly focused on classes? 00:05:02.079 --> 00:05:04.000 Do you manage to have some fun on the side? 00:05:04.319 --> 00:05:07.759 Yeah, I'd say it was somewhat focused on classes, but I was on a floor that was pretty rowdy. 00:05:07.759 --> 00:05:16.240 I was definitely like the studious kid on the floor was Burton Third Bombers, known for partying pretty hard, which wasn't really my vibe, but I was sort of on the edge of that. 00:05:16.240 --> 00:05:19.759 You know, at the end of freshman year, they give out these awards to each freshman. 00:05:19.759 --> 00:05:21.839 Every freshman gets an award, which is like a funny name. 00:05:21.839 --> 00:05:30.079 And I had my appendix out and said the award was to Peter for um, you know, they successfully removed your appendix, but the stick up your ass was inoperable. 00:05:30.079 --> 00:05:31.920 So that was the tension. 00:05:31.920 --> 00:05:37.600 I was on the party floor and enjoying it to some extent, but I was still pretty uptight, I guess, as far as that floor was concerned. 00:05:37.920 --> 00:05:39.600 That checks out, that checks out. 00:05:39.600 --> 00:05:40.160 Uh yeah. 00:05:40.160 --> 00:05:47.199 The floor itself was definitely quite rowdy, and Peter was probably the one who was like most diligent when it came to

studies. 00:05:47.199 --> 00:05:51.439 It's funny, I feel like Peter is one of those people who's very much an old soul. 00:05:51.439 --> 00:06:00.399 He's kind of always known what he wants to get out of life, and in general, he's a very ambitious and I'd say structured and analytical thinker. 00:06:00.399 --> 00:06:13.759 I remember one of the funny stories while all three of us were roommates together freshman year, I had ordered this little like charcoal grill off the internet because I was really excited about putting a little grill out the window and like making hot dogs. 00:06:13.759 --> 00:06:26.000 I remember both Ilya and I got into this argument related to the grill, where Ilya was worried that the grill would be somehow overnight like off-gassing a certain amount of uh CO2. 00:06:26.000 --> 00:06:42.720 And so I remember Peter actually did the math on the diffusion rates of inert carbon from this charcoal to say that it would probably kill us in something like 700 years or something like that, of sleeping in the same room with the charcoal grill. 00:06:42.720 --> 00:06:45.680 So I I think that's sort of the like flavor of Peterism. 00:06:45.680 --> 00:06:50.879 Like he goes deep from first principles to understand what kind of the real ground truth is. 00:06:50.879 --> 00:06:56.879 And I think that's something that he's just like always done and continues to do, even to this day, with his various businesses. 00:06:57.360 --> 00:06:58.319 That's hilarious. 00:06:58.319 --> 00:07:04.959 Um, do you remember when Peter first got interested in tech versus what he was studying? 00:07:05.279 --> 00:07:05.519 Yeah. 00:07:05.519 --> 00:07:11.439 So Peter that first year, like both of us were kind of tinkering around with a lot of like slight software projects. 00:07:11.439 --> 00:07:20.319 But I think he first grew a lot more serious about kind of like building a startup when he and Ilya took this class called uh the founder's journey. 00:07:20.639 --> 00:07:28.720 Yeah, beginning of junior year, we started to talk about like, oh, it would be fun to start something together because we were reading these Paul Graham essays. 00:07:28.720 --> 00:07:30.639 Calvin in particular had put us on that. 00:07:30.639 --> 00:07:34.399 And then Ilya and I took a class, Founder's Journey. 00:07:34.399 --> 00:07:40.399 And the concept of the class was just a founder comes in, talks about what they built and what their story was. 00:07:40.399 --> 00:07:46.240 And the very first week was Adam Smith, who had founded Zobni in box backwards. 00:07:46.240 --> 00:07:51.040 Yeah, he just told us his story how he raised like almost 50 million dollars and like built this really cool company. 00:07:51.040 --> 00:07:57.759 And I was like just totally in shock of like, wow, this person is so impressive, and which he was. 00:07:57.759 --> 00:08:03.199 And afterwards, Ilya was like, you know, let's just go talk to him, let's ask him if he wants to come back to the dorm for

beers. 00:08:03.199 --> 00:08:04.959 And I was like, Don't do that, that's embarrassing. 00:08:04.959 --> 00:08:05.839 Why would he do that? 00:08:05.839 --> 00:08:10.000 He's so like there's never no chance in hell that he's gonna come and want to hang out with us. 00:08:10.000 --> 00:08:11.600 And Ilya's like, screw it, I'm doing it. 00:08:11.600 --> 00:08:15.759 And I was like, Oh my god, like nervous in the corner, feeling feeling shameful. 00:08:15.759 --> 00:08:18.560 And Ilya goes in the ass, Adam, and Adam's like, yeah, sure. 00:08:18.800 --> 00:08:24.240 And I remember just all three of us like hanging out with this guy who, in all respects, seemed like a very normal guy. 00:08:24.240 --> 00:08:28.079 Like he could have been just like a older upperclassman or something. 00:08:28.079 --> 00:08:31.759 But I think that was kind of the first moment where it's like, oh, maybe we like can do this, you know. 00:08:32.000 --> 00:08:36.480 It was just like such a humanizing, normal guy kind of experience. 00:08:36.480 --> 00:08:38.639 I was like, maybe I could do this too. 00:08:38.639 --> 00:08:44.799 You know, it like took this thing that we had put on the pillar of like founding company and just like brought it to like a very digestible human level. 00:08:44.799 --> 00:08:49.279 And so I think that's probably around the time that we were like, okay, we can do this. 00:08:49.279 --> 00:08:55.440 I think then it like really started percolating and our like confidence that we could do it was pretty transformed. 00:08:55.440 --> 00:09:01.759 And then MIT has this thing, January is like an independent activities period where you can just do something of your own kind of design. 00:09:01.759 --> 00:09:06.960 And so we converted my dorm room into a little hacker space and set up a bunch of desks in there. 00:09:06.960 --> 00:09:09.440 I think we had like six desks in there, yeah. 00:09:09.440 --> 00:09:11.200 In a tiny, tiny dorm room. 00:09:11.519 --> 00:09:14.559 He basically converted his dorm room, which already is not very big. 00:09:14.559 --> 00:09:17.519 It's like, I don't know, 170 square feet, something like that. 00:09:17.519 --> 00:09:30.639 He converts it into a hackerspace where there are just now instead of like his one desk, there's kind of like a desk with like two chairs there and like maybe a separate desk under his bed, which he like lofts a bit. 00:09:30.639 --> 00:09:38.799 At any given time, there were probably like three or four people hanging out in Peter's room working on various ideas where it's like, okay, we're just gonna like hack on ideas. 00:09:39.039 --> 00:09:41.279 And through that whole month, we're just like building stuff. 00:09:41.279 --> 00:09:46.559 You know, by then the wheels were in motion, I think that would cause us to leave school, go to Y Combinator, all those things. 00:09:46.799 --> 00:09:48.799 So, yeah, I mean, let's start talking about some of the ideas. 00:09:48.799 --> 00:09:53.279 I mean, the first idea I don't think is the one that you ended up on, but what was the first idea and how'd you come up with it? 00:09:53.600 --> 00:09:56.480 First idea was bookstore, uh, book XOR. 00:09:56.480 --> 00:10:06.320 The idea was to give professors a place to host their like class materials, and then they could kind of get analytics back on like, you know, which parts of the doc students were reading. 