 Thank you for having me. In 2011, I founded Sigmund with a couple of my friends and roommates from MIT. And to give you a quick intro, Sigmund provides customer data infrastructure to businesses that put their customers first. So we help them collect data from all of their different customer touch points, unify that data, and send it downstream to all of their MarTech tools. So since starting the company, we've raised a little less than 300 million in venture capital. And we've grown to serve more than 20,000 customers all over the globe. And one of our toughest lessons learned to date is how to find product-market fit. So that's what I'm going to talk about today. How many of you are searching for product-market fit right now? Quick show of hands. All right, many of you. So statistically speaking, 80% of you are going to fail. And 80% of all startups fail to find product-market fit. And that means that four out of every five founders that just raised their hands is not going to succeed at finding it. And the one out of those five that does is most likely going to struggle through it. But that's just the stats. The reality is that failing to find product-market fit is a bottomless, emotional free fall. And when you're failing to find it, it becomes this obsession. You see mirages of it, and it literally makes you sick with worry. In 2012, at the height of our search for product-market fit, I went to the hospital twice for panic attacks and lost 10 pounds in three weeks. And I haven't had a panic attack before or after. So failing to find product-market fit really feels like the depths of despair. And I'm certainly not the only one. Sam Altman, who was the president and head of Y Combinator for many years, was previously a startup founder. And during his search for product-market fit, he actually got scurvy because he wasn't eating properly. So the search for product-market fit is not a healthy activity. But it is an important one. And so we saw mirages of product-market fit and lied to ourselves about it for sure. And we always convinced ourselves that we were one small feature or one small tweak away from getting product-market fit. So this is a picture of myself with my co-founders Calvin Ilya and Ian in the midst of our false hopes in 2012. But back then, we were just four guys in an apartment failing to find a product-market fit, but convincing ourselves that we were right on the cusp. We were writing a lot of code, but we had no customers. We did have one thing going for us, though. And that was that we were a bunch of cockroaches. And in the search for product-market fit, it's very important to be a cockroach, to stretch your startup's runway as long as you possibly can, so that you can test more than one idea. So you can test two, maybe three ideas and get more shots on goal. And so to make this process explicit, step one is that you build, launch, and test several product ideas to attempt to find product-market fit. And to improve the odds as much as possible, you stretch your runway as far as you possibly can. Step two, something magical suddenly happens. And this is product-market fit. And it really does feel magical. And we'll talk more about that in a minute. And then third, product-market fit suddenly turns you into a uneroach, if you will. And rather than just surviving, your product-market fit actually makes everything slightly easier. Investors are suddenly interested. People suddenly want to work for your company. People actually care. Customers want to help you improve your product. And at this moment, your focus as founder and CEO really shifts from product-market fit to go-to-market growth. So today, we're going to dive into those three stages in more detail with some stories from our early days. And first, we'll deconstruct what it takes to build a category-leading company. We'll shine a light on what bad product-market fit feels like. And then we'll talk a little bit about what good product-market fit feels like. And I think that third piece will hopefully be most useful for all of you. So why talk about category-leaders? Well, why do we care about them at all? Well, the reality is that they can be 10 or 100 times larger than the next closest company. So if you look at Salesforce as a great example of this, they're 10 or 100 times larger than the next closest real CRM company in sugar CRM. And so that's why we're here to discuss how to build them. We want to build really big companies. But how do you actually do that? What is it that makes a category-leader? What allows someone to dominate a category like that? Well, almost by definition, being category-leader gives you access to the largest set of customers and the most profitable set of customers. And what that means is that you can build a platform on top of that. And that platform allows you to create a moat, where other people are building products and marketing and selling to your customers. So what's required to build a platform if that's the key to category dominance? Well, generally, to build a sustainable and compelling platform, you need to have a SaaS business with at least $100 million in revenue, with line of sight to $1 billion. And what that means is that there's enough meat there for other people to build their business on top of your customer base. So in order to be able to category a leading business, you need a platform, and I need a platform, you need $100 million in revenue. But what is the path to $100 million in revenue? That's a big number. Well, Jason Lemkin, who started Saster, puts it like this. He says, from $0 to $1 million in revenue is impossible. From $1 