How to Win by Being Right and Contrarian: Lessons from Zappos to DoorDash

31:52And it's not just about one or two. Yes, McDonald's is one of the largest merchants, but winning a whole suite of them is actually more valuable. Google, the long tail keywords, more valuable than the head. And most of advertising was broadcast advertising. You start from the top. And this is a bunch of long-tail keywords. And I think a lot of great companies, whether it's intentional or not, try to do things very differently than what exists today. Apple, they weren't the first to many things. They didn't make the first personal computer. They just made the user. Their focus was always about the user experience, whether it was the computer, the phone, the watch.

32:48Everything they do is about having a great user experience. That was not the conventional thing that people thought about. It was about, in the early days of the PC, it was about processing speed and which processor was the fastest, et cetera, et cetera. When the phone came out, it was about the keyboard. We don't use the keyboard anymore. The Apple thing is fascinating. So is the BlackBerry thing, because they really hung on to that keyboard thing. But I remember when they came out with their new, I forget what the device was called. It was like their Apple killer. And I used it as a demo and I couldn't figure out how to get on the internet.

33:21And I was like, they're done. So user experience is so important. It took a demo to like figure out how to use the internet. Once you figured it out, it was great. But I was like, oh man, this is so not intuitive. Yeah. Let's go back to hiring and firing a little bit. Was there any moments sort of at Zappos or earlier with LinkExchange where you thought you had the wrong fit, but it turned around?

33:46That's a great way of asking that question. I would say probably not. In the early days of LinkExchange, we hired, and back to like conventional, and, you know, we hired the conventional right person. We were looking for a marketing person or looking for a marketing person that marketed to small and medium-sized businesses. They understood the internet. They fit all the specs that you would have in a job description. But they didn't fit the way we operated. They didn't fit what we valued. and it's easy to sort of limit that down to like they're just a professional marketer and they're mercenary not a missionary that's how tony would tell the story but at the end of the day they didn't value what we were trying to build they didn't have the same values that we had and which is why when zappos started tony spent a lot of time defining the values this is Tony Hsieh of Link Exchange and Zappos the values of the company and it's something that I talk to founders about if you don't have the same mission you don't have the same values you don't have the same operating understanding of the company it's very very hard to work together and then you're asking someone to change when they fundamentally don't fit to the company and how the company works.

35:16That is generally the issue. Then the other issue on the flip side, on the functional side, when your company is growing really, really fast and you see someone who is growing, but they're not growing as fast as the company, every single day, the delta between how far the company is moving or how fast the company is moving versus how fast or how far the person is moving, the delta gets bigger and bigger and bigger. So it gets very hard to turn around. And so we have this concept that a lot of people have at Sequoia. We hire for slope, not intercept. So experience obviously matters, but if you just hire for the experience and their slope is not fast, that's going to become a problem at some point.

36:08And so we really, really ask founders to think about what is the person's potential? What is their slope? And we'd much rather bet on someone with high slope and low intercept because in startups, in companies that are new, the experience is helpful. but you're doing something different and new and you need to reinvent. Bringing a playbook over from somewhere else is helpful up to a degree but usually many of the problems that you need to solve are just different and that's why you hear a lot of founders talk about first principles thinking and don't reason by analogy, etc. because you do have to solve the problem a different way.

37:01what's the difference between slope and potential it's just well slope is just the is is pretty much the same thing slope and potential but having a high high rate of learning high rate of you know sort of being able to move fast and so if someone has a lot of potential they're over here their potential is up here that slope when you draw that line is pretty high how do you gauge somebody's potential it's like you might be able to do this job you might be able to grow into it but like coming into it it's like they're not going to check the boxes they're going to look different the future does not look like the past but you have nothing else to go on but the past and you can see whether someone moved really really quickly in the previous situation where they promoted three times in the same year at a different company that was moving very very quickly do they have a sense of urgency you can test for some of these things you can ask about these things but yeah you don't know whether they have the potential to grow into the next stage of the company until you put them in that position.

38:02Often I think the challenge is the company is moving so fast and you're hiring someone that has done the job before and you know that they can do the job for the next year or two. And that's maybe good enough. If you hire someone, you have to let them go in three months. That was a messire. There's almost no reason why that person was not going to be able to do the job for three months. They either were not a cultural fit, they didn't have the right skills. If they stick around for a year or two, it's usually because they've done the job before and they're picking up a bunch of low-hanging fruit.

