A Decade of Learnings from Y Combinator's CEO Michael Seibel

Okay, great to see everyone. Um, I thought what I would do that could be the most helpful is teach you a little bit about what we've learned at YC over the years. Um, to give you a sense of YC, we've now funded over 2,000 companies, about 4,000 alumni. Um, our companies combined are worth about $100 billion. We have about 100 companies that are worth over $100 million and 17 worth over a billion. And our top 100 companies have created over 28,000 jobs. Most people know the YC accelerator, but YC actually has a number of programs now, including a MOOC, the accelerator, a Series A program to help companies raise Series A, and a growth program to help companies expand post-product market fit. Even though we've had a lot of success, I would argue that we are far more successful in understanding failure. And so sometimes we call ourselves experts in failure. And what I'd love to do is talk to you about how we see YC companies fail. Typically, the story starts after Demo Day. So, after you've raised somewhere between $500,000 and $3 million. So, like any good list, I have a top 10 list. We'll start with number 1. Assuming that raising a successful seed round means that you've reached product market fit. This is extremely common. We tend to get a lot of founders in YC who've never experienced the frenzy of a normal Demo Day and have never seen all of these interesting investors actually wanting to talk to them. Um, what ends up happening is they think these investors are so smart, and they think these investors, because they like and want to invest in their company, their company must be amazing, and it's going to go on to become the next big thing. Um, unfortunately, That's rarely the case. The vast majority of the best investors have invested in whole dozens of companies you've never heard of who've died. And so it turns out that because you raise a check from a famous name, or because you're able to raise a lot of money, almost has nothing to do with whether or not you have a good product, or whether you have something that your customers want. So, in reality, we often have to shake founders after Demo Day and say, "You know whether you have product-market fit or not. You know whether you have negative margins or positive margins. You know whether you have to lie, cheat, and steal to get a customer to use your product, whether they're coming in every day. You know whether your product is actually solving the customer's problem. You are the biggest expert in your company. Don't let an investor convince you." that you're further along than you are. So that's number 1. Number 2 is hiring too quickly. I think that startups have this mentality that after they've raised a million or so dollars, they immediately have to get to 8 to 12 people. It's like cargo culting what a successful angel-funded company looks like in the Valley. This always surprises me because around that size, The primary job of a CEO has to start switching to management. And usually for a pre-product market fit company, the primary job of the CEO should be to focus on product market fit. You can see the disconnect. So we really tell companies to not cargo cult what they think is success. When success happens, it's going to punch you in the face, and you're not going to be able to confuse it for anything else. If you're not getting punched in the space with traction, you're not succeeding and you don't need all those people. Unfortunately, a lot of the advice that we have to give a year plus after Demo Day is you need to let some of these people go because you're running low on money and you didn't find product market fit. Other mistakes in this area are trying to take on too many problems or products at the same time. We often see companies will try to spin up product 2 when product 1 hasn't even hit product market fit yet. We also see that if you create the expectation inside of your company that you have product market fit with your employees and you don't grow, your employees start asking questions. And you really set yourself up to not be able to answer them well. So you want employees who are excited to help you find product-market fit, not employees who think, "Oh, this is a rocket ship that's just false-started." That's the easiest way to get your employees to leave. Number 3 is not understanding their business model. If we really look at a lot of YC companies nowadays, we have a ton of B2B companies, and the most common mistake they make is they don't really understand if they can afford afford the process they need to do to acquire customers. Oftentimes we see founders who want to be SaaS founders, but they're running an enterprise company, or founders who want to be enterprise salespeople, but awkwardly their product only costs a couple thousand dollars a year. And so a really big disconnect is not actually pursuing the strategy that, like, just interests you, You have to pursue the strategy that actually is commensurate with how much you're going to charge and who your customers are. Much of that is decided for you. So the next one, number 4, is not understanding when it's the right time to sell into a tech startup. So this is one that YC has a lot of experience in. One of the superpowers that not many people understand about YC is because we have so many companies who've done YC, In effect, you get warm leads to anyone who's ever done YC. So it's a lot easier for YC companies to sell to one another. It's not a surprise you have companies like Stripe or Mixpanel or Brex who are able to acquire tons of YC companies very early. What's interesting though is that this strategy of selling founders early has pluses and minuses. Oftentimes people will tell you to not sell to startups. They'll say things like, "Well, your customer doesn't have much money to spend. Your customer's likely