{"url": "https://www.youtube.com/watch?v=LPXw8HQ_5Rc", "page_url": "https://www.youtube.com/watch?v=LPXw8HQ_5Rc", "title": "Startup Advice from Sam Altman (Y Combinator) - Wellington, New Zealand", "show_or_channel": "Edmund Hillary Fellowship", "date": "2015-01-01", "duration_s": 1158, "type": "talk", "speakers": ["Sam Altman"], "confidence": "high", "notes": "Wellington NZ 2014", "people": ["sam-altman"], "id": "2015-01-01_edmund-hillary-fellowship-startup-advice-from-sam-altman-y-combin", "audio_duration_s": 1158.292625, "worker": "jc-w05", "transcribe_s": 15.1, "model": "large-v3", "segments": [{"start": 6.03, "end": 35.84, "text": "I'm Sam Altman. I'm the president of Y Combinator. We're based in the States, and we fund startups. We funded more than 800 now, so quite a lot. Some of them have been Reddit and Dropbox and Airbnb and Stripe and a lot of others. And we also try to build up the most sort of helpful network for founders that we can. We hope that by doing that, we can make it much easier to start companies and then help develop technology into really big companies that we all use and love every day."}, {"start": 35.84, "end": 68.58, "text": "That's my two sentences. Our version of this is that you should only start a company if you're in love with that particular idea. You shouldn't start a startup for the sake of having a startup. You should have an idea that you really believe is something that you want to do and the people you want to do it with. Starting a startup is sort of extremely hard, like unimaginably hard, but just barely doable. But it's so hard that if you don't believe in this and if you don't want to do it for 10 years..."}, {"start": 68.58, "end": 97.73, "text": "There are better things to do. And it's better to join a great company than to start a mediocre one. But all of that said, it's doable. And I think we found that startups are much more doable, I think, than people think. It looks like now we have enough data to say this confidently. More than 1% of the startups that we fund go on to be worth more than a billion dollars. And they go on to change industries that they're in."}, {"start": 97.73, "end": 133.14, "text": "They change. And this is not just in software, but everywhere. And I think New Zealand is one of the top up-and-coming areas that we see for startups that we're funding. Startups are staying here. They're coming to the US. They're not wherever. But there's clearly the environment here to support it. So if it's something you really want to do, you should do it. But if it's not something you really want to do, then it is way more pain than it's worth. Well, here's just one data point from what we have."}, {"start": 133.14, "end": 162.48, "text": "No company that we've ever funded that has been founded by one person is in our top 30 most successful companies, even though we fund a fair number of single-founder companies. It is possible for sure, and there's many great examples around the world, but these things, again, are so hard to do, and the psychology ups and downs are so intense that if you have a co-founder or two, hopefully you're not all at the crash at the same time and you support each other."}, {"start": 162.48, "end": 191.81, "text": "One thing that we say is that startups usually, the expected value dips below the x-axis at some point. Things go really wrong. And if you're by yourself, you're just going to stop. And if you have a co-founder, you will sort of like, out of obligation to each other, keep going. And you can sort of like cheat this law of thermodynamics somehow and both do it to support the other person. On the other side, we've never had a startup with four or more founders be really successful. It's one of these things where there's like exponential or I guess factorial."}, {"start": 191.81, "end": 228.35, "text": "Communication overhead. And it just gets way, way harder. So in our experience, two or three co-founders works the best. But I'm sure others. I think having something like a zero get built in a city is this unbelievable thing. It's very, very hard to do the first time. But having such a successful technology company creates this entire ecosystem of people that work there and then go on and do new things and invest in startups. And I think that's such an incredible asset."}, {"start": 228.35, "end": 264.02, "text": "I think that there are, in addition to the interpersonal pieces, there are two other things that we really think about that make a culture great. One is momentum. So, you know, winning teams tend to keep winning and teams that sort of take their eye off the ball and start losing tend to keep losing. And so one of the most important jobs as the entrepreneur is to make sure the company never stops winning."}, {"start": 264.02, "end": 289.92, "text": "It's very difficult to dig yourself out of that hole. And if you keep that going, it's just this ongoing tail end. The other thing that we see is the missionary-mercenary divide. I think most cultures end up being either a mercenary culture or a missionary culture. And the great companies are always the missionary cultures. If people believe in the mission, if you are sufficiently ambitious, if you present people"}, {"start": 289.92, "end": 320.14, "text": "We're going to make this big impact, and we're going to reach these people, and we're going to do this important thing that's going to be great for the world. That's what motivates the best people. And if you have a culture like that, where you do have this ambitious, important thing, it's easier to motivate people. In many ways, it is easier to start a hard company than it is to start an easy company. More people will be excited. You'll have a better culture if you're doing something that matters, if you're doing this big thing. If you have the kind of culture where you're telling people,"}, {"start": 320.14, "end": 354.93, "text": "You need to get into your desk at 9 and leave at 9 because we work 12-hour days. People might do that for a little while, but that is almost never. In fact, that is not what I have ever seen produce great results. The best companies do a few things across the board. Everyone's successful in a different way, but the commonalities, one is they talk to users very early. They don't try to keep the idea a secret. They're open. They actually listen."