From Pivot Hell To $1.4 Billion Unicorn

In just a few years, James Hawkins took PostHog from an idea hacked together right before YC's W20 deadline to a unicorn powering product analytics for thousands of teams. He joins YC's Brad Flora to talk about surviving six months of "pivot hell," why open-source analytics was the breakthrough, and how PostHog grew from fighting for its first users to launching full product lines—plus what he's learned about momentum, staying close to customers, and using transparency and humor to build a company that stands out.

• 0:05Speaker 00

  I'm here today with James Hawkins, CEO and founder of Posthog from the YC Winner 20 batch. James is here hot off the news of raising a 75 million Series Z round of funding at a $1.4 billion valuation, making Posthog the latest YC unicorn. Today, we're going to talk about how Posthog got started, how they pulled off what I think is one of the best YC pivots of the last decade, and how they use attention and humor to compete with not just other tech companies, but with everything out there for attention. Welcome, James.

• 0:34Speaker 01

  Thank you for having me.

• 0:35Speaker 00

  James, could you tell us what is Posthog? Explain it to us a little bit.

• 0:39Speaker 01

  Sure. So with YC, I always feel like this is my homework, being marked by the teacher. We help users to debug their product, to ship features faster through things like feature flags and so on. And we help keep all their customer and product data in one stack.

• 0:54Speaker 00

  And tell us a little bit about the state of the company, how big it is. You guys are pretty big these days. Give our viewers a snapshot of what's going on at Posthog.

• 1:04Speaker 01

  Sure. Yeah, so we're about 160 people. We're around 300,000 customers across like free, we have a lot of free customers unpaid. We have several thousand paid. We have, I think, 16 or 17 products in production or in development that are all at different stages. And we've been, I guess, at the moment, the main thing we're doing, and the reason why we did a lot of raising this year, we're pushing very hard on getting more products still out of the door. So instead of going upmarket, we're going just across more of the customer data that people are generating. And we're automating a ton of things through AI, which has been a long and arduous process. But we're starting to get to the point where I'm starting to get proud of what we've achieved so far. But I still feel like, ah, we haven't got started yet. There's a lot more to go,

• 1:50Speaker 01

  which I don't think is a feeling that will ever go, probably. But yeah, it's really a lot of fun. And we're just starting to scale up pretty aggressively now. We're thinking of starting at 70 people. This year, and so we're supposed to be about 200-ish by the end of the year or something. So we're starting to get properly and took on a scale up maybe at this point.

• 2:05Speaker 00

  So you've got this team, you've got all these customers, you have all these products that you're developing and improving on. But what was the initial product that Posthog started with? What was that, the first thing that you guys brought to market as Posthog?

• 2:16Speaker 01

  Sure. So yeah, the first thing that worked for us, we had a bunch of pivots before, was self-hosted product analytics. Because we pivoted so many times, and every time we had to set up product analytics over and over again, and we were just getting frustrated having to implement it. I think we found that there's an awful lot of really strong competition. But we felt all the products were built for product managers to force engineers to go implement. But it's reasonably technical, especially my co-founder is a very technical co-founder. I was just getting annoyed at like, I can't, I want to write SQL, I want to see the data underneath. I would like to keep this in my infrastructure so I don't lose data due to ad blockers. So we thought, okay, we think there's room for a self-hosted alternative. And we also, we spent some time talking to potential customers. And we found that quite a lot of people had self-built their analytics stack.

• 2:59Speaker 01

  And we're maintaining this kind of janky system, which we felt if we can productize something like that, because new infrastructure and we could see this working quite well. So yeah, we started with just open source product analytics on Hacker News. We spent like the last four weeks of the batch desperately trying to get that out of the door. And then we actually, we pivoted, we iterated, I would say afterwards. And now we're all cloud-based with multi-product, but that was the thing that landed for us finally.

• 3:24Speaker 00

  I think at Posthog, well, there's this meme among founders of pivot hell.

• 3:29Speaker 01

  Yep.

• 3:29Speaker 00

  And it's a thing that YC founders talk about a lot, where you're going from idea to idea to idea. And you're trying to find something that's going to work. And you guys not only came out of pivot hell and have built this awesome company, but what you built was literally forged in the fires of pivot hell.

• 3:44Speaker 01

  Yep.

• 3:45Speaker 00

  Right? Like while doing these pivots, you had this pain that only came because you were doing the pivots and trying to do them at a high level, setting up the analytics over and over again. Tell us a little bit about how you guys actually got to YC and started this journey of pivoting and finding startup ideas that might work.

• 4:02Speaker 01

  My co-founder. Quit. We used to work together. And when I heard news that he was leaving, I'd also been thinking about leaving and that kind of triggered me to go, okay, I want to do my own thing. So I made sure to grab him and instead of getting him to go off and work in another startup or whatever, he thought, hey, let's just do our own thing.

• 4:19Speaker 00

  And when he was leaving, was he leaving to join another company? Was he leaving to start something?

• 4:23Speaker 01

  Yeah, I think he was actually trying to get a job at Facebook at the time. He was doing an incremental kind of career move. And I think I showed up being like, I think I have this idea that we'll probably have product market fit for in a couple of weeks' time.

• 4:33Speaker 00

  Of course, of course, they are.

