Max Levchin of PayPal at startup school #2017

Hello, I am Max Levchin. So I never prepare for these things as a rule, and then the night before I start fretting and asking my friends, 'What should I talk about?' and send Paul messages like, 'Should I talk about this?' and Paul never responds. So he also didn't respond this time around, and I have no idea what I'm gonna talk about. So on the drive down, I kind of start taking notes while I drive, and I'm still here, which means that I didn't crash. But inevitably these things cause a real life-threatening moment. So normally I just sort of get into these events and I go, 'Well, I have piles and piles of material from my life as an entrepreneur that's generally pretty funny,' and I just sort of tell stories and people laugh sometimes, and I judge how long I can go by either the limit set by the organizer or how hard people laugh. More often than not, it's the latter that gets me to get off the stage. So this time around, I'll actually try to do something useful. Since I did risk my life by trying to type something with my thumb while driving, I promised Paul two things: one, if I run out of material, I'll just cut it over to you and you can ask me questions; and two, if I really feel like the place is falling asleep, I'll just kind of throw out two funny stories. But I really, really ask you not to blog or repeat them, because some of them are pretty embarrassing, and not for me.

So one of the things that one of my friends and a YC alumnus advised me last night is to talk about founders and co-founders, because that's the stuff that startups are made of, and it's so hard. People are having such a hard time finding co-founders, it's such a problem. So I thought, 'Alright, what do I know about co-founders?' I started four companies, all of them failed. The fifth one was PayPal, the sixth was Slide. I'm about to start another one, so I'm just one of these crazy guys that can't get enough of this stuff, and I should by now know exactly how to do co-founder picking. So as I got here, I realized that I have a funny story about one of the moments where I thought, 'Holy [__], I picked the wrong co-founder.' I was in a parking lot... I actually forget the name of the complex, but there's a kind of nice-ish complex right on the other side of Sand Hill Road. I'll remember the name, of course. But that's where Peter Thiel, my PayPal co-founder, used to live right as we started the company. And I lived just there, walking distance from there. So pretty frequently, when we'd be brainstorming about PayPal, he would either drive me to my apartment and then drive himself home, or drive himself home and I'd walk to my apartment because I didn't have a car. But we would frequent... I would drive to his apartment and drive myself, so whatever. One night, something like midnight, we were in my car, drove to his apartment, parked right outside the complex, and started talking about how we're gonna take over the world. It's about 11:30, and I'm just going on and on, thinking it's really exciting. And it's good things. By two o'clock in the morning, I realize that the parking lot is completely pitch black. I go, 'Well, Peter, get out. It's way past my bedtime. I gotta go.' Alright, so he gets out. I turn the key, but I realized that I left my lights on. Car battery's dead. And there's this moment of, 'Well, what do you do?' Peter looks at me, I look at him. Oh good, maybe he walks off at this point. It's 3 o'clock in the morning. I moved to Palo Alto half a year ago. I certainly don't know... I certainly can't afford AAA, so I wouldn't know where to start getting a replacement battery. Crap. I should not have started a company with this guy. Obviously he will abandon me in the middle of a dark parking lot at two o'clock in the morning. So, true story. This was probably the only time I actually doubted Peter as a co-founder. But I did have a moment to reflect. In fact, I don't actually remember how I got out of that particular pickle, to be completely honest. I may have just left the car there and walked home or something, because none of this is really memorable.

