Patrick Collison & John Collison - A Business State of Mind - [Invest Like the Best, EP.348]

22:53The other one that we think is exciting is marketplace-based business models, thanks to the internet. I think there's a few components, but one of the things you need is the liquidity of the internet that you get these Airbnbs and Uber, you probably need smartphones, not just the internet. So you need the tech foundations, but then you also need to make it easy for people to actually transact with each other and do two-way stuff. And, you know, Stripe Connect is the marketplace payment product that didn't really exist before that. And so I think you get these new kinds of businesses that are much, much easier than before.

23:21I think you're saying at some point, a change in degree is a change in kind, which I think is true. But I really like this question as well. And actually, if any of your watchers or listeners have suggestions, we'd be very interested. I think Atlas was this in that I think we know because they tell us that there are thousands of businesses that have been started in countries where high-tech entrepreneurship has not previously been as common or as popular that wouldn't have been started if not for Atlas. Steve Randy Wallman, who writes really interesting materials about economics and finance and so on, he had this blog post in 2013 or something that really stuck with me about how credit financing is fixed cost to the borrower, asymptotically fixed cost within some range or something, has the unbounded risk of destroying your business.

24:01Equity capital has unbounded costs, but does not come with the embedded risk of possibly blowing up your business. It's interesting to think about the possibility of a continuum there, to what degree it's necessarily dichotomous. If you think about a small business, effectively, their options run into that dichotomy. And if you just take a restaurant, why is it the case that, one, the restaurant probably can only get access to financing that might blow up their business? and certainly maybe they kind of convince some investors to write them their first check or something, but there isn't a whole lot in the middle.

24:32And I think one interesting story from the last 15 years on the internet is the proliferation of crowdfunding, obviously, different variants of that and so forth. So I think the question of, are there interesting new funding models that could be invented is a good one. Dominant assurance contracts, I thought were all interesting. They never really seemed to have taken off, but it feels like there might be kind of new points in the space. I think there's stuff in the domain of funding models. Something like a DAO feels like it should work, maybe not DAOs in particular. We do discuss different versions of this.

25:00I think standardization is often underrated. And one of the best things the EU has done is just standardize things that were previously for no particularly profound reason. In the sense of like an internet protocol as standardization, we all just agree that this is the way we do things. And so everything's interoperable as a result. Yeah, basically. The US is obviously with uniform commercial code. And I think the US minus that would be just like a way worse place to do business. But again, in a hard to articulate way, it sounds very arcane as well. You know, the product liability requirements at this stage versus that stage, blah, blah, blah.

25:36Pretty quickly, the listener is clicking on the next podcast episode. We've thought about, are there things that Stripe could help standardize that are not standardized today? We can just be a coordination point. Now, obviously, this YC did an awesome version of this with safes. How much less investing would happen at the early stage, absent the existence of safes? I don't know, but that seems plausible to me that 30 % or something, it seems credible that it's significant effect. Or maybe if someone else is standardized, if maybe we'd have standardized in a worse point in the space or something, we're going to make the whole podcast episode about this, but I think it's a really interesting question.

26:06If you think about the, I like the idea of the hundred trillion, where's the next 50 coming from? I'd love to hear what you've learned about business formation and small business versus reducing frictions for existing medium or large sized businesses and letting them get way bigger than they would, but for that outstanding friction. So even that like Chicago example sounds like a pain in the butt, even for a real established business. Maybe the faster way is to make it easier for big companies versus make more small companies. So what have you learned about big versus small and what matters in this goal to grow the GDP of the internet?

26:39In terms of where Stripe is going, a lot of where we started with was make it easier for a startup to do X. the lease tax example and just generally the billing system, a lot of what the Stripe billing team has been spending time on of late is companies like Atlassian and Cloudflare and all these very large public companies who are kind of moving their billing infrastructure over to Stripe billing so that, as you say, they can not just move the payment, moving of money, but all the business logic can live in Stripe so they don't have to build and maintain it themselves because the maintenance crushes people.

27:10We're continually surprised when we build things that we conceive of as being for the low end, how frequently they get enthusiastically adopted at the high end. So for example, payment links is, we want to lower the barrier to entry to Stripe, making a couple of sales, why do you need to build a payment form? Why can't you just sign up for Stripe and put in some details about your product and then just send a URL to your friend or your first customers or whatever. And then maybe when you reach your hundredth sale or your thousandth sale, okay, I'll do the work, build a proper thing here or whatever.

27:39Payment links is pretty enthusiastically adopted by large companies. for basically the same reason, in that it's like the Soylent Green principle, everything is made of people, and there are humans inside these large organizations, and those humans, even large organizations, have opportunity costs attached to their time, and if they can go and build some incremental new feature that aligns with the core purpose of their business, that's, in general, probably a more valuable thing to do than go and build this payment form. Whenever we've built no-code functionality, or just generally speaking, things that lower barriers to entry, where the original persona that we have or had in mind was somebody kind of starting out.

28:15We keep being struck by, we had a user at our weekly All Hands meeting last week. They're a pretty significant company. I don't know the exact employee count is, but I believe in the hundreds. His ask to us, we asked, how can we improve Stripe? He asked for more no-code functionality because they're strapped for engineering resources and they just want to be able to take things off the shelf and snap them to grid. So that's been eye-opening for me. It is also a dynamic where the very largest companies that we work with are often competing with startups and seeing new upstarts come along. And so the buying criteria are no longer as different as you might think, because they know that their customers might just go to work with a number of traditional grocery stores.

28:59And they're all competing with the Instacarts and the ships before they were acquired, all of these small upstarts doing delivery and stuff like that. So it's like, wow, the digital experience really matters. We got to figure out our app strategy. We got to figure out our delivery strategy, things like that. But that dynamic, I think, ends up merging the considerations maybe more than they otherwise would. The big part of the ethos of this whole Singleton idea, what got me interested in was the ability to study. Like you said, I love the Magnus Carlsen example. It's so perfect. Like he wins the trivia for jazz.

