The following is a conversation with Brian Armstrong, co-founder and CEO of Coinbase, the largest cryptocurrency exchange platform with 98 million users in 100 countries. I recorded this conversation before this week's SEC probe into whether some crypto listings are securities. Our world runs on money, and cryptocurrency seeks to build the next chapter of how money works. Coinbase and Brian are trying to do this by working with regulators. Bureaucracies resist change. The latest SEC probe is a serious attempt to limit fraud but also risks limiting innovation. Decentralization is a hedge against centralized power. Now, a quick mention of sponsors: Audible for audiobooks, Skiff for email, BiOptimizers for health, Fundrise for investing, and Athletic Greens for performance. And now, dear friends, here's Brian Armstrong.
Let's start with the fact that you're a programmer. What was the first program you've ever written? Or the first one you remember?
The first memory I have of programming was probably in middle school. I remember it was recess, and they had this time period where you could read books. The other kids were reading comic books, but for some reason I had gotten into this idea that I wanted to get into computers. I was playing with computers at home. So I got this book from the library called 'How to Learn Java in 30 Days'. I was reading this book at recess, and I didn't understand anything. I remember I went home and tried to get this thing working. If you've ever written a Java program, the first lines are like 'public static void main string args' and it's so foreign and difficult to get started. I was frustrated. I didn't understand anything. So the first thing I wrote was probably just a hello world app in Java. But I felt so confused about what was happening that I later learned a bit of PHP. PHP was more fun for me because it was like 'Oh, just print out what you want.' It didn't have all this complexity. So I got more into PHP and started building simple websites. I learned some HTML. That was my introduction to programming.
Yeah. You know, Java has a lot of out of all the Hello Worlds you can possibly write, Java is the one where I think it's the longest. Which is quite interesting because Java is often, at least for a long time, was used as the primary programming language to teach people how to program, or at least about object-oriented programming. I think most universities have now switched and high school switched to Python.
I'm not sure if that's the case.
Probably better. It's easier to learn. It lowers the makes it less scary.
And certainly none of them use PHP. I love PHP and I feel like it's a dirty secret I have to keep private to myself, like it's somebody I'm seeing on the side or something like that, because it's just not a respected programming language because I think there's so many ways you can write poor code with PHP.
Which is why it's not respected.
Yeah, it's a scripting language more so, although I of course Facebook built like a huge stack on top of it, a valuable company. But I still love Ruby to this day. Ruby's probably my favorite language. Python's great, too, but I just love the idea behind Ruby that it's like, let's make it easier for the human, harder for the computer, and make it a joy to be expressive. So, I was never the best computer scientist, but I was a good hacker. I could rapidly prototype products using languages like Ruby. Do a lot of computer science programs still use like Lisp and Scheme and things like that?
No. No, they do for That's like if you're hardcore. If you're legit, you're going to do some of the functional languages. I think there's a few others that popped up, but Lisp is a distant memory for a lot of people. That's like somebody has to like you go to a library you dust off the book. But Scheme a little bit. I think if you're starting there are courses about programming languages. Lisp might be one of those. You know how there are languages that nobody uses anymore? Like ancient languages. You might have to go to school in that same way for programming languages. Back in the day we used to use parentheses. I of course still use Emacs as the editor for most things that I do. And Emacs is you know a lot of the customization you can do is in Lisp. And that's the language probably when I first really fell in love with programming is Lisp. Because for a long time throughout the earlier history of artificial intelligence Lisp was the primary language, but it still had a life in the 90s and the aughts where some people would use it. Such a beautiful functional language, but it just somehow didn't pick up. That said, I should say push back. PHP I feel like it's still true that most of the web runs on PHP. Most of the back end is still PHP. So if you look at it's like the stuff that people don't talk about. It's what runs most systems in the world? What runs the most back end? What runs most front end? JavaScript you know HTML.
JavaScript I think the Stack Exchange surveys show JavaScript is the most popular language in the world I think, right?
Oh yeah. In terms of programmers and numbers of I wonder.
By survey of number of programmers on Stack Overflow.
Oh yeah, but those are also the cutting edge, right? Those are the people that are just excitedly writing code. I wonder if there are people that are just maintaining gigantic code bases.
I feel like the amount of Java out there just running industrial systems has got to be enormous. And then of course in the banking industry finance it's like even older stuff COBOL and whatnot.
