Building Y Combinator with Garry Tan: What It Really Takes to Build a Successful Startup

25:02When you have people who are actually deeply focused on product market fit, like, yeah, it changes your whole mentality. You don't get stuck in the same ruts. And you have people who can pick you up when you're down. You know, one of the more controversial things at Y Combinator that we've learned over and over again over 20 years is that it takes co-founders. And I think there are plenty of solo founders who have been successful, have made it. It certainly isn't the only way to do it. But if there's a high order bit, like having a co-founder who is truly your equal, who is as bought in, it changes the outcome of companies far more.

25:42So, you know, the people who can do it as solos are really, truly that much more remarkable. We're going to take a quick break. When we come back, Gary is going to tell us why Y Combinator values strong co-founders more than strong ideas. Friends, like people who have known each other for a long time, you know, who have been on camping trips that have gone wrong together, you know.

26:17Okay, we're back. And I'm speaking with Gary Tan, CEO of Y Combinator. You mentioned co-founders. I know in your process, you're looking for resilience in many ways. Like you're looking at kind of the personality of the founders, even if they're co-founders, what's the personality of them? What is your prerequisite list? Knowing that it's not perfect, but it's the one that gives you most shots of the goal in a way. Internally, we talk about this where sometimes you have people who are brilliant technically, but do not have the empathy or product side. And we call them Wazs or Wozniaks. And like the reality is like Apple wouldn't be Apple without Steve Wozniak.

26:56So they're necessary, but they're not sufficient. Like you also, you know, truly do need someone in the weeds understanding how products work and the user. Both of these things are incredibly important. Really, that's how we look at it. Like, we really want founders who have complete teams. You know, one of the more funny ones that is very intuitive, if you really think about it, but seems odd on its surface is it's actually really powerful when friends, like people who have known each other for a long time, who have been on camping trips that have gone wrong together. What is starting a startup if not loading up a boat and going off to the middle of the sea and saying, we're going to find something really great out there.

27:46You've got to find land. And so having really good friends who are complimentary startup companies together, that's amazingly important, actually. And you want people to be complimentary. Like, I mean, the classic thing is you need a hacker, a hustler and a designer. And the more of those things in as few people as possible is good. Like, and so I think this is also the age of the polymath. You know, some of the best founders in the world that we run across, they don't have boundaries. They contain multitudes and those multitudes allow them to be really like not just systems thinkers, but, you know, thinkers across many systems, perhaps all the systems.

28:29That's what like the best CEOs in the world are like. I'm on the board of Gusto, which is a YC company and just thinking of the three co-founders and they're the best, you know, for them, in your point, that's exactly what you're saying. One, there's an amazing resilience and bond and two, they're highly complimentary of each other. Yeah. And that's what it takes because you need to be able to like be a back to back to back, you know, like against the world. And, you know, when you have that, like nobody can beat you. Exactly. In many ways, timing could be everything in startups. It could be timing for market it could be timing for you know when you launch it could be timing for when you have your break and you have the opportunity come along and you're able to take advantage of it do you do you think it's you know overplayed timing like how do you filter for timing do you're saying this is a good idea but market is not ready for it so like maybe come back in a few years how do you think of timing i think like series a investors in particular do this extremely well, but I think at pre-seed and seed, you have to go by the founders and what they're saying.

29:33Certainly, like the idea that they're pursuing is a marker for the quality of thinking that they're capable of. So, you know, idea matters, how people are approaching things matter. But I think like when you're writing a 20 or$50 million check, it matters a lot more, that stuff. I mean, And you can sort of see this in the way late stage PE and pre-IPO stage investors are, you know, very much about the market and the P &L. Whereas like when you're early, you know, if you look at any of the graphs of revenue for any startup, you know, even leading up to the series A, like these days, you maybe you need a million to$2 million a year in revenue.