29:28This is great. When you think about the deep lore of companies and companies that you've studied, I'm thinking Dutch East India Company, that kind of thing. What very old companies have you been the most interested in, learned the most from, been the most curious about, and why? I think they're almost all interesting. The question is almost, is there something like your present circumstances that gives you reason to care and to find it memorable? I read a fairly lengthy history of DuPont, the chemical company, not that long ago, and it's super interesting. And as you pull on that thread and the history of the chemical industry and how it kind of came from the dye industry in Eastern France and Western Germany, and then the various changes that went through and how it gave rise to the pharma sector, and the whole thing is just super interesting.

30:14The chemical industry was one of the early technology industries in the sense that there weren't that many businesses with such uncertain R &D, where you didn't know what the properties of the chemicals you would discover would be. And so maybe it had some of the characteristics of some of the questions and problems faced by modern technology companies. And so perhaps it's interesting to see some of those parallels. It's interesting to see that, well, we republished this book by Mike Malone a few years ago, The Big Score in Early History of the Semiconductor Industry. I think it may be surprising and striking in that book is maybe your naive model would be that somebody discovers something, one of these companies, and they're like, wow, that's a really important thing.

30:52I'm going to go start a company around it myself so they can internalize the gain or whatever. Whereas what actually happens with every case, as far as I can tell, is the person discovers the thing, they're like, wow, this is awesome. We should totally go do this. We, the company they're at. And they try to convince their management that there's a real idea, we should take it seriously, whatever. And management refuses and is oblivious to it or doesn't think it'll work or is important or whatever. And then after two years of trying to convince everyone around them that this is important, eventually in frustration and out of a desire to just have the thing exist in the world, they go and do it themselves.

31:22But it's not awesome. I get to start a company. It's Christ. Okay, fine. And if I have to go do this myself, I will. But it's almost out of this sense of obligation to the thing, to the discovery. It was interesting to see in the early chemical industry, the same dynamic. And so this just seems to be not a thing about sort of Silicon Valley. Exactly. But this deeper thing about human nature, you extract things like that. But I wouldn't say that DuPont is one of Stripe's top influences. Tell you a funny story. Mom, when we were growing up, ran a company, her own business, called SQT. This still exists and still does.

31:55quality training, a kind of quality assurance training, ISO 9000, things like that. And in particular, there are a lot of, still are a lot of American multinational firms with operations in Ireland that they would have had a significant demand for doing quality training for their employees. But there's definitely a lot to learn from the, and it's like a business cliche, but Six Sigma, lean manufacturing, process optimization sphere, it's a lot to learn, But there are very important limitations. And I think it's not an original insight by us that a lot of people comment on the fact that so much of modern management, culture and theory stems from basically 1950s command and control manufacturing organizations.

32:40and we have aped their ways of working. And in particular, you're used to working with stuff that can be measured, which sounds kind of obvious, but I have inventory. It's at various stages of the production process. And for example, I can think about a balance sheet. The thing works or doesn't in a fairly binary way at the end of the process. Exactly. We can determine, was this part manufactured correctly or not? We're working with all this extremely legible space. And what we find so interesting about software engineering as a domain is no one can do it that well. You know, there's buildings going up around us.

33:14If you were to ask the guys building that office tower, is there a much better way to build a building? They're like, no, what are you talking about? You pour the foundation, then you do this, then you do that. We figured out how to build buildings. Then if you go interview someone at Google or Microsoft, or I'm not singling out any large company, you ask them what percentage productivity do they think they're running at compared to the total theoretical productivity, I think they generally tell you what, 20 % or something like that. And part of it comes from the lack of legibility in what you're working with.

33:42I sometimes joke here that whenever we talk about resourcing and how we should resource very initiatives at Stripe, people often want to talk in terms of heads. This is a thing that needs five heads or 10 heads for a year or something like that. And I always want to get them to denominations in kilograms. We need 500 kilograms or a thousand kilograms of engineers. And you're like, but that's absurd. That's the point. So is five heads or 10 heads. This thing needs a technical solution that could require one month or could require one year if you screw it up and do it badly. But again, we lack the legibility.

34:13We lack the meta understanding of software engineering as a domain to be able to have good ways of reasoning about that. And so I think the fact that everyone in the industry finds software engineering so challenging to measure, to do well, is part of the fun of it. I mean, the places where we do do it well. And part of the opportunity for stride. Exactly. Yeah, I was going to say. If the convertibility or elasticity between dollars and good software was straightforward, then the opportunity for startups generally, and Stripe in particular, would presumably be way more limited than it is, in that lots of companies can afford to spend way more on building good financial infrastructure than we can.

34:47This is the famous Buffett with a billion dollars, 80s with a billion dollars test. He talks about the test of a good industry is one where, imagine some rich guy gave his son a billion dollars to go blow on building a company in this industry. There are lots of places like real estate or whatever, where you can actually have a pretty good swing at it and they will make life harder for you. Whereas I think there's lots of people who have tried to put a billion dollars competing in this or other software industries. And it's really hard to just turn a billion dollars into high quality software.

35:17Just ask the US government. To that point, I think we've probably learned as much from because software, I don't want to say uniquely because I don't know enough about other sectors, But because, let's say, at least software to this intersection of creative work and mechanical industrial work. What's your analogy? Half building a bridge, half writing a bestseller or something. Yeah, yeah. There's a different version that you could use. We've learned a lot from science and learned a lot from creative domains generally. I think studying Hollywood from this lens is very interesting. I don't think there are directly transferable ideas, I think.

35:52What's an example about Hollywood that would be interesting? In Hollywood, I don't pretend to be an expert in Hollywood. You also can't just spend$500 million to make a good movie. Oh, gotcha. You really need the auteur. The whole thing is pointless unless someone has a vision. So often at Stripe, we're discussing something and maybe, yes, in theory, we should do something in this space. But the real question is, is there a specific person who has a vision here? And there's a hundred things that we would love to go fund tomorrow if there was a specific person with a vision. Of course, as Hollywood shows, and as is intuitively obvious, the vision's not enough.