00:10:06.639 --> 00:10:16.720 And the whole pitch for it was that students or professors could upload PDFs, typically like problem sets or class materials, and then anyone could comment on them. 00:10:16.720 --> 00:10:22.639 So uh it'd be the kind of thing where it's like, oh, we're both studying something, like I'll comment on it, you comment on it, we can like highlight it. 00:10:22.639 --> 00:10:25.279 Maybe we can start collaborating in terms of our studying. 00:10:25.600 --> 00:10:28.320 This has a product market fit problem, and the professors don't give a shit. 00:10:28.320 --> 00:10:33.279 And so, you know, we realized that, and then we pivoted to well, maybe they care about their lectures, right? 00:10:33.279 --> 00:10:40.080 Because like lectures are a lot of work, and maybe we could give students a way to ask questions because they feel kind of nervous and they could ask things anonymously, et cetera. 00:10:40.080 --> 00:10:42.000 So we we thought that would be interesting. 00:10:42.399 --> 00:10:49.200 We had been reading all of Paul Graham's essays, and kind of his chief advice was to build a problem that you understand. 00:10:49.200 --> 00:10:54.320 And so we said, okay, we're students, let's go try and like solve a student problem. 00:10:54.320 --> 00:11:10.639 And so one of the problems that we uh or the problem that we eventually landed on uh was this idea where the professor says something in the middle of a big lecture hall and it kind of goes over your head and you like look around and everyone else is confused, but the professor keeps on going. 00:11:10.639 --> 00:11:14.799 And we were like, okay, this like seems like a problem, like maybe there's a software fix for it. 00:11:14.799 --> 00:11:20.240 And so we went around and we interviewed a bunch of professors to see like whether this problem resonated. 00:11:20.240 --> 00:11:26.639 And one of the professors who we talked to uh was actually Robert Morris, an MIT professor who's involved in Y Combinator. 00:11:26.799 --> 00:11:32.080 You know, we we got his feedback on it and he thought it was a good idea, or at least he told us he thought it was a good idea. 00:11:32.080 --> 00:11:35.840 So that became class metric, and that's what we got into Y Combinator with. 00:11:36.080 --> 00:11:38.240 Do you remember, by the way, your your YC interview? 00:11:38.240 --> 00:11:39.360 Do you remember the details behind that? 00:11:39.360 --> 00:11:41.679 Was it like easy breezy, or did you get grilled? 00:11:41.679 --> 00:11:42.879 Like, how did that go? 00:11:43.360 --> 00:11:51.919 Uh,

we got grilled, but there was a funny dynamic where Paul Graham was like, you know, I always wanted something like this. 00:11:51.919 --> 00:11:53.919 And so I think that's what carried the day. 00:11:53.919 --> 00:11:59.120 But then after he said that, he turned to the YC partner, who's also an IT professor, and was like, Hey, what do you think? 00:11:59.440 --> 00:12:04.720 He looks to his left over at Robert Boris and he says, Oh, Robert, you're a teacher. 00:12:04.720 --> 00:12:06.480 Like, would you use this? 00:12:06.480 --> 00:12:10.639 And Robert Morris is just like, nah, probably not. 00:12:13.519 --> 00:12:17.360 We're like, bro, like we like we asked you when. 00:12:17.360 --> 00:12:19.600 So that was a real gut punch in the middle of the interview. 00:12:19.600 --> 00:12:20.879 It didn't matter in the end, I guess. 00:12:20.879 --> 00:12:25.360 And that's what we tried to build for like the first four months, like through all of my combinator. 00:12:25.360 --> 00:12:26.639 We raised like 600k. 00:12:26.639 --> 00:12:33.039 And then as we deployed it in the fall, as the classes started, it was just a total disaster. 00:12:33.039 --> 00:12:43.200 Like all the students opened the thing, they spent two seconds on our class metric application, and then they immediately switched over to browsing the web, buying stuff on eBay, checking their email, whatever. 00:12:43.200 --> 00:12:44.639 So it was it was disastrous. 00:12:44.639 --> 00:12:49.519 And it was pretty clear within like a week that having students have laptops in classrooms was not a good idea. 00:12:49.519 --> 00:12:51.039 And so that was the end of that. 00:12:51.519 --> 00:12:54.399 So, I mean, so so talk about that decision for a second. 00:12:54.399 --> 00:12:55.840 So you said that that's the end of that. 00:12:55.840 --> 00:13:00.879 I mean, you could have you could have just as easily gone back and said, we just had to work harder at it, we have to figure something else out. 00:13:00.879 --> 00:13:04.559 Like, walk us through the decision to just say, hey, this isn't working and let's move on. 00:13:05.440 --> 00:13:05.919 I don't know. 00:13:05.919 --> 00:13:06.879 It was pretty obvious. 00:13:06.879 --> 00:13:08.000 Like, I don't know that there was that much. 00:13:08.000 --> 00:13:09.440 It was just like it was so bad. 00:13:09.440 --> 00:13:12.399 Like, literally, we were apologizing to the professors afterwards. 00:13:12.399 --> 00:13:15.759 So it wasn't like, oh, the product just needs to be improved. 00:13:15.759 --> 00:13:20.559 It was like the premise of like laptops in the classroom is like not a good one. 00:13:20.559 --> 00:13:26.159 And like I think we've just emailed all the professors and we're like, you should ban devices from your classrooms. 00:13:26.159 --> 00:13:27.840 Um, that's our takeaway. 00:13:28.159 --> 00:13:31.440 Don't don't just stop using our product, but but get rid of all devices. 00:13:32.480 --> 00:13:38.879 If we'd been better founders, we probably would have tried to find something that worked with the current idea. 00:13:38.879 --> 00:13:42.159 Because frankly, class metric wasn't like a total disaster. 00:13:42.159 --> 00:13:48.559 But at the time, we were just so burnt out on this, and we're like, oh man, we like kind of suck at this. 00:13:48.559 --> 00:13:51.759 We're not good founders, we don't understand the market. 00:13:51.759 --> 00:13:54.720 We want to switch off education entirely. 00:13:54.720 --> 00:13:57.600 And this was actually like kind of a contentious thing. 00:13:57.600 --> 00:14:01.360 I think it was definitely a little rough before we kind of landed on our next idea. 00:14:01.360 --> 00:14:13.279 Ilya, Ian, uh, the fourth co-founder who we had brought on for design, and I were all like a little bit burnt out on this idea of trying to sell to college professors, and Peter was still kind of in on it. 00:14:13.279 --> 00:14:18.320 And eventually we we kind of just voted and we were like, okay, we're gonna like try new ideas. 00:14:18.639 --> 00:14:25.360 We moved on to analytics, which is like a bit of a honeypot or a tar pit idea, I think, for founders who are lost. 00:14:25.360 --> 00:14:31.120 And you know, one one investor asked for their money back, so we wired that back, or two, two, actually. 00:14:31.120 --> 00:14:34.480 The rest were like, yeah, we invested for the team, like have fun, pivot. 00:14:35.039 --> 00:14:37.600 Have you talked to the two investors that asked for their money back? 00:14:37.919 --> 00:14:38.639 No, I haven't. 00:14:38.639 --> 00:14:38.960 Yeah. 00:14:38.960 --> 00:14:40.320 Big bummer for them though. 00:14:40.559 --> 00:14:42.320 I kind of wonder what they're thinking. 00:14:42.799 --> 00:14:43.360 Yeah, yeah. 00:14:43.360 --> 00:14:54.879 One of them had 200k in, which I think would have been yeah, somewhere in the like 40 to 80 million returned in the end. 00:14:54.879 --> 00:14:55.919 So whoops. 00:14:55.919 --> 00:14:56.879 Whoops. 00:14:58.639 --> 00:14:59.840 We'll be right back. 00:14:59.840 --> 00:15:07.440 And now, back to the interview. 00:15:07.440 --> 00:15:11.440 So analytics, it's not working great. 00:15:11.440 --> 00:15:13.200 Like, what wasn't working about it? 00:15:13.600 --> 00:15:16.799 It turns out that the existing analytics tools are like pretty good. 00:15:16.799 --> 00:15:19.440 So we just like never really achieved product market fit. 00:15:19.440 --> 00:15:21.759 We always thought we were like one feature, what iteration away. 00:15:21.759 --> 00:15:24.799 I guess unlike the classroom tool, the signals weren't quite so obvious. 00:15:24.799 --> 00:15:28.799 And so we actually did spin our wheels for a long time, like over a year. 00:15:28.799 --> 00:15:29.360 Wow. 00:15:29.360 --> 00:15:31.679 Um, like trying to grind through that. 00:15:32.080 --> 00:15:34.080 So you're working on it, you're grinding away. 00:15:34.080 --> 00:15:36.080 I like like what do those conversations look