to die. You're gonna have high churn rates." So on and so forth. What I would tell you is it depends on what you're selling. We've seen a long history of YC companies be very successful in selling to startups initially if they're selling key components that if they work, they will not be ripped out. If your payment system works, you're not going to rip it out. Also, when you sell to startups, you're often pitching a founder. Founders have the ability to control budget and make decisions quickly. So there are pros and cons to selling to startups, but I'd argue that it's been extremely successful within YC. And it's a great way to essentially build a brand as being an innovative tech company that you can then use to sell to larger customers later. What I will say though is that there are some advantages and disadvantages to not selling into an early-stage startup. The advantage is clear. They have more money and less likely to churn. It's more likely to be a real business. The disadvantage that some people don't realize is that if you're selling to a larger company, you're often selling into an executive. You don't know what their power, what their budget, what their decision-making ability is inside of the company. Also, you don't know the products that that executive used previously that they might be inclined to continue using in their new role. So this is something you should always consider. When is it better to enter into a company? If you can get in early, it often is better, and you can often grow with the company. But if you're not solving a problem that an early-stage company has, you have to go later, unfortunately. Next, number 5, is, is assuming investors will be a large differentiator. The best advice I got as a startup founder was very simple. An A investor gives you money, signs paperwork when you want them to, and shuts the fuck up. That's an A. There's a lot of room below an A. There's not a lot of room above. Founders often believe that their investors are going to do far more for them than actually they end up doing. And I've thought about this a lot, and I've realized why. It's actually really simple. There's only one skill that an investor can iterate and improve on in the short term. It's hard to know whether you're a good investor for years, but it's easy to know whether you're good at this very quickly. And that skill is closing. So oftentimes, as a first-time founder, you will experience investors who are incredibly good at closing. It often involves promising you the world. And it's one of their amazing skills. Unfortunately, it's extremely rare that investors will practice the ability for helping. 'Cause once they're on your crap table, it's on to close the next deal. Their motivation is to get into as many good deals as they can. So, I often see a disconnect with YC founders who get promised the world on demo day, 6 months later people are replying to emails slowly. And so I just like to tell these founders, if you don't go in with much expectation about what your investors will do, you won't be disappointed when they don't reply to your emails. Emails. And this is more common than people talk about. I'll also say it on the other side. Once you are successful to any degree, there is a large amount of advice and help that people will be willing to give you for free. I've experienced this twice in my startups, and I see it every day with YC startups. So, someone doesn't have to be on your cap table to help you. And success creates more success. All right, number 6. The next big problem we see companies have post Demo Day is not establishing best practices around hiring. So I don't mean about the number of people they hire. I mean about setting up an intelligent interview process that good candidates are actually going to enjoy going through. Having good open communication about equity. Which is something that I see a lot of founders really flubbing pretty hard. Not setting clear expectations about what an employee's role and responsibility is going to be. Not talking about the mission and culture of the company. And then most importantly, over-believing in their ability to hire great people. I've never met a founder who says, "Oh, I'm not good at hiring great people." I always meet founders who say, "My team's the best." And not every team, clearly not every team is the best. I often ask founders, especially when they're struggling with money, do you have any non-essential employees? Non-essential. And by that I mean, it's like if this employee left tomorrow, you wouldn't be crying in your bed. An early-stage pre-product market fit company should be trying to minimize their non-essential employees. And you should not believe you're great at hiring, which means unfortunately you're probably gonna have to do a lot of firing. If someone is not an essential employee within 3 months, that's often a great sign that you didn't make a good hiring decision. And that's hard to hear because it's a lot easier to give them more time, to change responsibilities, so on and so forth. But you're not gonna be great at hiring. Everyone is not great at hiring. Google's not great at hiring. Understand you're gonna make mistakes. Number 7, not establishing best practices around management. This is something that I see extremely commonly. It turns out that early-stage management isn't that complicated, but if you don't do it, your team performs poorly. And so with this, typically what's missing are consistent one-on-ones between managers and employees. Some type of all-hands meeting. Getting employee buy-in on strategy and tactics. I can't tell you how many companies run with two co-founders who go into some magical room and come up with some magical idea and then the rest of the team has to build it, no questions asked. I always say to myself, if you're an amazing hirer and you're bringing the smartest people