Venture capitalists have a reputation for wielding ruthless insights, judgements and criticisms of companies, tools and talent.
But how does a former-CTO-turned-VC assess companies? And now that he’s the one with the money, what does he look for - and expect to find - in a company’s engineering team?
Enter Jason Warner on the Dev Interrupted stage.
Jason is the most unique of investors: He made the jump from being a CTO to managing his own $725M fund at Redpoint Ventures. After leading GitHub to one of the best exits in the history of Silicon Valley, Jason is now dedicating his full time to picking the tech, teams and products he thinks will define the next decade and beyond.
In this freewheeling conversation, we cover everything from Jason’s start in tech carrying boxes while interning at IBM, his tenure as GitHub’s CTO, the reason full stack engineers may be a dying breed and why Redpoint Ventures avoids “yoloing” investments, even if it’s a lucrative practice.
We hope you enjoy the conversation as much as we did.
Episode Highlights Include:
- (2:53) Jason's role as CTO of GitHub during its acquisition
- (15:20) Decline of the full stack engineer
- (21:38) Jason’s first investment in the “AWS of crypto”
- (24:15) How Jason judges engineering teams now that he is a VC
- (33:32) The best approach to rolling out massive org change
- (40:29) Why it pays to “Yolo” investments (and why Redpoint avoids doing it)
Starved for top-level software engineering content? Need some good tips on how to manage your team? This article is inspired by Dev Interrupted - the go-to podcast for engineering leaders.
Dev Interrupted features expert guests from around the world to explore strategy and day-to-day topics ranging from dev team metrics to accelerating delivery. With new guests every week from Google to small startups, the Dev Interrupted Podcast is a fresh look at the world of software engineering and engineering management.
Listen and subscribe on your streaming service of choice today.
Transcription:
Dan Lines: Host
Jason Warner: Managing Director of Redpoint Ventures
---
[Music plays]
Jason: [0:00] Do you understand why you've made technological choices and the tradeoffs that you've had to do to get those. So, you know, if you're an app company, and you're struggling to get people to sign up and use your app, and you're talking about a fully distributed systems backend, using Kubernetes, and ultimately scalable to tens of millions of people, you've lost the plot. You’ve not-you're not getting enough people. Focus on keeping your costs down and your infrastructure simple and get more people to use your app on a regular basis.
Producer: [0:27] Hey, everyone, it's Connor Bronsdon Dev Interrupted executive producer. So you might have heard that back in September, we held our first ever live virtual engineering leadership conference, Interact. With over six hundred engineers and engineering leaders in attendance, Interact was a big hit with our audience, and produced some amazing interviews like Maria Gutierrez, his talk on scaling engineering processes, which has become our biggest episode of all time. So, I'm incredibly excited to announce that Interact will be making its return on April 7th, you can register or learn more on our website, which is provided in links. We hope to see you there. Now on with today's episode.
Dan: [1:01] Hey, everyone, welcome to Dev Interrupted. I'm your host, Dan Lines and today I'm joined by Jason Warner, Managing Director of Redpoint Ventures and former CTO of GitHub. Jason, thanks for joining us.
Jason: [1:15] Thanks for having me.
Dan: [1:17] Yeah, awesome, man. So, a little background on you that I've written up. So, you spent four years as a CTO at GitHub. You've been an angel investor. You've been a VP of Engineering for Heroku. You know, these small companies that nobody knows, just kidding, you've done a little bit of everything. Most of our audience are engineering leaders themselves and I know they would love to hear kind of the story of your journey from being an engineer to an engineering leader, investor, and now a VC. So, what's your background? How did this all happen?
Jason: [1:49] So how I got into engineering is serendipity basically, grew up in Connecticut, in the States, didn't have a computer in the house for a long time. I wasn't one of those kids and started coding at eight. Essentially, what happened is, in the town that I was-where I was going to high school, IBM was there. And they would hire high school kids to basically carry printers and computers around, high school Co-Op kids. And they got a tax break for that. And when I applied, because my auto shop teacher said I should apply. Because, you know, that's a way for kids to break out of circumstance. I was the only kid that applied that didn't have a computer or a car that they got on a 16th birthday. So, this is exactly why these things are-were-these types of internships exist. So, I literally got my start in computers by carrying them around IBM's, building for a tax break.
Dan: [2:43] That’s awesome.
Jason: [2:44] And then I slowly started asking people, What the hell do you do with these things all day?
Dan: [2:48] Life is interesting how that stuff happens. And now look, where it's led you! You did the four years or so at GitHub, tell us about that because that's just an unbelievable company and there's a $7.5 billion acquisition by Microsoft, like, what was that like for you?
Jason: [3:08] I mean, obviously, it's a-it was a ride, in all the good and bad senses, but in the good sense is it was a ride, it was fun. GitHub there's not a developer on the planet who doesn't know what GitHub is.
Dan: [3:20] Yeah.
Jason: [3:21] And so when you-when you think about the ability to influence or impact the entire field of software, there's very few places in the world that could have had as large an impact, not in terms of maybe dollars, in terms of what we were making at the time, we were making quite a bit of money but there are only four entities in the world when I joined GitHub that could fundamentally change software. It was Microsoft, Amazon, Google and GitHub. That was it.
Dan: [3:45] Yeah. Yeah. At a planetary level.
Jason: [3:48] Yep.
Dan: [3:49] Everyone on the planet. What were some of those ups and downs that you could share with us?