I've been actually looking for somebody to interview who represents COBOL and Fortran, like who's the figure still holding the flag. I did, you know, with Java, founder of Java, creator of Java, creator of Python, creator of C++, but nobody wants to hold the flag for COBOL and Fortran, even though some of the most important systems in the world still run on those.
Yeah. Uh, like power systems and infrastructure systems, which is fascinating. And ATMs and stuff like that. Like a lot of stuff that we rely on, that just works. And the reason we don't change it because it works well. Is written in language that people don't use anymore.
That'd be a cool series of interviews. Get the stuff that's like tech that was invented 40, 50 years ago, but still is being used widely. I mean, Emacs is an example of that.
Let me ask the big question of, what are cryptocurrency exchanges and what's Coinbase? How does it work? Before I'll ask even bigger questions, but it's just a nice kind of palate cleansing question of what is Coinbase?
Coinbase is a cryptocurrency exchange, brokerage, custodian. Basically, we're the primary financial account for people in the crypto economy. How they buy crypto, how they store it, how they use it increasingly in different ways. We can talk about that. So yeah, we want to be the way that a billion people hopefully access the open financial system globally.
How does it work? What, uh, what's cryptocurrency? You know, there's Bitcoin, there's Ethereum. What does it mean to be an exchange? What does it mean to store? What does it mean to transact? How does it What does Coinbase actually do?
Okay, so, you know, basically in any given market, there's some people who want to buy, some people who want to sell, and you keep an order book of all those prices. And then, if someone's willing to buy for more than the lowest price someone's willing to sell, then you get a trade to execute. That's kind of how an exchange works underneath. And a brokerage is kind of simpler than that even. You don't have to look at the whole order book, but you just go in and say, 'I want to buy $100 of Bitcoin or whatever cryptocurrency.' You get a quote, and if you like it, you can hit accept. The core things we do to make that seamless is we have to do payment integrations around the world to make it easy for people to get fiat currency into this ecosystem. We have to work on cybersecurity a lot. There's lots of hackers trying to break into our systems. We have to integrate with the blockchains themselves, which are periodically updated. And then we have to store the crypto that people buy securely. Crypto is like storing private keys. We've invented a lot of cool technology about how to do that securely. So those are some of the pieces that had to come together to get that early buy-sell experience to work. Coinbase actually has a lot of different products now. We have an institutional product, Coinbase Commerce for merchant payments, a self-custodial wallet, and we just launched an NFT product. So we're sort of like a portfolio of crypto products now. But the core thing we got started with and still the majority of our revenue is people just want to come in and buy and sell some crypto.
I'll ask you about wallet and NFTs about what is it called the stripe type?
Oh yeah, Coinbase Commerce.
I'll ask you about all that. But order books and exchange, what's the difference between that and stocks for example which there is also order books.
Yes, I mean stocks trade through order books, too. So do commodities.
So it's all similar type of situation. So when I want to buy one Bitcoin and I see Coinbase say the price of that Bitcoin is say $40,000. And I press buy, what happens?
Um yeah, okay. So you've gotten a quote. (laughter) Like, you know, when you press the button on your keyboard like an electrical signal goes up the wire. Now we won't get into that level. That's also important the timing, right? Because there's not price fixed. Yeah, that's true. It's giving you a quote. There's a whole concept of slippage. By the time the quote is executed, if the price has moved too much, we may reject it. But the simple version is we'll check the order book, give you a quote. It's good for some period of time or for some amount of slippage. Then we're initiating a debit to your payment method, whether credit card or bank account or stored dollars. We're debiting that and crediting you the crypto, and we're taking a fee for it. That's fundamentally what's happening underneath.
And then there's just some interesting slippage. How do you calculate the how much slippage is allowed? Like how do you know these things? Because order books are fascinating, the dynamics of that is pretty interesting.
Yeah, so there's a lot of people like traders who get super into this and high-frequency traders and arbitrage and all kinds of interesting topics. Flash Boys was like an interesting book on this whole thing. You want access to information the fastest, sometimes even putting your thing in the data center right next to the thing. We don't allow that colo stuff because we want it to be more democratized. But basically you give a let's say we wanted to keep it math simple. We want to charge a 1% fee. So if you're buying $100 of Bitcoin, we'll charge you $101. We've presented you the amount of Bitcoin you're going to get for the $100. Now, let's say 10 seconds later you hit accept. We go to fill the order. So it's going to be some error bound around that 1% fee. And if we think we're actually losing money on the trade, we'll often reject it. So some part of the fee the slippage is incorporated into that. Averaged over a large number of people. It's fascinating because even that little detail probably requires a lot of experimentation.