36:28You have some degree of execution ability and whatever skills are required in the domain and so on. There's an illegible trait of, do they have the vision? I guess, is the vision a good one? Where if absent, the whole thing is just somehow going to be a failure. There's something about the unquantifiability of taste and judgment and how that makes... Generally, I think that's tough for organizations to deal with because you can't build robust processes around that. But at least in Hollywood... But that corporate desire to build robust processes about things is an implicit belief system. And then like in science, a fact that really stuck with me is Gertie and Carl Corey ran a lab at the University of Washington in the couple of decades after World War II.

37:09Six of their students went on to win Nobel Prizes. Wow. University of Washington is a good university, but you can't tell a story as to it wasn't the single best university. So they didn't just skim the cream of the very best students. It's not just about, I think, the characteristics of the students coming in. There aren't that many Nobel Prizes won. Nobel is here. I think we're all won in either medicine or chemistry. That's two prizes a year. And so I think it's this incredibly vivid demonstration of how there's something that they knew or that was somehow embodied at their lab that was transmitted to these people that was incredibly important that made them collectively, the students in their careers, vastly more successful on average than the rest of that population.

37:50And again, I think this is something that is hard to quantify. It's hard to even put your finger on what it is, but it's this really rich demonstration of the potency of tacit knowledge. I think for complicated intellectual work in general, that tacit knowledge that has to be transmitted through mentorship is the rule, not the exception. And a lot of our intuitions, like for example, our desire to enable tenure at Stripe is substantially informed by this, where if we're in a superficial domain where, I don't know, you can come in basically knowing it already. And then after two years, you're not that much better than the person who started last week.

38:22That would be one thing. And I think might prescribe one set of personnel and HR practices, but we're in a domain where I think the returns to experience and the returns to knowledge are really extensive. Having six years of context in our domain gives you a tremendous advantage over somebody who just arrived. I think we're informed by some of these other places and maybe just on the Hollywood point, if I don't know if you've ever looked at this book, not knowing much about Hollywood. I found it very interesting. There's an oral history of CAA, the agency, that I found to somehow be a really good introduction to the whole thing.

38:54I also love the Hoban's idea of packaging. Just that as a technology. Packaging doesn't happen in business all that often, and I always wondered why not. There's a lot in Hollywood that I think is super interesting. One other random one to talk about is it's not published in the book, but you can get a samus.pdf if you read around your front. It's Perspectives on McKinsey, which is McKinsey. I've heard of this one, but it's one I haven't read. I'm sure it's in the dark web somewhere. If you have enough Bitcoin at your disposal. The professional services, just so I was reminded as you discussed tenure, professional services firms really take training seriously.

39:26Law firms, accountancy firms, consulting firms, and they take it much more seriously than I think most Silicon Valley companies. I think that's something we definitely looked at is how do you have people who are experienced with Stripe do a good job of spinning up people who are, but there's a case team leader at a consulting firm, And they are really responsible for both the work, which is done by the more junior people, and for those people's development. And they have to do so much of it because a lot of people graduate out that they get really in the practice of it. But I think there's something really interesting to study there of companies that, by virtue of their business model and by virtue of just how they work, have to take training up new people incredibly seriously and just spend way more time on it than your traditional tech company where people have a four-week boot camp and then off you go.

40:08As I was thinking about Stripe on the way up here, what is it about the business that's different or unique? I won't pick on any one other developer-facing tooling company or something, but there's plenty of other companies that built products originally designed for developers that didn't achieve this status of almost Stripe as a state of being. It has this ambient quality about it that you can just see it in the way people talk about Stripe as a company, which of course has a lot to do with both of you, but also it's achieved this thing as an entity separate from the both of you. And I'm curious why you think that is and how much of that is a lucky random side effect?

40:44How much of that was really deliberate based on decisions you made or the kind of company that you wanted to build? Because whether you wanted to or not, it has come to embody a certain era of technology business in a way that if you looked at other companies that also build similar-ish products, let's say for developers in the early days, did not achieve that same thing. And I'm curious to what you would attribute that difference. I think part of it is, We're a self-serve product for a fairly ubiquitous need in a domain where it's actually hard to be self-serve. If the audience is creators and founders and business owners, they tend to be kind of exacting in their judgments.

41:21They kind of have to because they are with respect to their own thing. And so I think a lot of the other developer-oriented companies take Databricks, unequivocally an awesome company. But Tommy, our youngest brother, mentioned that he's seeing a physiotherapist. The physiotherapist mentioned they were using Stripe. I presume that physiotherapist is not using Databricks. And so I think some of it is probably just the nature. We all share if they were, though. Yes, yes. Yeah, really want to meet this person if they are. The fact that we work with Amazon and Ford and Hertz and these giant companies, Databricks, of course, also serve the enterprise, but that we also can be at the low end is part of it.

41:56People at Stripe really care. Catchphrase at Stripe of really, really, really caring. I think the returns to care, to sweating details and so on are also unquantifiable. And part of the problem with them is it's brand thing. And so you can, in any narrow individual case, you can coast on not caring. If Apple sent you a poorly designed invitation to something, you would still go because the overall halo of credibility that they've accrued over multiple decades will carry them through even a single poorly executed thing. But that means that the local incentives to care are actually not that strong.

42:34This collective action is public goods problem within the company where you can benefit from everybody else's work. And so I think the aggregate long-term returns of Caringenot are super high, but the individual local returns are low. And that creates this collective action problem. Most companies think fail to solve it. And I don't particularly credit us. I think this is deeply embodied by a very large fraction of the first hundred people at Stripe, but it became this thing with a tremendous amount of momentum and endurance. And people at Stripe today are still maniacal, although that padding issues and spacing issues and consistency issues where you use this word over here, but that word over there, it adds up.