to $10 million is improbable. And from $10 to $100 million is inevitable. And the reason that $10 to $100 million is inevitable is because by the time you get to $10, you have so much momentum behind you that you're going to make it eventually. $1 to $10 million is this brutal grind for the founders. And because as Saster says, you have a real business, you have real customers, but you can't yet hire a world-class executive team. And $0 to $1 million is impossible because that's this product-market-fit stage. That's where 80% or more of companies fail. So today, we're going to focus on that $0 to $1 million, that impossible stretch. And that's crazy, right, that four out of five companies fail at that stage. But it's actually even worse than that. So investors and founders often talk about the value of failure that will learn something from having failed. But it turns out that the research doesn't actually back that up. The odds of improving in your search for product-market-fit don't actually increase if you fail to find it the first time. However, if you have found product-market-fit once, the odds of you finding product-market-fit the second time go from 22% to 34% or 50% improvement. Still not a great overall success rate, but a significant improvement. And my interpretation of that is that there is no learning encoded in failure, but there is learning encoded in success of knowing what it does feel like to find product-market-fit. So when we were struggling to find product-market-fit in 2011 and 2012, I felt this very acutely. I had never seen what finding product-market-fit actually felt like. I didn't know what good felt like. And so I didn't know how to throw out the failures fast enough. I only had failures on one side of the machine learning training model. I didn't have an example on the positive side. And so I hope to give you that example today. So let's go back five years ago and walk through our story of finding product-market-fit. Segment is customer data infrastructure today. But back in 2011, we started as a classroom lecture tool. So the idea was to give students a button to push to say, I'm confused. And the professor would get this graph over time of how confused the students were. We were super excited about this. We were in college at the time. We got a bunch of professors excited about using this tool. Or at least we thought excited. We got into Y Combinator with this idea. We raised 600k coming out of demo day. We were super fired up. We'd invested lots of time in hundreds of thousands of lines of code to make this product successful. And then as the fall semester started, we deployed this into the classroom. And this is a picture from Berkeley where we were deploying it in the classroom. And a few professors were sort of obligingly willing to try this out in their classroom, I would say. And we took that as a sign of product-market-fit. They're excited to use our product. The reality was a little different than that. We actually went and stood in the back of one of these classrooms and looked at what students were doing. And this is what they were doing. It was a disaster. It was absolutely awful. They were all on Facebook and Solitaire and doing anything except actually paying attention to the lecture. It was a total disaster. We had literally just received investment from all of our investors. They had just wired us the money. And we didn't know what to do. So we called back all. It was horrifically embarrassing. We called back all the investors and asked them, what should we do with the money? It turns out that we don't have product-market-fit at all. And our investors said, well, we invested for the team. So go find something else. We said, OK. So we spent about a year trying to build an analytics tool. The idea here was that we should have been able to see what was happening with those students in the classroom by looking in a web analytics tool. We should have seen that they were disengaged and not actually using the tool. And we hadn't been able to see that. So we would go out like a good lean startup company and talk to occasional customers out there and ask, like, oh, do you have problems with analytics? And they would feed us an occasional problem with, oh, I don't quite have this report right. And I wish I could get this other dashboard. And we took that as a sign that they were really interested. They would ask, hey, can you keep me posted on your product updates? We said, OK, great. That must be product-market-fit. We must be getting close. They're really interested in learning more about our product. So we returned to our office and we started coding again. Six months later, now in San Francisco, we had done almost nothing but code. I had taken one sales trip and had met a few customers who had a few complaints around the edges with their current analytics tools. And we thought that we could build the features to solve that. And so we thought, OK, we're almost there. A few more tweaks, and we'll get to product-market-fit. And we really tricked ourselves here. We really were misleading ourselves with the level of feedback that we were getting from these potential customers. We used other little positive interactions with prospects to sort of buoy our hopes for what might happen. And every once in a while, a stray person would come by our website. This is a live chat conversation that we had with a customer. The three is 3 AM. This was the middle of the night. And we thought, oh, this is really exciting. This person has questions about what it is that we're doing and who we are. But the reality is, if you look at this