With that, I wanted to start making a list of my top mistakes as an entrepreneur. I was gonna make a top 10 list, but the drive from San Francisco, where I now live, is too short, so it's a pop-something list because I didn't really number them and I don't remember where it ends. But it gets progressively darker and more embarrassing. So this is kind of an unvarnished account, and I will give you the point and then offer a rebuttal. I've been quoted saying that one should always look for a co-founder, and I think PG is a proponent of that notion as well. Lots of people will tell you it's lonely at the top, it's really bad news to be alone. It's actually not entirely true. I've had a chance to reflect on that because PayPal was pretty successful and I had a co-founder, and Slide was by some measure successful financially, by other measure didn't come close to PayPal. We'll talk about that in a moment. So I generally thought it was kind of mixed results, and I didn't have a co-founder. So I thought, 'Well, clearly correlation, sample size of one for each category, there he goes.' It doesn't quite work that way. One of the more memorable moments where I actually completely stretched out the thing about Peter being an [__] for dropping me in the parking lot like he did was this one time we had about nine weeks of cash. It was May 2000. Peter at this point... I actually know it's just history, by the way. This was all new to me at the time. I was like, 'Holy crap, this guy didn't see the future because he was very convincingly arguing for basically the end of the world in capital markets, like any minute now it's gonna crash.' And of course, as some of you will remember, 2000 was a bad time. Company capital markets couldn't raise money. So like, 'Wow, this guy has a crystal ball, he's amazing. I gotta stick with him no matter what he does in parking lots.' And of course, now that I've known the guy for over ten years, he actually just always predicts doom and gloom. At any time, you see on record, it's like, 'Well, Peter Thiel says the end of the world is coming, the real estate is gonna collapse, it will be a dark depression, possible people with pitchforks will enjoy.' And at some point he'll be right, and that'll really look amazing. But between now and then, you can just continuously predict badness. But in 2000, the writing was actually on the wall, and we were really just feeling the heat. It was bad news with nine weeks of cash. We tried to sell to Yahoo. I met Zod, who I didn't really know at the time, a super nice, very mellow guy. The first opening question he had to me said, 'How many hackers do you have?' We had like seven engineers, and he's like, 'Just hackers?' So I'm like, 'Oh, that's it.' So anyway, Yahoo didn't buy us. We were running out of money, and we were just starting to feel the burn of fraud. So we were working insane hours. I walked into Peter's office one of the nights, and he was like, 'Hey man, how's the deadline for this and that?' And I said, 'Yeah, you know, well, what does it matter? I feel like we're gonna run out of cash, and you're right, the markets have tanked, and we're toast.' He's not a warm fuzzy guy at all. He sort of looked at me from his desk, completely exhausted, just flew in from Japan where he was trying to raise money from the Yakuza. I kid you not. Incidentally, the Yakuza wired some money before we figured out the real Yakuza, and then we said, 'Oh, we can't take your money,' and they said, 'No, no, we already wired it to you.' Like, 'How did you wire to us? We didn't really give you wiring instructions.' Like, 'We're Yahoo's?' True story. I look it up: Hikari Tuition is the name of this shell company. I wish I could make this stuff up. So Peter looked up from his desk, fresh from Japan, basically said, 'Look, don't worry about it. We will raise money. You'll be fine. I give you 100% confidence that we'll be able to close this round. Don't run out of cash. Just launch the damn thing already.' And this was probably the single warmest experience I've had with him as a co-founder, because I actually went home. Yeah, exactly. He could have given me a hug or something, but no, all he could say is, 'Don't worry about it, just launch the thing already.' But the job of a great co-founder is not to commiserate with you. The thing that makes you feel great about having a co-founder isn't some guy you can go, 'Oh man, we are [__].' A great co-founder is actually in the moment when you're at your lowest to tell you, 'You'll be fine.' Not like, 'You'll be okay,' like your mom telling you you'll be okay when you're failing out of college. You're still gonna fail out of college. You should actually drop out and start a company, but you're denied. So the job of a great co-founder is to provide that platform of support when you most need it. At that time, I actually walked out of his office thinking, 'You know what, I don't know if he's right or wrong, but that just gave me all the reason in the world to go stay up one more night and finish that thing and launch it.' Whatever it was, it worked out great. He did in fact close the round, and we didn't really even have to take a valuation hit at all. But the thing about having a co-founder, or the thing about having... sometimes people tell you equal parts co-founder or somebody with the same amount of equity or whatever, really what it boils down to is that moment of getting emotional and practical support. And if that person is not your co-founder, or that person is a board member or a confidant or a YC compatriot or whatever, that's what's important. You can actually start a company entirely on your own, and I think you'll be alright. But not having anyone like that next to you can be really, really detrimental, because at the very top, when you're responsible for it all, no one can hear you scream. You go home and go, 'We are [__],' and there's no one to tell you that you're not, and know exactly what the parameters are. That's really painful. And even if you have a great girlfriend, boyfriend, wife, whatever, and they'll tell you you'll be okay, it's exactly the same thing as your mom telling you you'll be fine if you fail out of college. You can take it next year. No, your company will fail, and that's kind of the end of the world for that particular moment in your life. So it's an important thing to know.