43:14Is there a detail you're most proud of, of this type in Stripe's history? Well, the one that people often cite is we show you your API keys in sample code. And a lot of errors in Stripe will tell you what we thought you were trying to do, not just what the actual mistake is. And it's funny, even though some of those were implemented 10 years ago, We'll still see tweets occasionally where we saved somebody a few hours because not only do we tell them their thing didn't work, we told them what we think they were most likely trying to do instead. And to be controversial here, I think it's very harmful to bring consumer product thinking to a space like Stripe or I think B2B space generally, because I think there is implicit or explicit.

44:00You just have users by the wheelbarrow. You treat them in aggregates. You talk about mouths. Everything is statistically tested. On the active users for those who don't see my technology. But again, people in the Valley would say mouths as their way of referring to the people who are using the product. But you are just, I'm not a number, I'm a free man. You're a number. Everything is statistical and everything is linked in with dark patterns. It's like, oh, complete your profile. Everything is nudging you towards some business outcome. And what we really try to constantly clobber people over the head with at Stripe is these are people starting a business who are generally really talented.

44:36and they have some great idea and they have a vision for it and they're making it happen. And they have a pretty sophisticated understanding of their space and business and finance and everything like that. And you can just see it if you go talk to any of them. We have another value that we try to really inculcate is talking up to the user. When you're writing product marketing copy, and Patrick and I will spend a lot of time reviewing that, you're not trying to bamboozle them with some highfalutin statements. You're trying to explain to a busy business owner who, again, go back to Charlie Munger's multiple mental models is constantly switching between product and business and finance, all these things, why this thing will solve some specific need that they have.

45:14But it's not hucksterism. It's not sloganeering. It's having a technical discussion with someone who really knows their shit. Patrick mentioned this weekly all hands. We every week have user at this fireside, we call it, where we'll just for 10 or 15 minutes, again, they're busy. We'll interview them about their business, how they started it, how they got there, how they're using Stripe, what they'd like to see do better, you know, let's do better, everything like that. I think that's important to stay connected to this because, again, I think the classic technology product management domain is a bit, I would say, it's not respectful enough of who the users actually are and how capable they are.

45:53I think that's maybe too normative. Maybe it works fine for those businesses, but not our business. Exactly. The thing that we have to take seriously is any signup could, in principle, be in not that many years in the future, 5 % of our revenue. And so allowing... Because they got really big, not because we got really small. Allowing any of them to fall off the wagon and just shrugging our shoulders. If the distribution of possible sizes of our users was normal, I think you could shrug your shoulders. And if 1 % of them, something goes wrong, then, okay, that's unfortunate. We'd like to be less, but it's fine.

46:24But our tails are really heavy.

46:30It's something with a lot of ketosis. And worse, the concern you have to have is if something is going wrong for these customers, maybe it's actually biased and correlated so that it's the ones who are more successful, who are more likely to encounter the issues. And so therefore, I think your intuition has to be, to John's point, kind of exactly the opposite of the consumer domain, where when you hear about a single case of something going wrong, your ears really prick up and you get interested because, huh, maybe it's the companies that are selling a ton in this new market. What we had in the early days of Stripe was businesses implementing these multi-sided ecosystems were consistently running into compliance problems and payout problems and mechanical problems, whatever.

47:07And an attitude of, well, not that many of our customers are doing that, shrug, would have meant that wouldn't have discovered this incredible opportunity that has been a huge deal for us. We're big into anecdotes here. There's probably the correct amount of data in the world. People have realized that data is important, but sometimes they underestimate anecdotes because it surfaces the questions that you should be asking. Data is good if you have a specific question and you want the answer to it. But I mentioned about our mum dick growing up. Our dad, for a while, ran a hotel, a small hotel on the shores of Loughdurg in Ireland.

47:38And what I find funny is you do not need to run a complex user research study to know if you're doing a good or bad job in a hotel. You can just walk around and go, oh my God, this waiter is terrible. These people have been waiting half an hour for their food. You just walk around and the questions you should be asking, slap you in the face. Sometimes literally. Exactly. Depending on how bad things have gotten. Faulty towers. But in the tech case, again, people do a lot of the data analyses. People talk about management by walking around, but they talk about it in an internal management context.

48:10You got to do the management by walking around with the customers and in the actual market. How much more likely do you think it was that you became entrepreneurs, given that both your parents were? Is that something we can simulate? What makes more businesses? My dad was an entrepreneur. I'm an entrepreneur. Your parents both were. You both are. Maybe there's stats on this. We can just answer the question. But for you personally, how much of an influence was it that they were both doing that and you watched them? You got to do the twin studies, find some twins separate. Exactly. Yeah. I remember when we were playing in the garden when we were like six and eight or something, we used to often play.

48:42We don't internet at that age. So we had to play in the garden. and we'd play it. I don't know if we were starting a company, but I guess running a company, I guess just because that's what our parents, our dad was a fireman, we'd have been firemen. It's hard for me to say no, it had no effect. When I think back to those anecdotes, I never, I think, were first engineer in childhood. And even if it is somehow attributable to our parents, I don't know, maybe it's genetic, maybe we lack the risk-averse gene or something, but we've underdeveloped prefrontal cortices or something. But yeah, it's probably not really coincidental.

49:10There's also probably an element, you probably remember seeing this with your dad, But you know, the way starting a business goes in and out of fashion. And famously, after the social network, everyone decided they want to go start a business. But at least I think a lot of what we saw growing up was, honestly, maybe even just the work ethic. There's other things in the customer centricity and things like that. But if you just take a single thing away, it's absolute just hours on the problem that you have to put on it. It's maybe useful to see firsthand. And I think that probably subconsciously seeps in.

49:37And again, I think you probably saw it with your dad. And I think maybe if you haven't seen that front, maybe you're operating with a disadvantage if you haven't just seen that firsthand to really ingrain it with you. We work quite hard on Stripe, but I think we work less hard on Stripe than our dad did at the hotel or our mom did at this training company. And just we are not at the office for most of the weekend. He was at his office necessarily. And so even just that, we get to take weekends. I've seen you talk a lot about your relationship with your parents and their businesses and everything growing up.