like? 00:15:36.080 --> 00:15:39.519 It's I mean, you just you basically you're obviously talking to investors, they know it's not really working. 00:15:39.519 --> 00:15:40.720 Inside, you know it's not working. 00:15:40.720 --> 00:15:42.159 Like, what are the conversations like? 00:15:42.480 --> 00:15:43.919 We weren't really talking to our investors. 00:15:43.919 --> 00:15:48.879 We were scared to go talk to our investors because like we didn't really have anything impressive to show. 00:15:48.879 --> 00:15:56.159 Users like weren't you know, it's like very rare for someone to come to the website because we weren't launching anything, we weren't writing blog posts. 00:15:56.159 --> 00:16:00.720 We were just kind of like grinding, trying to get to a product that we thought was impressive before we would take it out. 00:16:00.720 --> 00:16:03.200 So we were like afraid to launch, which is a terrible idea. 00:16:03.200 --> 00:16:07.440 We were just kind of like in our hidey hole in our apartment working for a long time, which is bad. 00:16:07.440 --> 00:16:10.399 And you know, it was starting to feel kind of dire. 00:16:10.720 --> 00:16:16.080 So I think for most startups, you can kind of like go along for a while and like pretend that you're doing well. 00:16:16.080 --> 00:16:17.679 And that was the case for us. 00:16:17.679 --> 00:16:22.639 Like, we were like, oh yeah, we're like one feature away from hitting product market fit. 00:16:22.639 --> 00:16:30.000 And I think at the end of the day, well, obviously we didn't know anything about like finding product market fit if we've been honest with ourselves about it. 00:16:30.000 --> 00:16:37.919 Uh, I think by the time that we got to December of 2012, frankly, it was this kind of like elephant in the room that we couldn't ignore anymore. 00:16:37.919 --> 00:16:43.440 And we finally like looked at it and said, like, wow, we are just not very good at this. 00:16:43.440 --> 00:16:45.759 Uh, we basically burned all this capital. 00:16:45.759 --> 00:16:47.840 We have almost nothing to show for it. 00:16:47.840 --> 00:16:50.080 We have no active users, no one who actually cares. 00:16:50.320 --> 00:17:01.679 And I think there's like a the Airbnb story, I think, is like poison for founders in this situation because they did grind for like two years or something, and eventually they took off. 00:17:01.679 --> 00:17:06.319 And that is just like, I don't know anyone else who's ever had a founding story like that. 00:17:06.319 --> 00:17:08.559 Like, it's just it's not how it works. 00:17:08.559 --> 00:17:11.200 Airbnb is the like exception that proves the rule. 00:17:11.200 --> 00:17:15.200 And like I got I got married that fall, and we went off on our honeymoon. 00:17:15.200 --> 00:17:19.839 And I was like, through this period, I was like going through panic attacks, like in the in the hospital. 00:17:19.839 --> 00:17:20.720 It was bad. 00:17:20.720 --> 00:17:23.359 Um, it's like, I think this thing is like

not gonna work. 00:17:23.519 --> 00:17:26.000 And I remember that being like a pretty dark time. 00:17:26.000 --> 00:17:32.240 Like all four of us, like normally our mode would be like, oh, you wake up in the morning, you start coding, and you like start figuring out what to do. 00:17:32.240 --> 00:17:34.400 That week, like we basically didn't code at all. 00:17:34.400 --> 00:17:40.720 We just kind of like sat on the couch and argued with each other about what we should do next. 00:17:40.720 --> 00:17:48.720 Because I don't know, it's funny how hard a startup is, like for someone who is used to flying very high. 00:17:48.720 --> 00:17:59.599 And I think all of us came from a background where it's like, oh, you do very well in school, you like study the material, you understand how to pass the tests, you pass them, like you're kind of always just set up for success in that way. 00:17:59.599 --> 00:18:01.519 Whereas startups are like totally different. 00:18:01.519 --> 00:18:05.599 It's like if you don't do something that users care about, you're done. 00:18:05.599 --> 00:18:13.039 And I think it was it was a very humbling experience for us where it's like, oh, it doesn't matter that we're like these hot shot kids from MIT or whatever. 00:18:13.039 --> 00:18:15.759 Like we were just bad at making things people wanted. 00:18:16.079 --> 00:18:33.279 And eventually Ian and Calvin, mostly Ian, I think, were like, you know, I think that there's a big idea behind this little open source library, analytics.js, that we built like a year and a half prior for ourselves to distribute data between different analytics tools from class metric. 00:18:33.279 --> 00:18:38.640 And so they, you know, they had like an analytics.js manifesto that they wrote as like a GitHub markdown. 00:18:38.640 --> 00:18:39.920 And we debated that. 00:18:39.920 --> 00:18:41.039 I thought it was a terrible idea. 00:18:41.039 --> 00:18:42.400 I tried to kill it. 00:18:42.400 --> 00:18:47.200 Uh, and then, you know, tried to kill by convincing everybody to just like launch it on Hacker News and see what people think. 00:18:47.200 --> 00:18:49.200 And I thought that'll kind of kill it and make it go away. 00:18:49.200 --> 00:18:52.160 We had maybe like five or six months of runway left. 00:18:52.160 --> 00:18:55.759 Uh, and so we we put it on Hacker News and it just like blew up. 00:18:55.759 --> 00:19:02.319 So that was the that was the product market fit moment, and then we were off to the races trying to build that out as fast as we could. 00:19:02.720 --> 00:19:04.160 Do you remember the comments on Hacker News? 00:19:04.160 --> 00:19:08.480 Like, was it all positive or was it, you know, there's typically uh, you know, also a bunch of haters. 00:19:08.480 --> 00:19:10.559 Yeah, that's a good question. 00:19:10.799 --> 00:19:14.160 I remember being like mostly positive, which was kind of surprising. 00:19:14.160 --> 00:19:19.680 Uh like I just pulled it up 326 points on December 12th,

2012, 12, 12, 12. 00:19:19.680 --> 00:19:21.759 This is a cool idea for a project. 00:19:21.759 --> 00:19:24.480 It was like shockingly positive for a hacker news thread. 00:19:24.480 --> 00:19:28.240 And we got like thousands of stars on GitHub and thousands of email signups. 00:19:28.240 --> 00:19:30.319 Like it was pretty shocking. 00:19:30.720 --> 00:19:34.559 So our story actually is maybe somewhere in the middle, which is kind of funny, but like totally different. 00:19:34.559 --> 00:19:36.799 We um we were working on Weebly. 00:19:36.799 --> 00:19:38.720 It just kind of wasn't working out. 00:19:38.720 --> 00:19:41.200 Um, you know, it's the same thing, like one more feature. 00:19:41.200 --> 00:19:44.319 And then we came to the conclusion that it was a dead idea. 00:19:44.319 --> 00:19:46.480 So we tried something else. 00:19:46.480 --> 00:19:48.720 Um, we tried a MySpace profile creator. 00:19:48.720 --> 00:19:52.799 Then we tried an ad customizer, like an ad optimizer. 00:19:52.799 --> 00:19:55.200 But in the meantime, we're like, hey, let's just not shut this thing down. 00:19:55.200 --> 00:19:56.559 I actually don't think I've ever told this story. 