30:14And that's a lot of revenue. But if you look at it on a graph compared to like the kind of revenue you need when you IPO, it's like, it's the thin edge. It's like the thinnest of edges, like 99.9 % of the work is ahead when you are like just starting out, even when you get to$2 million a year in revenue. So, you know, the wild thing is, how do you actually go from zero to one? I want to talk about the craft you teach at Weiss. You talked a lot about the experience, the support, the teachings of being bold, going after your targets. You know, this podcast, I created it kind of in many ways to show that there's no one way to build.

30:55If people could listen to the YC story and say, okay, so I come in, you give me all these skills, you accelerate me, and then I'm coming out with best practices. But really, it's not just best practices, I'm assuming you're giving them, but also in a way to index on what makes them unique. So I'm curious if you have examples of how that process comes to life and how do you see that in the founders that's come along? Honestly, I think we just give away all the best practices on the website. Like, you know, YC Startup Library, like we have a YouTube channel and we have a special show on there called Startup School that is just, you know, all tactical, all best practices.

31:33This is what works. This is what doesn't. And we give it all away for free. So I actually, I mean, we're in the age of LLMs now. Like the how-to does not matter. I think how to think and how to have that core loop that turns you into James Bond, I guess, is like, you know, in order to actually change. Like I, you know, when I was running Initialized Capital, it was, I had left as a partner of YC. I left to start my own venture capital firm. I was doing my own pre-seed investing. And I too thought like, you know, if I just meet them every week, like if they work out of my office, like I'll just change them.

32:10I'll be able to help people speed up. And that was never true. Like no matter how hard I tried, but when I sent them to YC, they became different people. Like they became more formidable. They asked questions in a better, deeper way. They understood what was going on. They sort of were less likely to delude themselves or lie to themselves about things. And so, I mean, some of the more wild examples, you know, I always think about Brian Armstrong of Coinbase. I mean, the interesting thing for him was, you know, he had started other companies before. He started a tutoring business. He did marketplaces.

32:51He worked at Airbnb. So we saw the inside of that. when he first started Coinbase, the idea was not that baked. I think he initially started a Android client for the Bitcoin client. And then after a while, he realized, oh, it's going to take too long to actually sync the blockchain. So he said, you know, what could I do? I mean, again, it goes back to empathy and understanding of the user. Like it requires a little bit of empathy to realize like, well, technically speaking, it's a feat to have a Bitcoin client working on an Android phone. But when you actually use it and you actually like watch how your users use it, it's like no one's going to leave their phone unlocked overnight to like sync up to the blockchain every night.

33:35So, okay, that's a problem. What do I do? And so the most obvious thing is like, okay, well, I need to now put it, you know, web hosted the way Git, you know, is hosted by GitHub as a hosted website. So that was like the next iteration of Coinbase. But, you know, he came and did YC, he was giving away a Bitcoin to anyone he would meet, which was wild. I think a lot of people kept that and are pretty happy about it. But even then, nobody used it. They signed on and they never clicked anything else. And the majority of people signed on, opened an account, and then did nothing with it. And you do some more introspection, you try to understand, Like, let me sit down with users, ask them what's going on.

34:21I mean, it sounds very simple to do, but like the sheer fact that he did it, like it led him to these simple observations that were like very actionable and totally the correct thing. He realized getting Bitcoin is hard. You know, if you had to go to another website called Magic the Gathering Online Exchange and Western Union, some money to Japan and you'd get a Bitcoin and you realized no normal person, like weirdos who are really into this very obscure thing are going to do that because it's interesting and cool, but normal people are never going to do that. And so that's how they actually unlocked, like, you know, what is one of the core pieces of the business today.

