Originally aired on December 13, 2021 @ 12:00 PM - 12:30 PM EDT
This week is Cloudflare's Founder Spotlight on Cloudflare TV, featuring dozens entrepreneurs from across the tech industry and beyond!
This session features Calvin French-Owen, co-founder of Segment. Segment is a customer data platform (CDP) that helps companies harness first-party customer data. Their platform democratizes access to reliable data for all teams and offers a complete toolkit to standardize data collection, unify user records, and route customer data into any system where it’s needed. More than 20,000 companies like Intuit, FOX, Instacart, and Levi’s use Segment to make real-time decisions, accelerate growth, and deliver compelling user experiences. Segment was acquired by Twilio in 2020.
to find the rest of these featured conversations — and tune in all week for more! Founder Spotlight hub All right, we are live on Cloudflare TV. I'm Evan Johnson from Cloudflare's product security team. And with me is amateur DJ, white t-shirt aficionado and ultra marathon runner Calvin French-Owen. More importantly, the co-founder and former CTO of Segment, which was acquired by Twilio, I think like two years ago, almost exactly. Last November. One year ago, one year ago. Oh, okay. One year ago, almost exactly by Twilio for I think the whopping total of $3.2 billion. Huge acquisition. And Calvin was there from the beginning of it all the way through. Calvin, do you want to do an intro for yourself? That's a little less... What? I don't know how I can top that intro. White t-shirt centric. Well, I got to say, your role at Segment was enormous and changed over time. I worked with you there. I joined as a security engineer when the company felt still really small, like a tiny, tiny little startup still. And you were one of my favorite people to work with, not just at Segment, but in my career so far, because you've brought so much energy to like, kind of even just meaningless interactions of like a sprint demo or whatever. It was always a blast working with you. How did you feel like your role changed over time? And like, from the beginning as when you and Peter and Ilya and you guys were all cockroaches, as Peter put it in his talk, to like a real corporation? Yeah. Man, it really evolved a lot. I think that's actually one thing that most founders don't realize is how different your job will be if the company is succeeding. And oftentimes, actually, when we'd be doing the same thing over and over again, that was like a classic sign that the company was not doing well. So I think when I started out, basically, it was kind of myself and Ilya building all of the back-end infrastructure for Segment, because we were the two guys who had the most back-end experience, had like actually gone to school for CS. And then my co-founder Peter was the CEO, so he managed all the finances, like a lot of talking to customers and kind of led on overall company building and fundraising. And then Ian took all the design and the front-end. And I think one thing I hadn't appreciated at the time is how much of kind of a full-stack team we were. This summer, I was just helping out at YC and coaching about 100 of the startups there. Wow. And it's funny how many of them are like, oh man, I need like design help, or it's hard for me to build this feature without like the right engineers, or maybe they have like one technical person on their team who's able to stuff. And I look back on it, I was like, wow, like the real reason that we could build stuff so quickly was just that we had four very technical people who could kind of fill all layers of the stack, at least for what you need for a, I don't know, MVP. And that was really where things started. I think over time, as we started growing the company, my role shifted a lot. We brought on Tito, who was our VP of engineering and eventually became the chief product design officer who basically like oversaw the management and growing and recruiting of the team. I then did a lot of work on kind of our engineering brand, like helping make sure, well, basically kind of running the Cloudflare playbook of putting interesting stuff up on the blog, sharing stuff that we built internally, making sure that people knew like, hey, Segment is the right way to build a data pipeline. And if you're interested in coming and working on those sorts of problems, here's how we solve them, which helped us hire our initial engineers. And then I kind of worked on a bunch of different product teams. So stuff like helping land our warehouses rewrite that our team kind of took this built quickly system and then transitioned it to a much bigger system, which I was amazed with their progress there and like how they took the system of hundreds of customers who were all running on this infrastructure, which had to be exactly right row for row. And then basically like 10X its reliability. I worked a little bit on some of our delivery pipeline systems with Tom, Alex Noonan and Ashil, worked a little bit on the visibility view that you see when you log into Segment. And then wrapped up working mostly on our developer platform, getting people to actually write code that runs on Segment, which I know is