00:19:56.559 --> 00:19:58.000 We're like, let's just leave it online. 00:19:58.000 --> 00:20:02.799 And somehow, about 12 to 18 months later, the thing just starts taking off on its own, completely out of nowhere. 00:20:02.799 --> 00:20:06.160 And we're like, hey, you know, maybe this isn't such a bad idea after all. 00:20:06.960 --> 00:20:08.240 Wow, that's cool. 00:20:08.559 --> 00:20:10.799 Um, but um, I mean, did you have a plan? 00:20:10.799 --> 00:20:13.680 Like, if that didn't work, what what what were you guys gonna do? 00:20:14.160 --> 00:20:16.240 Uh we were discussing all kinds of different ideas. 00:20:16.240 --> 00:20:18.400 We'd had uh like office hours with PG. 00:20:18.400 --> 00:20:20.240 We'd given him a bunch of different ideas. 00:20:20.240 --> 00:20:23.519 He was like, So he was just like, that's a terrible idea. 00:20:23.519 --> 00:20:25.599 So yeah, we were we were kind of at wit's end. 00:20:25.599 --> 00:20:27.279 Like it was kind of the end of the road. 00:20:27.680 --> 00:20:29.039 Do you remember that conversation? 00:20:29.039 --> 00:20:31.359 Like like where you were, like were you guys all sitting together? 00:20:31.359 --> 00:20:32.400 Was it pretty heated? 00:20:32.400 --> 00:20:33.599 Like how did how did that go? 00:20:33.920 --> 00:20:38.079 Yeah, it was in our apartment slash office on Russian Hill at Thousand Chestnut. 00:20:38.079 --> 00:20:40.720 That was the argument about Analytics.js. 00:20:40.720 --> 00:20:43.680 And then that was also where we were sitting all day when it was blowing up. 00:20:43.680 --> 00:20:47.039 And so we kept we kept building out of that apartment for another six months. 00:20:47.359 --> 00:20:47.599 Yeah. 00:20:47.599 --> 00:20:48.400 How how did that? 00:20:48.400 --> 00:20:51.119 I mean, so you so you post on Hacker News

and then it just blows up. 00:20:51.119 --> 00:20:53.599 Was that the moment you're like, wait, this is where this is gonna be it? 00:20:53.920 --> 00:20:55.680 I mean, it was clear that we just needed to go work on that. 00:20:55.680 --> 00:20:58.640 So we just like pretty much immediately started working on that. 00:20:58.640 --> 00:21:02.319 But we we we almost took like a holiday from vision setting. 00:21:02.319 --> 00:21:10.640 Like we were just like, we were so exhausted of just like, oh, here's a vision, let's go build the vision that we were just like, I don't know, man. 00:21:10.640 --> 00:21:13.839 Like, let's just build whatever what are people asking for? 00:21:13.839 --> 00:21:16.240 People want more integrations, people want a hosted version. 00:21:16.240 --> 00:21:17.279 Cool, let's just do that. 00:21:17.279 --> 00:21:22.720 Just like an absolute state of exhaustion and like emotional, like end of the road. 00:21:22.720 --> 00:21:27.440 And so that's kind of what we did for like six months is we like we tried not to have a vision. 00:21:27.440 --> 00:21:30.160 We just tried to build exactly whatever people were asking for. 00:21:30.160 --> 00:21:33.359 And then over time, the vision you know solidified. 00:21:34.160 --> 00:21:43.680 I think the moment that we got even like a glimmer of traction, it felt like enough of a rope that we should start pulling on it. 00:21:43.680 --> 00:21:45.839 And so that's effectively what we did. 00:21:45.839 --> 00:21:54.559 I think for me, I came out shell-shocked enough that I didn't even think this would be a real business until like eight months in when we signed our first enterprise contract. 00:21:54.559 --> 00:21:59.839 But I think for all of us, it was like, okay, there's at least something working here, which is more success than we've ever had today. 00:21:59.839 --> 00:22:00.640 Date. 00:22:00.640 --> 00:22:05.599 So we might as well keep seeing where it goes and figure out if there's a way to get to a bigger, bigger business. 00:22:05.839 --> 00:22:09.440 But it was a it was a very different motion and it was kind of a cool feeling, actually. 00:22:09.440 --> 00:22:16.319 Like product market fit for us almost felt like we'd been pushing this boulder uphill for a really long time, and but we were in control of it. 00:22:16.319 --> 00:22:21.519 And then it was sort of felt like the market like took control of the boulder and it started rolling downhill. 00:22:21.519 --> 00:22:23.680 And on the one hand, it was very exciting. 00:22:23.680 --> 00:22:28.240 And on the other hand, it was like, oh my God, like I don't, I don't really control this thing anymore. 00:22:28.240 --> 00:22:33.200 Like it's not where I want to go, it's like where the market wants to go, which was also like a feeling of loss. 00:22:33.200 --> 00:22:39.599 So it took like months for us to like kind of emotionally get realigned and to like, okay, this is actually exciting and this is cool and promising. 00:22:39.839 --> 00:22:40.799 What so you said it? 00:22:40.799 --> 00:22:42.799 I mean, when you launched it, you had six months of runway left. 00:22:42.799 --> 00:22:44.079 How low did you get? 00:22:44.079 --> 00:22:47.920 I think we got down to four weeks, maybe three weeks. 00:22:48.880 --> 00:22:49.200 Yeah. 00:22:49.200 --> 00:22:52.079 And then called our investors and was like, hey, I think it's time for us to raise. 00:22:52.079 --> 00:22:52.880 And they're like, oh, cool. 00:22:52.880 --> 00:22:53.839 Like, how's it going? 00:22:53.839 --> 00:22:55.920 Like the update, we're like, how much cash do you have left? 00:22:55.920 --> 00:22:57.200 I'm like, oh, like three or four weeks. 00:22:57.200 --> 00:23:00.319 And they're like, holy shit, like, you got a fundraise now. 00:23:00.319 --> 00:23:01.440 Like, what is going on? 00:23:01.440 --> 00:23:06.000 And then, yeah, from there it was, yeah, truly off to the races. 00:23:06.480 --> 00:23:10.720 We got, I think at one point in time we had like three weeks of cash left too outside of YC. 00:23:10.720 --> 00:23:15.680 You know, at that point you start making contingency plans about how you're going to pay rent next month because it's coming up. 00:23:15.680 --> 00:23:18.559 It's like next month, and you guys have those kind of conversations. 00:23:19.200 --> 00:23:20.559 Uh I was very naive. 00:23:20.559 --> 00:23:22.880 We were just like running out of wall. 00:23:22.880 --> 00:23:24.640 Yeah. 00:23:25.759 --> 00:23:26.720 Oh, that's awesome. 00:23:26.720 --> 00:23:27.519 So yeah. 00:23:27.519 --> 00:23:33.279 So that so then at that point, I mean, if you raise money, you got that pressure off your back, how quickly did it start taking off? 00:23:33.759 --> 00:23:34.720 Not that fast. 00:23:34.720 --> 00:23:39.759 Like we basically grew linearly forever in terms of number of users, which is very strange. 00:23:39.759 --> 00:23:41.119 I still don't understand why. 00:23:41.119 --> 00:23:47.119 And what we figured out over time was a lot of our growth was driven by moving into larger accounts over time. 00:23:47.119 --> 00:23:55.440 Um, so like bigger companies instead of, you know, a small startup paying $20 a month, like in the end, Procter and Gamble paying like $7 million a year. 00:23:55.440 --> 00:24:04.720 And so, like, there was a lot of that kind of shift in account size over time, which was both bigger companies and larger contract value because we were grossly undercharging. 00:24:04.720 --> 00:24:06.160 We were terrified to ask for money. 00:24:06.160 --> 00:24:09.920 And in the early days, customers would even email and be like, hey guys, you got to charge more. 00:24:09.920 --> 00:24:12.480 I like don't trust that you're gonna be around that long. 00:24:12.480 --> 00:24:14.480 Can you please like you should raise the price? 00:24:14.799 --> 00:24:17.680 Do you remember the first moment you're like, this is gonna be big? 00:24:17.680 --> 00:24:19.440 Like, this is really working.