}, {"start": 354.93, "end": 384.61, "text": "A common finding mistake is to talk to users because someone told them they're supposed to, but be so convinced they know the right answer to not actually listen to what their users are saying. You have to build something that's so good that users spontaneously talk about how much they love it. That is the test, and that is eventually how you keep growing. The second thing that correlates really well is obsessive customer service. The founders themselves are constantly talking. This is part of talking to users."}, {"start": 384.61, "end": 412.08, "text": "And then the third thing is a very short cycle time. So the way to make a great product is to make it better every day. And if you make it, say, 2% better every day, that compounds exponentially over time. And the best companies get this very tight feedback loop. They have an idea, they test it, they change it, they test it, they change it, and they keep going in. This, over time, produces incredible results."}, {"start": 412.08, "end": 441.6, "text": "And then the last thing I'll say on this point is an obsession with the details of the product quality to a level where they almost, where it's almost like stupid does tend to correlate with success. So founders that like, you know, spend a lot of time on their documentation pages and their jobs pages and that if, you know, like there's this famous YC lore story where Drew was supposed to be speaking, Drew Houseman, the founder of Dropbox, would be speaking at a YC event."}, {"start": 441.6, "end": 477.74, "text": "A new version of Windows in Sweden, Service Pack 2, something like that, went out that broke Dropbox for seven users, and he just canceled the speaking event to go fix it, because it was bothering him that much that it was broken for this tiny number of users. So there's this obsession with quality of the product, which is what makes users love it, that turns out to be really important. Here is one really important point, which is..."}, {"start": 477.81, "end": 506.02, "text": "That people get more ambitious over time. And when the founders of Airbnb first came to us, they were delighted they had three users. And if we told them that someday they were going to be running a $50 billion company, whatever it's going to be, with more nights than any hotel chain in the world, they would have looked at us and said, you're crazy, we don't want to work with you because you don't know what you're talking about. So I think there's this pressure in the last couple of years."}, {"start": 506.02, "end": 534.99, "text": "To have an immediate vision about how you're going to have this $100 billion company. And no one honestly has that in their first days. Google almost sold the company for a million dollars. They would have if the offer came through. Zuckerberg very, very nearly sold Facebook to Yahoo for a billion dollars. So it's okay to like..."}, {"start": 535.44, "end": 565.23, "text": "Become more ambitious as you go. In fact, that's what should happen. So what we look for is people that have a kernel of an idea that we believe can develop into something huge. But you can't focus on that right away. You have to focus on delivering something great for a small number of users first. The way we talk about this is that you can either create an initial product that a lot of users really like or that a small number of users really love. Obviously, you'd like to make something that a lot of users really love."}, {"start": 565.23, "end": 593.57, "text": "But those ideas get taken by big companies and you can't compete with them. And so you don't have to have a takeover the world plan on day one. In fact, very few of the companies that we've been involved with that have gone on to do it thought they were going to initially. The founder had something they were really passionate about. They had a plan, as Scott was saying, about how they had some defensibility. And it just grows and grows and grows from there. But at the same time, you can be extremely ambitious from day one."}, {"start": 593.57, "end": 622.56, "text": "Over time, right? So here's an example. SpaceX. Elon had already done a couple of companies, so maybe he had the right to be really, really huge vision. But from day one, it was, we're going to Mars, and that's where this company, like, that's the success case. And I'm sure that Mars is not success. But product one, so yeah, you can have a vision like that, but product one was not, like, let's build the Mars Colonial Transporter. No, you don't start with that. You can't, that's the thing, right? That's a bad idea. So, like, in fact, we just funded a rocket company."}, {"start": 622.56, "end": 659.09, "text": "The next YC batch. And they would like to go to Mars too someday. But their initial project is to launch 20 kilo satellites. And that's sort of what I meant by you can start with something less ambitious. I mean, generally, I think it's difficult to classify companies by a particular country. The one thing I tell founders anywhere outside of the US as a general statement is that it's okay in the US to be sort of..."}, {"start": 660.72, "end": 687.66, "text": "Arrogant is the wrong word, but it's okay to be ambitious, I would say. I think it is more forgivable to stand up as a complete unknown and say, I'm going to try and start the best search engine in the world, or I'm going to try and start a rocket company that's going to take me to Mars so I can die there, or I'm going to start a social network, and I know you think I'm crazy because it's just starting with college students, but someday everyone in the world is going to use it."}, {"start": 688.37, "end": 711.34, "text": "And people aren't afraid to fail. If they fail at something big, I think Silicon Valley is even a little bit too forgiving of that. But it is one thing that has worked really well for US-based companies. So we wouldn't take that too far, because I think in the US we take all that stuff a little bit too far. But I would just sort of, that's my general advice any time I'm out of the States."}, {"start": 723.89, "end": 753.6, "text": "One thing I believe very deeply is you should never start a company in a particular area because someone on the panel told you that you're well suited for that. That is just a recipe for disaster. So I will try to not name specific areas because I think anyone in the world can now compete in any area they want. One of the really amazing things about the level of globalization that we've had and the depth that the internet has reached is that anything you can do in the US, you ought to be able to do in New Zealand."