Jason: [3:53] Well, I mean, obviously, like a bunch of the ups for the acquisition and feeling of that and releasing some of the-the-the features that we did, after we joined, you know, GitHub action is-is still one of the highlights of my career. The last thing that we incubated in the office of the CTO, which I ran, towards the end of my tenure at GitHub, we incubated Copilot, which you might know is the-kind of the artificial intelligence code editor, that was a lot of fun, so those are always some amazing highlights. Working with some of the smartest engineers and some of the smartest people in the world, and including sales and marketing, GitHub’s had some remarkable people and any-any of those functions was great. You know, in the lows, it's a startup. The lows are, it's like-it's chaotic, and it's crazy. And this happens or that happens to this person or that person, or all those sorts of things. It's no different than that, it just so happens that the fairy tale outcome of the acquisition was written.
Dan: [4:45] At GitHub, when you're working on like a feature that-do you ever think like, “Wow, I'm gonna” I mean, this is what I would think, “Wow, I'm gonna impact the planet, for better or worse.” Did that stuff run through your mind or you're just like, “I don't know, screw it, like, let's just do the best we can.”?
Jason: [5:02] You know, there's, there's two ways to look at that one, which is you have a responsibility to do something with it. And that's actually how I looked at my tenure at GitHub, which is, if GitHub does not succeed in becoming one of the most important software companies in the world, for this entire one-hundred-year Epoch that we might be in, it's a failure at an epic level. And I looked at it as a responsibility to make sure it didn't happen from a corporate perspective, but then also for developers, to give them something wonderful. So, I never thought of it as like, “Oh, my goodness, this is gonna be so much fun.” I thought of it as like, “Oh, my goodness, we have to do something with us, because we're one of the few people that can positively impact developers lives.”
Dan: [5:41] That's awesome. What was like the-the biggest learning experience that you took away from the time at GitHub?
Jason: [5:48] I thought, there is a bunch of different, like, micro that can become macro learnings at GitHub but, you know, when I-when I started, at GitHub, the functions I ran were product, design, engineering, data in for security, and I had support initially. If you think about the size of that organization, and what it looks like, a lot of my learnings were-for organizations to scale appropriately, it’s literally about the people and the people at the top setting out macro directions and being able to hold people accountable for the for the execution of those, then from a company perspective, one of my biggest takeaways is even at GitHub scale, if you've developed something people might not use it, might not find it, and they might not understand what you're trying to do so you have to scream it from the rooftops over and over again. To this day, when I have, you know, it's five years is going-going up five years since when I joined GitHub originally.
Dan: [6:43] Yeah.
Jason: [6:44] And I'll still have people say, “Oh, it's kind of like SourceForge.” I’m like, “Have you not been paying attention for the last three and a half-four years on what we've been doing?” And it's-it's not on them, that's on the company to yell that. And so even if a company as large as GitHub can't get people to think past being SourceForge, 2.0? What would it mean smaller companies are going to find?
Dan: [7:04] What does that mean? Is it kind of just you got to get everyone going in the same direction? Or is it about repeating yourself over and over and like, let's keep, like continuous communication like those-those are some of the things that come to mind for me.
Jason: [7:17] It’s about repeating stuff over and over.
Dan: [7:18] Yeah.
Jason: [7:19] I think internally and externally. And there's I don't know where the sayings have come from. But I will say my way, which is I basically say, to be heard, you have to say something seven times and to understand something, you have to hear it seven times. So that means that you might have to say something forty-nine times for an individual to just hear and understand what you're trying to say. And that is inside your company, which has the most fidelity about what you're doing. Imagine what you have to do to the outside world.
Dan: [7:46] Yeah. You mentioned that something about accountability in scale, lots of, I guess, kind of departments that you were overseeing. Like, how many people are you talking about?
Jason: [7:57] Well, when I, when I joined, it was roughly I want to say was around seven-hundred total people at GitHub and probably like five-fifty of them reported to me, when we got acquired, we're right on a thousand people, and around eight-hundred of them reported to me.
Dan: [8:12] Yeah.
Jason: [8:13] And at one point, we topped out, I'm not sure how to get how big GitHub is right now, but we were closer to around twenty-five, twenty-seven-hundred people at one point, because we acquired some internal divisions from Microsoft and brought them in.
Dan: [8:26] As a CTO, how do you think about accountability at that scale, you know, we have a lot of listeners, they got growing engineering teams, or you know, we'll get into it, but hey, we're receiving rounds of funding, we're gonna double triple our engineering organization, accountability for that, from what I've seen it, it becomes different when you're a little bit bigger. Any thoughts on what that means or how to hold, I guess, an organization accountable at that size?
Jason: [8:52] I think you've got to set out waymarkers, which what I mean by that is say, “Hey, we're going to go do this, and we're going to do it by this date.” or “We're gonna hit this revenue number.” or “We're going to release this feature, and it's gonna be able to do this.” And if you can't, you have to give people ownership over those things and you have to expect that they're going to own it. But the flip side of ownership is accountability, and so they have to be accountable. And so, when you say hold people accountable, the flip side of that, what I'm really-what I really want to say is they're not-they're the owners of this, if they can't claim accountability on this, they're no longer the owner of it. So that might sound Brutalistic [sic] but that's just reality, if people can't claim ownership and accountability for the thing, they're not leading the thing anymore, they're-and they're not leading in the first place, so they're not going to be even, quote unquote, “leading” it anymore. So, I think when you when you want to hold someone accountable, you have to give them the thing. You have to tell them what it means if they succeed and tell them what it means they don't they have to be on board with that ride.