• 4:34Speaker 01

  Of course. Yeah, it's only a hop, skip and a jump away and then we can just, you know. I had like a, I'd created a list of problems I'd encountered professionally, thinking like, okay, I should solve a problem I've had myself. And we started going through the list that we thought we wanted to bootstrap initially. So the two of us were just like working out of coffee shops. One of the main problems we found was every time we went to see a customer, we'd have to buy another round of coffee. And so we'd often end up going like, whatever the coffee equivalent of waste it is, basically. So eventually we did this for about six months and we worked through an average of one idea every like five or probably average of about five or six weeks.

• 5:07Speaker 00

  Do you remember what any of those ideas were?

• 5:09Speaker 01

  Yeah. So the initial thing we tried to build was a sales territory management product. So I used to run a sale, I was VP of Sales. And I felt that like in sales, you're literally wasting over 90% of your time because deals don't usually win. They don't usually close. Like they usually just float around and gradually get killed off. And we thought if we get better at not focusing on deals that are closing, but on the ones that aren't closing, going a bit more ruthless, using statistics to pull them out of people's pipeline and change the territory for the sales team is what I thought might work. And it was just kind of really complicated to build. And I think because I'm much more technical than a traditional sales leader would be, I just think we built them kind of complicated. And I don't think our go to market was right or anything. Yeah, we had this idea, we built it.

• 5:51Speaker 01

  I think the thing that was important or the harsh lesson I learned was my co-founder was building it, I was just trying to get interesting. Prospects to buy it from us. So I got through, I got about 15 sort of Series B, C, D and sales leaders queued up saying they wanted to use this. Well, they said they wanted to use it. We got the first version done in a couple of weeks, send them the link to create an account of the 15. 14 of them didn't even click the link, one clicked the link and then didn't create an account. And we're like, it was like, first good, we should have read the MONT test was the, we hadn't done any of the user interviews correctly. But also it made me feel that building for salespeople or sales leaders in particular, I feel the signal to noise ratio is pretty poor.

• 6:33Speaker 01

  Because they're really quite willing to hop on quick calls and are very, very positive and kind of very friendly people. A little bit later, it kind of dawned on us that we think we'd be better because we're not very good at product. We're terrible at product initially because we're kind of bad. We need to deal with people that are what they say is closest to what they'll do. And we felt that like natural problem solver people like engineers or customer support people probably would be a better audience for us to work with. Because out of all the many variables you can get wrong, like it might be like, oh, your product's incredible, but you're terrible at explaining what it does. You'll still fail. And we thought like, okay, that one of the variables we can remove from this equation is the person being a little bit easier to read, basically. And so that was a big part of why we wanted to build something dev toolie later on was that sort of realization.

• 7:18Speaker 01

  But yeah, that was the first thing that we worked on.

• 7:20Speaker 00

  And so when you guys did Y Combinator, the product that you had was another pivot was having developers take surveys about technical debt. When they would submit a poll request. And so you'd already done a few products by that point. You came into YC with another one. Did you guys have a, like, how did you kind of keep your spirits up? Or how did you track your trajectory? Did you feel you were getting stronger or better or closer to the target as you were going through these pivots?

• 7:46Speaker 01

  I would like to think that was true. I'm not sure it was, we didn't reflect on that. We should have asked, like, are we actually getting better each time around? Like, I think we should run a retrospective, basically. I don't think we actually were getting that much better each time. There are definitely some things that were consistently very important to us, though. Like, for example, we went full gas on each idea. Like, we didn't sort of like call it in, like, we would go in person to every single customer that was remotely interested. We were going in person to go meet. We started in England and then came to San Francisco to do YC. Like, we'd catch like a bunch of trains and buses to get to a customer in some random place. And like, we would put the effort in to go really push it. Because then kind of what happens when we started charging for all of these products. You would find that, oh, I'm not failing this because I don't have a relationship. I'm kind of friendly with this customer at this point and they still don't want to buy from us. This is a really good sign. This product isn't solving the problem.

• 8:31Speaker 01

  This really demonstrates to me it's nice to have. Like in my previous job, we sell enterprise software for millions of dollars a year. And I can remember multiple times thinking I had to work kind of as hard to sell a like $50 a month SaaS subscription as I was to sell like a $100,000 a year plus deal in my last job. I know that I've tried really hard. And so I can rule out one of the other variables, I guess, is like, are you actually trying as hard as you could be? And we were. I think that helps us feel confident enough to go, this just isn't working. So yeah, we just, we worked very hard. And I can remember it feeling weird, like on LinkedIn, for example, I'd talk about like, hey, this is what we're up to. And it would be like one week it was like some dev tool, next week it's a CRM thing. But we just didn't care, like we just threw ourselves into each problem.

• 9:13Speaker 00

  There's an aspect of shamelessness that you have to have.

• 9:16Speaker 01

  Yeah.

• 9:16Speaker 00

  To just kind of keep up the momentum and keep trying and not get bogged down.

• 9:19Speaker 01

  I think it's easy to feel kind of lame almost, like, yeah, you have friends who are like in investment banking or they have these like proper careers and we're doing something that just seems so kooky. But like, that's the bit, I think that is a specific reason why most people don't run companies. Because that first bit is like, is so existential and all over the place. I think we just have faith that if we just keep making the input we're looking at is like, are we making very regular, real progress or selling the ideas? We'd often at the start of the week be like, okay, by the end of this week, what is the list of things? We will feel good if we get them all done. And it would be something like, okay, we want to do like 10 custom meetings, build these three big features. And we just held ourselves to account on the like very short cadence all the time. As people, I think we got much stronger at things like product much later on.