more closely, this is a six-minute conversation they left after I asked a question without even saying goodbye. And that is not what product-market-fit looks like. And this is what it this is when no one cares. This is what it looks like. So by this time, we realized that this analytics product was not going to work. It was December 2012. We'd been working on two broken ideas for well over a year and a half. We had 100k left in the bank, so we had burned half a million dollars. We decided to go back to Y Combinator. This is just outside their office. And we went for a walk with Paul Graham around the little cul-de-sac there. Paul Graham leads Y Combinator or led Y Combinator at the time. And as we finished delivering our story and our update, he said, so just to be clear, you've spent half a million dollars and you have nothing to show for it. I thought that was a rather brutal way of putting it, but extremely true. So pause there and let's rewind all the way back to that first week of Y Combinator. And in that first week, we'd been like, what? We should have analytics on our classroom lecture tool. We should figure out what people are doing on this thing. And so we looked around and we found Google Analytics, KissMetrics, and Mixpanel. But we couldn't figure out which one was going to be better. And so we decided to build this little piece of code that could take the data from our website and send it to all three of those analytics tools. Now, we built this. We used it as 50 lines of code. Six months later, we improved it a little bit more. Six months later, we improved it a little bit more. Eventually, we open sourced it. People started starring it on GitHub, occasionally submitting pull requests. And so by the time we got to this decision point in December 2012, we were looking at this and we had up this tiny little JavaScript library with these 25 stars on GitHub and a few pull requests or what other idea? And so my co-founder Ian was like, you know what? I think that this could be a big business. I think Analytics.js could be a big company. And I thought, that is literally the worst idea I have ever heard. It's 500 lines of code. It's already open source. I do not understand in the slightest how that could conceivably be a large business. So we fought about it all day long. And I went home and I was racking my brains trying to figure out how to kill this idea. And finally figured it out, came in the next day and was like, all right, guys, here's what we're going to do. We're going to build a beautiful landing page. And it's really going to pitch the value of this open source library, a hosted version of it. And at the bottom, we're going to have an email sign-up form. And we'll post this on Hacker News and we'll see how people respond to it. So we agreed to do that. We built this beautiful landing page. We posted it up to Hacker News. So I started paying attention to other things. And much to my surprise, it went straight to the top of Hacker News. Got hundreds of upvotes. We got thousands of stars on GitHub. We even had people reaching out to us on LinkedIn demanding access to this beta. This is a guy who's now an MIT professor. He says, I'll give you feedback and tolerate bugs like you wouldn't believe. So full stop. This is what product market fit looks like. Every single thing in your business goes crazy. It's not metrics slowly moving up over time. It is literally everything going crazy simultaneously. So with our lecture tool, we'd been searching in the dark for the next feature, the next thing that would make it have product market fit. But now we had people cursing us out because we only had seven integrations and they wanted this other integration. And people were actually just building them themselves and pull requesting it into our GitHub repo. With our analytics tool, we had those sad, unanswered questions in live chat. But now we had people reaching out on any available channel demanding access to this beta. So don't let yourself be misled. These early indicators are not what product market fit looks like. This sort of explosion of interest that happens, this is what product market fit looks like. So let's be incredibly clear that product market fit does not feel like idle interest. It doesn't feel like a glimmer of hope from a random conversation with a prospect. It feels like everything in your business is suddenly going haywire. It feels like a proud rush of adrenaline that people actually care. It feels like actually a little bit of a loss of control because customers are telling you what they want. And it's no longer necessarily as much about your vision. You simply cannot mistake product market fit for not product market fit. And if you have the question, do I really have product market fit, I think the answer is no. But the hard part is being honest with yourself about it. And of course, that didn't stop us from making this mistake. I thought class metric for sure had product market fit. It didn't. I thought for sure segment.io, our analytics product had product market fit. It didn't. And even on our third attempt, when we finally did find product market fit very suddenly and explosively, I thought for sure it was a terrible idea. So it just goes to show you how difficult it is to find product market fit or maybe how stubborn I am. So if you take one thing away from today's talk, I hope it's that you have to be very lucky to become one of the five that makes it through this gauntlet of finding product market fit. You need to be brutally honest with yourself about what the world really wants and not what you think it wants. So thank you.