38:35But I look at Tony at DoorDash. Many of the people who work for him have worked for him for a long period of time. He likes to grow people from within. He obviously hires from outside as well. But many of the people who've been around the company have been around for a long time. Amazon, many of the senior VPs, that were around Bezos had been around the company for a long, long time. Now that Jassy's were, you know, they're now a CEO, he's trying to develop people also for a long period of time. If you look at great companies, they're very, very good at developing talent both from within and hiring from outside.

39:16Is it a red flag if the turnover is too high then? Is it like a sign that people don't know how to hire, right? I think it's a red flag and you look at it and you don't want to... So the issue with metrics is people start managing to the metrics. And so if you start telling them, hey, your turnover is really high, they start managing to the turnover number. And that's another thing. You have to manage the turnover of regrettable. What is a regrettable turnover? Why are you not able to retain your best people? and if you have unregrettable turnover then why did you hire these people in the first place and so you if you break the problem down that way you can't find our reads on what the issue is with the turnover i really like thinking about it in terms of regrettable versus unregrettable turnover um a lot of people when they're hiring they sort of we need this job i want to know somebody who's done it at the next level because that's where we're growing to so they're trying to anticipate where they're growing.

40:20And one of the problems that I hear commonly is that people run into, well, that person might have done the 100 million to 300 million growth, but they had a different team than we have. They have a different system, different resources. How do you think about that when it comes to hiring? My observation is that when you try to hire someone that is from a company that is one chapter or two chapters ahead, they tend to work out better. And part of that is because they're only one or two chapters ahead of you. If you try to hire someone who you're over here and you're trying to hire someone that's 10 chapters ahead, that usually is a problem because you don't know whether they understand the problem a few chapters before.

40:59It's been too far away. On your particular point about the system, I do think that if you hire people who are 10 chapters ahead and they have a whole system in place, even though they took it from 100 to 300 in a product division inside of a large, large company, they probably had a lot of guardrails that allowed them to not make a bunch of mistakes. And so you have to dissect whether that operating system works the same way as your operating system. Most startups don't have an operating system. So the thing that you have to sort of assess is how well can this person operate when it's just more nebulous it's just less clear are they going to put the the operating system in place dive into this story behind zappos from sort of the initial idea in the early days just high level five minute version to the the exit with amazon maybe from the start zappos was this is this weird idea we i can't even believe well we funded it because that this was uh this was originally founded by Nick Swimmer, he had called and left a message on our answer machine.

42:14This is how old and back we were going in 1999. And Tony and I were running a seed fund, a venture fund, an incubator called Venture Frogs. And Nick said, hey, I have this crazy idea. I'm a webmaster at AutoBitel, which is an auto website. I'm the webmaster there. I think it'd be great to just start something for shoes because I went to one store, I couldn't find the right size, went to another store, I couldn't find the right color, went to another store, they didn't even have the style that I'm looking at. And he had already searched the web and found a whole bunch of websites where people were basically trying to sell specific shoes on the internet.

42:58And we thought it was the craziest idea. And we almost, I think one of us had our finger on the delete button to that voicemail. And then he said that you might think this is crazy, but mail order is already 5 % of sales in the U.S. for shoes. I'm like, okay, 5%, still small. But the shoe industry is$40 billion, 5 % is$2 billion. And it just wasn't rocket science to understand that the internet was going to be bigger than mail order. And so we took a meeting, we decided to make an investment, and we funded the company$500 ,000 at a time when we were at VentureFrogs. I had to then leave to go work at a company called Tell Me Networks.

43:49And Tony joined Zappos relatively early on as an advisor and then became co-CEO and eventually CEO. but the most important thing back then was because it's an e-commerce company and we're going from 1999 to 2000 it was growing from 2000 to 2001 i was like oh we have a situation and the day after 9 11 the company had zero on sales zero you went from whatever it was which was small to zero and that was the first crucible moment um well using the term because sequoia uses termed Crucible 1, what we're going to do about that. So we shook the company, and we basically went back to basics and wanted to build a company that was profitable.