35:01It's, you know, making it easy to be fiat to crypto rail, right? Make it easy to get Bitcoin and whatever crypto you need to get. It's just, I feel embarrassed to, you know, even explain these things in that they sound very simple, but, you know, the implications when you're right are very, very profound because these things do end up changing the way people use technology. It changes commerce. It changes the way people live. It, you know, ideally does solve very fundamental problems if you can do it. That's the hope. That's the thing we want to like see increase in the world. So for me, like when I, when I hear you talk, I'm like, hey, a lot of this know-how is is out there but it's in the shape of a theory and it's it's the theory maybe yeah one percent would forgive it credit to like that's the worth of that theory but like the application of it it's almost like you go and you can't really learn product just by going to school you have to build it you have to spend time with customers with yourself you have to like go for the hurdle so i'm like in my mind i'm trying to do this a b test like there's gary at like initialize capital and then there's gary at yc and to your point gary that was very hard to kind of you weren't able to kind of get them into that accelerator mode you wanted but in yc you could is it because it was immersive there just was no way out it was like 24 7 you're in that so like your transformation is almost like a given that level of intensity is it the intensity of the program?

36:33Yeah, I think the program is intense. And on the flip side, like our realities are socially constructed. So, you know, when I was just meeting them, like, you know, I like to think having a meeting with me is a big deal. But honestly, like, you know, 30 minutes, an hour with me every year, I mean, on a relative basis, one way to look at that is like, you know, how long does it take to have true like transformation with a human therapist, meeting them one hour a week takes 10 years, right? Whereas think of like the most intense experiences you've had, like you change everything. You move to San Francisco, you, you know, actually can't hang out with friends and family.

37:16Like, you know, it sounds weird to say that, but like, you know, you actually shock the system. We actually somewhat explicitly say this, you know, now that you're in YC, you're in the sprint of your life. You know, demo day is 90 days out and this is your time. And so someone's going to ask you to help them move this weekend and it's your best friend. And you know what? Like last week you were going to do that, but this week you're going to say, I'm so sorry, I can't. And they're going to understand because you are in the sprint of your life. It's literally learning how to run fast. You know, our realities and our beliefs are socially constructed and you can't really rebuild your reality and your, you know, how you think about the world doing the thing that you were doing before.

38:02This is another reason why we tell people like, please, please, you know, when you come here, work with your co-founders and your team out of, you know, a live-work loft or your, you know, two, three-bedroom apartment or a rented house, like create your own culture. You cannot create a new culture and speed up to the speed you need to and ask the right questions. If you're going to an office and there's a water cooler and there's a happy hour that night, you know, you have to actually be in a fundamentally different situation doing something that you've never done before. And then you'll do it.

38:36And then once you do it, you get that first, you know,$100 ,000 a year revenue contract. You know, you raise your first seed money like you're, you know, doing it. You have to radically change everything about what you're doing in order to change your mindset. And then that mindset will forge you into a founder. Well, this gives a great lens into the ethos of the program in many ways. So this one is slightly personal for me. So I call it the valley of despair question. You know, as a former startup founder myself, an angel investor today, like one of, at least for me, one of the hardest parts I see in the startup journey is when your product is doing okay.

39:18Sounds familiar, yeah. It's not doing really well and it's not tanking, right? It's like people are using it. They're actually getting good feedback. Like the graph looked like, you know, it's moving up and to the right, but very slowly, very gradually. And you don't have as much time and you feel like, you know, you're not sure. Do I have it? Do I don't have it? And you iterate, but the graph is not moving there. And I remember like at one point I asked Bill Grosser when I was building it. It was like, hey, how do you know? Should I fold? Do I continue? Like this could take years. and his answer was like you don't i can give you examples of companies who persevered and they made it and then some companies at the same time like the whole stupidity was filled with a very similar criteria yeah what's your perspective on it because i feel like a lot of companies are in this stage of like slowly growing people are using it i have something there is something there otherwise they won't use it but it's not taking off and it's not flaking it's like in the middle i think uh some of it is like the serenity prayer right there are things that i can change their things I can't.