always a scary thing from a security perspective, but an awesome thing from a customer perspective. And then I actually wrapped up working on this cross-functional effort with our digital, well, we called it the digital team, but eventually it bridged product development with sales and marketing. And the idea was to focus on our self-service growth. That makes sense. Yeah, it sounded like to me and giving that overview and my impression was the same while I was working there was in the beginning, it sounds like your role was anything and everything that needed to be done as a engineer on a small team fighting for survival. And then at some point you kind of grew and Tito joining also helped you kind of get involved with the crisis of the week or whatever else needed to be done. It was really, it sounds like actually your job didn't change too much because at the beginning, whatever needed to be done, put Calvin on it. And then towards the end, it was still whatever needed to be done. You were the guy who led these important initiatives. Yeah. I mean, I was definitely there for a bunch of it, but I think it would be wrong to say it was only me because frankly, there are a ton of people who stepped up to do different parts of, well, improve the infrastructure in various ways. I mean, like when you look at the fingerprints that like Albert or Daniel St. Jules, Ashil, Alex Noonan, or folks like Kevin, Hareem, Alex Mieh, like all of them, yeah, I think ended up taking the company much further than any of us other did. But I think the nice thing about having founders tackle those sorts problems is often, like as a founder, you kind of just want to do what's ever good for the company since your personal outcome is very tied to whether the company does well or not. It kind of doesn't matter otherwise. So I think that's where a lot of companies actually kind of employ founders to take on those different like fire of the week or whatever tasks. That makes sense. That makes a lot of sense. Yeah. I was always really impressed with the, well, you mentioned Cloudflare and how there were some similarities in the blog and I also thought that there were similarities in like the way that they launch products where I know we try to just launch things really quickly, get it out the door and then improve it really quickly, make it good, get it in front of customers, get feedback. And that's what's important to us. And I know that was what was important at Segment as well. And then kind of closing out your arc, like at what point were you like, all right, I'm ready to go do something else now? Yeah. So as the talks for the acquisition started getting more serious, kind of at that point, I started thinking like, oh, what is it that I want to do next? And whether it's become part of Twilio or go do something else. I think for myself, I saw it as a good kind of end of a chapter. It was like, basically I'd spent this 10 year period building Segment. And now I was just excited to see other business models, other things I could work on and other things I could focus on. Because I think I've also been reflecting on this a bunch over the past couple of months, but I think the most rewarding part of Segment was seeing an awesome group of people and working with them day in and day out to build something that we were proud of. But that said, the problem itself wasn't actually the thing that interested me most. It was the thing I enjoyed working with other people. It was a great business. And I think we had a super high caliber team. But collecting customer data, I think there's other things that I would prefer to work on versus that problem. Yeah, for sure. I mean, also, it was something you worked on over a 10 year period of time. And at some point, you're just wanting to see a different problem. So I would, if I was saying this right before the call started, if I was in your shoes, I would need a good few years to decompress from just being focused on that one problem for so long. Everybody's brain just needs to see different things, I think. Well, you said you're working, you did some, I don't know the word, advising or you were involved with YC and all of these newer companies. What is your message to founders and people who want to be founders? I've seen like a ton of different startups springing up from the segment alum team, and they all seem to be doing really well. And so there must be some secret sauce in the water. Yeah. So I think the number one thing that we learned as a founding team is that the narrative that you tell the world isn't necessarily true. You can't really like, you can't force your idea on the world. Instead, you have to listen and understand the market and really understand customers' needs, and then kind of work backwards from that. And I think this is something that you hear a lot at Amazon. They're pretty famous for their like, start with the customer, and you kind of write this like press release and FAQ first, and before you build anything, and then you work backwards and figure out what you actually need to do to build that to satisfy that customer need. And for us, for the first, I want to say 18 months of segment, we basically took the opposite