The corollary to this story, and as I said, I'll get progressively less funny and darker... picking a co-founder. But I do have some additional zingers, I think. Picking a co-founder that you have a hard time dealing with, or don't quite have the same value system, or that you don't want to impress on occasion, those are the sort of things that I jotted down to try to figure out what I thought about when I was trying to decide who will be my co-founder for company X, Y, or Z. The way I described the original meeting and conversation... I met Peter maybe a hundred yards away from here, so this is a particularly auspicious spot. He was this nerdy guy who was giving a lecture about something I had no idea about: financial systems and currency trading. He was obviously not a computer science guy, but he's obviously a nerd. And we formed this kind of weird friendship, courtship between two nerds that were trying to basically measure dicks, trying to impress each other. We spent the next couple of months hanging out a lot and throwing math puzzles at each other in cafes, basically like, 'Well, I can do this,' and 'Well, I'm a chess master.' I played chess in high school, I was the 20th highest ranked chess player in America when I was a teenager. And these were the things where we were kind of sizing each other up and figuring out what exactly this guy can do. But by the end of that weird process, we basically knew that the other person was someone we ultimately really respected. So whatever crazy flaws he had... I keep coming back to the parking lot story because I don't have enough material... it was always something that we could rely on, where we would basically be like, 'Hey, no, that guy is really cool. I remember when we met, that was really important.' To give you a counterpoint, I once started a company with a guy I thought was incredibly impressive. He finished business school, he had run a company, he had all kinds of accolades. He was like five years my senior, which when I was 19 years old, I was really impressed with. And we start a company together, and we were selling something to the enterprise. God knows what it was, it was a really bad idea. I actually don't remember exactly what it was. I saw him on the phone once trying to sell something, and I thought, 'Holy crap, he doesn't know what the hell he's selling.' He is explaining something, and he's either lying or he doesn't know what he's talking about. At this point, probably incorrectly, I note that I'm a little bit older, I try to correct his pitch while he was on the phone, writing down like, 'Hey, here's how you pitch this thing.' To his credit, he actually kind of didn't really say anything. He finished talking, he hung up, and he just exploded. And I realized, 'Whoa, he thinks he was completely right. There's absolutely no business that I had teaching him how to do his job because he was a sales guy, he was a successful sales guy.' And there is absolutely no way to ever run a company with a guy who would explode like that at me. So we broke up, and the company obviously failed as a result. But the only salient point of the story is when you ever have that moment with your co-founder where you go, 'Oh, this isn't going to work,' like the weird tang of 'we might have to break up,' when you think about a relationship getting divorced, now that you know it will never work out. I have experienced that enough times in my entrepreneurial life now to know that this is actually good advice. A lot of times, you'll feel really wrong and really bad because the person you're working with, and you're trying to get by, the weird encounter you just had is someone you really believe can add a ton of value to the enterprise, somebody your venture capitalists have just put a big bet on, and you're the young and inexperienced co-founder. That doesn't matter. If that's the thing that makes you go, 'Whoa, this isn't the right person to be with,' it's not... my favorite movie quote of all time is, 'Whenever there's any doubt, there is no doubt.' It applies to just about everything in life, but it particularly applies to people. When you can't really believe anymore that this person is the right co-founder, run away. Even if it means returning money to your investors, it will never work.