• 10:00Speaker 01

  And once the postdoc had sort of taken off, but I think we were just like very, we just, yeah, put our heart into things. We're very fast. That was kind of what we needed, I think, realistically.

• 10:08Speaker 00

  You guys were working through these pivots, maybe getting better, maybe not, trying to be wholehearted, trying to be consistent, trying to be really committed to the bit, maybe as a comedian would say. When you started with the open source product analytics, what became the final pivot? What signals did you get early on or signs that maybe this is it? Maybe we're really on to something. It was a month before YC Demo Day. Tell us a little bit about how that unfolded.

• 10:31Speaker 01

  So I think we had a slightly higher level of excitement about this idea ourselves. I think we had had a couple of ideas where in our heart of hearts, we weren't that interested in them. Like we're kind of doing them because we felt like we had to get something to work. Whereas I think this one, we had an office hours here and I can remember and we're like, well, we're kind of thinking about building this thing that will do two. At the time, we thought there are two things that are important. We thought we want to capture all user data completely automatically. Like we don't want anyone to have to set up event tracking manually. We didn't even realize that someone had already built this. And that didn't turn out to be particularly important. But then the second thing, we're like, we want to be able to self-host this thing so that it can save our infra. And we came out the office hours going, ah, that's not the right framing. It's open source product analytics. Then it's not just about it being your infrastructure. It's also just like a whole vibe.

• 11:14Speaker 01

  Like I think immediately we were like, we feel very confident. I could see that really landing on Hacker News. I mean, both, my co-founder and I both read it and it just felt like the kind of thing you would see on the front page. So the level of excitement was a little bit higher for us. I do think it's a bad idea though to, like sometimes I see other founders that like, I'm waiting for an idea that I'm excited about. But I think the reality is for us, it's, we had to just build a bunch of things to find the thing we're excited about. What we didn't do was like sit cross-legged on the top of a mountain, waiting for inspiration. So by trying a few things, we started going, ah, this idea just doesn't feel good, but we know that because we have tried it.

• 11:46Speaker 00

  You had to find something in the middle of what people want and what you're excited about.

• 11:49Speaker 01

  Yeah, exactly. And so I think doing stuff gives you much stronger, I think it's like a Brian Armstrong quote, but like doing stuff gives you much stronger signal, more information. And then I think when we started booking interviews, it was a little bit easier to get people to talk to us. But the big deal was on Hacker News when it was just on the front page and had like, it wasn't the best ever. Like at the time, it was pretty good compared to what happened before.

• 12:08Speaker 00

  It was a good launch. You had a good Hacker News launch.

• 12:10Speaker 01

  But if you look at it like sort of 2023, 24, 25 vintage Hacker News launches, it was very mediocre, but it was enough to get an initial audience.

• 12:18Speaker 00

  I think, because I go back to people's little Demo Day slides a lot of the times when helping batch founders figure out what they should say. So I pull up the post-talk slide every now and then, and I believe according to the slide, and I think it's true, it was the most upvoted dev tool Hacker News post of the year. So pretty good. I remember being in the office hour. You probably shared the idea with a bunch of people, but I remember the first time I heard it and it just made a lot of sense. There were a bunch of product analytics companies out there, but no one had really gone all the way to make this truly developer focused by going open source. And so it just sounded like something someone should try. And I wasn't that surprised that it hit the way it did and was really excited for you guys at the time. So, okay, so you had this, you were starting to get some users.

• 13:02Speaker 00

  You had to go out and raise money for this thing at Demo Day and you'd only been working on it for a few weeks at that point. What was that like? Like, how do you get your conviction up to go, you know, pitch investors about a product that's really only a month old at this point?

• 13:13Speaker 01

  It was very mixed. We, at first, we felt really good. We were like, okay, it feels like we're gonna have this done in like two days or three days.

• 13:19Speaker 00

  Slam dunk.

• 13:20Speaker 01

  Yeah, I think we're coming in going like, yeah, we had a really solid launch. I think we were still buzzing a little bit after like, oh, it's such a feeling of kind of relief that suddenly we've got something where it just instantly, the day we launched, it went from being a push basis to a pull basis, getting people interested. Like suddenly it's just like, we're trying to service all these random problems or feature requests and stuff we can see coming in. And so we went out to raise, I think, very confident. Like, we thought we'd kind of be pushing quite a big valuation at the time. And I think we felt that we will have this done in just a few days. That the timing was horrible because it was 2020 in March and the same week, basically, COVID went from being this somewhat niche thing to changing the world and everyone just suddenly pulled out. So we kind of went from feeling like, oh, we've got like all the fanciest VC firms in the world are interested in this thing to

• 14:05Speaker 01

  everyone's pulled out and we're now doing like, we're trying to get angel checks for 5,000 bucks a pop to try and get all the way up to like raising a few million dollars. This is gonna take a long time. So it was very peaky. We also had a little stress over, like, we're in the US still. My wife is at home and pregnant and I was worried about getting stuck in America and my co-founder left because he was also worried about getting stranded and overdoing his visa and stuff. So there's just all this chaos kind of happening. We just kept plugging away, basically. I think it was very easy to, it would be very easy, I think, to get disheartened and it was painful, but we just didn't. I guess we just kept going was the main thing. And so we were doing angel checks,

• 14:43Speaker 01

  happily taking 5K off of people until eventually we got a proper term sheet and that led to another one. And then we got the round done very quickly and then the market completely picked up, like even just a few months later. I ended up doing like a series A like a month after the seed round had closed. But I think the statistics are something like. I think we met 160 different firms in the seed round. I think for like, if I added up the total number of firms I've met for the series A, B, C, D and E, probably like 20 in total, like 30 in total. It was by far the hardest round we had to raise. But also we sucked at, it's not just the market, like we sucked at pitching. We're trying to people please, I think, too much. And so we would be like, well, they'd be like, what are you going to do with the money?