Dan: [9:50] Yeah, I mean, probably an even better sign is that then they're telling you what it means to succeed or fail. I'm owning-thanks for giving me the ownership. Thanks for the direction. Now let me give you, you know, a few months go by or whatever, let me give you an update of what I think it means.
Jason: [10:06] That’s when you know the organization’s working well.
Dan: [10:08] Yeah, very cool. Today, you’re a venture capitalist, you know, you're not just making investments in the companies, you're kind of making bets on the future. Maybe you consider yourself a futurist or something like that. What are some of the developments, product trends, you believe will define the next ten years?
Jason: [10:29] So, ten years is a is an interestingly long, but also short time horizon, it's kind of fascinating in that way, it’s like, we'll see a complete company cycle, maybe two macro-economic cycles, interestingly. What I think is at play right now in the world is the world's obviously smaller, the internet is a thing, the geopolitical environment is, let's just call unstable to be like, there's a lot of stuff that's going on in the world, we don't really know what's going to happen. Everything's getting more sophisticated, faster, more complex, faster, information flow, faster. So, you know, trends that I see happening over the next ten years, are increased awareness and investment and things like compliance and security. I think that, you know, if-if companies don't have chief compliance officers now, they likely will in the future, I think a lot of companies are understanding that selling is about adoption first and selling second, someone's got to reach for you first and then they're going to find a value prop, and then they're going to want to give you money if they're finding utility out of you. And you know, in classic VC terms, you call that bottom-up to top-down selling motions, but like what it is, it's just an adoption curve, it’s a type of adoption. I think, outside of that too, then there's interestingly, a lot more emphasis on investing in hardware, again, and America in particular, owning its hardware manufacturer, which I think is obviously good. And then I think it's interesting, when you see the geopolitical environment of how we might have to invest in-in more sophisticated tooling for national security type things. But more than that, it's like understanding that we are no longer a single, you know, micro-geo unit called United States or a state like at GitHub, you’re a global platform, you have-you're regulated in many different ways, or you're attacked from people around the world in many different ways. And it's not just Palo Alto, or San Francisco, let-let alone California anymore.
Dan: [12:20] Yeah you kind of mentioned the bottom-up motion, right? Some people say like a PLG Earth motion a bottom-up motion. Why do you think that matters so much? Like a VC perspective, if I painted two companies, for you, I don't know, maybe a top-down and a bottom-up and some of the KPIs around revenue look the same. I hear a lot of VCs. “Yeah, I want a bottom-up company.” Why is that?
Jason: [12:44] It's pretty simple really. It's that a top-down is actually pretty predictable. If you can-we understand how to sell, we understand what a great enterprise salesperson looks like, and we also understand what the ceiling for that might be and also, like kind of, the long year over year chunky chunk, chunk, like what it could look like and then they're great businesses, if you can nail those. Salesforce is a really good example here of a very mediocre to subpar product with an excellent sales team. And you could just count on twenty year or 20% year over year growth. The difference is that a bottom-up paired with an excellent top-down can become a stripe. And they can do that really quickly, and it can-it's one of those just, if you see it in motion, you understand why it could fly off the shelf so quickly?
Dan: [13:31] Yeah, yeah. I mean, it's a lot of the things that we're doing at LinearB with developers. And the way that, you know, I think you painted it pretty well, but kind of the way that I think about it is maybe you have a traditional top-down company that could achieve a certain milestone within ten years. Some of these bottom-up companies, you know, you can accelerate that one, two, three Okay, three years in-four years in, we're already superseding that other.
Jason: [13:58] Yeah, the difference I talk about here is that, like a top-a top style selling motion company would look like a very clear steps, you know, okay, now we're gonna go 20%, and we're gonna go 20%, now we're gonna grow 20%. The difference is that it feels like if you can really nail bottom-up PLG and you really nail a certain segment of the market, and you have a big enough TAM market, you basically are flying through the dinosaur apocalypse, evolutionary phases, whereas yet the touchdown selling motion companies are still gonna chunk their way through. So, you could just *fwoosh* gone, we just accelerated about five years, what might take a top sound company, you know, twelve years to do or something like that in the same sort of timeframe.
Dan: [14:36] For our engineers that are listening-so you mentioned a few-few trends, compliance and security, kind of what I'm calling the PLG or bottom-up motion, and you brought up hardware; How do you think this stuff will affect engineers, developers how they work over this time period?
Jason: [14:55] When I started out developing, back when I said when I was eighteen over at IBM carrying printers around and stuff-
Dan: [15:00] Yeah.
Jason: [15:01] It was pretty-pretty nuts. Because it was-it was-it was hard and arduous, and the internet was barely a thing, and it was kind of annoying to program stuff, but basically, you could write an entire-an entire application for the-for the web or internet, in Perl or C++, or eventually Java a year or two later, or something like that. Now, the stack that you've got to use is just kind of insane, you know, all the way from the-the full stack, Ruby Rails plus whatever web framework you're going to use, plus whatever database you're going to use, and maybe you know, all the different infrastructure you got to put around it. I don't think that's going to go away. I think it's going to get harder in some cases, but I think what you'll will see is we'll see a specialization again, where people don't, you know, who will still have full stack engineers, but will still say, “Hey, I'm really on the front end, I’m the app side.” or “I’m on the infrastructure side.” And we'll see people start to talk about that more. I still think that we think of engineers as these monolithic people that are like, “Oh, you're an engineer, I need someone to redo my entire website.” like, no, no, I'm a back-end engineer. That's like a very different thing than a front-end engineer.