Speaking of that, one of my favorite mistakes, which I really paid for probably again and again in my professional career, not in a co-founder situation fortunately, but in employee situations. Frequently, you'll find yourself facing an employee, or sometimes a group of employees, God forbid, but typically an employee at a reasonably high level in a company, typically a really talented engineer that is very well-liked, someone who is just obviously very talented, people love them, they want to work with them, for them, they want to connect. That's great. And you know that person is a net negative. They're either just a nasty person, or they're creating a faction within the company, or they're causing rifts, or God knows what else. That is the time to let them go. The public service that I have to offer here is: no matter how hard you think it's going to be to get rid of that person, no matter how difficult you think that conversation is gonna be, do that now. It's the sort of thing: if you're gonna get divorced, do it now. If you know you're gonna get divorced, do it now. It's really hard. It's one of these things where it feels very gut-wrenching. Firing people in general is a huge pain. I recently read an article on Quora, which was anonymous, whoever you are, calling me a modern-day fascist. I think the word was misspelled, so I feel good about that. But because the person claimed that I fired people... something along those lines. I certainly have fired people, and firing people is always probably the single most painful process for me during running or managing a company. In particular, if you actually like the person. If you don't like the person, if they did something really nasty, firing them is like, 'Hey, guess what, you're an [__], get out.' But it's like, 'Hey, you are a great human being, I like you, it's just not meant to be.' It's a lot like breaking up with someone, except it's the person you probably spend a lot more time with than your significant other, because you're sharing an office, you know, smelling each other's hormones all day long, as opposed to just occasionally. So the situation there is really, really difficult. But you got to do it. However, I come with a message of hope and peace. I will teach you how to fire someone right now that will make both of you feel pretty good, or at least as good as you can under the circumstances. You sit the person down and you say, 'I'd like to ask you to resign.' It works. I have no idea why it works, but when people hear that, they kind of go, '[__], you're right, it's not working.' Write that down somewhere. It really does work. The most difficult firing I've ever done, I started with that line, and it worked out a lot better than I thought. Except for once, when the guy actually started banging his head against the table, which was really weird. That might also have to do with the fact that I had to break into his machine to prove that he didn't write the code he claimed he did, which was also a true story. I won't tell you anymore of that one.

On to some other classic mistakes of mine. Incidentally, all of these lessons... if it sounds kind of funny or non-sequitur or whatever, at least it's true. All these things I have committed several times before I learned the lesson. So I'm a smart man: a wise man learns from the mistakes of others. I've obviously not learned nearly as much as I could from my own mistakes, but I can at least record them. A classic engineering mistake that I have committed, and Peter in fact summarized it or distilled it for me, is confusing hard and valuable. When I was... I'd like to say your age, but I actually can't see half of you, and I'd like to believe that I am at least occasionally mapping well to the age of the audience. When I was really, really young, as in I just graduated college, I thought that basically a good use of my time would be to just bang my head against the wall to see if the wall would give. That's how Peter described it. What it meant is that I'd run into a problem and I think, 'Oh, that's really hard, and well, I think I'm smart enough to crack this particular puzzle, it'd be great.' Somewhere in there, I also convinced myself that as soon as I do that, money will pour down from heaven, or accolades, or some sort of great things will happen because I solved a really, really hard problem. Reality is, the causality typically works one way but not the other. Really, really valuable things are frequently pretty hard. If you want to build the next master password file for the internet, like Facebook, it's pretty hard. However, walking through walls, to give an example, is really, really difficult, but it's absolutely of no value to anyone. Unless you can really do it, the trying process will just break your very sporty body. And intellectually, it works the same way. To give you a true and honest account of the early days