approach. We just built a bunch of tech, but we wanted to figure out like whether people would use it or not. But we were mostly trying to build tech for interesting tech sake for problems that we thought existed. And when we repeatedly would give stuff to customers and show them these analytics tools that we built, just no one wanted them. And I think that was like a very deeply humbling experience for us. Because it showed kind of for the first time that in order to be a very good founder, you have to be great at finding and naming problems. I think if you look at the segment alum folks who are succeeding today, a lot of them came from teams where they had to do that, where they had to develop some new product from scratch. It didn't work, it didn't work, it didn't work. And then finally, they kind of hit that product market fit that allowed them to take off. Yeah, I've got Tejas and Josh Perl on later this week. I haven't talked to them in a while. I was keeping in touch with them when they were small. And now, I mean, they're, I blinked, and they're a real company now. So I've got the founders from Hightouch, who are a segment alum coming on later this week. And yeah, I hear what you're saying about the narrative. It's always, it's so important. And it's something that I see a lot of. I look at a lot of security companies, and it's something that I just don't see very much in the space. It's something that I think Cloudflare, and Matthew, and Michelle, and the entire exec team at Cloudflare does really, really well. But then how do you balance that with actually building something and shipping it? Because I feel like you can get really hung up in the details of, well, I guess you have to be able to balance it. But you can get really hung up in your narrative and telling the world all the stuff that you don't actually do. But then also, you need, like, I feel like the segment origin story of you just built this thing, put it on Hacker News, and it exploded. It's almost like the opposite. It's true. Yeah. Yeah. I think, I don't know, Will Larson has this good, Will Larson's the CTO of CalmNow, and he writes a lot on engineering management and how to progress your career as a technical engineer. He has this good line that oftentimes a cheap experiment is worth anything like N years of analysis. And I think that was really kind of what happened to us. We said, okay, we're going to test whether people want this thing. We're going to give it a good go and launch it. And if it gets picked up, we'll follow it. And if it doesn't, we'll try something else. And I think that philosophy actually serves most companies really well. When you look at the ones who are constantly innovating or constantly succeeding, like, they just move at this really rapid rate. And they're kind of always moving quickly, no matter what. And sure, they'll get a bunch of stuff wrong. But because they're focused on actually, like, trying to iterate and trying to launch, they learn really quickly. And so I guess the way I would balance it is saying, you want to have a clear direction or, like, set of problems that you're solving that you, like, know are problems. But then in terms of the solutions for it, yeah, it's worth experimenting and trying a bunch of different things. And as long as you're learning quickly, I think that's probably the most important thing. Yeah, that makes sense. Course correcting and being honest about what's not working quickly. And yeah, it is a hard balance between narrative and shipping. And then it's one that I think, like, me as a security person, I'm always the person, well, this isn't me personally, but the kind of tension about not shipping, you don't want to do it because it's new risk. And it's hard to get in that mindset where as a security person, you have to get in the mindset of it's ship or die. And that's how startups work. And so you have to take on new risk and do what you can to mitigate it. So I don't know, but there's always that tension between, I guess, shipping and security, which is way a tangent of your life as the co-founder of Segment. But actually, you were very security minded when you were there. I'm curious where that comes from. Was that just you being self-incentivized to keep the company secure? Or is that just an area of interest for you or what? I think it's mostly an area of interest, I would say. Well, probably a combo. I took a bunch of security classes in undergrad just because I thought they were interesting. And I remember learning about some of these low-level protocols, stuff like, I don't know, ARP or DNS or how IPE works. I was like, wow. It's like the Internet is really held together with this stuff. It's like all duct tape and it's not built from security from ground up. And so I feel like that was just something that was in the back of my mind as then I started building my own tools with Peter and Ilya and Ian. And I think my biggest fear was that there were a lot of things that... It felt like there were a lot of things that the company could do that would be totally fine. But one thing that I worried about constantly was because we housed all of this data for people, if that ever leaked, that would potentially