00:24:20.559 --> 00:24:27.839 I think probably the first time we hired a new sales advisor and I was like struggling to, you know, figure out how to do enterprise sales. 00:24:27.839 --> 00:24:33.680 And we went to meet with Nat at Zamarin, Nat Friedman, like real like Silicon Valley stalwart. 00:24:33.680 --> 00:24:40.559 And anyway, he was running his company Zamarin at the time, and they were using us, but not paying very much, $50 a month or something. 00:24:40.559 --> 00:24:44.160 And so we were going to a sales meeting, we were gonna upsell them onto an enterprise plan. 00:24:44.160 --> 00:24:50.799 And our sales advisor was like, Peter, in this conversation, you have to ask for $120,000 a year. 00:24:50.799 --> 00:24:53.440 And I was like, I don't know if I can do that. 00:24:53.440 --> 00:24:54.160 Like, that's crazy. 00:24:54.160 --> 00:24:57.039 We're chart, I think I think actually we're charging them $120 a year. 00:24:57.039 --> 00:24:59.839 And so it was like a thousand X increase in price. 00:24:59.839 --> 00:25:01.359 I was like, I don't think I can do that. 00:25:01.359 --> 00:25:03.440 And he's like, like, what is your sales advisor then? 00:25:03.440 --> 00:25:06.799 And he started walking away and I was like, whoa, like, okay, I'll ask, I'll ask. 00:25:06.799 --> 00:25:09.119 So we sat down and got to the end of the conversation. 00:25:09.119 --> 00:25:10.319 Now it's like this really compelling. 00:25:10.319 --> 00:25:11.759 Like, you know, how much does the service cost? 00:25:11.759 --> 00:25:14.319 And I was like, you know, 120,000 a year. 00:25:14.319 --> 00:25:18.559 And I turned just like completely beat red, just like bright, pulsating red. 00:25:18.559 --> 00:25:22.559 And he kind of looked at me and he's like, Well, how about $12,000 a year? 00:25:22.559 --> 00:25:24.240 And I was like, How about 18? 00:25:24.240 --> 00:25:26.480 And it's like, that's fine. 00:25:26.480 --> 00:25:29.759 And I walked out of that meeting and I was like very embarrassed. 00:25:29.759 --> 00:25:35.519 And then also was like, Oh my God, the value that we're creating is just way higher, you know? 00:25:35.519 --> 00:25:42.799 And so pretty quickly, then we got to a 36k deal and then a 45k deal and then a 150k deal, and then Angie's list at like 200k. 00:25:42.799 --> 00:25:44.880 And so I was like, where's the ceiling on this? 00:25:44.880 --> 00:25:58.480 Like, as we started to see those contract values creep up and we crossed like a million in revenue, that was like, okay, I now understand how this can actually be a big business because I'm I'm seeing the value per customer at like a level where that could be big. 00:25:58.960 --> 00:26:03.440 So, you know, considering the last pivot, you know, it seems like the idea was so fragile at that point in time. 00:26:03.440 --> 00:26:05.119 I mean, what do you kind of take away from that experience? 00:26:05.119 --> 00:26:09.200 Going through the pivots and then ending up on something that really, really worked. 00:26:09.599 --> 00:26:11.680 I mean, there were a lot of learnings from that. 00:26:11.680 --> 00:26:20.000 One learning for me was just like how important understanding the market as compared to like whatever your vision is, right? 00:26:20.000 --> 00:26:24.480 Like visions are worthless, truly worthless in that stage, right? 00:26:24.480 --> 00:26:32.640 They matter a lot when you're trying to like rally a company of people around like a direction to all pull the ores together. 00:26:32.640 --> 00:26:40.559 But when it's like four people in an apartment and you're trying to find product market fit, it's like the world could care less what you think about how the world should operate. 00:26:40.559 --> 00:26:41.599 People have problems. 00:26:41.599 --> 00:26:44.880 If you solve those problems for them, they're their problems, not your vision of it. 00:26:44.880 --> 00:26:46.960 Then it'll start rewarding you. 00:26:46.960 --> 00:26:52.400 And so, like just like 10x more humility at that stage, unless you've been steeped in the space for a long time. 00:26:52.400 --> 00:26:53.839 You've got no idea. 00:26:54.240 --> 00:27:03.119 You know, I know you know a lot of founders, I mean, probably every single founder at some point has to decide, you know, whether they keep working on an idea that's in struggle mode, try something new. 00:27:03.119 --> 00:27:05.200 You know, different startups find success in different ways. 00:27:05.200 --> 00:27:07.680 Like, how do you help people navigate that question? 00:27:08.079 --> 00:27:08.720 Boy. 00:27:08.720 --> 00:27:13.599 I guess for me, it's always come back to like the understanding of the customer and the market. 00:27:13.599 --> 00:27:26.559 And particularly in hard tech, people can get very carried away with the science or very carried away with the capability and not necessarily understand the like customer need that they're selling into. 00:27:26.559 --> 00:27:32.799 And so that's almost always where the where the failure mode is of sort of like grinding away in obscurity, right? 00:27:32.799 --> 00:27:36.160 Is like trying to solve some hard technical problem that might not even matter. 00:27:36.160 --> 00:27:46.640 So I guess that's that's where I always try to point people is like go understand the customer problem that you're gonna be able to like blow the competition out of the water through that understanding. 00:27:47.200 --> 00:27:49.200 How big did Segment eventually get? 00:27:49.440 --> 00:27:56.319 At the time of acquisition, we were about 600 people and doing like north of 100 million in revenue. 00:27:56.319 --> 00:28:03.519 And then we got acquired in 2020 for a little over 3 billion, and then continued to accelerate growth afterwards, which was exciting. 00:28:03.920 --> 00:28:05.039 Just paint the picture for us. 00:28:05.039 --> 00:28:07.920 Like, what were you doing the night before you announced the acquisition of the team? 00:28:07.920 --> 00:28:10.880

And like, what did you do the night after you did? 00:28:11.359 --> 00:28:13.599 Oh man, that whole period is a blur. 00:28:13.599 --> 00:28:19.440 I don't think I was sleeping very well because you have to live like a double life as like a CEO of a company. 00:28:19.440 --> 00:28:26.319 Like you have to keep executing, and yet with your small team be in discussions with indiligence with the acquirer. 00:28:26.319 --> 00:28:45.119 And so, on the one hand, like guns blazing, we're gonna go do X, Y, and Z thing, but like your emotional state is completely consumed with like landing the plane on the acquisition side because it's life-changing for everybody and like a completely different thing with like incredibly high stakes in every conversation. 00:28:45.119 --> 00:28:52.000 So I don't remember exactly, but like not sleeping well, and then I think it's like squawk box at like you know 4:30 a.m. 00:28:52.000 --> 00:29:00.000 Pacific, you know, in my like weird janky room downstairs with like a crappy collar shirt put on that I like you know never wore. 00:29:00.000 --> 00:29:02.400 So just like emotionally spent. 00:29:02.799 --> 00:29:07.680 Did you guys as founders ever, you know, I know, um at least in my experience, it's like busy, busy, busy. 00:29:07.680 --> 00:29:18.079 But did you ever have a moment of downtime where like just between founders, you kind of, you know, looking from, you know, math camp to roommates to like that kind of success, you're like, like we kind of did it? 00:29:18.400 --> 00:29:19.039 Yeah, yeah. 00:29:19.039 --> 00:29:27.680 Uh we've been trying to organize these founder trips every year, uh, where the three of us just go and kind of like zoom out and think about those and kind of reflect on it. 00:29:27.680 --> 00:29:34.000 Um, both for myself and I think for a lot of folks there, like the acquisition itself was kind of this like nice end of a chapter. 00:29:34.000 --> 00:29:36.720 We basically spent 10 years building this business. 00:29:36.720 --> 00:29:42.559 And I think one of the things that I'm most grateful for is seeing everyone whose careers were launched by segment. 00:29:42.559 --> 00:29:51.039 I actually think the coolest part for me is one thing that we did post-acquisition is we went around to the earliest employees and actually just thanked them. 00:29:51.039 --> 00:29:55.279 All three of us got on a call, folks who were still at the company and folks who weren't. 00:29:55.279 --> 00:29:59.759 And I think seeing the difference that it made in their lives is like one of the things that I'll treasure most. 00:29:59.759 --> 00:30:11.359 And so I think for a lot of folks, it was like, okay, we've kind of like taken this business to like a meaningful outcome for everyone, and now it's time to continue on to the next phase. 00:30:11.359 --> 00:30:15.519 At least for me, I didn't go on to work at Twilio, and neither did my co-founder Ilya. 00:30:15.519 --> 00:30:21.119 So it was kind of only Peter who was