But so, how do you protect against that? How do you protect? We'll talk about cybersecurity in interesting ways, but there's a lot of clever people trying to do clever things to earn an edge. How do you stay one step ahead?
Yeah, and it's kind of like a giant bug bounty out there because if you get it wrong, there's people who are going to arbitrage that. And we've had people sort of pen test our systems in really creative ways where they'll just fire programmatically with APIs, they'll fire off a million different quotes and look for one that's out of bounds and then take that money. And we get people doing all kinds of crazy stuff.
There's no silver bullet. It's a bunch of lead bullets, right? So, it's like, one thing we do is we just have good test suites. You're testing every piece of code that goes out. That's just common good best practice, but it's particularly important in financial services. Another thing we do is we hire third-party firms to try to audit this stuff and break in. Another one we do is we have a bug bounty program. We basically pay white hat hackers to find this stuff before the black hats do. And we've paid out lots of good bug bounties. So, try all the above. And occasionally you don't get it right and you lose some money and then you fix it and you keep going.
Let's talk about cybersecurity a little bit more. You mentioned using stolen bank accounts. So, that's another one. That's another interesting one. How do you protect against that?
Okay, so fraud prevention yeah, is a big topic. So, one of the things people do is that you use machine learning. You look at, to protect against it. What you want to do is kind of build up a labeled data set of all the different people who have turned out to be fraudulent and good actors and hopefully collect as much data as you can. Then you'll feed hundreds or thousands of these factors into your machine learning model and it'll come back with a risk score. An example of the kinds of factors people create is, obviously I don't want to disclose too many because it's a cat and mouse game, but relatively well-known stuff might be device fingerprints. What kind of device are you on and what fonts do you have installed? A lot of people who are farming lots of these accounts, they're using emulators and virtual machines. They're not like an average person on that device. One of my favorite metrics we tracked was called improbable travel velocity. We would track people's IPs. You might see someone who was one day in Austin, Texas and then like an hour later they were in London. That's very improbable. Sometimes people are using VPNs, so you got to be careful because legitimate people use VPNs too, but if it's not possible for them to have gotten on a plane and gotten there that quickly, then they're usually spoofing a device or IP. Another fun one is real users will type their credit card one number at a time. Scammers have a list and they'll paste in a whole number. So you can look at the number of milliseconds between keystrokes. There's all kinds of stuff people have come up with.
And even for travel velocity, you could probably incorporate VPNs, too, because there's probably a travel velocity for VPN switching, too. That's human-like. If you're using legitimately VPN for something else, that might be There's like legitimate uses to. Actually, you know, I feel embarrassed that I don't know this. Probably should. But the I'm not a robot captcha thing? So, that probably works in the same way. Like how do you move your mouse, maybe? Or how the dynamics of the clicking? But how does that even work that well then? And why can't it be faked? I need to look into this because it's a trivial captcha. It feels like should be very crackable and yet a lot of high security places use that. It's really interesting.
Yeah. It's another cat and mouse game. I think they've Yeah, but it's using a lot of similar signals like mouse movements, keystrokes, and then obviously all the stuff that comes over the wire with your browser. So, like what operating system, what fonts, what headers are being sent over. There's actually an old website, I can't remember what it was called, kind of like Panopticlick or Panopticon, but it was a proof of concept that they would just show you all the data that was getting sent over with your request and say that there's only one person in the world who has this exact set of data. It's you. So it's almost like a clever workaround to track somebody even if there wasn't a cookie involved.
Yeah, this is a fascinating world. We can't see anybody. You're in the dark and yet you have a lot of signal you have to figure out who's a real person who's not, who's a robot who's not. Let me step back. We're going to jump around all over the place. Let me step back.
Like your interviews. You get into technical topics.
So let's just use Bitcoin as a measure of time. You started Coinbase when Bitcoin was $10. And you just mentioned an incredible system with security, with transactions, everything is thought through. There's a lot going on, but what was version one? Back in those early days, the first prototype of Coinbase, what did that look like? Like what did it take to write it, to think through it, and make it work enough to at least make you believe that it's going to work?
