Stripe's Patrick Collison | Micro Pessimism, Macro Optimism

Hi everyone. Welcome to Grey Matter, the podcast from Greylock, where we share stories from company builders and business leaders. I'm Heather Mack, head of editorial at Greylock. Today, Greylock General partner Reid Hoffman interviews Stripe CEO and co founder Patrick Collison. They discuss Stripe's early days and its evolution into one of the largest financial services startups in existence. They also talk about the leadership lessons Collison has learned along the way and and how the company's micro pessimism viewpoint helps it stay focused throughout changing economic conditions. This interview was recorded in October 2022 as part of Gray Matter's I Conversation speaker series. You can watch a video on our YouTube channel, and you can read a transcript of this discussion in the content section of our website greylock.com blog. Both are linked in the show notes and if you aren't already a subscriber to Gray Matter, you can sign up wherever you get your podcasts.

So Patrick and I are beginning to make a habit of this. We were on stage last week at the Masters of Scale Summit, and we thought we would open this up. For those of you who didn't see last week's Master of Scale Summit, Patrick, who is a master of Scale, identified himself as a hamster of scale.

For context, you're like, why is there a hamster here? So I do think there's this thing that happens where, I don't know, I guess we conflate the size of the company with having figured things out. And I don't know, Laszlo Bach from Google, who ran HR and people there for a long time. And I'm sure he's a wonderful guy. Again, I don't know him, but I was very struck by the book that he wrote about Google culture, which I haven't read and maybe a wonderful book, but.

But you were struck by physically, right?

Well, it's just like, I mean, it's a book. I think the conceit of the book is like, how to have a culture as good as Google. Right? My kind of outside view on that is like, step one, to have a culture as good as Google is like, number one, have the business model that's the best business model in the world. And as good as Google, it's like, hi. It really becomes much easier to have a good culture when you have the world's best business model. In a sort of similar vein, I think Stripe has been fortunate to be hauled and thrust along by all these structural forces, but I do not feel I'm much more comfortable being denoted a hamster of scale than Anything beyond that.

It's one of the key things to always think about what is being an infinite learner. And, and so let's go to the deep part of where it is to be a entrepreneur kind of right now, which is this is wars, pandemics, markets. Oh my. After a 10 plus bull year market, all of a sudden we're in turbulent times. Because the turbulent times is not just questions around, okay, what does capital fundraise look like, but what does your customer flow look like? What are you seeing in Stripe, and what's the advice you would give entrepreneurs in these choppy waters?

Well, I suppose the simple thing today, just from a Stripe vantage point, is I think rates have changed substantially. And as a consequence, I guess investor behavior has changed a lot. And based on what we see in consumer spending and even business spending, that has not yet noticeably changed. Now maybe there were kind of, there are modeling effects where we thought that the pandemic pull forward was actually like some kind of structural new normal. And it turns out that actually, no, it is better modeled as a pull forward. But all the headlines you read about Europe, you're wondering, is there still an economy there? But not only is it still standing European, So fortunately so. But not only is it still standing, but still growing. And so just as a very basic foundational ladder, the economy seems to be in relatively good shape. We just turned a phrase at Stripe. We want to be micro pessimists and macro optimists. And over the years we're all trying to find ways to sort of inculcate that. And for example, when we were moving offices, I told John that I wanted to get kind of signs from other companies that we could put up around the office, like National Semiconductor or Atari or Osborne Computer or Sun Microsystems or SGI and so on.

This might be an old enough group that they actually recognize those names.

Well, just like, you know, Silicon Valley is famous for being such a productive and prodigious cradle for new technology technology companies. Whereas in reality, it's the world's most densely populated graveyard. And I think there's something so kind of gloriously macabre about the fact that the Facebook headquarters is not the Facebook headquarters, it's the sun headquarters. Right? And the Googleplex is not the Googleplex, it's the SGI plex. And we're like crabs just scuttling along and sneaking into each other's shells. But the millennials and the Gen Z's obviously don't know that. And so we're all trying to find these structural Things that we could do. John didn't let me do the sign thing. He told me that I fundamentally misunderstood how to motivate humans. But we're trying to do things like this. And so on the one hand, I think for many of the companies that are trading in public markets, and perhaps that's some folks here, I mean, I can't speak to that. Stripe is not public. I think it's super hard and I don't want to sound oblivious to that, but when I look on a 5 to 10 year time horizon, it seems to me that the really great companies that the Valley built over the last 40 years went through these cycles like Andy Grove's famous literature was written after and about these melding periods. And so again, I don't want to sound like some wizened expert here. Stripe is going through this right now. But if I disassociate myself from Stripe and just kind of think as an Observer Sitting in 2032, I suspect that on average this is very good for the Silicon Valley ecosystem.