44:34And so Tony ran the company at Breakeven and continued to grow the company for a long period of time until Sequoia came in in 2004, 2005 to make the first investment. And so most of the company was financed from a small seed round. Tony invested$10 million of his own money, but most of it was financed through being very, very creative, figuring out how to get merchants that we ordered from to give us credit and to increase the credit line. We eventually got a line of credit from Wells Fargo, but the company burned very little cash. And for a company that went from zero, when we sold the company, it had$1.6 billion in sales.

45:21that really didn't have a lot of equity financing was pretty incredible. You hear about companies raising hundreds of millions of dollars to get to$1.6 billion in GMV or in sales, and here's a company that basically raised$10 million. And it probably was undercapitalized, but it really did highlight that you can build a company by having CAC be profitable on the first order, by being focused on customer service and not focused on marketing. Talk to me a little bit more about that because you didn't have a huge budget for marketing. You had to get, you know, within the first month, you basically had to pay back yourself for marketing.

46:08How do you think about that, not only back then, and then what's different now in a world where there's a lot more money and maybe all the players aren't rational? about how they're spending money to grow. Back then, we probably thought it was a curse that we couldn't raise more money because you probably would rather run the business a little bit looser so we can grow a little faster, etc. But being tight led us to find solutions that was, again, non-obvious or divergent, to use your term. And the divergent idea we had was if we can only spend money that was profitable on the first order, We needed other ways to grow.

46:49And the focus was to become much better at customer service. We noticed that it was much easier to keep a customer and getting that customer to order more than it was to acquire a new customer. And then we just went down the line of what we can do for the customer. We tried to make the website load as quickly as possible. We tried to pick, pack, and ship within four hours. The solution to pick back and ship in four hours was to solve a operations flow issue and not batch the operations, which I don't think anybody else figured out in e-commerce back then besides Amazon. We decided to work very, very closely with UPS in particular, and then FedEx and USPS to figure out how to get the shipping rates as low as possible.

47:43We started shipping so that it was five to seven days ground shipping to eventually overnight shipping so that the customer got it the next day. We figured out logistics, reverse logistics, so we're bringing the shoe to your home. We heard lots of reasons why people didn't want to order shoes on the internet. We knew that before we invested. And we knew that we needed to bring the store to your home, have you try it on, and then return the things that didn't work for you, you did not want anymore. And by doing all those things, we provide a much better customer service. On any given day, the orders that we had, 80 % of it was from repeat customers.

48:33And we grew the business, some with marketing, but the majority by providing great customer service, having high repeat rates, having high LTV, and having customers basically tell everybody else how great that business was, how well they were treated, etc., etc. I read somewhere, and this was counterintuitive when I came across it, so correct me if I'm wrong, that your best customers returned the most shoes. Yeah, that was a very contentious discussion because every year the return rate would go slightly higher and it went slightly higher. But we kept growing. So one way to look at it is, well, it's not really hurting us.

49:17And then the other way is, well, the return rates are high and the most expensive process in the distribution center was managing the returns. You had the return shipping. you had to open up the boxes you had to look at the product and figure out whether it it's ready to go back on the shelf it's damaged i need to fix it or it can no longer be sold again because if you put something that cannot be sold away again on the webs on the website and make it available and that gets ordered it's freaking going to come back and be returned again we then did this analysis um and this is the type of analysis i really enjoy doing like all right let's take our best customers and let's take our worst customers and look at their return rates.

50:03And it turned out our best customers at the highest LTVs returned the most. And like that, that is crazy. How is that possible? Well, it turns out because they were, they understood how to use Zappos. They just ordered more. Yes, they returned more, but they kept more too because they would be willing to try things. If you never return anything, that means the only thing you buy are the things that you're super comfortable with. You know the size, you know the style. So we're trying to help people open up the aperture, try new things. So we did this analysis on the extremes and compared them and realized our best customers have the highest return rate.

50:46So then you say, well, the board doesn't like the fact that we have such expensive return policy and it's costing us a bunch of money so then the solution is not to reduce the return policy and make it less liberal you want to keep the return policy as liberal as you can but make the cost of processing as low as possible and so the first order issue there is exactly going back to first order issue the first order issue is not the return rate the first order issue is making the return process as cheap as and efficient as possible. It's so counterintuitive, right, when you think about it because you're actually like, no, we're okay with people returning shoes as long as they're buying more and then they're trying them on at home, which is exactly what we want and sending back what doesn't work.