Well, I definitely didn't know if it was going to work. I mean, it was kind of I felt like I was just following my gut. I was working at Airbnb, I was a software engineer there, project manager. I was working on some fraud prevention stuff. And I read the Bitcoin white paper in December of 2010. I started going to some Bitcoin meetups in the Bay Area. Met lots of interesting people there, crazy people, anarchists, really brilliant people, all the above. So I started nights and weekends trying to hack together a prototype. My initial thought was, SMTP is a protocol that runs email. Git is a protocol for version control. But people made Gmail and GitHub. Most people don't want to run their own email server or Git server. They just want a hosted thing that does all the security and backups for them. So the thought in my head was Bitcoin is this new protocol. There's probably going to be somebody who makes a hosted service that does all the security and backups for you and makes Bitcoin easy to use. So maybe I should make a hosted Bitcoin wallet or something. I was going to make Gmail for Bitcoin. A bunch of people told me it was a bad idea. Most of my smart friends said they didn't get what I was doing, Bitcoin sounded like a scam. Other people who understood Bitcoin told me it was dumb because if I stored all that Bitcoin I would just get hacked. So I thought, I'm not going to go all in and store everyone's Bitcoin. That would be too much right now. I have a day job. But let me just make a prototype and tell people this is a beta, don't put any real money in it, and see if there's interest. If I felt I was onto something, maybe I would go do this as a company. I really wanted to be an entrepreneur. I was 29, almost turning 30. I always wanted to start a company but I was an employee. So I had this prototype I was hacking together nights and weekends. I actually wrote a whole Bitcoin node in Ruby, which turned out to be a weird decision in hindsight because Ruby wasn't the most performant language. We've subsequently had to rebuild that many times. I had this hosted Bitcoin wallet, but I didn't have any users for it. I applied to Y Combinator because maybe if somebody there writes me a check, it will make it feel like a real company. I was trying to find a co-founder unsuccessfully. I was basically wandering in the desert. I had a lot of self-doubt because all my friends thought it was dumb and maybe Bitcoin would get shut down. There was definitely a feeling of wandering lost in the desert. Paul Graham and the Y Combinator group wrote me the first check after I interviewed. They wrote me a check for $150k. That was the first time somebody I really looked up to said this is worth pursuing. That gave me the confidence to quit my job and try it. I found the right co-founder after Y Combinator. We still didn't have any customers. I launched the hosted Bitcoin wallet. People were signing up, maybe 100 people would sign up and then nobody would come back. In Y Combinator they often tell you to talk to your customers and improve your product. So I emailed five of the users that had signed up and asked why they didn't come back. One guy said the app was okay for a beta but he didn't have any Bitcoin, so he didn't know what to do with it. I had a light bulb moment. I asked if I put a buy Bitcoin button in there, would you have used it? He said yeah, maybe. So we went about the process of getting a bank partnership, payment rails, an exchange, basic exchange functionality. The minute we launched that feature where you could just click buy, put in your bank account or credit card, buy Bitcoin, it showed up in your account. From that day forward the number of users started to go up. We finally had found product market fit after two years of wandering in the desert.
So you weren't even thinking about the buy the on-ramps. You were thinking it would be just the wallet, a place to store Bitcoin that you've already gotten. Okay. That's such a pain to do to have to work with others to convert dollars of any fiat currency into Bitcoin. So did you were you overwhelmed by the immensity of the task here or were you sort of not allowing yourself to think too deeply through this whole thing and just letting the optimism take over here?
You know, I was really looking forward to doing something crazy and a big challenge. I love kind of crisis moments like that where I'm very determined, especially when I get set on something and I'm just like, you know what? I'm going to figure out a way to make this thing work no matter what. So I reveled in that. I had read all these books about startups and every startup has major setbacks and nothing works. So I had no idea if I was doing anything right at all, but I was kind of loving the experience of it in a weird way. It felt stressful at the time, but I felt like I was on the right path somehow. I just kept going.
What was the darkest moment that you've gone to in your mind during that time? What were some of the tougher moments? You said self-doubt. Have you where did you go in your mind? Is there a moment where you're just laying there, this is hopeless?