50:08What's it like as adults? How has it shifted in interesting or surprising ways as you become adults and getting married and the whole deal? What has shifted most and what's it like now? Well, it's interesting to watch now with a bit more adult perspective to see where maybe our childhood perceptions weren't correct. And we used to make fun of our mom growing up for working a lot and in an admiring way. Now she retired nominally. Unsuccessfully. Exactly. She's had multiple failed attempts to retire. She tried to retire a few years ago and now somehow has found herself. I think she helped a hospital.

50:43Our youngest brother has a, today, relatively mild physical disability. So she became very interested in its treatment and how to do that well. And very long story short, ended up writing a book about successful treatment for cerebral palsy. And through that, learned that, well, there are these other conditions that don't have good treatment manuals for them. And so she's working with a Floyd, in fact, by a hospital in Minnesota, where they're collaborating to produce a whole series of books on the treatment of different physical conditions. And so I think five of those books will come out for the next 12 months.

51:13And this is her retirement. Definitely not in Minnesota. For us to watch as adults. You asked Emma about old organizations that are coming to learn to learn from. Was it the Royal Society who had the mission of reducing the gap between best practice and standard practice? Good question. I don't remember. Part of what Mom is doing is also working at Redenstamstaf philanthropic in Ireland, improving cerebral palsy care. And again, it's amazing the degree to which different treatments happen on a geographical basis for no real good reason. The condition doesn't know where you are. And so they have some of the leading treatments in the world for cerebral palsy are in Australia, where they've developed these interventions that work really well.

51:52And so they're now working on rolling them out. I mean, it's not just these. They're also doing some research and stuff like that in Ireland. But I get very excited about the idea of how can we advance the state of the art, but even just bringing things up to the state of the art in a lot of domains really has, I guess, the entirety of the lean manufacturing rollout in the United States in the 70s and 80s with just bringing things up to the state of the art. As you think about the future of Stripe, you guys are so lucky you have so much time to do whatever the answer to my next question is. What do you want it to be that it's not yet?

52:23And or what do you want it to be more of, like most specifically in the decades to come? An important thing about Stripe today is that we're always trying to remind people who join, because Stripe is, say, 12 years post-launch. so they might not appreciate this necessarily, that just measured against the goals we set out to solve when we started, there's still not only a lot to do, but in certain ways, we're even further from solving them than we were at the outset, where in the very beginning of Stripe in, say, 2011, accepting online payments or moving money online basically meant facilitating pretty basic, pretty straightforward credit card transactions.

53:00But now, over the last couple of years, there's been this incredible proliferation of payment methods, usually run by central banks, but all the different wallets and national payment schemes and so forth that have added. The good news is they've brought immense numbers, like billions of new people online and enfranchised them to participate in the internet economy. But the flip side is they've made the landscape so much more complicated. And we're now in this funny position where even very sophisticated companies are often inadvertently restricting their customer bases to be only a relatively small percentage of what they could be if they're sort of comporting themselves properly.

53:36So just assessed against kind of the most basic criteria of, is Stripe making it easy for a business to accept revenue from, let's say, at least 90 % of internet users? The standard for that has gotten an awful lot higher. And while Stripe is now way more powerful than it ever was in the past, it's interesting that just the complexities of the world are also advancing considerably. And that's before you get into all the other stuff that's happening around different tax things and regulatory things and data privacy things. Now you need to have an Android app and an iPhone app, and the complexity of the whole landscape is considerable.

54:07So we want Stripe to be a global programmable money movement engine for all of these different use cases and possibilities available to businesses everywhere in the world and capable of executing transactions on behalf of people anywhere in the world. And as we just think about the n-dimensional matrix of functionality entailed in that, we still have a long way to go. So, and then the second part of your question was, what do you want Stripe to be? What I'm getting at is, say, you're retired, hopefully better than your mom, and you're in your rocking chair, and you want Stripe to be described as X, Y, and Z.

54:44You'll be able to feel like you can rest easy, you've done a good job, because Stripe is blank or has done blank. Like that kind of looking backward. Have you read Tim Wu's book, The Master Search? I really like that book, and there's lots of random... I learned lots about the history of radio that I didn't know. And so it's always fun in that way. But when you think about it, as you reflect on it, you should be scared because basically the takeaway from the book is every other information network that has been started has started out with lots of innovation and the healthy ecosystem. Low barriers to entry.

55:17Exactly. Low barriers to entry and has ended up as a monoculture regulation captured oligopoly with two or three big players that get to extract rents. And that has happened with every other information network to date. But we'll see if it happens in the internet. That's kind of an unsettling note that Timu leaves you on. The internet is different. And the internet is, I think one could argue, a bigger deal than all those other technologies for lots of reasons. But the warning signs are there. And you see this a lot where a lot of internet commerce stops at national borders, which It should not be like the packets don't travel any slower as they cross them.

55:59The gravity equation in trade says that trade falls off with the square of distance. And that maybe makes sense if you're shipping beef. It goes off. It's complex to ship across countries. You have to refrigerate it, everything like that. It makes less sense when you're shipping software. And yet the gravity equation still exists in digital trade. And so, again, I think part of our hope would be there is a real plausible world where we end up with four or five big internet companies that are the internet companies that have a really outsized share of the online activity happening there. And it's not that it's illegal to start a company.

56:37It's not that it's impossible to start a company. But the barriers are so high to break through that no one is preventing you from starting at telco, but nobody does because the barriers and the costs are so high. that would clearly be a really bad outcome and should make us sad. And so I think if we could feel like we contributed to the next 70 years of vibrancy in the internet economy, which of course just becoming the global economy generally, where you're continuing to make it accessible for new entrants to come in and compete on a level playing field with all the existing folks, I'd be pretty happy with that.