there from the three of us founders until he left a year later. 00:30:22.079 --> 00:30:24.000 More after the break. 00:30:24.000 --> 00:30:30.240 And we're back. 00:30:30.240 --> 00:30:35.920 And so, you know, obviously segments not your only founder story. 00:30:35.920 --> 00:30:38.640 You know, how did how did you move into the next company that you started? 00:30:38.960 --> 00:30:46.960 Yeah, I started Charm Industrial in 2018, and it actually came out of a problem that I experienced as a customer at Segment. 00:30:46.960 --> 00:30:58.000 Uh at Segment we wanted to do the right thing from a climate perspective, and so we're buying offsets and you know, protecting forests, supposedly, in in uh Indonesia and uh the Amazon. 00:30:58.000 --> 00:31:01.200 And, you know, I was like feeling really good about that. 00:31:01.759 --> 00:31:04.720 You said you wanted to buy carbon credits, you want to do the right thing. 00:31:04.720 --> 00:31:06.720 Why was that important to you? 00:31:07.039 --> 00:31:09.359 I think my mom is a pretty big influence here. 00:31:09.359 --> 00:31:18.720 She was always clear with me that like it's great to do well and build wealth, but the impact that you're having on the world around you matters in how you do that. 00:31:18.720 --> 00:31:21.519 And this was part of that, right? 00:31:21.519 --> 00:31:27.039 Part of that was like trying to find ways that we could use the product to kind of help the world around us. 00:31:27.039 --> 00:31:28.559 I don't know that we ever really figured that out. 00:31:28.559 --> 00:31:33.440 It was a great business, but moving around data is like that, it's a business. 00:31:33.440 --> 00:31:42.160 And we we had like some nonprofit programs and so on where we kind of support folks that we thought had good missions, but minimizing the impact that we were having, at least on the climate, seemed like a good way to have impact. 00:31:42.160 --> 00:31:44.720 So that that I think is ultimately where the motivation came from. 00:31:44.720 --> 00:31:50.480 So we're buying offsets and we made the first purchase in 2016, I think, or 2015. 00:31:50.480 --> 00:31:54.640 And then in 2017, I like circled back and I was like, cool, like, you know, what happened? 00:31:54.640 --> 00:31:58.000 Like, did we have the effect that we wanted with that $20,000 that we spent on it? 00:31:58.000 --> 00:31:59.119 Like, you know, what happened here? 00:31:59.119 --> 00:32:02.079 And so we started digging into the nonprofits that we'd sent money to. 00:32:02.079 --> 00:32:09.599 And the deeper I got into it, the more I was like, oh boy, like, ugh, I don't think that this is having the impact that we wanted it to. 00:32:09.599 --> 00:32:14.559 Like 70% kind of went off into marketing fees that were lost at various levels. 00:32:14.559 --> 00:32:22.640 And so it wasn't until you got down to like, you know, the last 30% that actually there was any evidence that the money was actually going into rainforest protection. 00:32:22.640 --> 00:32:28.319 And then you were seeing these headlines like Amazon on fire, biggest for fire season ever in Indonesia, etc. 00:32:28.319 --> 00:32:33.359 And you're like, well, was the land that we put aside like, was it protected or did it blow up in flames? 00:32:33.359 --> 00:32:35.119 Cause like, I don't know, should we buy more? 00:32:35.119 --> 00:32:36.880 Like, what's the deal here? 00:32:36.880 --> 00:32:42.079 And the deeper I got into that, the more I was like, like, I don't, I have no evidence that anything could happen. 00:32:42.079 --> 00:32:55.119 And so I really wanted like a carbon removal product where I had trust that it actually had been removed and that it had been verified and that I could personally inspect some kind of ledger to show that it had been done correctly. 00:32:55.119 --> 00:33:00.720 So that was kind of the initial inspiration around charm was a high quality, verifiable carbon removal. 00:33:01.119 --> 00:33:02.480 So we're kind of moving in the charm. 00:33:02.480 --> 00:33:06.960 You know, how would you either describe it to your kids or actually maybe more interesting, how how do your kids describe it? 00:33:08.079 --> 00:33:10.480 Yeah, I have a seven, five, and two-year-old. 00:33:10.480 --> 00:33:15.519 The two-year-old prefers to talk about the zoo with his, you know, three words he can string together right now. 00:33:15.519 --> 00:33:18.079 Seven-year-old is starting to understand it though. 00:33:18.079 --> 00:33:21.680 We often, you know, there are many problems in the world of which climate is one. 00:33:21.680 --> 00:33:25.039 And it's challenging because it's a very long-term problem, right? 00:33:25.039 --> 00:33:27.920 There are lots of other problems that are extremely urgent. 00:33:27.920 --> 00:33:32.880 And climate is tough because it's one that's both very important but also very long term. 00:33:32.880 --> 00:33:43.920 And so we we try to describe to him, you know, as like there's a blanket around the earth from this gas, and it's just kind of getting thicker and thicker, and at some point it's going to become really problematic. 00:33:43.920 --> 00:33:50.240 And so I'm trying to thin that blanket by putting some of that blanket back underground in the form of bioil. 00:33:50.240 --> 00:34:03.519 And along the way, by the way, you know, you see the smoke, and sometimes the sky turns orange, and like we can prevent some of those wildfires because the stuff that we're putting underground is actually plant matter that would otherwise go up in wildfire uh smoke and so on. 00:34:03.519 --> 00:34:06.319 So that's kind of how I explain it to them. 00:34:06.319 --> 00:34:09.280 And, you know, carbon removal is important, it's a small part of the problem. 00:34:09.280 --> 00:34:17.199 First, we have to reduce all our emissions along the whole supply chain, and clean power is a big part of that with solar and wind and nuclear, geothermal, etc. 00:34:17.199 --> 00:34:21.440 But a big