I actually agree because it's actually partially the refocusing the question about how to navigate both difficult times and as well as the bull times is action. And if you can actually turn it to a strength, you get much more competitive differentiation for doing it. So let's dig a little bit more on.

You're forced to prioritize more, right? Yes, look, even I'm being forced to prioritize more and it's terrible in capital abundant times in, I guess, theater, it's. You want to kind of a. Yes. And culture. Well, you know, when capital is abundant, you know, your product developments can also be a. Yes. And culture. And maybe to some extent that's good because you get to kind of have some expansive ambition. But you know, even because our customers are on average presumably going to be investing, on average, investing less over the next two years than, you know, they were before. Stripe is some kind of overall averaging of that we're thinking harder about. Well, how do we make sure that we're appropriately calibrating what we're doing relative to them and we're making harder product trade offs than we had to three years ago.

Well, and I actually think that's embedded in what you're saying is you want to keep the same level ambition, but it has to be through a focused aperture rather than okay, yes, we can do that too. And we can do that too. And we can do that too. Everything from Microsoft all the way down to the Series A has to get more focus.

I think that's true. And I think there's something about, you know. So started Stripe with my brother John. And John reads earnings transcript in bed. And we used to be roommates and then we got girlfriends. And it would have been weird and all that, but for the first. And now a wife. Yes. For the first eight years of Stripe, we lived together. I'd be shuffling along to brush my teeth or whatever, and he'd be yelling at like, you'll never believe what they said in this 10q. I as a teenager was really into practice programming languages and Lisp and all this kind of crazy stuff, but at least aspired to being some kind of technologist. John was always super interested in business. I was early in Stripe's history, I pitched John on, well, we should just make Stripe free, because we'd grow much faster. John's like, interesting. Can you explain the business model embodied in that to me? And didn't really have one, but from a technologist standpoint of purity, I thought that was much more compelling. I think there is some kind of yin and yang where I suspect most people in this room self identify more as a technologist than a business person. And I think that's fine and good. And Silicon Valley is ultimately an ecosystem of technologists. Like, we're not New York. But I've certainly found over the years, and I've experienced it kind of myself, that I think it's very easy to become excessively technologist. And I think part of what. What's the difference between Steve Jobs 1 and Steve Jobs 2? I mean, you can kind of read that many ways, but I think Steve Jobs 2 was much more a business person than sort of the first iteration. Maybe part of it's the need to focus and all the rest, but I think part of what lots of companies, and again, I include Stripe in this is kind of the shift that we're having to go through is to not just acknowledge that kind of reluctantly. Okay, fine, we're businesses. I guess it's like to embrace the fact that there were businesses and think about things like hurdle rates and irrs and all the rest as part of the fundamental raison d'. Etre.

Yes. Although maybe less irrs and more. The question of similar to how we do compounding loops and growth curves and other things is also the compounding loops in economics. It's how the model comes together.

Yes. And actually there's a book, some of you probably have read it, it's kind of famous in some small little sort of factions, the Outsiders. Okay, I'm seeing Some nods and smiles.

Yeah.

And basically you realize you can nerd out about data pipelines and, you know, the optimal way to do iterative data computation thing you can nerd out about. Like, the optimal way to be thinking about return profiles and the outsiders is I think that the best single distillation of this.

Yeah. So. And speaking of nerding out, we'll come back to why Lisp was the language that you selected in the things, but.

It'S the best one.

Okay. Speaking of, we'll put the nerd out on the side. Micro pessimism, macro optimism. How does that cash out some? Like, what are some of the ways that you operate culturally? Make decisions, plan for both ambition and downside? What's some of the way that that cashes out?

The tension I was trying to navigate is I think there was just like a fundamental contradiction in the startup mindset and this cognitive dissonance where on the one hand, you have to be incredibly attentive to your problems. Right. And to recognize the myriad ways in which your product totally sucks. Right. You know, if you're oblivious to that, if you're kind of cheerfully whistling along kind of Lego Movie style, you know, everything is awesome. I haven't actually seen the Lego Movie, but. But I heard that reference once. But if you're, you know, just like the book. Exactly, exactly. Yes. Yeah. If you're oblivious to this stuff, obviously reality is going to kind of give you a rude slap in the face pretty quickly. I think the challenge and the tension is if we're kind of rubbing our own and kind of each other's faces in this, it's easy to become kind of to have that drag in your spirits. Right. And you simultaneously need to have this attitude that, okay, everything is terrible today, but we really have conviction that we can make it fantastic in two years time, in four years time, in six years time. And so this kind of micropessimism, macro optimism, part of what we're trying to do is to give people kind of acknowledge this tension and to give people structural permission to be extremely critical of that which exists today. Because I think otherwise, people feel a bit tentative. Like, I don't know, are they criticizing my baby? Can they say that our dashboard sucks or this API sucks or whatever the case might be?