57:09I love that. I think it's Peter Thiel's notion that diagram of the investing perspective, the best things being seems like a bad idea and is a good idea. The intersection is where you want to live. And I'm curious, I think last time we talked in this format, you said something about the most impressive thing about the big internet companies is that they've ridden multiple waves. They weren't one trick ponies. They successfully evolved. I wonder how you square those two things. The waves maybe seem obvious. Maybe one might've been crypto. Now it might be AI, might be something else, material science or biotechnology or something.

57:42I don't know. How do you think about that zero billion Jensen Huang idea with the Peter Thiel idea up against AI is obviously the thing we should focus on as you think about Stripe's future? I think the nature of the problem for a Facebook or for a Snapchat or a Pinterest or something is actually quite different because they're serving a consumer sensibility. It's ultimately a kind of entertainment product. There's utility and messaging and so forth, but some significant portion of use is entertainment. And there's a lot of, in principle, fungibility across different pursuits there. And on some level, watching Instagram stories is competing with Netflix.

58:23Maybe it's even competing with ChatGPT. I don't know. Because consumer taste by its nature evolves, I think that necessarily puts one into this mode of wave writing. Whereas Stripe operates at a lower level of the pace layering diagram, You have fashion and infrastructure and culture and these different sedimentary layers evolving and changing on different timescales. And at the infrastructure layer that Stripe operates on, things change more slowly. That's bad news in the sense that it's not possible to have these insane meteoric appreciations in the overwhelming majority of cases. But the flip side is, if you get things right, you can compound for a very long time.

59:06And I've always found it striking that the internet really never grew in any given year at an insane rate. It's just been this incredibly durable 30 % a year compounding for several decades. 1.3 to the 40 just ends up being a large number. So I think Stripe has more of those dynamics. And you can see this if you look at businesses like Visa or MasterCard or something where they're obviously conceived in a pre-internet world. And yet the wave of the internet, it obviously changed their business in various ways, but it didn't fundamentally restructure it. And with everything on the horizon, it's still the case there are going to be businesses and transactions and exchanges and associated challenges and complexities.

59:49that money can be exchanged for goods and services. They come with that, exactly. So I think Stripe, because of its position in the ecosystem, needs to strike the right balance between spotting the opportunity in these waves. And we're excited about the fact that basically all of the AI companies are implementing revenue models at the outset of necessity because it's expensive to run inference and training and everything. And basically all of those companies are doing it on Stripe. That's hard to find an exception. And not only are they using Stripe for the most basic payment, but they're using all the higher level functionality on top of that because they have sophisticated needs and they are sophisticated companies.

1:00:21Let's jump back for a second to just your previous question as sort of what we want Stripe to be. And we've talked a lot, I guess, over the course of this conversation about prosperity and that growth and why we care about that and how we think Stripe can be implicated and all that's obviously important and true. But the thing we haven't said or touched on as much that just in your invocation of the rocking chair and sepia tones and so on, looking back, something about beauty and craft, if Stripe is a monstrously successful business, but what we make isn't beautiful, and Stripe doesn't embody a culture of incredibly exacting craftsmanship, I'll be much less happy.

1:00:56I think the returns to both of those things in the world are really high. I think even beyond the pecuniary or financial returns, I think that just the world is uglier than it needs to be. It's a free lunch where one can just do things well or poorly, and beauty is not a rivalrous good. We can do this in architecture, but my intuition is that more of Stripe's success than one would think is downstream of the fact that people like beautiful things and for rational reasons. Because what does a beautiful thing tell you? Well, it tells you the person who made it really cared and you can observe some superficial details, but probably they didn't only care about those and then everything else in a very slapdash way.

1:01:37And so if you care about the infrastructure being holistically good, indexing on the superficial characteristics that you can actually observed is not an irrational thing to do. There's a reason there's a very intricate filigree on the$1 bill. Yeah, right. And then I think the kinds of, like there's something about the talent selection, the people who like good APIs and good architecture, like good software architecture, tend to surprise things being done nicely and well in other domains and so on. So I think there's this bundle of - It's not just a nice warm and fuzzy thing. It's very rational.

1:02:08Yeah, I think there's a bundle of self-reinforcing loops such that it is a sensible thing to do. I think people behave quite differently in spaces that are beautiful. Banks had it figured out 100 years ago where bank buildings should be nice. We're just continuing that tradition. They lost their way. If someone really convinced me that the financial returns to craftsmanship and beauty, which I believe are greater, are not greater. You'd still do it. I'd still do it. Life's too short. If you think back on Stripe's history and had to identify the most important, impactful, memorable conversation between the two of you, what would it be?

1:02:42individual conversation, I mean. I remember a conversation we had, I remember even the restaurant we had at in the summer of 2010 when we'd been working on Stripe for seven months. And the question was... Right there. No, no, no. This was whether you were going to drop out or not. Oh, yeah, yeah, yeah. Or I guess we were going to drop out, but John was the one more on the fence. And it was this relatively somewhat dispassionate conversation as just how big a deal could Stripe be. And we just gotten excited about it in the beginning. It's fun and cool to solve this problem. But then we're talking almost for the first time about like, well, where could it really go?

1:03:19If this is not just a little random side project, what could it become? And that was, I think, the first time we'd ever sort of really taken a step back and talked about in much more inchoate and larval form, some of these same ideas that we're discussing here. I really remember that one. The flip answer is a conversation back in 2009 when we were walking home from dinner and just decided to start working on Stripe. And as I said, you know, how hard can it be? Famous last words. I think a more serious answer might be in that conversation on Valencia Street, the wine bar, where you were morose and you were just down maybe two or three years into Stripe.

1:03:55Is this even working? Is this even worthwhile? Not growing fast enough, things like that. I think that actually maybe gets to one thing we're very lucky to have is a good team dynamic, where I think it's really, it's just like a lot for anyone. I don't know how you do it. It's a lot for any one person to have on their shoulders. I think two people will naturally have various cycles that they go. And I, of course, just like, what are you talking about? It's great. You know, it's going to be fabulous. But I think just because he was maybe in a more pessimistic humor at that point, and I was in an optimistic humor.