part of it is industrial, like cement, steel. 00:34:21.440 --> 00:34:23.840 Um, these are like really hard to decarbonize areas. 00:34:23.840 --> 00:34:32.000 And so we should figure out a way to re-engineer our industrial processes to remove carbon in their processing instead of emitting it. 00:34:32.000 --> 00:34:41.679 And so charm industrial is char farm industrial because the idea is you take the biomass off the field, you leave behind the char, and you take the bio oil as an industrial feedstock. 00:34:41.679 --> 00:34:47.039 In the future, we hope that the sort of liquid biomass that we make that we can, you know, use it as a feedstock in those. 00:34:47.039 --> 00:34:50.639 I think that's probably like a 2030s kind of proposition for us as we come down the cost curve. 00:34:50.639 --> 00:34:54.639 But yeah, today we just take the biowale and stick it underground. 00:34:55.039 --> 00:34:55.920 Like walk me through. 00:34:55.920 --> 00:34:59.119 I mean, you come up with the idea, you know, you decide it's time for you to do something new. 00:34:59.119 --> 00:35:02.559 You know, what were the early signals that gave you the confidence that was the right direction? 00:35:03.119 --> 00:35:08.400 Uh, this one was more like I was convinced that this had to happen and that demand would materialize over time. 00:35:08.400 --> 00:35:11.840 That was based more on instinct than you know, knowing these industries. 00:35:11.840 --> 00:35:23.360 And it took several years of like smashing my head against a wall to break out of the tech industry networks into software industry networks into like steel, concrete, ammonia, you know. 00:35:23.360 --> 00:35:33.039 Like I don't think this worked that well, but like, you know, at one point I was on a sales trip to Chicago for a segment and I like made a stop in my Uber at like the headquarters of like CK Industries or something. 00:35:33.039 --> 00:35:37.760 There's like the company that runs most of the ammonia plants for fertilizer production in the US. 00:35:37.760 --> 00:35:42.559 And I like literally, I had like printed out, you know, flyers basically with like a memo. 00:35:42.559 --> 00:35:48.079 And I like walked in cold to the building and like handed it to a secretary and was like, you know, this is for X. 00:35:48.079 --> 00:35:53.119 Um, so there's like I really smashed my head against the wall to like get into some of these places. 00:35:53.119 --> 00:36:10.800 Another one on the steel side that was very challenging was um uh we had an investor whose younger brother had been a roommate with someone in college whose brother had then inherited the family's rolled steel products business in Ohio. 00:36:10.800 --> 00:36:12.800 And so I got an intro through that chain to him. 00:36:12.800 --> 00:36:15.280 So we're like four steps deep already. 00:36:15.280 --> 00:36:19.119 And then he introduced me to like the Canadian Steel Association. 00:36:19.119 --> 00:36:26.079 The Canadian Steel Association thought this was interesting enough that they introduced me to their colleagues at Arcelor Middle to Fasco. 00:36:26.079 --> 00:36:34.559 And they thought it was interesting enough that after like three meetings, they finally introduced me to the direct reduced iron plant, which was my original target in Quebec. 00:36:34.559 --> 00:36:37.920 And then, you know, I was like, Oh, I'll be in the area. 00:36:37.920 --> 00:36:41.679 Like it's like northern Quebec, like you're not in the area. 00:36:41.679 --> 00:36:45.760 And so I just like flew out there and that became a productive kind of relationship. 00:36:45.760 --> 00:36:48.159 So I don't know, it was very hard to break into these industries. 00:36:48.559 --> 00:36:58.800 I mean, in some ways, oddly enough, it seems like a bit of the opposite to the segment story, but also it sounds like it, you know, maybe it was the customer pull that was really giving you the confidence like this is direction we need to keep going. 00:36:59.599 --> 00:37:05.199 Yeah, I wouldn't say that we had customer poll until 2020, but also that I was not internally focused. 00:37:05.199 --> 00:37:12.320 Like I had a feeling that this was like a strategic area to be in, but that it was gonna take a while to figure out what the product market fit was. 00:37:12.320 --> 00:37:18.400 And eventually we found this product market fit on carbon removal, and it's like it's honestly just grown faster than I expected. 00:37:18.400 --> 00:37:19.920 Like I'm always a little skeptical. 00:37:19.920 --> 00:37:23.199 I'm like, but how much voluntary carbon removal are people gonna buy? 00:37:23.199 --> 00:37:27.840 You know, like I mean, I was a customer of it, but like, are other people really gonna step up to the plane? 00:37:27.840 --> 00:37:41.840 And it was like, well, I don't know, like we'll probably get to 10 million in bookings, and then we got to 10, and I was like, well, no, like I kind of have visibility to 100 million in bookings, and then I'm like, well, we're at 100 million in bookings and like kind of have visibility to like a billion in bookings, and so you're just kind of like, I guess we'll just keep rolling with it. 00:37:41.840 --> 00:37:48.480 But importantly, I wasn't spending my time as founder on the grinding in obscurity on the technical process. 00:37:48.480 --> 00:37:53.840 I was trying to go knock down these customer conversations and things like showing up with a Chevron refinery. 00:37:53.840 --> 00:37:56.880 Like I was out in the field trying to understand what was compelling to them. 00:37:57.199 --> 00:38:13.599 One of the stories that Peter talked about that really struck me, you know, he talks about going cold into companies after he's already been a pretty successful founder of his own right, walking into their lobby, like, you know, a company in northern Quebec and just basically saying, like, hey, like I want to meet with this person. 00:38:13.599 --> 00:38:15.440 I mean, how do you imagine that

going? 00:38:15.440 --> 00:38:16.880 And like, what do you think it is about him? 00:38:16.880 --> 00:38:18.719 And how do you think those conversations went? 00:38:19.039 --> 00:38:25.760 I think our early failure, where we came in with such expectation for like, oh, of course we're going to succeed. 00:38:25.760 --> 00:38:27.440 We succeeded at everything else before. 00:38:27.440 --> 00:38:29.599 Oh, of course we're going to make something that people want. 00:38:29.599 --> 00:38:33.519 Oh, of course, this idea that I have, like, that's the way the world should work. 00:38:33.519 --> 00:38:39.760 I think being hit over the head so many times with like, hey, the world doesn't necessarily work like that. 00:38:39.760 --> 00:38:48.800 The world works a certain way, and it's kind of up to you to figure out how to fit into it and to build a company that makes it work a little better. 00:38:48.800 --> 00:38:54.239 I think that lesson hit all of us pretty hard, and especially Peter on the product market fit side. 00:38:54.239 --> 00:39:02.400 And so I think since then, like frankly, Peter's biggest superpower is like figuring out the go-to-market path for a company. 00:39:02.400 --> 00:39:07.119 And I feel like there's lots of problems out there, and I think he's gonna go after them. 00:39:07.360 --> 00:39:15.840 You know, from an outsider perspective, I think it's actually fascinating about how the three of you in some way post-segment explored, you know, climate tech and you know, kind of kind of various solutions. 00:39:15.840 --> 00:39:19.760 You know, Ilya with fire, you with heat pumps, you know, Peter was obviously doing charm. 00:39:19.760 --> 00:39:22.800 I mean, do you have any insight into like, was there something in the water? 00:39:22.800 --> 00:39:27.280 Like, what do you think it was that got all three of you interested broadly in this space? 00:39:27.599 --> 00:39:34.639 I mean, I think for Peter, he's always wanted to solve like one of the big problems of our time uh and have more of an impact. 00:39:34.639 --> 00:39:47.840 Uh, and I guess stepping back, even thinking about kind of like post-segment life, I think for all of us, we really want to try and solve some problem where we look at it, we're just like, oh man, like this feels like a worthy thing to swing for. 00:39:47.840 --> 00:39:52.480 And so I think for him, obviously, climate change is like one of the biggest problems of our time. 00:39:52.480 --> 00:39:55.920 And he feels like he has a way to solve that. 00:39:55.920 --> 00:40:02.880 I think it makes a lot of sense that he would like get deep into it and try and understand, like, hey, how can I build a bunch of infrastructure to solve this problem? 00:40:02.880 --> 00:40:09.039 I think he's definitely had an impact on me and Ilya as well, I would say, just from like talking about this so much. 00:40:09.039 --> 00:40:18.880 Like, I feel like I've learned so much from both him and his wife, Erica, about the climate space