of PayPal: the company was founded under the bright name of FieldLink, which was meant to connect field workers, link them as a matter of fact to their enterprise systems. This sounds unfamiliar? Discussed? Nobody really wanted the damn thing. But we wanted to build a system that would take your PDA and give you a true, fully secure VPN into the enterprise. That's what PayPal was founded to be, because I wanted to solve really hard problems in security. But security actually is a pretty difficult thing in its own right, but I didn't think it was hard enough, so I wanted to put a VPN powered by RSA encryption, which is a relatively difficult mathematical structure, onto a Palm Pilot, which at the time was powered by the DragonBall chip, which is a Motorola 68000. Maximum speed is 60 megahertz. Those of you who don't know what 'M' stands for because you were born in the gigahertz days, it's really, really slow. So signing, for example, or generating a key... forget signing, generating the key in a one-time operation in an RSA cryptosystem on a DragonBall chip took about 43 minutes on average. So you look at your Palm Pilot, you wanted to read your little corporate email from the field, you wait for 43 minutes, and you get through. Okay, so I built it, and I was like, 'Hey, Peter, we're gonna start a company around this.' And Peter was nuts enough to say, 'How? Sounds pretty good. Can you improve it? 43 minutes is a little long.' 'Yeah, I can totally do it.' So we spent a couple of months, and Peter put in a bunch of money, and we're kind of cranking at it. 'Yeah, I think I can do like 20 minutes. I figured out some really clever things.' He's like, 'Are you mad?' I drove up and down Silicon Valley pitching this to various enterprises, and to Palm, and they were like, 'Yeah, if you guys can do it, it'll be great, but DragonBall is gonna be replaced by the next chip, DragonBall Z, and that thing is gonna be powerful: 24 megahertz.' Alright, so 17 minutes. So it was definitely looking really grim. So I went to Stanford and found a guy who's now a good friend of mine, who was at the time an assistant professor, now a full professor, named Dan Boneh, probably one of the brightest cryptographers working in the field today. I told him all about this stuff, and he said, 'Have you thought about elliptic curve?' 'Oh [__], you're right, I completely forgot about that stuff.' Elliptic curve cryptosystem is a different kind of cryptography. Look it up if you're interested, but it's much, much faster because the math is a lot simpler. Discrete log problem beats factoring products of large primes any time. So I went and implemented my own elliptic curve, and then I found a library that somebody did, and it was like, 'Man, we could do it in like two minutes,' which actually would have been okay. Except I went from one hard problem to another hard problem. But no one really wanted to read their corporate email on a Palm Pilot. It just wasn't a market for it. It was well ahead of its time. Moreover, there were less than 1 million Palm Pilots sold at the time, and they were mostly in the hands of consumers, and they were not allowed in the enterprise. No security officer was going to allow them for the next ten years. So this whole thing was just a grim dream upon grim. So long story short, if you have a really, really hard problem you want to solve, ask yourself the question again and again: is it of any value to anyone? And if the answer is no, don't bother.

I will break this up now with a short... I can tell by the nodding of the heads that this particular one didn't resonate, but it's probably the single best one I have to offer as far as usefulness goes, so maybe it'll stick. But I do have a moment of comic relief for you. One other thing that I learned at PayPal and at Slide, but at Slide in particular: don't let your office be a dump. True story. If you think that having mismatched IKEA desks is really cool right now, but as soon as you get funding you'll fix it... if you're working really, really hard, you won't have time. As soon as you get funding, you'll be busy getting more people into your company. And whatever state of the office you're in right now, if it's a dump, it's just gonna be worse than a dump when you move office. You're gonna take all the crap that you've amassed in dump number one with you, and you'll make a bigger, more elaborate dump number two. So case in point: you don't have to go with Neiman Marcus, but be careful with the whole dumpy office thing, because it really sticks. I remember looking back when we were moving out, having been acquired by Google, from the office that we had, the last office that Slide had, which was office number four, I think, like, 'Wow, it's a really large dump we have here.' So I was kind of bummed out about that. I think we could have done a nicer job. Next one, I'll do a better job, I think.