Dan: [16:05] Well, I remember like, not too long ago, being a full-stack engineer was kind of all the rage.
Jason: [16:12] Yeah,
Dan: [16:13] But I don't know, maybe like, you're saying, hey, that trend might be dying out.
Jason: [16:16] I still think there'll be full stack engineers. But I think you'll start to see people specialize a little bit more and be open with the idea that they’re specializing a little bit more, because operating at a Roblox scale as a back-end, infrastructure engineer is very different than operating as a full stack engineer for a series A company, which is very different than operating as a, you know, front-end developer for, again, like a series D, pre IPO company, they're all-they're all very different things and you might find that you're specializing in a very different way. I also think that one of the other trends that we will see for developers in their lives, are that they are getting better tools, they’re just getting more of them. And they're going to do more specific things we’ll have more generalized platforms as well, but like more specific and general at the same time. And so, this choice might become almost paralyzing. And we have way too many ways to do the same thing. And maybe that goes away. Maybe it doesn't, I do think that like with any market downturn, you have consolidation. So, like, you know, projects that shouldn't have been some companies will probably go away, but there’s still going to be a lot-a lot of choice. So, we're going to have to find ways to navigate way too much choice.
Dan: [17:27] Are you looking into investing into developer tools or companies that are focused on there or open source? Like how do you?
Jason: [17:35] I invest in developer tools, open-source things that should be developer tools, I, my-my big areas where I focus what I call infrastructure, so infrastructure, and security, compliance. And infrastructure is a massive category, by my definition so infrastructure could be FinTech infrastructure, a new primitive for the entire financial system, or a new primitive for the new data systems. So, it's a very large, like Databricks is an infrastructure company, as far as I'm concerned, or Stripe would be an infrastructure company, by my definition. So, like, I'm giving myself a very wide category, but what I'm-why I'm intentionally doing that is because I'm saying that they're new primitives for the way in which we're going to work.
Dan: [18:16] That’s interesting that you're kind of extending the boundaries of what-it's like infrastructure in a specific sector or something like that.
Jason: [18:24] Well, the easiest way too, for people understand why I might be interested in in general is, I'm a distributed systems engineer by heart, having scaled some of the largest systems in the world, these companies that I would be interested in, also have that same problem. And for tackling something that requires you to build out like massive, massive systems in the backend because you've one, become that successful but two, you’ve become that complex, that's likely to be a new primitive for the modern era.
Dan: [18:51] Thinking about like the developer tools, something you said, I mean, there's so many of them out there, I think, what you actually said is to get one task done, there's many, many different ways to get a particular-like solve a problem. What do you think it that will be like for developers? Like, do you think that complexity or I guess that choice is going to be increasing over the next ten years? Or do you think it's going to come the other way, like, it's too crazy, there's too much out there. Let's scale it back as a community.
Jason: [19:23] I think it'll always increase. But I believe that we will have phases of expansion/contraction. So, what if you look back historically, forty years ago, I believe each phase is kind of represented by an explosion of tools and options, then a consolidation on the winners. Then the next phase is built on those with a more-another explosion, but at a higher level of abstraction, and further consolidation on those winners, blah, blah, blah, keep fast forwarding. I think that will happen and will continue to happen because I think that's the way that innovation does-does continue. And I think like whether or not you would define us as in an expansion or contraction phase likely depends on your viewpoint. I believe that we are in an expansion phase, but due for a contraction.
Dan: [20:09] A little bit now about your time at Redpoint, I know you're still relatively new. How long have you been there, seven months? Yeah, seven months in. And so-so you've come on as a partner?
Jason: [20:24] Yep.
Dan: [20:25] In a new $725 million fund called Omega Four.
Jason: [20:30] Yep.
Dan: [20:31] Can you tell us a bit about the strategy of the fund, like what made you choose to join Redpoint? As opposed to some, I assume you had a lot of opportunities of where to go.
Jason: [20:43] So Redpoint is an older firm, they are one of the first quote unquote “internet funds” out there. But obviously, no one calls themselves an internet fund anymore. They're just a fund and a firm. You know, same way you don't call yourself a mobile fund anymore, you're just a fund, and Redpoint, historically, invest in things that I find very interesting. So, it's early-very early investors in Heroku, HashiCorp, Stripe, Twilio, Snowflake, Netflix, you get the idea, like historically relevant companies, amazingly large, amazingly influential, and again, primitives in their own right, on the internet. We just invested in a couple of other ones that are doing incredibly well in this modern way, which one being like Cribl, which is doing it at the-it's like a newer version of, like, log visibility and Splunk type stuff, if you think about it, and Ramp, which is like revamping the payments, systems and the credit card and back office for all these companies in the world. And in my-my very first investment is one called Alchemy, which is in the crypto infrastructure phase. And it is literally the primitive for the entire crypto ecosystem, like there is no other, and it is AWS for crypto. And that's what Redpoint wants to invest in. And I was super attracted by that plus their historical like, you know, the reputation. I was also attracted by the fund strategy, which is different, it's not a huge fund. And what their strategy is they want to be a multiples-return fund. So, you know, one, two, three, five, ten X return, as opposed to some others which want to play an IRR, or basically, they want to have markups. They want to get markups from other-other firms and say, “Hey, we're returning-we're returning theoretical capital at this rate.” I prefer to think of it as just, you know, dollars return, you gave us this much we return this much to you.