and like where there are problems and where people could use better solutions. 00:40:18.880 --> 00:40:26.800 Like right now, he's doing the carbon removal stuff with charm and also helping run Revoy, this electric freight startup. 00:40:26.800 --> 00:40:28.960 Like, but between all those things, it's crazy. 00:40:28.960 --> 00:40:32.480 So I would say that's definitely he's rubbed up on us in a good way. 00:40:32.719 --> 00:40:37.440 And then you also joined Revoy, uh, which is a company that we've invested in. 00:40:37.440 --> 00:40:40.559 Can you describe what Revoy does and how you decide to get involved? 00:40:40.880 --> 00:40:46.320 Yeah, Revoy was founded by uh my friend Ian, who was another floor mate of ours at MIT. 00:40:46.320 --> 00:40:50.800 Uh, he was actually our next door neighbor when Calvin, Ilya, and I were roommates. 00:40:50.800 --> 00:40:53.599 Really brilliant idea that he had for Revoy. 00:40:53.599 --> 00:40:58.800 So the idea is it's an electric dolly that sits in between your standard diesel tractor, a standard trailer. 00:40:58.800 --> 00:41:03.039 No special hookups, like just you would normally hook up a trailer to a tractor. 00:41:03.039 --> 00:41:11.360 And what's cool about it is that you just have like a load cell in the kingpin, and it basically tells you whether the truck is trying to accelerate or not. 00:41:11.360 --> 00:41:24.159 And purely on the basis of that, instead of burning gas, you can transfer 94% now of the mechanical power is coming from the electric dolly, not from the gas burning in the in the truck. 00:41:24.159 --> 00:41:25.280 It's pretty incredible. 00:41:25.280 --> 00:41:28.480 Like effectively, the truck is like barely idling at this point, right? 00:41:28.480 --> 00:41:34.079 If you're hooked up to a revoy, you're doing 120 miles per gallon in your semi-truck instead of seven. 00:41:34.079 --> 00:41:39.679 So it's really incredible piece of technology and very quickly in the right markets. 00:41:39.679 --> 00:41:47.039 And we think with like 25 to 50 revoys, we can get to a place where it's gross margin positive, like breaking even with diesel. 00:41:47.039 --> 00:41:49.280 And that it's a very cost-sensitive industry. 00:41:49.280 --> 00:41:55.599 And so I think if you can even start chipping off like a few cents per mile, you start to be actually pretty compelling for carriers to use. 00:41:55.599 --> 00:41:57.920 And from a safety perspective, in many ways, it's better. 00:41:57.920 --> 00:42:03.760 Your stopping distance is shorter, just because you have the regenerative brakes on the revoy, another axle in addition to the truck. 00:42:03.760 --> 00:42:08.159 And you also get faster acceleration, twice the power at your fingertips. 00:42:08.159 --> 00:42:10.239 So drivers love it from that perspective. 00:42:10.239 --> 00:42:21.280 They kind of get this better maneuverability of like accelerating onto the freeway without getting, you know, hit from behind or jamming someone in the on-ramp lane, as well as yeah, shorter stopping distances in the case of like an accident or coming downhill. 00:42:21.280 --> 00:42:29.519 You have a regenerative braking that's like taking a huge amount of energy into the battery instead of a long downhill where you're just burning brake pads and you know, could potentially have a brake failure. 00:42:29.519 --> 00:42:33.599 So there are safety and driver experience benefits as well. 00:42:34.320 --> 00:42:38.960 Yeah, I've had a very small view of some of the work you've been able to do help out there. 00:42:38.960 --> 00:42:43.679 And, you know, I have to say it's been super impressive, just super methodical, very focused on execution. 00:42:43.679 --> 00:42:48.000 What are some of the main lessons from segment and charm that you think you've been able to bring to Revoy? 00:42:48.480 --> 00:42:55.199 I think from the charm experience, the number one learning, which has been quite painful, is just like deleting the part, is really important. 00:42:55.199 --> 00:43:00.559 Like if you can eliminate a part, you lower the maintenance cost, you lower the failure mode, so you increase reliability almost always. 00:43:00.559 --> 00:43:03.119 You know, it's one less thing that has to be designed and kept track of. 00:43:03.119 --> 00:43:06.800 And I just can't overstate like if you can delete something, you must delete it. 00:43:06.800 --> 00:43:09.599 Um I'll give you one example for reware right now. 00:43:09.599 --> 00:43:13.599 We have a we have turning on the rear axle, which is really convenient when you're trying to hitch it up. 00:43:13.599 --> 00:43:17.280 But you get a lot of extra friction, you need a lot of extra weight. 00:43:17.280 --> 00:43:21.519 There's like all these components that then are managing the fact that that axle can turn. 00:43:21.519 --> 00:43:22.960 Anyway, so we're experimenting with that. 00:43:22.960 --> 00:43:28.960 I'm very excited about this because if you delete the ability to turn, you delete all these gears, you delete all this friction, all these like other pieces. 00:43:28.960 --> 00:43:34.239 So it's like anything that you can find a way to delete is like so transformative. 00:43:34.239 --> 00:43:37.440 So delete the part as a mantra now. 00:43:37.840 --> 00:43:44.639 What do you think is special about Peter in terms of his strengths, in terms of what he like uniquely contributes as different than most other people? 00:43:45.280 --> 00:43:52.159 Peter is able to go deep in such a broad range of subjects, like maybe more than almost anyone I know. 00:43:52.159 --> 00:43:54.239 It's ridiculously impressive. 00:43:54.239 --> 00:43:59.760 He's like quickly able to cross disciplines and dive into a bunch of different areas, like maybe more voraciously. 00:43:59.760 --> 00:44:01.679 Than anyone else I've ever seen. 00:44:01.679 --> 00:44:16.239 And so I think finding all of those in one person who has that range, but also has that ability to then go and just like execute in like a very deep, thoughtful, and inspection-driven way, I think is very rare. 00:44:16.239 --> 00:44:18.960 Because I guess I feel like I talk to a lot of investors. 00:44:18.960 --> 00:44:22.960 Like, I think investors have the range, but like not often that sort of like operational cadence. 00:44:22.960 --> 00:44:28.559 So I would not bet against Peter just in terms of his desire to go and get things done. 00:44:29.199 --> 00:44:33.760 The joy in life, in some sense, is like the struggle and all the learning that comes along with it. 00:44:33.760 --> 00:44:34.320 I don't know. 00:44:34.320 --> 00:44:36.559 I feel like there are big problems that need to need to be solved. 00:44:36.559 --> 00:44:38.639 So fun to go tackle them. 00:44:39.039 --> 00:44:39.840 Well, that's awesome. 00:44:39.840 --> 00:44:40.559 This was really great. 00:44:40.559 --> 00:44:41.679 Thanks for sharing your story. 00:44:41.679 --> 00:44:43.599 Yeah, thanks for having me. 00:44:43.599 --> 00:44:50.000 This podcast is brought to you by Leapforward Ventures, an investor in early stage startups. 00:44:50.000 --> 00:44:58.559 If you like the show, the number one thing you'll want to do is sign up to get notified when we release new episodes that includes transcripts and key takeaways from each story. 00:44:58.559 --> 00:45:01.280 Head to leapforward.fm to sign up. 00:45:01.280 --> 00:45:06.800 I'd also love to hear your feedback on the episode and who you want to hear from next. 00:45:06.800 --> 00:45:12.719 Just shoot us a text at 415-915-3050 to get in touch. 00:45:12.719 --> 00:45:17.360 This episode was produced by Theo Balcomb and Kim Nederveen Pieterse. 00:45:17.360 --> 00:45:19.119 Craig Eley is our engineer. 00:45:19.119 --> 00:45:21.360 Reese Laudano made our cover art. 00:45:21.360 --> 00:45:24.719 Music is by Jim Brunberg and Ben Landsverk of Wonderly. 00:45:24.719 --> 00:45:28.159 I'm David Rusenko, and this is the Leaf Forward Podcast. 00:45:28.159 --> 00:45:30.000 See you next episode.