1:04:24And if you're just naturally will have highs and lows, you'll tend to cancel each other out and act as dampeners for each other, or even just act as someone to talk through the issues with and stuff like that. I feel very fortunate that I think not that many people have real partnerships that they can work on something worthwhile on on the multi-decade horizon. And again, maybe if more of those partnerships were formed, we get more interesting multi-decades opportunities. Because again, I think we think with Stripe, we spent the first 10 years getting the table steak stuff out of the way. And now the next 20 years, you can do a bunch of the really interesting stuff where you have the platform and the ability and the fighting fit army to actually go execute on it.

1:05:06But a lot of people just don't get to that point. The same vein, setting aside maybe the very earliest years, is there a defining moment in the business's history that comes to mind? Very continuous business. Um, everything we do starts very small and then just compounds for a long time. And Atlas is now a double digit percentage of all Delaware incorporations in the U S and that's like including multinationals incorporating subsidiaries. That's all Delaware incorporations. Exactly. So of the true startups, I mean, I don't know what the exact number is, but it's a lot. We launched Atlas and I think 2014 or 15.

1:05:43And so it's just 80 years of compounding to now be really big. But everything we do has that shape. There's a product that I'm really excited about this year that I'll leave nameless for now. It's doing super well. And I think it's going to be a huge deal for our users in a bunch of different ways over the next couple of years. We've been working on version of this product since literally 2014. I think after a decade, it's going to be, say, next year, if that's its 10th anniversary. It's going to be phenomenal. But it just took iterating for that long. We're not a business of discontinuities. You think about the way you spend your time today.

1:06:17and you could go show your day to the 25-year-old version of Vue, in what ways would that 25-year-old version be most surprised about how you spend your time now? For you, presumably meeting with customers. Yeah, there's a lot of customer time, which I find very energizing. And I did a lot of it back then as well. But I think maybe I would have been surprised by how continuous that. It's also like who they are. Yeah, that's one real big surprise. And again, it converts over into how we spend time and who we spend time with. We've been really surprised the degree to which basically large companies have ended up needing Stripe.

1:06:55And so in our Series A pitch deck, we were explaining why you should care about a business that sells to developers and startups. And so we had a slide in there where we were like, obviously, you know, Amazon will never use Stripe. But there's a lot of developers out there. There's a lot of startups. And so this can still be a big deal. And of course, we've now been working with Amazon for five years, and it's a super productive partnership. I would do all sorts of things with them. And I spent a bunch of time with our counterparties over there and lots of other similar large companies in a similar way.

1:07:22And that's maybe part of, again, how the product thing works, where I think Patrick probably spends more time internally with the technology teams, where in the abstract, how should things work? What do we want a well-designed global money movement system to be? And I'll probably spend more time with customers there saying, I really need you to solve X for me. And those two can meet in the middle when it comes to product development. They're getting like feature requests from Jaguar, Land Rover, or from Hertz, or from Heathrow Airport. Versus Instacart. Yeah, exactly. Or Ford, you name it. It's the complexion of the customer base.

1:07:58A lot faster than that. Yeah. It surprised me where, as John says, these companies want to be and have realized the importance. Mark Andreessen wrote The Software is Eating the World piece in 2009, and The Wall Street Journal published it. I think it's very interesting that The Wall Street Journal published it in the sense that in what way is that a contrarian point of view? And it's hard to put ourselves back in the mindset of 2009. The idea that software might be important was, in fact, this combative sentiment to express in The Wall Street Journal opinion pages. When Stripe started out, the idea that we were focused on the needs of startups was certainly a detractor in the eyes of established enterprises, whereas now it's inverted so completely where every company of any size and significance and of the businesses that have retained any degree of agility, which is most of them, they all have plans for how to take advantage of the possibilities of the internet and the possibilities of ML and the possibilities of these global customer bases and so on.

1:08:57And they know that, well, if they don't execute it, some crop of startups is going to come along and sort of do it in their sector for them. And so now when we go and talk to the CIO of some giant retailer or some large manufacturing company or something, the fact that we're working with actually the world's most innovative, at least new companies, is a very strong kind of positive in their eyes. This funny inversion where the same trait went from being a detractor to really being a selling point. And they want to hear what are the upstarts doing so that they can go and roll it out themselves.

1:09:32And honestly, I've been impressed where TakeHertz, as an example, we were discussing them yesterday, they're doing a really big deployment with Stripe. They're really excellent to work with. We're doing a bunch of stuff with United Healthcare and Optum and... Big companies. Yeah, yeah. And I think a lot of these companies, to a greater extent than people appreciate, have actually hired some really good talent. Maybe 20 years ago, they didn't have people who really get software. Now, in a lot of cases, they do. And a lot of our engagements with those businesses are just really productive. One of the coolest things in my life recently is that I have kids of a certain age now where the kids are starting to do things that make me very proud for the first time.

1:10:07They're developing agency and they're doing things where it blows your hair back. Like, oh my God, I can't believe they did this. It's made me really love asking people this question about pride, not in themselves, but in other people. I would love to hear from each of you, what about the other makes them most proud? You're talking to Irish people. Exactly. Introspection and vulnerability. It's coming from Irish people. It gives me a license to ask the question. Well, look, I mean, this is very relevant for this discussion. John has been serving as our interim CFO for most of this year. And I think it has always been true that John is a scholar of business.

1:10:47And when we were living together in an apartment in San Francisco, I'd be shuffling to bed and he'd be yelling out to me from the bedroom. You'll never guess what they said on his earnings transcript or whatever. John, I'm brushing my teeth. There's a way for both of us where I'm normally engaging in the business as co-founders with one layer of indirection. And so we're very lucky to have Stefan Tomlinson joining as CFO in a couple of weeks as of the time of this recording. He'll in fact have started, but it's been very fun for me watch John actually temporarily inhabit the role of CFO. And I think it's been significantly to Stripe's benefit.