51:31Yeah. I think we changed the whole industry and it originally was just a competitive response. We had a liberal return policy in 30 days. Other people copy 30 days. Then it was 90 days. Other people copy 90 days. And then we made it 365 days. And not everybody can copy that. Some did. But by doing that, it would put a stake in the ground that we want people to try the Zappos experience and have a great experience. Why do you think it works for shoes but doesn't seem to get as much traction when it comes to clothing? No, I think it got a lot of traction for clothing. It was an understood problem by the time we got into clothing that you wanted to have high returns.

52:14You're going to have high returns and therefore you need better processing for clothing. I think for clothing, for shoes, and for a variety of soft goods, return rates are high because you need to try it on and try it out. I'd also point out that shoes, the difference in sizing really does matter. I mean, you wear a small or a medium, both could probably fit, and you may look more buff when you wear small. It's loose-fitting when it's a medium or whatever size you want to use. but for shoes the difference between seven and seven and a half or 12 and 12 and a half is big difference and did you guys do things and i don't know i was never a zappos customer so sorry but like did you do things where it's like hey last time you you're ordering a 10.5 or 11 last time you ordered a 10 we're just letting you know as an effort to like reduce returns like maybe you hit the wrong button yeah we we let people know those things we also let them know that this brand You're used to buying this brand, like Asics running shoes, and Nike fits a little looser, so you might want to up the size.

53:26We had this measure of, like, does this fit true to size, or does it fit a little looser or a little tighter? You guys also had, like, an incredibly unique culture. What went into the thinking behind that? The culture of Zappos was very special. And it was something that we wanted to preserve partly because Tony Hsieh had this experience at Link Exchange where we didn't define the values of the company. We started hiring people that were conventionally right for the positions in the job. And one day he woke up and realized the company that he had founded and built was no longer the company he wanted to work at.

54:08And he wanted to make sure that the company that had this special culture, that was Zappos, kept the culture. In the early days, it was not very well defined. What is the culture of Zappos? Well, you know, that was more nebulous in the early days. And then the company grew and people asked, well, what is the culture? We don't have it written down. How do you know someone's a culture fit? Well, does Tony like them? Does Fred like them? Does Alfred like them after the interview? And I don't like the term, the word like. Like could mean so many different things. It was just not well defined. And at some point, Tony decided to just ask the company, what do you think the values of Zappos are?

54:55What are your personal values? And what do you not like about Zappos? And just parsing through that, we got a lot of responses. And in his book, Delivering Happiness, he talks about the fact that we started with 38 different values. Yeah, these are all Zappos values and combined them, shrunk them. And we ended up with 10 core values. The problem with 10 is that there are probably a handful of people that remember all 10. Most people remember the first one or two or three of them. And the first one was to deliver well through service. It was, in some sense, the most important core value. But we meant it in a way not just to deliver well through service to customers, but also to employees, to our partners, our business partners, and to investors.

55:48You had a different management system as well. Talk to me about, was it holistic management? Holacracy. Holacracy. Holacracy, yeah. So the management philosophy was different, partly because we wanted to hire people that could just self-run and self-sustain. And so there was an element of allowing people to do what they love the most and pair them with other people who love other things and make it sort of homegrown. I think Tony's perception of his job was that his idea of creating an operating system for the company was to basically build a greenhouse, was his analogy. And he's going to try to help every single person grow as tall and as strong as possible.

56:45And so he very much focused on allowing people to focus on their strengths and hire other people for their weaknesses to supplement them. Is there an ideal growth rate, do you think? Like, can you grow too fast? The answer depends. I think the easy, if you're growing so fast and the customer experience starts to degrade, you're hurting yourself. And so, yes, you can grow so fast that it's just a complete mess. you're losing customers because you're not servicing them correctly. And at Zappos, we were very mindful that whatever customer experience that we delivered, it was the best customer experience that we can deliver at that growth rate.

57:29And I think if you ask Tony Hsu at DoorDash or Brian Chesky at Airbnb, they would probably say the same thing. It's like, we want to grow as fast as possible with the constraint that we want to make sure that there's a great customer experience. And maintain the internal culture. Yeah, and maintaining the internal culture is one way of thinking about it, and you may have to grow the culture over time.