40:18And, you know, please give me the serenity to know the difference. I think you have decent amount of control, but you do not have total control. Yeah, I agree with you. Like, I think the danger is to just keep doing it. I mean, I guess maybe it's just to what end, right? You can always take the graphs that you're on, like you're basically on sort of some sort of glide path. And then sometimes it's like, you know what, if we just keep doing this, you know, maybe we're not growing more than 2x a year. You know, growing 50 % a year, for instance, is like pretty good. Like if you can get to scale, like maybe it takes, you know, it is still this sort of exponential, but, you know, those early years, it might take you five years, eight years instead of two years.

41:04And that's quite common. Like one of the things that is wild about YC is that about 50 % of the companies that go through the program do end up raising a series A, but about a quarter of them raise it in year five and six. So this is a very long road. And then if you can grow and the dynamics make sense, and you're down for it, like, you know, hey, stick with it, right? The other thing is like, I think that you can divine, you can use your senses and like, you can use logic and narrative to at least try to make sense of your situation. And there are just a lot of different things you can do. You can change your product.

41:46You can change the way you sell it. You can change the way you charge. All of these things are very connected to one another. They're more connected than not. I guess the concrete example for me was when we were working on Posturus, our growth totally stalled out when Instagram came out. Posturus was a dead simple blogging tool. You could post any type of media, especially photos from your brand new iPhone. And for a while, we grew 10x year on year to be one of the top, the biggest like social sites. But we stopped growing the day Instagram came out because instead of emailing their photos to Posturus, they started posting to Instagram.

42:24And we had a choice there. You know, what my co-founder ended up doing was turning it into a paid version of Google Groups. He ended up selling the company for 20 million. It was a great exit for really everyone involved. Not huge, but it could have been much bigger, maybe a 50 or a 500x sort of outcome if we had charged money. And so, you know, all of these things make a difference. There were enough people who would probably be willing to pay, you know, on day one, like I think we would have been making at least a million, maybe$2 million a year. That would have more than paid for payroll. And if we could have, you know, taken the profits and reinvested it into growth, like, you know, we didn't have to stay on the venture path.

43:08We could have taken, you know, taken the off ramp in the moment, like we made different choices. Right. So, you know, I think smart iteration involves, you know, even just constructing a model in your brain of like, what are the dynamics of your business? Like, what are the inputs and outputs? You know, what are the variables? Capital is one of them. I guess the funniest analogy is like, think of your business as a game of StarCraft. And so sometimes it's a resource problem. And then how do you overload that resource? Like, what can you do? You sort of go down the list and you can like very logically diagnose, you know, what's going on.

43:45And then also be very crisp about like, what are things we can do and what are things that are beyond our control? And then, you know, hyper laser focus on the things you have control over. And that includes product, includes pricing, includes, you know, decisions about the team, includes, you know, what's your roadmap? Like which personas are we going to invest in versus not? Which personas are most likely to yield fruit versus not? I mean, there's a lot in your control. I love that. You have a very unique vantage point. And when you think of the future, both in terms of founders, in the sense of, you know, I'm a big believer in idea of a full stack builder in the future.

44:25Like as we You have this ability to build across and you go back to, you have empathy, judgment, creativity, vision, and then you might not need all the other functions. You can get them all baked into you. And then the other one is just like emerging needs or technologies. I know you guys pull out like a hypothesis every year or every batch. There's like, here's the stuff we were thinking about for the future. Yeah, request for startup. Yep. So when you see both that idea of like the ultimate entrepreneur or both like the spaces to play in, give me a lens into how you're seeing the future play out for those.

45:00I mean, you know, the requests for startups are just really us, you know, hearing a lot about, you know, all the YC founders are sort of in the idea maze. And then those are just sort of our, you know, semi-educated guesses at what might be the next thing. The reality, though, is like, and this is true for any investor, like, we're just guessing, right? But also, you know, founders are guessing too. So the best thing is to guess well and be right and then give yourself as many shots as possible at being right. In this new AI age, you know, I think the two things that are most important are agency and taste.