Anyone here still watches or watched South Park? Good. I've never actually seen this, so I am retelling you a story I have in fact no real knowledge of. But one of my more honest investors, and I will keep their name out of this conversation, one time took me aside after a board meeting that he was present for and said, 'Have you seen the South Park underpants gnomes episode?' I said, 'No, I don't really watch TV. I don't have time for that. I work so hard.' He said, 'I don't give a [__], watch the underpants gnomes episode tonight.' Of course I didn't, but I asked for a summary. He said, 'Look, basically your business model is the underpants gnomes model: collect underpants, ???, profit.' So the fancy way of putting this is: typically people, when they start companies, follow a vision. They kind of wake up one morning and say, 'Holy crap, I have a problem I really want to see fixed,' or 'I know somebody else's problem they want to see fixed.' And that's a great aim, that's where you want to go. There's a part that is very easy to get really attached to, namely, 'I'm gonna go and write some code because that's what I know how to do. I'm really fast, I'm really good at it. Go, go, go.' But it's that middle part, the question mark in the underpants model, where after you've collected all the underpants, or whatever it is you're gonna do, you have to tie that together enough times where it winds up being profitable, or winds up being a business. It doesn't have to be profitable; you might get acquired for $20 million, half a million per engineer. So that middle part is called lack of clarity. It's the inability to have the discipline to condense your vision into a strategy. It is probably the single most annoying thing about every potential angel investment that I evaluate, that I typically pass on, or always pass on as a matter of fact, when the founder cannot say, 'Hey, the distillation of my vision into a specific strategy is this.' My head turns off and I go, 'Underpants gnomes.' I've never seen the episode, but I'm confident those of you who have know exactly what I'm talking about, and those who haven't, by now should go and watch it. It's clearly a brilliant distillation of American capitalism.

So how did I wind up with the underpants? This is the sad part. I know I should know better, right? I'm on stage giving you advice. What the hell am I doing if I don't have the distillation down? The thing that plagues second-time founders, and this is potentially relevant advice to a lot of people, but it's at least interesting hopefully, the thing that plagues people pretty often is fear of failure and fear of judgment. I certainly for one thought it would never be a problem for me, but after PayPal, for a few years I actually lived in a strange world where I thought it would be more important to live up to the shadow that my team and I cast with PayPal than to have that middle part really, really figured out. It's actually pretty easy to do, especially if you have success in your past. You can raise money at a huge valuation. You can see things that people are too polite to call you out on, to tell you that this is just another pan strategy, and you can kind of weave your story through without that middle part ever really being fleshed out for a very long time. Eventually you have to deal with it, and I certainly had to and dealt with it. But it doesn't really matter exactly how it applies to me personally. The lesson to be learned there is: starting a company or running a company requires you to become oddly unemotional. That means dumping fear of failure, dumping fear of judgment, dumping the desire to impress anyone. Nothing really matters in running companies other than passion for the product. That's the only emotion that you should allow yourself to have. Everything else is just extraneous, it's garbage, it doesn't really help you at all. If you focus on that, you'll realize that the holes that exist in your vision or in your story are really big and bright, and you have to fill them with content, or with substance, or with meaning. But if you spend a lot of your time going, 'Wait a second, how will this make me look?' you will inevitably wind up in a lot of dead ends. So this is probably the single most important lesson I can offer, especially if you're contemplating your second startup, whether your first one failed or succeeded: be mindful of emotion that doesn't serve the purpose of building an amazing product with the passion that you have for it.

All of this, I realize, is probably a little too balanced around, and it's a list that I constructed while I was trying to figure out what I can say that doesn't embarrass anyone too badly, except for myself of course. The final point that I have to offer, actually that I wish I could take credit for, but really came from my wife when I was going over the stuff with her. I thought through these same mistakes ad nauseam during Slide, during PayPal. We've been together for a long enough time now where she'd see me miserable over my own errors and errors of my peers and errors of my co-founders enough times. So the most important thing you can actually advise yourself or anyone else: ignore your mistakes. What is your problem? You're good at some things. Just go do that. The number one thing you should worry about is: am I doing the things that I'm good at?