Dan: [22:26] So now that you've been on both sides, like you know, previously you were growing businesses, you're the CTO of GitHub, you're an important position trying to increase the value of the company. And now you're on the other side, trying to think of like, who should I invest into? Can you tell us a little bit when you think about what is a healthy company to you? I would love to see both sides of the coin there.
Jason: [22:50] Well, I think healthy at each stage matters. So, if like you're a seed company, what is healthy look like or you're an A or B? It's very, very different. But-But let me step back and say like, with Heroku and GitHub, I'm of the opinion that if you can capture all of the attention, and the zeitgeist as I call it, like you become mainstream enough where you-every developer wants to use you or every one of your user sect wants to use you, and you can't figure out how to monetize it, that's a very much “you” problem. And so, my goal is really to capture as much of the attention and Zeitgeist as possible. So, I want to when I when I think about crossing the chasm, in any way these companies, it's not about the revenue flywheels like that will come and I get that. But it's more about like the fact that like the signups are happening. And the utility and the signups and retention are there and just growing, growing, growing, growing, growing. And anytime any of those things flatline that freaks me out. So, I want to look at signups I want to growth, I want to look at usage, then I'm going to look at revenue. So, revenue to me is a trailing indicator. And if you're a bottom up, top-down selling motion company, so that’s why I like those businesses, I just really understand them really well. And traditional tops-down selling motion companies, you have all the usual suspects. But the most important one that I think about is do you understand how much it costs you to acquire customer? And what's your-your churn? So, I want to understand, is this a leaky bucket? Are we spending money just to lose some? Or do you know how much it costs? And can we make a little bit more efficient? And then what does your machine look like?
Dan: [24:15] When you're looking at a company now from an engineering standpoint, do you ever look at the engineering team for certain signs? Or is there anything that would tell you “Hey, this looks like a great engineering team.” or “I have some concerns.”?
Jason: [24:30] Yeah, I mean, I think it's usually pretty, pretty straightforward to understand whether or not again stage dependent on what kind of engineering organization they are. And again, like infrastructure companies are different than SAS companies are different than app companies in some ways. But like this is a very simple frame is “Do you understand why you've made technological choices and the tradeoffs that you've had to do to get those?” So, you know, if you're an app company, and you're struggling to get people to sign up and use your app and you're talking about a fully distributed systems back end using Kubernetes and ultimately scalable to tens of millions of people, you've lost the plot, you've not-you're not getting enough people. Focus on keeping your costs down and your infrastructure simple and get more people to use your app on a regular basis. And this kind of goes to my point about like, hey, what, to the engineers, the engineering team understand what they're building, why they’re building it, and you know what the tradeoffs are.
Dan: [25:26] And the types of companies where you describe looking at signups, I assume you're looking at usage, like signups to usage and you want to see it growing and you're saying, “hey, the revenue side of it, we can take care of that.” Have you always seen, especially in the developer space, if you get lots of developer attention, lots of developer usage, there's always a way to then like, kind of monetize the business? Or are there any examples where it's like, it didn't happen?
Jason: [25:55] So there's lots of examples where it didn't happen. But I don't believe that to be the case-I don't believe that to be a flaw in the strategy. I think it's an execution issue. So, without naming names, because you don't want to do that to the companies, there are some companies that totally cross the chasm for the Zeitgeist and we use them on a regular basis, but they're not-they haven’t monetized it and they haven't figured out how to monetize it. And that's a shame. But they're not that dissimilar from how you do it in any other space, it's just, you've got to find a way to extract that value on a regular basis. Now, the problem is that most people don't understand why they got this success in the first place. And so okay, that-you know, you could, if you don't understand why people reach out for you, and why they come back day after day, you don't understand how to sell the value, and then how to, more, create more future value with new product, new expansion, or new capabilities.
Dan: [26:47] That's interesting, because if you would think you would understand that, but I guess sometimes you could, I don't know, create a great product, maybe get a little bit lucky.
Jason: [26:57] I think luck plays a massive part in early traction for some companies. And if they don't take time to be a little bit introspective, to be honest with you and say “Why did we get so lucky? What was it? What like, what knob did we turn? Or what lever did we pull? Or button did we press?” Then they're just gonna start spaghetti-ing stuff later, because they're gonna try to recreate that magic.
Dan: [27:17] I would think if there was a company in that state, let's say we have millions of developers using but we can't figure out, I don’t know I guess, a revenue model or like, figure out what the value is; Have you seen those types of companies maybe bring in different, I don't know, more sales oriented people or different-the, hey, let's fit, fit, come and figure this out? We got something good here, but.
Jason: [27:37] It’s easy, sales-sales is a very easy solve in that case. If you want to just lay around enterprise. Sales you’re gonna add, you know, role based access controls, you're gonna add a couple of compliance check boxes, and you're gonna let the vehicle sell this, and it usually works decently well, the challenge is, how can you monetize the credit card side of it, or the online side of it, and that's typically where a lot of them break down, and, and then not lose their soul as a bottom-up, you know, developer oriented or user oriented company and not just become an enterprise company. Not that I'm saying that's a bad thing. But it’s different, it’s a different company at that point. I think it's pretty common to bring in a very seasoned product engineering leader to do the bottom-up and try to try to accelerate kickstart that, and then you bring in a COO or a president, and then a CRO, and then you try to optimize the enterprise sales.