58:06you have a company, it's growing. Every year you do a strategic plan and a financial plan. You figure out what your next year's revenue is going to be. You're hiring a plan. You're marketing a plan. You have all these plans. You should also have a plan for culture and how to grow that. I think the best companies think about that. How does our company change? How do we grow up? How do we become one year older and better and faster at the same time? how when you're having so much success scaling like going from a million dollars in sales to a billion do you not sort of get complacent i think that because you were dominating the market at that point in time you were the leader clearly in that space you're talking about zappos who yeah well in general in general i think in in for every company the if you read jim collins's book why the mighty fall or it's the first sign is the hubris of much success and it we you know at sequoia we're just never complacent but you only really work not to be complacent we actively work not to be complacent and almost every single company to make sure that you're not complacent successful companies just think that way at Sequoia we talk about you're only as good as your next investment at Sequoia at Zappos we talked a lot about about how easy it is to ruin a good reputation you just the next order if that's not good for a customer you've built all this effort to build a good reputation with a customer and one bad order on an important order, like their wedding day, and you ruin that reputation.

1:00:06And so I think a lot of companies have different ways of thinking about that. And Brian Chesky thinks about what's the next thing? What's the transformation that we're going to create? What is the next thing for the customer experience that's going to be great? Tony Hsu thinks about the next product to power the local economies? What's the next set of people that we're going to help? And I think, you know, if you sort of go back in the history of any business, you notice that customers are just enormously impatient. They're enormously unsatisfied. Everything that is new and novel becomes standard.

1:00:53and base has said this line about how customers are just always always dissatisfied i want to kind of go deeper on that it's almost like there's a natural entropy to create more sediment inside an organization toby luke in our interview used the word sentiment like you start building up the bureaucracy and entropy being like that's the natural result of growth and scale and then you actually have to apply a lot of energy to make sure that it doesn't happen what are the early signs that you see that of sediment yeah well there are a lot of uh examples of this when you you know what the reason i always take a snapshot of when i when i go into a company for the first time to see how fast they're moving.

1:01:44And then I tell them, today is a great day. Hopefully, you will move faster than today. And I hold that as a bar. And if you don't have that mentality, it will start to slow down. And so what's the sentiment? The process becomes, instead of making fast good fast decisions the the thing that people value is having a good process that slows you down to make sure you have the right decision you have to make high velocity decisions at scale um and i look for that like are we able to make high quality fast decisions as quickly as possible? Or do you let the process determine the decision? How do you judge a decision's quality?

1:02:35You can't necessarily at the moment in time. You can think about things at the moment in time as how reasoned, how deep have you thought about the problem? How much research have you done? And you can sort of at least say, well, you researched the problem, you understood it deeply, and you've picked a course and hopefully that's the right decision. And in hindsight, you can always look back and look at whether those decisions are right or wrong. And I don't think enough companies look at that enough. Instead of projecting next year's revenue and next year's strategic plan, part of the strategic plan should be about looking backwards.

1:03:15What decisions did you get right? What decisions did you get wrong? Why do we need to course correct? You guys also had an element of, I think it was even specific back before, became popular, but getting 1 % better. Talk to me a little bit about that and what it meant, practically speaking. There were two things at Zappos that I pushed very hard and popularized. One was the power of and, because I found too many people trying to figure out ways to do either or. And it wasn't like, okay, we're going to ship product more quickly, but the website's going to be slower. Those are, you got to do both.

1:04:01And if you think about the little things that we do at Zappos, it was about just compounding 1 % every single day. And I used to write on the whiteboard 1 plus 1 % to the 365. because if you take a dollar and compound at 1 % every single day, you get this ridiculous result, which is you get$37,$38. And so when I hear about making a leap forward, $1 to$38 is a big leap, and it's on the scale of figuring out what's a 10x idea. It's even bigger than a 10x idea. I hear founders talk all the time about, I only want 10x ideas because those are the things that move the needle. And it turns out just simply compounding 1 % every single day moves the needle even more than a 10x idea.