be kind of this company ending event. I think it was something that as we grew bigger and bigger too, the target on our backs just got bigger and bigger because suddenly instead of having a small amount of data for a handful of customers, we were helping some of the world's largest enterprises. Yeah. Storing all that data, definitely big target on your back. And when I joined Segment, I joined as the first security engineer there. And Albert had joined and I joined around the same time. And they said, Evan, your desk is going to be right here and Calvin's desk is going to be right here. And so I was very nervous having the CTO sit directly next to me because of your interest in security. I was like, oh God, what have I gotten myself into? But you're not a very stressful person to be around. So it wasn't so bad. But I did take note that you were very security minded and everything from finding the AWS keys and making your dashboard gamifying it to whatever else needed to be done really. It was definitely great having that support. I'm curious in terms of the growth arc of the company, like you went from being a tiny company without product market fit and growing and growing and growing. Was there ever a point where you were like, okay, we're fine now, like where the company is not going to implode at any moment? Or were you ever like that? That's an interesting question. I'd say sort of yes and no. Like there was definitely a period where Segment got big enough that it felt like, it's kind of like the analogy I like to use is sort of like a boulder where for a long time, you're like pushing this boulder up the hill. Yeah. If you ever stopped pushing the boulder, it's just going to roll back down and then you're done. I think we got to a certain point where there were enough systems in place where the company was growing at a sufficient rate, where the market was still big enough ahead that it felt like, okay, if any one of the founders stops pushing the boulder, it's going to do just fine and keep going. I think that idea lived in my head at the same time as figuring out like, how could the company be even better or continue to be kind of world changing? Because I don't know, as someone who's really excited about the company, whether you're a founder or early employee, or just someone who's very passionate about work, I think work tends to dominate an outsized part of your mindshare. Like when I'd go to bed and thinking about like, oh, how could Segment be better? When I wake up, how could Segment be better? Showering, how could Segment be better? And so I think it never really felt set or like it was just on rails, but a lot of existential risks of the company went away, if that makes sense, where you never felt like, oh, okay, everything's going to go to zero. That case got removed. But instead it was more like, oh, maybe we'll get beat out by a competitor in five years. Gotcha. So it stopped being like, we could run out of money, or we could not be able to grow, or we could, it stopped being more immediate risks and being, I guess, like the time it takes for the risk to realize itself started being bigger. And maybe that's a way to look at it. That's interesting. And I'm curious, what times were the easiest being a founder? The easiest? There's got to be something that's easy about it. Yeah. I don't know if it was the easiest, but I think the times that I look back on and reflect most fondly, we're seeing, it was mostly seeing teammates step into either a larger role or a point in time where they were tackling a really ambitious project, which they'd never done before and actually succeeding at it. And I think one thing that was a big part of the segment DNA, which I love and hope gets preserved for the rest of time and definitely would be a trait I'd want to bring to a new company, is this idea where you give people a really big challenge and you trust them and take a bet on them. And most of the time people will rise to meet those expectations. I think that was something we did well at Segment that when I now learn about other companies who have a good culture of experimentation, a lot of them make it okay to take those risks and they expect their people to take those big bets. Totally. Yeah. It's in line with the hire for somebody's slope rather than their, rather than, what is the other way? Their Y-intercept. Y-intercept, yes. Slope, not their Y-intercept. So you can hire these people who are fast learners, who are curious, who want to do great stuff. And if they won't figure it out immediately, they will pretty quickly. And it usually works out if you've hired good people. Yep. Sometimes I'm sure you can hire good people and it doesn't work out, but I mean, sometimes problems are really large. Okay. So I'm curious in your, I'm looking at my notes here about what I wanted to grow you on. In the world of startups and new startups and I guess, is there anything you're really excited about? Or I know you're security minded, anything in the security space you're really excited about? It doesn't have to be security though. Yeah. To be honest, I haven't followed the security space super closely recently. Let's see. I'll throw you a laundry list of