So shifting a little bit or focus out some. One point. You said the Internet is like the world's most vibrant city and the equilibrium has yet to be reached. Where are we on the evolution of the Internet? Are we still very early days? Are there places where we need to.

Be making corrections at our event last week, Masters of Scale, Hamsters of Scale. And Reid asked, we're going to rename.

It Hamsters of Scale. You heard it here.

Reid asked about technologies I'm excited about. Obviously there's blockchain and drones and AI and IoT and all the things. And I said one of the things I'm really optimistic about and that we see burgeoning around us is vertical SaaS.

This is the right audience for that comment.

I'm seeing at least one not there. Great. So every sector of the economy, not every single, but essentially every sector of the economy is still fundamentally analog, bespoke, handcrafted, and not in the nice kind of attractive handcrafted sense, but just in the horrendously efficient sense. You know, all the tools that we benefit from in kind of how we, how we do business, I think 97% of the rest of the economy has yet to benefit from. You know, we did this video with this company back in Ireland last week. They're called Herd Watch and it's literally SaaS software for farmers to manage their herds. Certainly today, you know, before Herd Watch and the likes, you know, that the cows are, you know, they're not being tracked in Google sheets and they're, you know, the cows are not on slack. But there's tremendous efficiency gains and, you know, not just in the logistics of the farm, but like how you do market making and matching and price discovery and, you know, like actually quite fundamental things. Now there's Farm Logs, which is another company some of you might be familiar with. They do sort of analysis of, you know, crops and optimal distribution and, you know, pesticide use and try to lower that and everything else which is both healthy and cheaper. It turns out if you get into this, it's like a whole burgeoning ag tech sector and so many different ways which software can be brought to bear. But again, we're at a very early point in this distribution. Carlotta Perez, Sigmoid I don't quite know how to quantify it, but if I had to just offer a point estimate off the top of my head and again, based on what we see from our customers and all the different platforms are going out trying to serve this, I think we're still in the single digits because it's all different sectors in the U.S. but then of course, it's not just the U.S. or Western Europe. You have to take stock of the full global extent of that on a truly global basis. I think we're barely off the starting blocks. The Fact that Jerome Powell correctly presumably is running the Paul Volcker playbook, I think should not in any way diminish our optimism about the 5 to 50 year prospects of the technologies being worked on.

No, I think that embodies the micro pessimist, macro optimist point from before.

But just like on that, I feel like maybe I was too cavalier earlier. Like I do not speak with any wisdom. So just consider every answer. In fact caveat with that. But especially in this one like stripe started in 2010 and exactly has benefited from the longest bull market in post war U.S. economic history. Right. And so you know, we Stripe have not lived through hard times, but certainly from all the people I've spoken with, all the reading I've done and just trying to get some sense for what might loom in our futures. We once hired someone from Amazon who was describing how after he started at Amazon, the stock plummeted by 70%. God, that must have been very painful. It's like actually no, I could deal with that. What was very painful is when it then fell another 80%. That's when I started to feel bad. And that was Amazon. Right. And obviously these moments in these companies histories don't tend to garner as much attention when the kind of ex post facto Bradstone books and so forth are written and so much and all as I'm extremely optimistic. I think that these you've lived through them. I have not. But my sense is these cycles can be pretty intense.

They're super intense. And one of the things that is important to recognize is that it's hitting everybody. And so your differential ability to switch swim through it, Whether it's Amazon, PayPal, other things, is actually in fact the thing that gives you strength. I'm going to ask one more question and then open it up to the audience just to give some breadth. So you know, Patrick is comfortable expositing the hamster view on a wide variety of subjects, which is good. So progress studies, why did you do the Atlantic article? What's the key thing? We just had a dinner on this earlier this week. What's the lens of why this is important? Because this is a room of people that will agree with progress studies but just may not be familiar with it.