Well, there's a couple moments I'm remembering. For whatever reason, I had this big chip on my shoulder. I really wanted to do something important in the world. I could have a good life and work for some good companies and write software, but for some reason I never wanted that for myself. That probably would have been healthier. But I was willing to sacrifice a lot for that. I remember one of the darker moments was when we probably had maybe five employees. A bunch of bad things happened all at once. First of all, we were all sleep deprived, which exacerbates everything. The site would go offline in the middle of the night and I'd get paged, sometimes two or three times a night. So in that environment, you can get discouraged. One bad thing was we had a bug on the website and thousands of people on Reddit and Twitter were pissed at Coinbase because balances were showing wrong. They were saying, this company is over. I'd never had this feeling of a thousand people mad at me at the same time. I'm a pretty chill guy, most people don't get mad at me.
Can we pause on that? That's so interesting. So you were saying here's a dream, I'm trying to create something, and now forever the reputation of this dream is ruined. It will never be recoverable. It's over. That kind of feeling.
Yeah. Well, I didn't know at that time. I was like, is this the end? Everybody hates us. So is it over? Nobody told me before starting a company that a bunch of people will hate you for this. That's very counterintuitive because most companies are trying to do good things in the world. Even if someone is trying but failing, I'm generally rooting for them. But that's not the case at all. Most founders I've known have gone through this too. They're very surprised at the amount of hate they get. I think it's actually a muscle you can build, your tolerance to it. Because you feel terrible at the center of the storm, but then you go talk to your family and they have no idea about the negative press. They're busy in their own life.
Can I once again linger on that? There's an interesting person I'd like to bring up just as an example. Bill Gates. He gets a very large amount of hate on the internet. And there's something about him, this is me talking, not you, that he seems out of touch about that hate. I believe, at least in my understanding, that with the resources he has, he's trying and is actually doing a lot of good. And yet there's a gigantic amount of hate, conspiracy theories, and stuff like that. And feels like that's the case because he's somehow out of touch with people. So I wonder how you stay in touch with the voice of the people without being destroyed by the outrage. Is there any wisdom you have to that?
I don't know about wisdom, but I've thought about this too. You want to always be open to feedback, especially from people who have your best interests at heart. You can become isolated from it and surrounded by yes people. Maybe Xi and Putin are in situations like that, I have no idea. But if you listen to too much and try to please everyone, you'll never get anything done. Most of the best leaders are people who can act when they believe they're doing something net positive for the world and humanity, and they don't really care if they piss off some proportion of people. Almost anything you do of significance in the world today is going to piss off 5% of people, maybe 49%, maybe 60%. So you never want to become so surrounded by people who work for you and will say yes, that you think you're a genius. That's how you become a dictator. But you also can't care so much about what people think because then you'll never do anything authentic to yourself. One other thought: I've thought about this a lot. Why do people generally hate on Zuck and Bill Gates? They don't really hate on Elon. Well, actually Elon has a lot of haters too, but it's a different thing.
This is measured. I was looking at some surveys. I think Zuck is the most loved and hated. He's the most hated. Then Bill Gates, and Elon is down there, in the double digits but low double digits. So it's interesting. You just look at the stats, ask yourself why.
Right. So I ask myself this sometimes too. I don't claim to know any of these people well, but I've met them briefly and my impression is that they're all smart people trying to do good things in the world. So why is it that some are really hated and some aren't? It's a complicated question. Obviously Zuck and Facebook got blamed for the whole election thing. Social media has gotten a lot of pressure. But one thing I've noticed is that a lot of these people have a little bit of Asperger's. Sometimes people with Asperger's don't emote in the same way. I think it's almost a form of bias against their cognitive type. That person doesn't emote right, so I don't trust their intentions. Another thing I've thought about is that sometimes I think some leaders...
Well, let's take a break and come back to that. But I think we have a good sense of the early days. Thank you for sharing.
You know, like maybe Zuck or Bill Gates, they can come across as a little bit PR rehearsed, like they're basically giving the PR-approved answers, whereas Elon just says whatever he thinks, to a fault. So even if people hate what he says, they're like, at least I believe it's authentic.
So I've always thought about that too for myself. I'm like, how do I? You can screw it up on both sides. If you just come out and say whatever's stream of consciousness, you'll often end up pissing off people on your team or tripping over some regulation. There's all kinds of things about running a public company – you can't say certain stuff. But if you're too PR-approved in your answers, nobody trusts what you're saying. So this is something I think about a lot. I don't think I have the right answer, but I'm trying to find that balance.