things right now. I think in general, I'm very excited about data visualization. That's one area on the long ways to go. Like in some ways, I think it's no accident that a tool like Excel became really popular because it's a very natural way for people to think of data, right? It's like you've got columns and rows and you can visualize them and build models and do whatever you want. But unfortunately those tools don't work very well for today's datasets. If you have more than a few thousand or tens of thousands of rows, it's hard to analyze within those. And usually then you'd reach for a data warehouse and maybe a tool like Looker or Tableau or Mode. What I really think should exist is some tool that bridges those gaps pretty effortlessly. So you can go between a database of a large dataset, get results streamed back to you, visualize it, and almost play with data like it's Play-Doh. So that's one area that I've kind of been digging into a little bit and built a prototype. And there's a couple of companies in the space who I think are doing really great work. You built a prototype? I built a prototype. Oh, it's so bad. I was trying to learn CSS and all the colors are wrong. You got to protect your eyes so you don't stab them out when you check it out. But it was fun. It was more just like a get used to working with different types of tabular data. There is this open source project. Oh, shoot. I'm blanking on the name. It was something I built a little while ago. I'm not going to remember it, but it was basically like a cross-platform representation of sort of like Parquet files, like columnar datasets that you could store and analyze in memory. Nice. Yeah. Oh, Arrow. Arrow is the name. Okay. You said that's on your GitHub? Oh, it's just an open source thing. It's Apache Arrow. Oh, okay. Okay. Yeah. Let's see. Other than that, I'm pretty excited. There's a lot of work that's happening in the energy industry right now. That's pretty exciting. We were talking a little bit earlier about Fusion, but I think the Fusion space itself is starting to be really attractive as a way to generate clean energy. I guess the super high level view is that Fusion, or today when we think of nuclear energy, we typically think of nuclear fission where you're breaking apart heavy molecules. And when you break those bonds, it releases energy that drives a steam turbine. Fusion is the opposite process where you bring molecules together, you subject them to a lot of basically pressure using magnets or electronics. And when that happens, if they get close enough together, they fuse, and that also releases energy, which is the way the sun works. And so for a long time, we've been trying to generate energy from Fusion, and we've never gotten more energy out than we get in. And so have been a bunch of big government projects to generate that. But I think the result is that everyone's centered on this one single design called the Tokamak, which most scientists agree is the best reactor design to generate actual power. What that means is that there's all these other ideas that people have for generating Fusion, whether it's accelerating two particles together super quickly, or shooting a laser at one so then it explodes and then fuses with another, or I don't know, keeping a set of magnet or a liquid metal that you then kind of pulse that then generates enough pressure to create Fusion. All these other kind of rando reactor designs that really no one has pursued because there wasn't a lot of research and grant money in it. Today, maybe there's like 15 or 20 startups who are all trying out this other 80% path to see whether anything is promising there. Oh, wow, that's really interesting. Yeah. Geothermal is another really interesting energy one. I didn't know this before, but actually the total amount of like thermal energy in the Earth's crust is 20 billion times our annual consumption. And so if we're able to tap into that somehow, then potentially we could get this clean limitless energy to allow us to, I don't know, terraform, go to the stars, desalinate, whatever we want to do. And so there's this one startup who's basically focusing on making it easier to drill. Where today when you think about like an oil and gas drilling down, they use just a mechanical drill bit. And most of the time, if you're drilling deep, you're like bringing the drill bit up and bring it back down and repairing it and doing that most of the time. But there's this one startup called Quaze, which is actually sending down an energy beam using millimeter wave technology. So super high powered beam that then breaks apart rock by disrupting the lattice formations and its molecules. So if it works, it has the potential to make geothermal super cost -effective anywhere, which would be really cool. That's fascinating. This is a whole giant space that I know nothing about. So it's, I don't know if it sounds like science fiction to you, it sounds like science fiction to me, but it's amazing that this kind of research is going on and there are startups in this space. Well, we've only got a few seconds left, Calvin. I want to thank you for coming on and I really appreciate your time and it was great hearing from you.