• 15:27Speaker 01

  And it's like, well, you know, we'll like kind of get some revenue, we'll build some features. It's just very vanilla feeling. We had a lot of people rejecting us because they felt we were too early. But the reality was, it's like, well, we're dealing with a lot of these investors have never invested. At that time, they never invested in open source at all. Like the vast majority hadn't because it was quite, it was much more unusual then. And I started realizing that I don't even believe that we'd want to spend money across all these areas. Like, we actually just want to go full all in on the open source project for a while. Again, I think that's very bad advice for something that doesn't have competition that's making lots of revenue. So have to be very careful with something like that. But we started getting much more opinionated and that's what started that eventual landing. So, hey, like, you're out of your mind if you think we're going to go big on like hiring a big sales team or trying to go into the enterprise or something. Like, it's just about building this like nice, big inbound kind of community.

• 16:11Speaker 01

  We think is the, if we can't do that, nothing else will follow. And we need to really establish that properly. So, yeah, we got much punchier.

• 16:17Speaker 00

  It's funny how that is. It's almost like investors are pretty agnostic on what the plan is, so long as there's a plan.

• 16:23Speaker 01

  Yeah.

• 16:23Speaker 00

  And that you're really passionate about it and you've thought about it and you're ready to go execute on it.

• 16:28Speaker 01

  I think there's a Mark Benioff quote, which is something like, it's better to be different than right. I think that it's almost equivalent here where I think it's, if you're trying to raise money, it's better to have a plan than it is to necessarily be like somewhat correct. Like be either very clearly right or wrong because you'll learn something from that.

• 16:44Speaker 00

  So that round was really hard. You talked to 160 investors. It took months to get the money?

• 16:49Speaker 01

  Yeah.

• 16:49Speaker 00

  Okay. Fast forward to this latest round that you guys just raised. Obviously, it's five years later, things are very different. The company's in a very different place. You have revenue and customers. What's it like to raise a $75 million round? You know, how is that relatable to folks that maybe are out there trying to raise the first money for their company? What's that experience like?

• 17:09Speaker 01

  Yeah, perversely it's easier. Like, we had, we spoke to one investor for this round. We raised from P15, this one called Chalendro that is the main partner, but also Onav. Is involved as well. The round before he tried to take part in, we only had just gotten to know him though. So we kind of felt like we want to spend a bit longer with you before we raised with you, basically. We did spend a little bit more time in them because we liked him as a person. They offered to do the next round preemptively and at first we thought it was a bad idea. Because we just felt we don't know how we would spend this. We haven't spent like the round before, the round before almost. But then at the same time, what was starting to happen was we were getting increasingly confident on AI.

• 17:49Speaker 01

  What actually happened was I went on vacation and read about some of the motives for why some of the co-founders built OpenAI in the first place. And I thought the reasoning was really interesting. It was humans have a brain. It's a physical object. We can therefore build one, basically. And there's no technical reason we can't build ADI, essentially. Then I just, and this is nerdy reading, but I started thinking like, what are the specific differences between your brain and LMs today? And I kind of concluded, there's a lot of things that could be built, but I don't understand why they don't exist yet. So I therefore think that AI will, I don't think it will flatten out. Like, I think the rate of improvement may vary. But I think over a long timeframe, it will be incredibly important to the world. So I got much more bullish on its importance. And then thinking at work, I'm like, okay, I can definitely see a path. I just spent some time reflecting with my co-founder.

• 18:34Speaker 01

  I'm like, okay, how could this shake out? Basically, like, should we change our strategy if we both believe this is this important? So at the time, we were thinking about building like a CRM, a support platform, all kinds of other data, customer data oriented products. And everything was kind of working, but we felt actually we can do so much more in the realm of product first. Like, we can build kind of product autonomy out. So we want to go deeper within our products that help you understand what to build. So we wanted to work on, like, at the moment, we're working on like a desktop app, for example, that ships pull requests based on your customer data. So we'll look across all your session recordings, all your analytics, all your error tracking, your LM traces. Go, hey, these are the issues we can see in your customer base. We've literally fixed them whilst you're asleep. Here are the pull requests. So instead of like, Please build me this feature.

• 19:18Speaker 01

  It will be, the flow is pool based. So it's like, hey, we've built these features based on what's happening. You can just review, close, edit, merge them, depending what you want to speed up development further. And we just were like, hey, let's just go, we want to be able to go all in on this idea. And so we raised to have just like the comfort and the confidence. So a lot of it's psychological, like, I don't think we'll manage to suspend it. But we want, that's kind of the point. And it's like, this means that we can just fully embrace our product strategy and take a bigger, and just like increase the size of swing that we're taking.