1:11:19And so we're both extremely excited about Stefan. In the same way that I'm trying to find a productive equilibrium with our product leaders where we need to empower them. And a lot of those people are really fantastic and it would be a mistake of me to be getting in their way, but want to find some way to have some end of productive meddling. I think similarly for John. I've been encouraging him to make sure that we retain sufficient productive meddling on the finance side because he's very good at it. When we had all that discussion about quality and craft and everything, that's all Patrick. And in particular, he has a remarkable eye for it and cares about it.

1:12:00It kind of has to be good because something will be wrong in the universe. He'll be itching. He'll be sitting awkwardly in his chair if the stuff isn't good. And we'll put in the hours. We'll be there very late the night before our conference, ensuring that the talks are in good form or whatever. But I think how that bleeds over is people really want to do great work because of the expectations that they know Patrick has of them and that now they have of themselves. and I think again as we talk about culture there's a really nice cultural thing at Stripe there's tons of really talented people and again people who are meticulous in their craft at Stripe but again I think Patrick has set an expectation and a culture there where people maybe start out wanting to do excellent crafts because it is expected by him of them and they end up wanting to do it themselves I was describing a capability of Johns there's lots of those but maybe another view of Stripe would be that every organization, you're always trying to balance these dichotomies.

1:13:06I think lots of things are dichotomies and trying to find a reasonable intermediate point, but maybe an important one and somehow one that we straddle is I'm grim and austere and joyless complaining about the misplaced pixels and all the rest. And John, everyone who knows him, is cheerful and extroverted and sociable and is bringing 100 of his closest striped friends to the bar next door on Tuesday because it just opened and we need to inaugurate it. I think for an organization of humans that do great work, you actually do need both where you're not going to achieve super high expectations absent the standards, but it also needs to be a group of people where you want to spend time in it.

1:13:49Look, I don't think either of us actually are sort of on either end of the continuum here. And we're probably even less different on this continuum as individuals. Similarity highlights the contrast or something. Exactly. Where we play off each other or something as a duo. But I think there's an interesting way organizations need both of those. The more Irish appropriate question might have been, what does the other do that makes your eyes roll the most easily? Oh, well, that's the same answer. Need to add another hour to the podcast for that. I think it's been incredibly fun. I know we're coming up on time.

1:14:19I had the pleasure of doing this with John before, and I have a traditional closing question, so I get to direct it at you this time since he's already answered it. What is the kindest thing that anyone's ever done for you? I should have known this was coming, but I didn't prepare an answer. Everyone else has prepped for this, so you get to think about it. Apart from my parents, my principal at our secondary school, which is the high school plus plus that I attended in Ireland, which was a public school and a very good public school, at a public school that actually wasn't the closest public school, but our relentlessly resourceful mom thought it was better than the most approximate one and managed to find a way to...

1:15:00Basically, we had lived in its catchment area when the school was founded or something. She... Utility bill from 1991. Exactly. So she found some way to persuade them to take us. Anyway, the principal consistently let me violate all sorts of policies and rules and turned a blind eye to many things over the years in such a way that I had a fabulous time in my five years there. And at first it was I just wanted to read my books during class. And I thought I was being very deviously and effectively furtive, having nobody noticed. No one knows. Exactly. In the small little classroom, there's no way this teacher could spot this.

1:15:38But I learned subsequently that he had just quietly instructed all the teachers to not complain about my ignoring, I guess, the classwork and just reading my book. I did the homework or whatever, took the test. I wasn't paying much attention in class. And then at this stage, I wanted to take some time out of school, mostly to spend it on working on this programming language. And he found a way to categorize this so that it worked in theory. He's still in school. Yeah, so actually, I did a lot of it physically in the school in part so he could argue, well, he's at school. He didn't ask me if he's going to the classes.

1:16:10Then at the end, I wanted to go to college in the US and not do the standard examination that everyone in Ireland does and so on. And again, he kept finding ways to make it work in theory. Very tragically, he died a few years after I left school, suddenly at a premature and early age. And so I did thank him, but I always felt regretful. I never felt that I had sufficient opportunity to thank him. It's amazing how much this category of answer comes up. The two major ones are this one of someone noticed something about me and did something customized and shaped me in a certain way or bet on me when they didn't need to.

1:16:44In some ways, the same thing. It just comes up over and over again. I hope more and more people do that for others. I think you're doing a huge service by highlighting those cases to a bunch of people who are in the position to do it. I don't know if this was your intent in asking this question or if this is now a tool-aside benefit. I don't know where the question came from. So I do this as a hobby, still as my favorite hobby. Effectively, I'm focused on investing, but this is something I love to do. And I didn't know how to do it. I'm not a journalist or anything. And so I was really into the mythology of Joseph Campbell's hero's journey or whatever.

1:17:17And it's all this same story for all these great historical myths. And so I thought I would ask a question in the podcast that covered each of the nine stages of the hero's journey. And one of them, let's stage three or four, is random helpers come and assist the hero on their quest before they get to the belly of the beast. And so the question was just, what's the kindest thing? But I had these nine questions and the rest of them sucked. No one Never cared. But that one always got an awesome answer. And so that's the only one that lasted. So that's where it came from. I wish I had higher intentions.

1:17:46Oh, that's really funny. Well, I think you're now, yeah, creating more kind moments. Yeah, yeah, yeah. Well, it's a testament to how interesting just the core thing that you have worked on and will continue to work on is that we didn't even get to in two hours. AI and all these other amazing things happening in the world of technology. But I'm really glad that we focused on the core thing because that's what got you here. And that's what you're focused on for the future. I'm deeply appreciative of the time and all the insight and all the detail that's gone into it. So thanks so much for your time.

1:18:14Thank you. This is really fun. If you enjoyed this episode, check out joincolossus.com. There you'll find every episode of this podcast complete with transcripts, show notes, and resources to keep learning. You can also sign up for our newsletter, Colossus Weekly, where we condense episodes to the big ideas, quotations, and more, as well as share the best content we find on the internet every week.

1:18:50Outro Music