I suspect everyone in this room is united in the belief that progress, and by progress I mean just like the overall broad society wide betterment that we've all been beneficiaries of over the last at this point several centuries, that it's possible and that it's important and that it's Contingent, right. I mean, if you thought that we were just on some neat escalator ride to success, I mean, probably there'd be no need to start a company so someone else would do it. And so I think the kind of intrinsic, irreducible necessity of human action agency in Ireland, where I grew up, people are very culturally aware of this because as recently as 1960, Ireland was a destitute, autarkic, insular Catholic theocracy. My parents remember getting electricity and grew up in subsistence farms and all the things like the big literature about Ireland being the sick man of Europe and what's going on and so forth. And Ireland, basically, Ireland had bad ideas. We thought that being kind of closed off and walled off from other economies was a good idea. We thought that excluding women from the workforce was a good idea. We thought that having the Catholic Church run our country was a good idea. There were lots of things. And then a couple of people made the case successfully that things should be changed. And Ireland had this explosive growth over the course of essentially my childhood, where it was just kind of starting around maybe 1985, and then it kind of reached the peak, I guess, post the financial crisis or, excuse me, pre around 2008. So I grew up in the midst of this. But as a result of that, the kind of sociology in Ireland was everyone really knew that it sucked to not have the right kind of development ideas and that things can be so much better when you get the circumstances configured the right way. And then I came to the US I guess the US has been a beneficiary and kind of rich and developed for so long that there seems to be more of some kind of shrug, right? And maybe even the opposite, where you worry about too much development or too much growth or what have you. And we're not at the end of history yet. We have not developed all the ideas we need. We have not cured most human disease. We have not solved climate change. We have mental health. There's lots of problems. And so trying to find some common hashtag that could generalize beyond just say we should have more startups. What's the unifying sentiment there that's more expensive? And so we just wrote an article, we were trying to propose a label.

For that, and I recommend it. I myself sometimes say, yes, I'm a progressive. I believe in progress other than vertical SaaS. Which of the technologies do you think is most underrated in the way that it's going to have an impact that entrepreneurs in the room should be thinking about? Definitely AI and say a Little bit more.

I've actually been an AI pessimist up to now. What's the difference in AI and ML is a bit of the sort of the kind of continuous redefinition where ML is the stuff that is actually useful, AI is the stuff that's not useful. And so it's kind of tautologically true that AI is not useful. So acknowledging that. But in as much as you think there's kind of something more to this whole thing than good classifiers or whatever, I think that has not really been borne out to date. The aggregate ARR of the AI sector, I mean, of all the GPT3 and other kind of the Dall E and other image generation apps and midjourney and so stable diffusion and all these things, and summed to all the revenue of the GPT3 and other LLM users, Jasper and copy AI and all those companies. I don't know what the actual aggregate revenue is, but I'm pretty sure it's sub a billion dollars. And maybe it's. I don't know, but maybe it's sub 250 million. So just like the market is saying, it hasn't really mattered thus far. Right. I think that there are lots of things that don't matter and are relatively economically inconsequential until suddenly they're not. It seems to me that we're reaching that tipping point with all of LLMs and various of the sort of image generation, video generation, sound generation, speech generation, speech transcription, all this stuff together. I mean, there was the paper in 2017, attention is all you need from a bunch of folks at Google bringing into the world the Transformer model. And it kind of seems that they might have been right and that maybe Transformers are all you need. And the fruits of scaling them up seem to be super high. And So I think 2023 will be actually the year of AI. And I say that as someone who has been on the substantially skeptical side to date.

I totally agree. But before I ask the next question, I want to see if there's another question here. Okay, great.