1:05:00Where do people tend to go astray, right? So you have a great core product, and I'm just imagining you start building up a team, you start getting out of your focus. Like, where do companies go wrong? A number of places. if you just look at great companies, they compound at a very high rate for a long period of time in their core business. Google for a long period of time was search. Amazon for a long period of time was e-commerce. And there's a difference, I think there's a difference between act two and category expansion. So in Amazon, they went from books to music to electronics to a whole bunch of, they just went category by category.

1:05:41but it was the same real like core business which is getting product to the distribution center pick pack and ship the products that you want and shipping it to you um and it could be books it could be music it could be electronics it could be shoes it could be um tools it could be a hammer but but that's one core business and all i think founders start their company because they have novel and compelling insights into the world and specifically for their company. And they like the new shiny penny. I'm not saying that you shouldn't focus on the new shiny penny that is adjacent to your business.

1:06:19I'm just saying that you shouldn't focus on the new shiny penny that is like way out, away from your business. Eventually, all companies need an act too. AWS was an act too. That was very different than the e-commerce business. But even the way that Amazon thought about their web services business. It's like, well, we're very good at building distribution centers. Originally, the distribution center that we built was for physical things. And here, we're building a distribution center for electronic things, for bits. Like, even the way to sort of think about it was quite interesting. They didn't think this was a, this is a high-margin business.

1:07:05They thought of it as like, okay, this will be a low-margin business too, and we like low-margin businesses. And that was Bezos' famous quote, right? Your margin is my opportunity. I like the idea of ACT-2. For a lot of companies, ACT-2 is the founder steps down, professional management sort of comes in. Talk to me the difference between, as you see it, a founder-led company and a professionally-led company. And I use the word professionally loosely because founders are professionals, but I just want people to get the distinction in their head. Well, there's a lot being said about the difference between founder mode and manager mode.

1:07:40And I would go back to the question. The question is, shouldn't it be and? Shouldn't you want, don't you want to be both? I have a concept of fire and ice. The best company is above this huge fire. They have this entrepreneurial spirit and they run hot because the idea is so compelling and interesting. But they also have this icy side. It's just cold facts. It's management. It's about getting to the details. And the best companies have both. And it pains me to see that you have to choose one or the other because the best companies do have, they figure out how to make sure that both are in the company.

1:08:23What are the different strengths, I guess, than maybe that's a better way to think of this question between founder mode and management mode? What are the weaknesses and strengths of each? Like how do they work symbiotically together? I would say that there are many different modes of a founder or a manager. The beginning days, you're a creator. You're finding zero to one. You're creating something that the world has not seen before. And the mode there is to be a creator. And instead of calling it a founder, you're creating something that the world has never seen. once you create that well sometimes it gets into you know sort of operator mode and found many founders are fine operating and making their creation better it's about understanding the inputs and outputs how do i sort of make the system how do i make the system better i create something i need to sort of go to market with it how do i systematically make the go-to-market happen and you're trying to just operate the thing in a way that sort of scales the business and many founders are very very good at the scale then you get to a mode where you need to manage you're trying to figure out resources resource allocation and it's like should i invest a dollar in the core business or should i invest that dollar in a new project and sort of having a framework on how to make management decisions like that that's what a lot of um companies at scale but the ceo needs to figure out and then there's just a level of leadership around how do you lead the organization and think about it from an organizational standpoint so some of that is manager mode and some of that is is is um founder mode but let's just break it down to what it is and not talk about one contrast over the other.

1:10:22Well, I want to go back to something you said earlier about crucible moments. What are they and how do you identify them? Crucible moments is the term we use at Sequoia for things that are basically very important type one decisions. And I've had to deal, as an operator before Sequoia, We didn't call them crucible moments, but I've gone through a fair share of crucible moments. You know, link exchange. Do we want to sell the company to Microsoft? At Tell Me Networks, when we're burning$60 million a quarter, how do we get through that? And how do we pivot the company from a consumer business that wasn't working into an enterprise business?

1:11:14These are decisions that change the trajectory of a company. At Zappos, it was, what do we do after 9-11 when sales went to zero? What do we do during the financial crisis when credit became very, very difficult and we were borrowing$60 million? We had$60 million of debt. We did nothing wrong, but the banks didn't have the liquidity anymore. They're calling the loan. and how do you go through these things? At Airbnb, there have been a number of crucible moments as well, one of which started very early on that Brian Chesky talks about a lot, which was the PR crisis with a guest trashing a host, and that led to host guarantees, but there were many of those.