More and more with the internet, there's a premium on authenticity, just like you're saying. People really appreciate that. So for leaders, it's a challenge: how do I make sure I'm authentic, but also don't say stupid stuff?
And so that's an interesting thing. I've noticed that having interacted with a bunch of leaders, you have to be careful how much you surround yourself with PR folks.
Because the best marketing folks are extremely good. They understand that great marketing is authenticity – showing and revealing the beauty – as opposed to PR out of fear: don't say that, don't say this. That pushes you toward a bubble where you can't express your quirks and weirdness and the cool things about what you're doing. I find, especially with tech, like even Coinbase, the way to reveal the beauty is not only by showing what you can do with it, but by showing the great engineering underneath. Let the nerds shine. It doesn't have to be a commercial of a happy family sending flowers. It can be gritty, real stuff. That's a general observation. But you said you were talking about dark moments and that there were people on the internet pissed off that the site was down. You said there might be something else.
So they understand exactly what great marketing is. It's authenticity, revealing the beauty. As opposed to PR out of fear. If you do that, you start living in a bubble and can't express your quirks. I find, especially with tech, like Coinbase, the way to reveal the beauty is not only by showing what you can do, but by showing the great engineering underneath. Let the nerds shine. It doesn't have to be a commercial of a happy family using Coinbase to send flowers. It can be gritty and real. Anyway, that's a general observation. But you said you were talking about dark moments, and that there were people on the internet pissed off that the site was down. You said there might be something else.
Yeah, so sleep-deprived, a bunch of people on the internet were pissed at me. The balances were messed up. People were tweeting the company's over. Just give the money back. Then somebody posted my cell phone number on Reddit, saying if you need to get a hold of the CEO, whatever. I'm in the office late at night, trying to hack and fix the bug. We all need food, and my phone is blowing up. I answer a call, and it's a guy trying to deliver food. I answer, 'Is this Brian?' 'Nope, wrong number.' Click. Next call – a reporter from Japan. 'Nope, wrong number.' Finally I get the food, we survive. I remember a point that night where I just curled up in a ball on the floor and cried for a little bit. I let myself wallow in self-pity, took a nap for five minutes, and then said, 'Let's fucking solve this.' I stopped being a little whatever and got back up. Sleep deprivation combined with the stress and pressure of the site going down and everyone wanting it up – that pressure was mentally tough.
Sleep deprivation combined with the stress and pressure of the site going down and everybody wanting it up – that pressure was mentally tough.
What was your source of strength during that time? Like somebody patted you on the back and said we got this.
Yeah, it definitely helped to have a co-founder. There's an old saying: it's better to be in a great relationship than to be single, but better to be single than in a bad relationship. Co-founders blow up a lot of companies, but when you find the right one – I was lucky to find Fred Ehrsam – that was very important. There were moments where I was at wit's end, and he was like, 'Dude, let's rally.' He carried the team a couple of times in key moments.
What advice would you give to startup founders about this particular stage? About surviving the five-employee stage?
Yeah, if you're pre-product-market fit, the best advice I have is that action produces information. Just keep doing stuff. Paul Graham had a great line: startups are like sharks – if they stop swimming, they die. So even if you're not sure what to do, just do anything, because it will produce information. People liked it, they didn't. That was very true for me. There were times I just did something instead of debating it endlessly. I shipped it, and the minute I shipped it, I knew we built it wrong, but now I had an idea of what to do next. I only would have had that idea if we'd gone through the exercise. Another analogy: you're at the base of a mountain shrouded in fog, trying to figure out how to get up. You can only see three or four steps ahead. So you have to take steps into the unknown. When you take three steps, the next three steps are revealed. Sometimes you hit a local maximum and have to retrace, or come up to a cliff. But most people don't take the steps into the unknown because it's scary – what if I fail? But the risk is often overestimated. If you go to a startup and it fails, you're even more valuable to your next employer. You can raise a seed round, pay yourself a salary, try for two or three years, and if it doesn't work, get another job. People overestimate the risk and never start because it seems crazy and all your friends think it's silly.