• 19:47Speaker 00

  Well, and it sounds like what you're saying is you're able to fully explore the implications of the original idea of Post-Hog, which is understand what's happening in your product and what your users are up to. And there's all this stuff downstream of that. Like, once you understand it, then you can do these things and take these actions.

• 20:01Speaker 01

  Yeah, because we started thinking like, okay, it's a bit like if you're trying to, if your customer is like a painting that you're trying to understand, if you have one data type, it's a bit like seeing a painting, but only being able to see the color blue. You're not going to get a nuanced understanding of a customer. Like, it looks like they're not using this feature. We should email them to cross-sell them the new feature we've just built. But at the same time, your other data might be telling you something conflicting, like, they're really pissed off in support right now. And this is a terrible time to automatically send them an email or something. So we kind of felt like, and that's what a human would do today, is you kind of look across your tools. And we thought, oh, we can basically just build something that does this. And so we set off trying to build basically a product manager.

• 20:34Speaker 00

  That's a physical thing, we can build it.

• 20:35Speaker 01

  Yeah, we don't want to build something that has like, that we can't conceive of. We want to build something that exists, but just like a better version. And we're like, well, there's a human product manager, has product market fit. Like, companies buy those. That is sort of where the idea came from. So instead of just building, because normally our strategy had been, what are the products have product market fit, have customer data? We will build all of them, it's more convenient. But then we're just like, well, also, like, you could consider the team members and thinking about, sales has product market fit in companies, so does support, so does engineering. So can we build those things now? Because it's now possible. I think it's going to be correct that that works, but we will see.

• 21:06Speaker 00

  All right, all right. You mentioned to me before we started talking today that you're, this is like, unexpectedly, maybe, you're having more fun than at any other point in the company so far. And that is a little counterintuitive to people. A lot of founders say, oh, the early days are the most fun, and they kind of look back nostalgically. Why do you think you're having more fun now than ever before?

• 21:27Speaker 01

  The work feels much more leveraged now, which is quite entertaining. I think before, you're, so much of our focus was trying to get attention from people, because you have no attention by default, it's, the world is very noisy. And so I'm just like sending emails and LinkedIn messages into the ether, basically. And it just feels like I'm doing an awful lot of stuff that has no... Impact. Obviously, when you do that, like a tiny little spark, it then needs to be fire. Whereas now, it feels like we're playing a computer game. We've just unlocked like rocket launches or something. So it's like, oh, we can like go absolutely nuclear on this particular idea and like really fully build it out in a way that would be very hard to. We're like lucky enough to be able to build this because it involves having to have like 15 or 16 products and also a bunch of AI stuff.

• 22:08Speaker 01

  And we need to be able to like have the time to build this because let there's been remarkable work at like Intercom where they built Finn. And what we're trying to do is analogous to that, I think, but it's not making Posthog work with kind of a couple of products. It's trying to make it work across like 16 of them. I feel like we're going much deeper than we've ever been, but it's enabled us to make a bet like that. Like I think similarly, if we're starting from scratch, trying to build what we're now trying to build would be probably impossible, I think. Like SpaceX or something is a much more grandiose idea. Like on day one, it's not building like the transportation for Mars. It's like day one's like, we need to build actually relatively unambitious satellite launching business. I think we've been, we've sort of done quite a lot of the homework that means we're able to now do these, take the bigger test, basically.

• 22:51Speaker 00

  I wonder how many products Posthog would need to have to be as grandiose, as ambitious as SpaceX. It's an interesting comparison. You know, you mentioned like having fun getting this work out there. Like I know building in public is something that you're keen on. What does that look like as a CEO? Is getting this attention, getting that spark and putting it out there and trying to start a fire? How do you think about that? And doing it in the open where people can see what you're up to and hear about it and follow along?

• 23:18Speaker 01

  Trust was the original reason for it. Like we just felt that even if you're offering, like most of our users don't pay anything, like the open source project, no one paid. But it still takes up, there is a cost. Like it's trust in terms of what are they doing with my data? And also just time with setting this tool up. I think especially software is getting more, like it's getting quicker to build. There's more software appearing. It's getting just more and more competitive. I think it's become a step change, more competitive with AI. And so we felt that, okay, the fundamental thing we have to achieve in marketing is we need to highlight how we're different because it's in a busy industry. And then we need to build trust with users and like the foundation of trust we felt was transparency. So instead of just having like a one pager landing page. We're going to like really fully explain everything we can about who we are, what we're trying to do. I think that gave us the trust we needed.

• 24:00Speaker 01

  And we thought about kind of, I looked through other Hacker News launches and just sort of summarized like, okay, these are all the pieces of critique I can see. And so we're going to make sure we have transparent answers to these types of, the questions would be things like, they don't have a clear business model. I don't think this is sustainable, for example. There's like this sort of inherent suspicion of ideas. And we thought, okay, let's just like think through these types of questions that we know that our audience developers would have and we'll just address them on the website. So we like, you know, we didn't have a pay product. We said, hey, this is how we think the pay product is going to work in future.

• 24:27Speaker 00

  And that was from like day zero, you, you put some of those things in.