Dan: [28:28] I asked you a little bit about, you know, maybe what you look for in an engineering organization or an engineering team. What about specifically, if you're, you know, thinking about investing into a company, what would you look for? And I know it depends on the size, but what would you look for from a great CTO the position that you have been in.
Jason: [28:47] So I think CTOs and startup companies are, they get a rough go a lot. And what I'm looking for there is I'm looking for like, honestly, someone who's super pragmatic, super thoughtful, driven, but also kind of an athletic person, and what I mean by that is like a very common term in San Francisco, but what I mean by that is, they can bend and start thinking about the business or they can flex over and start thinking about technology. They don't necessarily need to be the best at any one of those things, but they do have to spike in one area. And on that one area, you want them to spike on is you want them to be excellent technologists. You know, that doesn't have to be that they're world class. Obviously, you want them to be, but they have to be in their domain capable of identifying world class engineers, and then motivating world class engineers. But CTO is a little bit harder to evaluate for what it's worth than other ones, CEOs. CROs or CMOS.
Dan: [29:40] Now you're kind of more rare or unique that you kind of have this engineering background and you’re a CTO and maybe a technologist and a lot of engineering skills, but now you know, it seems you're on the business side or maybe business savvy or investing savvy. What are some of the traits or behaviors for people listening, that maybe are on a very technical path, but would want to be like you someday? “Okay, let me get over to the business side or investing side.” Like what are some of the organic traits that you think you have that allowed you to do that?
Jason: [30:15] Sure. So, I appreciate the comment. And I also do you think that if I had to go back and do it all over again, I'm classically built, more like a CEO for what it's worth, in terms of like the way I think about the business, I've always said that I never wanna take a CTO job, in which I kind of step into the CEO seat and understand the business intimately. And it's the way I work or the way my mind works. But I think-let me describe myself as an engineer, I'm a distributed systems engineer, you know, infrastructure, distributed systems. But I always consider myself a very average engineer in terms of lines of code, the module, the functions, I was an average engineer. But the moment I became a tech lead, I was better. And the moment that I became an architect or a manager, I was better still. And as my role continued to progress in terms of tech influence, and tech thinking, I was much, much better. And that's the difference between programs. I heard this defined to me this way that's different between programming and engineering. So, I'm not a programmer, I'm-or an average one. But I'm an engineer and I build things I built-what I build is systems, and I build large distributed systems that operate, and that basically, if you apply that that's the same thing as organizations, just organizations are organic, distributed systems that actually have a higher loss ratio.
Dan: [31:37] So they're just going to go to that point. So, you know, it seems like you have a good mindset for systems, distributed systems. Can you describe maybe how you use that mindset to figure out a company or an organization is? How does that translate for you like what goes through your head?
Jason: [31:53] So there are a couple of things. One is you're constantly thinking about, there's always a picture in your head of your organization and as its evolving, and it’s constant kind of like a rotating planet, to a degree. You're like watching it in your head, and you're watching information flow, and you're watching the way it's supposed to be going through the organization or where hotspots are in terms of like, hey, this isn't going well over here, and you're kind of like trying to zoom in, zoom out constantly. I'm not sure how to describe it, other than that's exactly what's happening in my head. It’s the same thing as a system, when you're kind of evaluating it holding all that information in your head as you're building it. It's like, okay, I can totally see how, once this is live, this part of the system is going to get creaky. Let's go investigate what's going to happen at scale here that that's the same sort of thing that's happening. But then you're also applying, Alright, so now if I want to increase the dial on X,Y, or Z, maybe a new product, or new line or new capability, or I want to like increase the quota on sales as an example, or put some pressure on marketing to make sure that we're getting growth numbers higher than we are right now, what's going to happen? And it’s the same sort of model in your head, what's going to happen, oh, we don't have the right leaders in place. If I do that, very likely, I'm going to stress this and it's going to break, here's what's gonna break. Okay, so let's go figure that part out so we can pull the model together.
Dan: [33:12] That totally makes sense. You know, on the engineering side, of course, you could like think-it could be a thought experiment, you could think about scalability and your design, but also you can actually test it in engineering. We can do load testing, we can do scalability testing, we could see what breaks, let's say, before it goes live to prod, is there any way to do that, when you think-think about a business to kind of do a stress test without just a thought experiment?
Jason: [33:40] There is and there's-but you have to layer on a couple of different approaches to this.
Dan: [33:44] Yeah.