Yeah, Paul Graham had that line. Startups are like sharks – if they stop swimming, they die. So even if you're not sure, just do anything, because it produces information. That was true for me. I just did something instead of debating. I shipped it, and the minute I shipped it, I knew we built it wrong, but now I had an idea of what to do next. You only get that idea by going through the exercise. Another analogy: you're at the base of a mountain shrouded in fog, trying to figure out how to get up. You can only see three or four steps ahead. So you have to take steps into the unknown. When you take three steps, the next three steps are revealed. Sometimes you hit a local maximum and have to retrace, or come up to a cliff. But most people don't take the steps because it's scary. They think, 'What if I fail?' But the risk is overestimated. If you go to a startup and it fails, you're even more valuable. You can raise a seed round, pay yourself a salary, try for two or three years, and if it doesn't work, get another job. People overestimate the risk and never start because it seems crazy.
It's just basic fear. It's the same kind of fear if a guy sees a cute girl at a bar – the fear of coming up to her. 'What's the actual risk, exactly?'
She'll say, 'No, thanks, not interested.'
And I guess the risk is that it's going to be mentally difficult to deal with rejection. So just like it's mentally difficult to deal with failure. If you have a bunch of ideas and you implement them and realize they're not good, that can be difficult to keep pushing through. But that's life. You're supposed to persevere through failures. The risk is low.
And then the whole time through the fog up the mountain, you're looking for product-market fit.
Yeah, that's right. You know you have it when the usage of your product keeps growing without any marketing dollars. More people keep going back every week or month. You're basically watching your stats. Nothing is working, you see little wiggles of false hope in your metrics. You just keep talking to customers, improving the product. Talk to customers, improve the product. And try not to run out of money, so be really scrappy. If you're lucky, you hit some threshold where the thing is good enough, or you hit on a use case, and it organically starts to grow. Then you have a whole different set of problems once you hit product-market fit: how do we scale, hire people, raise more money? The problems totally change.
Well, you were there through the whole thing. So that's the other question that's fascinating. Again, back to the girl at the bar, how do you hire people? How do you find good friends, good relationships? In this case, how do you hire good people – engineers, executives, all of it?
One thing is I've done a lot of reps on hiring at this point. Coinbase has about 5,000 people. Probably the first 500, I interviewed every single one. But you have to remember, on average maybe 10 people went through the process for everyone we hired. So by the time we had 500 employees, I had done like 5,000 interviews. I was very burnt out. Some days I did seven interviews in a day. You might not get burnt out, but it's a different kind of interview.
Very different, very different.
Yeah, very different because you're – most of your interviews lead to rejection, which is also exhausting.
So first of all, most of your interviews lead to rejection.
Which is also exhausting.
Yeah. And there's a whole part of the interview which is about candidate experience. Sometimes you know it's not the right person, but you want to make sure they have a good experience. If you're exhausted and on your sixth interview and you wrap it up quickly, you're going to create a detractor who says, 'That company or Brian was rude to me.' So I had to work on that in the early days. I needed to make sure when people came in, I made them feel comfortable, asked warm-up questions like, 'How was getting in the office?' and not just run through an assembly line.
But also there's a moment when you early on know it's not going to be a good fit, and you still have to land that plane.
There's also a moment when you early on know it's not going to be a good fit, and you still have to land that plane.
And all that kind of stuff. That could get really, really exhausting. So, yeah, anyway.
Yeah, so we tried many things over the years to make interviews more efficient. We usually get them down to 25 minutes. For hiring a big team of contractors, I've seen people do 10-minute interviews. You can interview a thousand people in a week. But for full-time employees, I usually do 25-minute interviews. One thing we've done is put a Google Form online with basic hurdles. We used to have brain teasers in the early days, but we don't do that anymore. We do normal interviews and references. The kinds of things I ask: I think about what we need this person to accomplish in the role, get really specific. I'll ask them to tell me about a time they did X, or the hardest problem they've had to solve in Y, and what they did specifically to overcome it. I'm also looking at culture fit. Are they concise communicators? Can they give a clear answer and stop talking? Some people ramble for 5-10 minutes. Some are interrupters. I'm looking for humility too – I ask about a time something went wrong, and I see if they were part of the solution or still holding on to blame. So I'm looking for: can they get the job done, will they work together on a team, can they communicate effectively, and do they fit our cultural values.