}, {"start": 753.6, "end": 780.98, "text": "That said, I really want to echo the point about the advantage of being in a small country. I think that it's so much easier to do something disruptive in a lot of different spaces here. A really ambitious startup in New Zealand could probably get a government regulation change, which in the US is just like laughably impossible. Having a small but concentrated customer base that you can test with."}, {"start": 780.98, "end": 819.04, "text": "You can iterate on it. You can get to this point where you build a product that people love before you unleash it on the world. It's incredibly valuable. So I do think that no matter what you decide to do, you should use the advantages of being in a smaller contained ecosystem, which are huge as much as you can. It is changing for sure. And it is easier to raise money now online or without in-person meetings than ever before. But I think..."}, {"start": 819.04, "end": 846.26, "text": "The correct way to think of fundraising is like really high value sales. And just like any other sort of sales, the human connection really matters and doing it in person is so much better. So one thing that happens at Y Combinator a lot is that companies will come to the Valley for a period of time, raise money because it's like 100 times, literally 100 times easier to do it in person, and then go back to whatever country they're from."}, {"start": 846.26, "end": 877.54, "text": "It used to be that most investors wouldn't do that. They would say, I only want to invest in companies that are going to stay within 20 miles of Palo Alto. But that has totally changed. That is long over now. But I still think it is, for sure, a lot harder to raise money via email. In fact, so hard that other than something like AngelList, I don't recommend trying it. And I think it's just worth it. If you can't raise money from investors here, which hopefully you can, then I would just..."}, {"start": 877.54, "end": 918.62, "text": "You know, be willing to fly a lot. We will fund nonprofits directly, and we've funded a lot, and some of them I think are totally awesome. But I'd say generally we want to do things that have a huge positive impact. And as Scott was just saying, I think most of the time those are for-profit entities. I think in the current world, and you can argue whether or not capitalism is good, I don't think it is, but the best way to make a change, and if you look at the big improvements that have happened in the last hundred years, most of these come..."}, {"start": 918.62, "end": 946.72, "text": "Google certainly did not understand the impact that making all information instantly accessible would have on the world and it's been a massive, massive change. If we can get a nuclear energy company to work, that will have a massive, massive change. But the converse of that is that"}, {"start": 946.72, "end": 977.28, "text": "I won't invest in a company I believe is bad for the world. Twice I have passed on a company that I thought I was going to make money on for sure because it was bad for the world. Once it was so bad it sort of self-imploded and the other time I would have made a bunch of money but it wouldn't have been worth it. I think the biggest problem that the world, the macro problem the world is going to face is that there's going, left unchecked, that we're on a trajectory to have massive, massive wealth inequality."}, {"start": 977.28, "end": 1007.34, "text": "And slowing economic growth leading to mass unemployment. And, you know, it turns out that economic growth, I think, like everyone wants their life to get bigger every year. In a no-growth world, it's very zero-sum of everyone's fighting. I think the only way that the sort of developed world will continue to function is with massive economic growth. Business is what's going to deliver that. And so I think just continuing to drive economic growth is probably the most important thing for ongoing prosperity."}, {"start": 1007.34, "end": 1050.19, "text": "Peace and everything else that we can do. I had a rule with myself that I would only deal with the most current crisis at a time when things got really bad. Where people get really debilitated is when they think about the 30 things that are going to go wrong and that are going to kill the company. And that leads to paralysis and inaction. And the way to get through it is to say, you know what? I'm pretty smart. I've gotten through a lot of bad stuff. I'm going to solve those other 29 problems. I don't know how. I'll figure out that later."}, {"start": 1050.19, "end": 1078.96, "text": "But here's the thing that might kill the company today. I'm only going to worry about that. And if we live to fight another day, then I will have earned the privilege to worry about a set of things. And focusing yourself on this most urgent fire and giving yourself permission to ignore these other fires until tomorrow and not just sort of thrashing and context switching between them is really big. And then the other thing is just keep perspective. Terrible things have happened to a company in the past."}, {"start": 1078.96, "end": 1108.62, "text": "Terrible things would happen to the company in the future. This probably won't be the one that does kill the company. And if the company does get killed, it's just a company. It's not life. It's not family. There are more important things. So the worst-case scenario here is terrible, but not world-ending. And you can probably solve it. Founders are incredibly resourceful, and you can almost always figure out some way through it and survive for another day."}, {"start": 1122.26, "end": 1151.63, "text": "I actually studied AI in school, so I'm happy to answer this. I think it is probably the machine learning is probably the fastest going concentration in every good CS program in the States. And I expect that will just keep that growth rate will continue. And I think there's just there is such strong belief among new CS students that this is the most important thing to be working on, that that's going to just keep going."}, {"start": 1151.63, "end": 1157.49, "text": "I mean, probably AI will kill us all, but until then, we're going to turn out a lot of great students."}]}