• 24:30Speaker 01

  Yeah, like before we launched Hacker News. I think the other thing we realized from a content and like a marketing perspective was, I think experts writing about things is more compelling often than non-experts. And if you look at like the front page of Hacker News, there'll be like incredibly technically savvy people talking about very complicated topics quite often. And I thought, oh man, I'm like, I'm not clever enough to get there, but I need to figure out a way to get us visible in Hacker News. It's like the probably the most popular place for our audience. And I thought, well, I am, the thing I'm actually an expert in is our business. And so I can write about what we're learning and doing. And so I feel like a really natural topic. So we like, we wrote a blog post also during the Bicy Badge, like shortly after the launch of what it's like moving to San Francisco.

• 25:06Speaker 00

  I remember that post, yeah.

• 25:08Speaker 01

  And again, that was also viral, but I think I built a lot of trust because it humanized us underneath. I was like, put photos in of like, we climbed up like Twin Peaks, for example, in the night, at night. I think people realize like, oh, there are human beings, you know, in the same way that someone might be really aggressive online, but as soon as you meet them in person, they're much more chill. It's the same sort of feeling and we're trying to make that clear, like with all our products today. You can see like the literal engineers that we have every single team member on the team pages. You can see all the, so you can look at like product analytics, you can see the handful of engineers building it. You can click their bio and read about like their pet cat, for example. So that you're like, oh, there are people building this. The humanization, I think is really important to stand out, basically.

• 25:40Speaker 00

  Yeah, it lets people know someone's at home. You go visit this house, there'll be someone there and you'll have a neat interaction. I actually, I, I think about your, your moving to San Francisco post quite often over the years since, because if you look at it, there's nothing that remarkable in the post. But you got it on Hacker News, people read it, they talked about it, it was discussed. And I think you're right, it's because you just showed some of your humanity behind this product that people could use. And everyone's looking for ways to relate to, to folks like themselves. I remember when I read that, I thought, wow, they can. They can climb up Twin Peaks and write about it and get developers to want to check it out. They've got, you know, this is going to go somewhere. Building in public is one of the ways that you guys have really connected with developers.

• 26:23Speaker 00

  I think another aspect of that, though, is humor. And bringing a sense of merriment, amusement, and maybe even a little chaos to how the post hoc message gets across. One of the manifestations of that was present all over San Francisco this last year with the crazy billboards that you guys put up. In my opinion, like, there's billboards all over the city. When people move to San Francisco, they're shocked by all the tech ads everywhere on all the billboards. But I don't think I've seen anyone do quite what you guys did with your marketing. And so I thought it'd be really fun to just take a look at some of these together and get your take on what's going on here. So I guess this first one is about session replay ostensibly, but it's also like an ad for tomato sauce. And it's promoting the sweet taste of understanding.

• 27:08Speaker 00

  Where did you guys come up with this? And what made you think to compare tomato sauce to session replay, for example?

• 27:14Speaker 01

  There are a few things. I think by design, we wanted it to, it had to be different. So our grand theory of billboards was how interesting your brand is. If your brand's normally here, everyone's a bit more interesting on billboards. Like, if you just look at, companies seem to come out of their shell a bit more on billboards than they do on their websites.

• 27:31Speaker 00

  Why do you think that is?

• 27:32Speaker 01

  I think they're trying to stand out, basically. But I think it's a bit like, again, this is going to sound harsh, but I think it's a bit like watching a first grader play football, where they're going to dabble, but they're starting from basically being a first grader. Where it's like, well, we're already pretty weird online, so our billboards need to be bizarre. Like, they need to be totally out there, so we should lean harder into this. Which we think will work, because the whole point is it has to stand out in the environment. And we need people to, we want people to talk about and see it. And we're not trying to get conversion. Like, you're unlikely to decide to install our SDK whilst you're dragging your car down the freeway. Instead of trying to please everyone, we're not going to care about conversion. We are just going to try and raise awareness. And then the other thing I think that had struck us was, from actually doing a bunch of social stuff,

• 28:14Speaker 01

  where like, oh, I can spend the time to write a useful post based on something we're learning, and that will work reasonably well to get some, like, if it's Helpful, people will share it. But somewhat depressingly, if we write like, if I just write something I think is funny, it can sometimes go a thousand times further in terms of reach. So if I'm just trying to raise awareness, it has to be funny. And that's actually the primary thing we want to get across. And so, yeah, we just really wanted to like crank as far as we could over on the like weirdness, slightly funny scale. It is a real skill. I think we've been trying very hard to, I think a lot of corporates or whatever do like corporate try hard. And it's very hard to not come up, because you're obviously trying hard.

• 28:52Speaker 00

  Well, your stuff is almost a parody of corporate try hard. You've transcended corporate try hard.

• 28:57Speaker 01

  It's a very, like on Hacker News, for example, people talk about like on Hacker News, it's like trial by fire, basically, where there's unbelievable level of criticism. Our marketing channel in Slack is like that too, where people are like, your joke is just not funny. It has to be better. You look like you're trying. So it's, we have a very heavy feedback culture in marketing.

• 29:14Speaker 00

  Let's, let's talk about that. So before a post hoc tweet or billboard gets out into the wild. What is the incubation?