45:40And then agency, you know, can be stoked. I think that YC, the program, the experience, the community specifically is designed to increase the agency in people. people go in low agency, sort of powerless, somewhat hapless. I mean, that's just the natural state of things is like, I don't think that any of our organizations where we raise our kids or the default path for anyone encourages or even allows agency. And then the harder part is actually the taste part. It's very hard for us to teach taste, right? But that's also important. And it's just funny because in the age of AI, the core loop is prompting and then evaluating the response.

46:27So prompts and evals. And that maps exactly to, like, I think what is happening across all of society now. Like, you can increase your agency, but man, taste. That takes a long time. And like, that's the part that, you know, often we're trying to find. This was wonderful. I've learned a lot. Thank you so, so much, Gary. Thanks so much for having me. This conversation was a treat. Let's jump right into my takeaways. First, the real signal of a great founder is how they spend their time. Gary draws a sharp line between what he calls playing startup house and actually building one. Playing startup is about chasing PR, being visible, investing in the appearance of success, while doing a startup is obsessing over your users, sitting with real customers all the time and iterating on your product daily.

47:21This is one of YC's core filters, founders who spend their time chasing optics versus chasing the truth. Second, part of chasing truth is making sure you evaluate the quality of your data. Many founders lie to themselves. They get stuck in a delusional loop of positive signals that aren't real. As Gary said, people lie to you in your user interviews. It's not malicious. It may not even be inauthentic, but eventually you can only rely on user actions, the real data. The logs don't lie. The Stripe account doesn't lie. If you're not seeing the results you're expecting, keep asking yourself why until you find the truth.

48:07Third, agency can be taught, but taste takes time. YC can help increase a founder's belief in their ability to act. But judgment and taste, what's worth building, what good looks like, those are much harder to teach. We live in an age when you can learn or teach yourself so many skills, but can you tease out the human meaning in the data? Can you understand what will resonate with users? That takes time and experience. Fourth, having a co-founder matters a lot. This is something that YC considers important when reviewing applicants. The bond and resiliency of the team is more important than the power of the startup idea itself.

48:51Not only that, the diversity in capabilities matter a lot. In Gary's words, you need a hacker, a hustler, and a designer. And if your founding team has that, that can really give you an edge. Lastly, one of the most emotionally difficult phases in the journey of the builder is when your product is good, but not good enough. User growth exists, but it's very slow and not taking off. This is sometimes referred to as the value of despair. It's the hardest phase to get out of or walk away from. So what do you do? Don't mistake motion for progress. Just because it's not broken, it doesn't mean it's working.

49:35Be brutally honest about what's in your control. You can change pricing, products, sales motion, positioning, so focus there. Don't waste cycles hoping for luck. Know when to pivot versus persevere. Sometimes you can grind it out, you can iterate and you can hope for the growth curve to appear. But other times, the market is just not there. Try to see the difference. At the end of the day, for every successful founder at the top, there are thousands who persevered, but were not successful. Sometimes you just have to have the combination of grace, luck, and timing to seize the right opportunity when it comes along.

50:14I'm Tom Erkoin. Thank you for listening. I learned a lot from this conversation, and I hope you did as well. Don't want to wait around for season three? Check out some of my courses on LinkedIn Learning. We've got some great lessons on product, AI, and leadership. Building One is a production of LinkedIn News. Our host is Tomer Cohen, LinkedIn's chief product officer. This episode was produced by Max Miller. Our associate producer is Rachel Karp. We're engineered and mixed by Asaf Gadron, and we get additional production support from Alicia Mann. At LinkedIn News, Sarah Storm is senior producer.

50:49Dave Pond is head of productions and creative operations. Maya Pope-Chapelle is Director of Content and Audience Development. Courtney Koop is Head of Original Programming. Dan Roth is the Editor-in-Chief of LinkedIn. If you know a product leader we can all learn from, send us a line at pitches at linkedin.com.