I've done a bunch of hiring for this podcast and at MIT. I was always looking for simple questions that can reveal a lot. I used to ask, 'Do you think it's better to work hard or work smart?' I had this idea that people who say work smart don't actually work smart. The textbook answer is work smart, but people who haven't done anything say that. My belief was that to discover what it means to work smart, you have to work your ass off and fail a lot. So I was suspicious of people who said work smart. But I've learned that there are people who are exceptionally efficient even at a young age, so you can't disqualify based on that. You have to dig deeper. Some of the most interesting people unapologetically say work hard, and they're usually the ones who know how to be efficient. I've always searched for questions that can reveal something profound in as brief a question as possible. Then there are basic attention to detail and brain teasers depending on the role. Is there a question you find yourself leaning on?
You see, you said 'how did you solve a hard problem in your past' – that's one.
We started with brain teasers at Coinbase, but we got away from that relatively quickly. I think it does show how someone performs under pressure, but it's not a super reliable indicator. Some people are good at typical work, but being put on the spot in a live interview is not typical. People get nervous. So I'm a little skeptical of brain teasers. There's a whole question about universities getting rid of entrance exams, so if you're hiring right out of university, the degree sometimes doesn't mean what it used to. I've heard some companies are creating their own standardized testing for college grads.
That's fascinating. It's not just about you trying to hire a great team, it's also to help them find the right place to work. It's a two-way street. All right, so once you found product-market fit, how did Coinbase become what it is today? Let me ask an engineering question actually. From the Ruby wallet days, what were some of the interesting challenges? Were they engineering, regulation, financial, hiring, lawyers?
Post product-market fit, a lot of it is scaling and building out an actual company. I was still writing a lot of code for a while. We had about 25 people. One of our investors came by and asked, 'Brian, how much of your time are you spending writing code?' I said maybe 50%. He said, 'How much hiring?' I said 20%. He said, 'I think you need to flip those numbers. The company is not going to scale. You're the CEO, you don't need to be writing code every day.' That took me a while to internalize. I was still holding on too much to decision-making. Even to this day at Coinbase, we have to push down decision-making. So you need to make sure you have enough money to survive a downturn. We were periodically profitable during up periods, but crypto would go down and we were unprofitable. We had to manage our psychology and balance sheet. We had to professionalize services that had been thrown together by 20-year-olds – cybersecurity, finance, etc. Our finances were a mess. I remember after raising a Series C, I put $25 million in a different bank account because I had heard horror stories of startups thinking they had money but then going insolvent in three weeks. So I used that as a cheap hack until we could hire a real CFO and finance team.
What was the ordering of hiring? How many engineers early on? Did you even have a CFO for a bit? What was the landscape of hiring as you built the company?
The first person we hired was for customer support because we were staying up till midnight doing it. Then we got more engineers. The sixth hire was a recruiter – that turned out to be a great force multiplier. We needed legal and compliance badly because of questions about legality. We needed more senior people. It was hard to find legal people who work with crypto; nobody had more than three years of experience because it had only been around a year. We found people from adjacent fields. There's a certain personality type willing to join early companies because they have no structure. You can't commit to a team or boss. Hiring is one of the hardest things in the early stage. You need to find people crazy enough to join you on the journey.
So one of the interesting things about Coinbase is you're very focused on the mission. That simplifies things, makes hiring easier, makes working at Coinbase easier. What is the mission of Coinbase?
It's to increase economic freedom in the world.
And what is economic freedom?
Economic freedom is a term economists use to measure countries. It looks at property rights, free trade, stable currency, ability to start companies, corruption, etc. Organizations score countries by it. The cool thing is that it positively correlates with things we want: higher growth, happiness, better environment, income for the poorest. It negatively correlates with things we don't want: corruption, war. So if you give people good property rights and rule of law and allow them to trade, it encourages them to do good stuff and society benefits. I spent a year in Buenos Aires, Argentina, which went through hyperinflation. There's a pessimism that creeps into countries without economic freedom – people don't try hard because they don't think they'll keep the upside. In America, historically, people try more because they believe if they do good, they'll keep part of it. When I read the Bitcoin whitepaper, I had a hunch that this technology could inject good financial infrastructure and economic freedom principles into countries all over the world. Crypto is well-suited for economic freedom: property rights through a 12-word phrase, global free trade, stable currency through Bitcoin and stablecoins. It's a secret hiding in plain sight that can create economic freedom for everyone.
I didn't know about Argentina. Why did you end up in Argentina?