• 29:23Speaker 01

  Yeah, it's a very small number of people will be involved, like probably like three or four. And it is a very high trust, direct place. And so we're really trying to get something that we think is genuinely, it might be sarcastic, but something that's genuinely funny to get onto a billboard. Or a genuine, like just something that would make us laugh or that we would want. We are kind of doing them for ourselves. Like, is it something I think would be quite funny to see? It's sort of the bar. It feels like you're competing against B two B software. Because like there's all these like AI billboards that are quite similar to each other. But I think the reality is if you want people to pay attention, it's like, no, you're competing with like what they're listening to on the radio or what they're doing on the phone or whatever else. And so the bar is so many times higher. So it's sort of like, it needs to be like in the, at least in that realm.

• 30:06Speaker 01

  It's not just about being like, doing a slightly better, incrementally better job than other software companies. And which is like impossible, I would know in Neo is interesting as Mr. Beast or something. But like, the mentality is like, oh, I wish, we need to be playing that game at least a bit more. We felt that we would make fun of adverts in general. And so we thought like, okay, we'll do like a very like Americana style.

• 30:26Speaker 00

  That's right.

• 30:27Speaker 01

  Ads, but we'll make it like, it's like a 1950s thing, but for like AI.

• 30:31Speaker 00

  So tastes better than spaghetti.

• 30:33Speaker 01

  Yep.

• 30:33Speaker 00

  And you're done up here like the host of a like a 1950s food show.

• 30:38Speaker 01

  Yeah.

• 30:39Speaker 00

  Talking about your B2B, B2B SaaS product. Okay, and so why, why 1950s Americana? Like, what, what was the appeal there?

• 30:47Speaker 01

  Yeah, I think if you sort of say like, what's the most stereotypical kind of billboard ad you'd imagine? I think you sort of would imagine you would go back to like.

• 30:53Speaker 00

  What's the origin of billboard?

• 30:54Speaker 01

  That's almost when you would look at them from a positive perspective, I think of like, oh, like back in the day, ads were kind of cool. They're like funnier, better thought out, whereas now they've just got a bit.

• 31:01Speaker 00

  Death to humans.

• 31:02Speaker 01

  Yeah, yeah. Or it's just like big words on a billboard with like no design element to it at all. And we thought like, actually, let's try and make something that's kind of a little bit like warm feeling. And pretty stupid, like I was wearing like a frilly frock.

• 31:13Speaker 00

  And what about the decision to put yourself in all the ads?

• 31:15Speaker 01

  So I was like, I don't, will anyone even realize this is me? Or they think they've got some like random middle aged guy to turn up these ads. Either case is probably fine. Yeah, we did a lot with hedgehog. We did do some with hedgehogs as well.

• 31:26Speaker 00

  Yes.

• 31:26Speaker 01

  But we felt again, it's just not quite as edgy. It's, it's hard to make it not feel like too professionally done almost where it's like kind of kooky to have like a very average person on the ads or whatever, instead of like something that we clearly put tons of time into.

• 31:40Speaker 00

  Too hard on yourself, James. I mean, I agree. Look, I think these are great. My wife gives me so much grief. I get so excited every time I see a postcard ad around the city and my wife is like, oh, it's spaghetti guys again. But it lets people know that someone's home.

• 31:53Speaker 01

  Yeah.

• 31:53Speaker 00

  That there's actual people behind this and that they really care a lot. You know, even things down the details and the typography, like it sends a message. We are going to make a product that's also this well thought out. And it's also trying to get you what you need to the same level.

• 32:07Speaker 01

  Yeah, I think there's just like a level of signaling that like you know what you're doing to pull off an ad like this. Because it, and it does show a level of savviness. Like the website is similar in thought process of like, we're trying to demonstrate to so many people who isn't yet using our product that, oh no, they kind of get technology or they like kind of just get me as a person a little bit better than like. To face this corporation.

• 32:28Speaker 00

  Let's talk about your website. I think you've got the best website of like any B2B company maybe in the YC portfolio. I don't get in trouble for saying that, but it's incredible. Your website's awesome. What was the genesis of going in that direction? Because it wasn't always like that. You mentioned a little bit ago that early on you guys had a lot of back content, a lot of lore.

• 32:46Speaker 01

  Yeah.

• 32:47Speaker 00

  On Posthog early on, but the website itself was still pretty typical. Static website with some information for a SaaS company. But now it's this whole experience simulating, you know, using a computer in Posthog land. Where did this come from?

• 33:02Speaker 01

  I think a few things. I think we kind of since like 2021-ish, the website's our sales team. Like there isn't a single customer that won't go through the website before they buy the product. And especially because of the nature of our audience, we're very self-serve. And so like, well, we should invest in this. Like it's the, we don't have to hire a sales team. It turns out in the first few years, or we had, but it was tiny, kind of like two people. And so, well, I think we can therefore justify spending like a ridiculous amount of effort on the website. The cost isn't that high, but like we can put, it's kind of the energy really, because I think this will make us stand out. So I think a lot of people quickly 80/20, something you can extremely quickly get to a polished website now. Like in a day, you've got a polished website. So everyone just does that and they're like, cool, I'm done now. This is 80/20, then move on. And I'm like, well, like marketing to me is like, because we're in a busy industry, it's about standing out.