into your company, why wouldn't you want their opinion about what you're building? Furthermore, someone's always gonna be more motivated to build the product that they had some say in figuring out what it was going to do. So actually having a good process for getting that buy-in is extremely important. And then the last one is not creating an environment around transparency, around money in the bank and KPIs. If you're not transparent with your employees about how successful your company's doing or how much money you have in the bank, One, it's likely they're gonna try to find another job that's better. But two, they're gonna suspect the worst. We saw this at Justin.tv work the opposite way. We were always extremely transparent about our cash on hand, our months of runway, our traffic, our revenue. And there was a moment where we only had 2 months of runway left. We were generating about $750,000 a month, We were burning about $1 million a month. We had $500,000 in the bank. And instead of employees leaving left and right, we sat everyone down and we basically asked a simple question. Are we going to cut to a $100K a month burn, which is gonna give us 5 months of runway? Are we gonna break even? Are we gonna get profitable? And everyone said in unison, "We're gonna get this company profitable." 2 months later we were profitable and we ended the year with $8 million in revenue and $1 million in profit. And so when you're transparent with your employees, they will actually rally to the cause even when it looks like you're in your darkest hour. When you're not transparent with your employees, they're not gonna rally, unfortunately. And they're probably gonna be pretty vindictive. The next one is one that we suffered at Justin.tv. This was not clearly defining roles and responsibilities between the founders. So it's very common that in the early stage everyone does everything and you don't really need strict roles and responsibilities. But after you raise money and you have a couple of employees, suddenly there are some hard decisions to make. Who's gonna lead product? Who's gonna lead tech? Who's gonna lead sales? Who's gonna be responsible for recruiting? It's often the case that teams will not make these decisions, and then every decision, large or small, will have to go into some kind of founder committee to get resolved. And I think one of the best things that good startups do early on is define responsibilities and make it so that only the most important decisions end up being something that's debated. The other part about this is being willing to see your co-founder try something that doesn't work. Or being willing to see an employee try something that doesn't work. You have to be OK with that. Your ideas aren't going to work most of the time either. So really sitting down and having a hard conversation about who does what and who's responsible for what in the beginning is very valuable. Number 9. Almost done here. Not having level 3 conversations within the founding team to relieve conflicts. There will always be conflicts within the founding team. There will always be performance issues. There will always be need for change in roles and responsibilities. This will always happen in every startup. Great startups have a system to have hard conversations. Bad startups either bottle it in or get into constant fights. And so having some type of system where the founders are going to meet and work through whatever the bullshit issue is of the day is extremely important. I think the most important thing here is to actually create a space where people feel comfortable providing really on-the-point feedback without feeling like they're attacking or feeling attacked. And it takes a lot of effort to create an environment where someone can be extremely honest to someone without it being perceived as attack. But most companies slow down significantly if the founding team doesn't have these types of conversations. And there will be problems. And so this is another strategy to get through them quickly. All right, number 10. This is the biggest one, and this is something that cripples YC companies specifically a lot. Assuming the Series A will be as easy to raise as the angel round. It's interesting that a lot of founders look to Series A investors for advice on when they're ready to raise a Series A. It's tricky because you're getting advice from an extremely biased source. The thing that I hear founders tell me time and time again with B2B startups is, "I can raise a Series A with a $1 million run rate." Time and time again. Like, I hear this probably every week. So, they think, "I'm gonna take all that angel money, shove it into getting a million-dollar run rate. I'll be at 6 months of runway or less and Series As will just come knocking on my door." They say this often parroting Series A investors who go on stages like this and say, "Oh yeah, if you have a million-dollar run rate, we'd love to talk." I think the thing that we've realized is that for most investors who are considering writing a $5-10 million check, that is the first moment that they might be interested in talking to you. It's really interesting. That's the moment where they want to basically preempt or front front run anyone else, but they don't want to waste their time. So put another way, that's the moment where they have maximum leverage in the negotiation. And they have even more leverage if you've gotten to that $1 million run rate with not much runway. You've given that Series A investor maximum leverage. That's not how I like to fight battles. And so oftentimes I tell my founders to think about this slightly differently. I like to tell them, "Hey, you should think about this like a video game." And if you have to fight a level 20 boss, and you got 3 options: you'll grind up to level 10 and get your ass kicked 20 times in a row, you'll grind up