Jason: [33:45] And also it depends on what kind of organization you are and where you are in your curve, and also where-how-what kind of leader you are and what kind of trust you have with the organization. So, like, if you're a top style, militaristic, just listen to what I say leader, and your organization has no agency whatsoever, and then kind of-you can do whatever you want to do, you just like ram something through, right? It could be brittle, like the change could completely break and people are gonna churn but that's how, you know, that's your approach. My favored layered approach to these things was one is you're constantly signaling future intent. You're constantly saying, like, I've been thinking about this, you know, in all likelihood that you're going to build something in six months, and you're slowly warming the organization up to that point, it's kind of like pre-selling your partner that you want to go on vacation in Hawaii in June by starting talking about it in October of the previous year. It's the same-same concept. Then the other is I have this concept of canaries, which are people throughout the organization in every single department at every single level that I talk to on a regular basis to get feedback and also test out thoughts. Like alright, so like, “Hey, I've been thinking about the marketing growth like in, okay, our growth team, is it like I have the impression that we’re stalled. Like has is that what-is that your impression? Or should we try another experiment, blah, blah, blah.” Back and forth back and forth that person. I'm not gonna meet with that person every week but I'm gonna meet with that person on a decently regular basis. And it's not a formal meeting in most cases, it's kind of like, I have a lab calendar to go find this person in the coffee sometime this week to have this discussion. I'm trying to give I'm trying to get back I'm trying to see, hey, is what we're doing working? Like where we are right now working or what's broken? Give me some insight. Also, like, hey, I've been thinking about this, what's your immediate gut reaction? And then, if you-say you want to go roll out a massive, massive strategy change and org change, two months from now.
Dan: [35:37] Yeah.
Jason: [35:38] Like it's already failure if you say you want to roll a massive strategy change and org change tomorrow, and you've not done any of this work, that's a failure already. But let's say you want to do it two months from now, you're going having conversations leading up to this saying, like, Hey, what are you what's your gut reaction to this to start, and they’re gonna noodle on it, they’re gonna give you some feedback, then you're gonna go back to them two weeks later, you're gonna say, “I've been thinking about this, and I can't get it out of my head. Like, I'm thinking we should go do this. And I need to know where we fall down when we do this.” And they're going to give you some more feedback. And then you're going to use that to retune your model and test it in your next set of conversations with everybody that you're going to go do the exact same thing with, like, what I just described too, is a shit ton of work. It is a shit ton of work, you're having lots of conversations time, time and lots of time, a lot of the same conversations over and over again. But it's valuable, valuable information.
Dan: [36:27] So we kind of talked about, okay, we're going to talk to a lot of people, we're going to put time in there. The other side of it is like data, reading data, or like describing an organization through data like how metric driven were you as a CTO?
Jason: [36:40] So it depends on the area. So, like, I'd want to know what the signups per day are, I can only look at certain high levels. And I have to assume that other people are taking care of the lower-level ones. And I would go and verify that with folks, like “Hey, walk me through how you do your SRE dashboard. Or I want to know about the Bolton's-I want to know about your security dashboard or the faults in production dashboard.” something like that, just to just to have a sanity check. But like, at the highest level, you're talking about business metrics. So, you broke them down like okay, user signups, user churn, revenue, you know, all that sort of stuff. And you got a level down, you're trying to figure out some other stuff. So, like, hey, it's really important to me, like, what's our ratio of men to women? What's our ratio of underrepresented? What's our pay bands? Are we out anywhere? Have we looked at geo pay bands? Like you're trying to use that you can't-that's not gut feel. One person could say it to you and the hair on the back of your neck could go up. And then you got to go verify with data. You know, you can’t just trust that someone says something is important. But it could be a signal to you to go look and dig in and get the numbers. And then you want to like kind of market comp yourself too like, “Hey, we're roughly this many people, what's our ratios? All that sort of stuff? Like how out of whack? Are we-are we efficient, are we inefficient?” That sort of stuff.
Dan: [37:46] When you're considering making an investment, I know we talked about the metrics of user signups. And I'm thinking of like a bottom up, you know, interesting SAS or too-tools or platform company. Are you just like saying, okay, the metrics look good, the signups are increasing are like, Are there any other metrics that you're looking for in that PLG motion? What is the thought process that you go through when I guess evaluating, step by step? Should I do this? It's a big decision.
Jason: [38:18] It's a big decision, especially if you're, you know, we at Redpoint in the growth phase will typically invest somewhere between twenty-five and a hundred million dollars. So, a large number on the growth side of the fence. And you know, what we're looking at is we're looking at a couple of different things, but it can be basically boiled down to, from a metrics perspective, are things increasing? So, you know, just-just look at the graph, there's like a very classic, and-
Dan: [38:44] Which way does it go?
Jason: [38:45] Yeah, but if like-is it increasing? And that's like-then you could look at raw numbers and make sure that like, the valuation makes sense, and for its time and market, all that sort of stuff but you want to understand that first and foremost. And the second, like, this is not a metric. But this is a really, really important question for the way that we at Redpoint invest. Is this an important company? Does this company incredibly matter for its space? And will it continue to matter? Because then you solve a lot of the growth issues if it has some by just being important. And then you get enough time to figure it out. But if, you know, like, how many social camera apps or social audio apps were released in the last two years, that went all the way up and all the way down within one hype cycle?
Dan: [39:32] Yeah.
Jason: [39:33] A lot of them. So, they're not important companies. They were important or even important, like functionality, they were ephemeral things. And we're looking at things that are new primitives, things that other people are going to build on. And that's a very, very different view than, you know, will it-will it hit a hype curve, or not?
Dan: [39:51] Just kind of trying to break down the thought process. Seems like if I need to summarize like maybe, you have some metrics, especially around growth and signups that you probably want to see going in the right direction. Like maybe that's step one, but then you're also trying to confirm, is this just like a hype thing that's gonna, I don't know, burst and go down? Or is it really gonna change the world? And is it like, okay, if it's a yes and a yes, then hey, we're go, or is there anything else that I missing in that process? It’s super interesting.
Jason: [40:22] No you’re basically hitting it.
Dan: [40:23] Yeah.