Okay. So you have a very unique situation of founding a company with your brother. I'm curious. I think there's a lot of founders in this room. Probably many of us have co founders. I do. Do you think it was easier or harder? And then also you've gone through so many phases and your company's 12 years old. What have you learned in terms of working with other senior leadership and founders that make that kind of a different special relationship and things that that have helped make you successful as a company, as a co founder group. Okay, I'll try to give a very attenuated answer because I don't know, you could probably bore someone for hours and talk about all the co founder dynamics. But first, building a company is hard. And it's obviously hard in the case where things don't go well. And the thing they don't tell you on TV is it's super hard. Or they didn't tell me is that it's super hard. Even in the cases where it does go well. Greg Lamond, who I think was a Californian, but the Greg Lamond quote, it doesn't get easier. You just go faster. And one builds companies at the kind of like you hold constant the rate at which you can tolerate problems and everything else is kind of the variable that moves. And so you're just constantly in this zone of kind of maximum pain. And if you're below that threshold, then you take on more stuff and you can get yourself right to that peak. And so given that, I've certainly found it extremely helpful and stabilizing to have someone else who's in exactly the same boat as you, who's been with you through the whole thing, has full context but a different perspective. Otherwise it ends up. I have tremendous admiration for the people who are doing this solo. And look, there are some phenomenal entrepreneurs who are doing it solo. I mean, market Facebook is a perfect example. And evidently empirically that can work well. But it seems to me that it'd be quite a bit harder. And so I guess my kind of learning from that is I think where they exist, those relationships are precious and worth really kind of investing in and fostering. Early on, I think both of us were much more insecure. Not to some crazy extent, but we very much conflated kind of stripe and our leadership and our own personhood. And those things were all wrapped up together and I think kind of naturally so because I mean, it was a thing we had just made, right? And so they weren't even. They didn't have kind of particularly separate identities. But I think something that's been very healthy and that I think, if I could speak to six year ago me that I try to emphasize more is you are not your leadership style. I now think of myself as, I don't know, like when I go into work, I put on a mech suit. Not in terms of, you know, offensive capabilities and smiting people and not literally right, but just in terms of if someone said that the mech suit sucks, I Would not feel personally attacked. It's just like, oh, I guess, you know, that little laser blaster is kind of shitty, you know. You know, I had a conversation with one of our leaders last night and he gave me long list of things that he thought I should do better and I agreed with all of them and I fully kind of intellectually acknowledged that like such a long list does exist. And you know, I could add more things to it, but I think six year ago me would find that a much more uncomfortable experience. Whereas now my mindset is much more. He's giving me like excellent upgrade ideas for the mech suit. And so I think, I don't know, I'd be trying to yell all that to a six year ago me.

Well, that's an excellent bridge to a question that we are going to be closing many of our sessions with. And so I'm going to ask you that question, which is if you were to be yelling to your six year ago self, whether it's five, seven, whatever, what would you most tell yourself about like what knowledge would you impart? Not with the knowledge of the future, but what would you tell yourself to do differently? More of this, less of that.

I mean there are lots of specific things, but actually I think the meta one is from which most of the others follow is you are not your leadership style.

But when you say you were not your leadership style, does that mean the fact is you always be learning and seeking to make yourself better and not internalize it, but treat it as an external thing that you're a skill set that you're improving or correct? Yes. Okay.

Yeah, yeah, yeah. It's like, you know, you're not. Sorry, analogy. Mech suit is probably a bad one. Much too pugnacious. Yeah, yeah. Trying to think of a good example of something that people do create but it's not obviously sort of part of their identity. And you know, maybe you could say their code or something, but people do get very possessive with their code. So I don't know what the right analogy is. But yes, fundamentally that idea that it's a. Well, both, it's a separate set of skills, but actually it's also a choice. And I think we kind of fall into these patterns and again I include myself where you find a thing that kind of works and like now that's you. Right. But actually, I mean it doesn't have to be you. And I've been very lucky where there's other leaders at stripe who have a maturity to this and a hard won wisdom that I've learned a lot from. And I've even witnessed their evolution and changes. And that's been, you know, there's an implicit gauntlet there where it's like, man, if they can grow and develop that much, like, shit, what am I doing? You know, so. But anyway, all of that kind of bundled together.

Yeah, no, look, I think it's excellent advice and I actually think one of the things you embody is it's not just your mech suit. It's how are you making improvements in your team's mech suits together? Like, how are you playing? Like, how are you playing the sports together? How do you get your role increased partnering well with them, et cetera.

Okay. Actually, all this philosophical mumbo jumbo, not that useful. So my philosophical mumbo jumbo, I mean, so maybe a more tactical, useful prescription is actually when I meet companies that have product market fit, and I think when a company gets to that stage, like pre product market fit, you should be like scrappy and iterative and just focusing on the product exploration in the space. Once you've product market fit, I think all the kind of lean startup intuitions don't carry over and you become super focused on building a leadership team. And just like the question is always, who's your leadership team and how well is it working? And that's a complete inversion from before. So maybe actually six years ago, me, I would say focus more on your leadership team.

Well, as you can see why I always love talking with Patrick and every single event I do, I call him and ask him, hey, will you come on stage with me? So if those of you who are last last week masters of scale, you saw that There Will Be Future, please give him a hand.

That concludes this episode of Gray Matter. You can find the video of this interview on our YouTube channel, and you can read a transcript of this discussion in the content section of our website greylock.com blog. Both are linked in the show notes. And if you aren't already a subscriber to Gray Matter, you can sign up wherever you get your podcasts. I'm Heather Mack. Thanks for listening.