to level 20 and maybe have a 50/50 shot of winning, you'll grind up to level 30 and you kill the boss every time. Which one would you rather do? Different founders have different preferences, but more often than not, Founders rather go into a Series A conversation with high amount of leverage so that they can get the investor they want at the terms they want in the least amount of time. Usually our guidelines for B2B companies is trying to be in the $150,000 to $250,000 a month range. That, and close to profitable. That puts you typically in the maximum leverage category. Which significantly increases the chance that going out and raising a Series A will result in a Series A. Most founders don't quite understand how rare Series As are. And the number of companies that go out and try to raise Series As versus the ones that actually succeed, it's a very low conversion rate. The other thing that's tricky here is advice. Your angel investors often don't give great advice on this subject. They're often parroting the advice that they're seeing from investors. So just because your angel investor thinks you can raise a Series A, once again, don't put on your blinders. You understand how much leverage you have. You can look at your numbers and you'll know how much leverage you have. The other thing is don't be assessed what you read on TechCrunch. During my entire 20s, I saw stories on TechCrunch every day about some Joe raising $10 million for some company that was clearly gonna fail. It had no traction at all. Every day. And I thought to myself, "What the fuck? Like, I can't raise $2 million, and this guy's raising $10 million." After we learned fundraising better, after we raised our Series A and Series B, after I became in my 30s, I realized something very different. Suddenly all of my friends who struggled to raise in their 20s were easily raising in their 30s. And I look at these friends and I'm saying, "You're better, but you're not that much better than you were back then. So what's different?" In reality, two things are different. Investors, I don't think for a good reason, significantly bias for second-time founders. Significant bias towards second-time founders. And there's something you have to understand. Second, if you've been in the Valley for 10 years, a lot of times you're pitching people you know. This is something that always blew my mind. It's a lot easier to try to get $5 to $10 million from someone you know than someone you're meeting for the first time at a VC office. And so, because someone raises doesn't mean they're doing well. Because someone raises doesn't mean their business is gonna work. Because your competitor raises, it doesn't mean that they're better than you. You don't know all the circumstances behind that fundraise. And so because they raise, it doesn't mean you can. And so I often tell people, you can try to think that maybe you're special. Some people are special. My former co-founder, Justin Kan, he can probably raise with a haiku. He could raise a lot of money. But most people aren't special. They need leverage. And so when you go into a fundraise, the better bet is to have as much leverage as possible. These are the 10 things that most YC companies get wrong, especially the B2B ones, post-Demo Day. The last thought I'll leave you with is that I actually think most of our startups have a pretty good hypothesis, a pretty good thesis on why their company can be big. I think that most of them fail not because their hypothesis or thesis was wrong. I think most of them fail because either their timing was slightly off or because of some of these problems, they didn't have time to iterate their product so that they can find the real solution to the problem that they're trying to solve. If you take the case of Justin.tv and Twitch, Twitch sold for $1 billion, but most people don't really understand the backstory. You know, Twitch started as an online reality TV show. And if you look at the story, it started in 2006, and by 2012, it was worth $24. By 2014, it was worth a billion. So that entire time, we were trying to figure shit out. The entire time. It took us a long time, 6 years. To actually figure out what the market wanted, how to solve the problem we were trying to solve. If your plan is you're gonna— you already have solved the problem, or you think you've already solved the problem, or you just need 6 months to figure out the solution, beware. It might take a lot longer. And so if you keep a small team, if you iterate quickly, and if you don't believe the hype, and you actually understand whether your product-market fit you can actually take the time you need to solve the problem. And I tell this to YC companies at the end of every batch. You know, it might take you 1 year to get it. It might take you 3 years to get it. It's a lot easier to raise your Series A post-product market fit. It's a lot easier to scale once you know what you're building. So stay lean until you have that moment. The last thought I'll leave you with is a really fun one. I always find funny, founders hate it. We have two kinds of founders calling in the emergency. One, when their company's dying, that's extremely typical. The other is when their company's succeeding. I think the funniest thing that happens in startups is success hurts more than failure. Because success comes with all these expectations. So not only is everything breaking and everyone's complaining and your customers are yelling at you and and your servers are going down. But now, like, you might have something. So you better, like, the expectations are now really high. So just remember that. Like, remember that unfortunately on this journey, like, the good times feel bad and the bad times feel bad. It kind of sucks. But hey, this is what you signed up for, and no one forced you to do a startup. So with that, thank you so much. Good luck.