Jason: [40:24] Which is also remember too, we're-we're about twenty years into SAS companies at this point.
Dan: [40:28] Yeah.
Jason: [40:29] So SAS companies are really well understood in terms of how they grow. So, there's no real investor challenge to understand that if a company is growing, you know, two X, then you know, one point seven X, and whatever, and, you know, its enterprise sales looks like this. And, you know, that's why these crossover funds or other-who basically don't know, the industry can just YOLO invest into them, because they understand what these companies look like, on a net revenue perspective and retention, blah, blah, blah, like, it's all just excel spreadsheets, when-what we want to do is we want to, like have more art and the investing as well, because then obviously, you're not going to get boxed out by someone who just says, well, those numbers look great, we're just gonna YOLO this in this investment in. A lucrative strategy in some cases, but it's also they miss-they miss the fundamentals, but they can see the ones that are the trends. So, like, we-we expect those crossovers or YOLOs want to come in to the stuff that we invest later. Because what we're trying to do is we're trying to find the ones that should be important if it works. So, if this is-if we're, you know, I guess maybe if anyone's listening to this, and they're talking to a Redpoint investor, you should be flattered, because what we're thinking is that you are majorly important company in the future, you have the potential to land at that. And we-that's what we wanted to mean, if Redpoint invests in you, we want to basically mean that we think of you as a new primitive on the internet, or-or in whatever sector that you are, and other people are going to build upon you. And then it should become obvious in the next round of the round after that, from other investors so that the Tigers or the, you know, the crossovers. there's a ton of them are like, oh, yeah, that is a great company. Yep.
Dan: [42:06] Jason, just before we head out here, I want to switch gears back to your career, and the obvious success that you've had, is there any book or reading or mentorship that has happened for you that you could tell our listeners, “Hey, there's a few things that, you know, really changed my perspective, or got me going in the right way”? Any tips that you could provide?
Jason: [42:31] Sure. So, I think, um, there's a book by one of the old 49ers head coach, it’s called The Score Takes Care of Itself, which I think has been recommended many times. I quite like that book. And the reason why is, I'm a sports person, I love football and basketball, and I play team sports. And when you think about what it means to win the championship, many people will think about scoring more points. And what Bill Walsh was saying in there is, if we build the right organization with the right intentions with the right principles, the score will take care of itself. But you can't cheat that process and still have that outcome. You can-you can-you can fake it. And you could maybe get lucky, but you can't build sustainable without that. That was a very influential book for me and mindset.
Dan: [43:14] The score in sports as a lagging indicator of what you're doing, you know, behind the scenes. Very cool. Yeah. Bill Walsh, obviously, one of the grades.
Jason: [43:22] So that was a very important one. The-another one, obviously, Andy Grove’s book, where he introduced OKRs, and stuff like that I'm not-I don't want to advocate for OKRs. It's not what I want. But I want people to read that and understand the mindset that he was going for in that book. And to push it forward. I think that's important. And I don't-I don't think anyone needs to buy this book. But I think that Jocko Willink’s book about Extreme Ownership is a very, very good one, only because I think if you're building a company, you have to find people who exhibit this sort of attitude. And you yourself have to have it. And then there's a whole bunch of other concepts that have really influenced me, and I'll just kind of list them off here real quickly. I'm not in a book style. But I think about startups, as I mentioned before, as like bossy distributed systems and think about physics a lot inside them. So, like energy, neither created nor destroyed, just kind of like transferred. So, you yourself as a leader have to bring the energy, you got to bring the vitality, you got to bring all of that inertia. You can't-if something is stalled, it takes ten times the energy to start it up again. But if that thing of the same weight is moving just a little bit takes a ton less energy just to keep it moving a little bit. So, if you think about concepts like that, it'll influence how you think about your business. So, it really is really important to me when I was building companies because of those things that one, if anything is stopped get it started as quickly as possible. If there's a small problem, fix it because all small problems become large problems. If something is moving, never let it stop. No matter what you do, never let it stop, get it going. You know if-it's all concepts like that, now I know people want very tactical answers. And I can give some, I can say, “Hey, there's this concept of boulders, rocks, pebbles, sand, and give energy and percent time attention appropriately to each one. And each one deserves it.” But then they're going to want to ask which-which thing is which in which category and optimize it like, this is the art of building companies. And if you don't fully understand that, find an investor who does and helps you work through those things, because that is where you're going to, like, become an OK, outcome maybe to zero or a mega company.
Dan: [45:32] Very awesome. Thanks for sharing those insights. The Andy Grove book that I'm familiar with, I think, is a High Output Management is that the one you-yeah the one you're thinking, okay, great. Alright. Jason, thank you so much for coming on the pod today. This has been an awesome conversation, especially kind of pulling the curtains behind how you think and some of your thoughts when you're investing at Redpoint. For people that are listening to the pod, is there anything that you would like to kind of promote or can they follow you or get in touch with you? Is there any kind of final pitch or statement that you'd love to make?
Jason: [46:09] Sure. Follow me on Twitter, Jason C. Warner, I say some pretty stupid things there on a regular basis. If you're building a trillion-dollar company, come talk to me a Redpoint. [Music fades in] And if you really want to talk about building companies too, I just I love talking about that on a regular basis. So, infrastructure companies, dev-oriented companies, crypto companies, I love all those.
Dan: [46:29] Amazing Alright, Jason, thank you so much.
Jason: [46:32] Appreciate you having me. This is amazing.