• 33:44Speaker 01

  And so there's no point doing 80/20. Like there's, it's just a missed shot. It's like, actually, all the respect for the website has come from like the last percent of effort. Where it's like, no, we're going to go like so insanely far past what is normal that it is remarkable, which means that people will talk about it. Because when we talked to users early on, it turned out all of them are coming to us for, people were showing up because they'd heard about us online or someone recommended us. And then we said, why did they recommend you? And it was all tools in one, low pricing, develop a brand and very technical support people. And we have literally just done those things. And but it's like, okay, we can't just keep doing them as it is. We should be trying to crank the dial up further. So it's like, instead of building like three products, we should build like 300 products. And how could that work?

• 34:25Speaker 01

  Or on the brand side, it's like, okay, post is kind of weird. It's like, it needs to be weirder still. Like it needs to be so weird that it polarizes. Every day we'll get a whole bunch of people talking about that website. And we monitor all our brand mentions online and it's going to be like, either this is literally the best thing I have ever seen. And we also get, this website is atrocious. But we feel like we're polarizing it largely in favor of our audience more. Like I'm sure there are some people who are like, what? But we're happy to. Pushed so hard in the direction we think is something cool for our audience that the people that aren't really our focus are gonna, we don't care if they disproportionately hate it, basically. And then partly, the other thing, the other phenomenon that was happening was it didn't just come from like a pure aesthetic perspective. We genuinely were trying to work out how do we build a website that fully documents

• 35:07Speaker 01

  and prices for like 15 or 16 products, but we also have a whole load of other things we've built that we think are useful. So for example, we have a cool developer jobs board. It's like a jobs board where the filters, instead of being just like location or pay range or something are like, what kind of laptop do I get is the majority of the company developers. The stuff an engineer actually would be interested in.

• 35:25Speaker 00

  Things they have to think about, yeah.

• 35:27Speaker 01

  Yeah, and we developed this as like kind of a fun project, but we're like, well, this is useful to our audience. So we have that in there. We have like a handbook that's quite popular for people to read and learn from. Like we learn how everything works, like how do we pay people, like when we let people go, like everything is there. And so that's like another thing. We have a newsletter with like more than 100,000 subscribers that's also in there. And so it just felt like the form of a website of just like a thing, lots of pages you scroll through. It felt like we're doing something more multi, much more multi-dimensional than that. That was, yeah, it was the two things. It wasn't purely aesthetic. It was also from experience, but we actually thought it would be better reflect what we're doing. Postdoc is quite a complicated company. Like, and again, it's tempting to be like, okay, we should just have like a single landing page with like one button you click and nothing, no other information. This is not what our audience wants.

• 36:08Speaker 00

  You're kind of just like spitting in the eye of the idea of a funnel. It's almost like instead of a funnel that you're trying to push people through, you're just giving them a great piece of cheese to dig into and enjoy and savor.

• 36:20Speaker 01

  Well, we were wondering what would happen to conversion rate and we were like, we kind of knew this is gonna tank the conversion rate. It's gonna get, we're like, this is getting much more traffic and the conversion rate's getting worse. And I was like, well, like if we ongoingly have way more traffic and it's worse conversion rate, if I could like double our traffic and half our conversion rate, I would take that. And so we launched it kind of knowing that's brutal would happen. But we did also think like, we think this is interesting enough that people will actually switch on for once online, rather than just like, sort of like trying to rush to get to the button. And like, we also thought like, I mean, if a developer can't work out how to click get started, when there's already like three of them on the page, like, you know, that marginal person who maybe would have not converted, like. I don't know, it just felt like you're playing like a weird game trying to appeal to that user. And so we think like of the marginal user, this will impress enough that they will convert despite the fact the conversion experience is very unusual.

• 37:05Speaker 01

  And so, yeah, the conversion rate did, it didn't actually tank. It was like maybe 10% ish off the top of my head, but we're going to do a post on it. It was worse. We did like one experiment to change the flow slightly. Now the traffic is much higher and the conversion rate is significantly higher too. And so it's performing better from a conversion perspective, even though it's like wildly complicated.

• 37:22Speaker 00

  My favorite part of the website, you have this unhinged merch shop that has some, some deep cuts in it. And my favorite of the deep cuts is visitors can buy a signed photo of James and Tim, the founders of Posthog, and it's sold out. Someone actually came and bought.

• 37:37Speaker 01

  I think we sold one, in fairness.

• 37:40Speaker 00

  I love the yacht rock aesthetic of the photo, by the way.

• 37:42Speaker 01

  Yeah, we had quite a few boating offsites. This was supposed to represent Tim alignment, but I've think about it on the way here. And I was like, I actually think, I think the reason that feels funny is you go like, this is to represent Tim and I being aligned and close together, but this is what aligned would be.

• 37:57Speaker 01

  We're close, but.

• 37:58Speaker 00

  Something else, this, yeah. All right. Well, thank you so much for spending time with us today and talking us through all this. I know we have so many folks out there that follow the videos that we put out that are looking for ways to break through.

• 38:12Speaker 01

  Yeah.

• 38:12Speaker 00

  And I think your point about like, you have to be remarkable is obvious, but overlooked so much today. And so thank you for sharing some of the ways that you guys have figured out how to try to stand out and be remarkable. And congratulations again on this latest round that you guys just raised. We're so happy for you guys, excited here at YC.

• 38:28Speaker 01

  Yeah, thank you very much. Yeah, it was monumentally important to us to go through the program. So yeah, tonight, if you're listening, like, go apply to YC.

• 38:35Speaker 00

  All right, thanks for joining us today.