1:12:02There was a fight to win Europe and a competitor called Windu that copied the Airbnb website pixel by pixel, or Dordash's decision to go after suburbs and not the cities. They didn't forego the cities, but we're going to start with suburbs first. Dordash's decision to focus on merchants and selection over speed. And you can have a situation where you can increase speed of delivery by just having a smaller radius. which then, you know, if I only showed you the restaurants that are close to your home, you'll get them faster, but it doesn't give you the breadth of selection that you may want. And these decisions change the trajectory of the company and their choices that many of the times they're one-way doors.

1:12:56Once you make this decision, you can't go back. And I think they're very, very important to get right. and so at sequoia we have a podcast about it um partly because we want to highlight how to how to founders and the management teams puzzle through these decisions and another characteristic about crucible moments is every one of them don't doesn't look like the other so you you're presented with a new problem and you're trying to figure that out for first principles how to solve that problem. That's the second or third time you've brought up first principles thinking. What does that mean? For me, it means start from a blank sheet of paper and you have a bunch of frameworks that you have that help you sort of triangulate the answer.

1:13:42Throw those out. Maybe you need a new framework for this particular situation. You're the king of mental models. To me, they're frameworks. We love them because it instantly helps us think about the world and narrow down the options. Some situations, you want to generate more options and more ideas. And in those particular situations, you don't want to narrow your thinking. You want to broaden your thinking. If we're on a board together and we get a problem, imagine, and it comes up, and how do you think through that in first principles thinking? Like, how would you do that in that setting? I'll give you an example from Airbnb.

1:14:23You know, when the pandemic hit, you can decide to do a variety of things. And there's all these pressures. There's investors who want you to raise enough capital so that you can go through and weather the storm. There are employees that want to know that they're going to be employed. There are hosts that want to know that even though guests can't travel, that they have a non-refundable cancellation policy, they're going to get the money. You have guests who want to figure out how to get their money back because they're not traveling. You have a business that went from a business that was growing 20%, 30%, 40 % a year to a business that declines and you lose 80 % of your revenue.

1:15:17There's all these pressures. It's a complete mess. and the solution that you had yesterday doesn't work for today because Airbnb before the pandemic was a lot of cross-border travel and today after the pandemic there is no cross-border travel. How do you think about this problem? Where do you start? It's a mess. Everybody is yelling at you from all these different angles and Brian just one of the most important things is to stay calm and look at the problem that we have at hand and figure out what the most important problems to solve first, second, and third. And he had principles that he had outlined.

1:15:57We used to have these emergency board meetings, and it was just always discussing what's the emergency this week, this week, this week. And then he leveled up and decided to say, hey, we need principles to decide what we're going to do. and he decided that we have a once-in-a-generation pandemic, one in a hundred years. That's outside the building. There's nothing that he can do to fix the pandemic. What he can do is to make sure that Airbnb survives for the next generation. And he had principles by which he wanted to make sure that when we get out of this, we're seen as someone who created Airbnb for the next generation and not something that wasn't going to survive this generation.

1:16:51Then he went to work. And he had these great plans of like, we're going to reduce our burn by doing all the things in this order. We're going to cut marketing. Nobody's traveling anyway. Easy decision. We're going to cut contractors. We don't have enough work for that. Easy decision. We're going to have to raise money, and we're going to do it in a way that doesn't burden previous investors. So we're not going to raise money at a low valuation. We're going to raise debt. Why do we need to raise the money? To survive this crazy situation where both the guests and the hosts want their money back.

1:17:33We have$3 billion of capital on the balance sheet at Airbnb, be, but we have$3 billion or$4 billion of customer deposits. We can give all of that to one side and one side will be pissed at us. Give it all to host, the guest will be pissed. If we give it to the guest, the host will be pissed. And we won't have a business when we emerge from the pandemic. So he goes out and braves$4 billion,$2 billion of debt. So two plus our three is five. We can cover both sides and then tell both sides, okay, we can pay both sides. Instead of paying it out, why don't you just calm down? You'll get this money.

1:18:16You know you can get this money. And let's see how the system works itself out. And so he had to sort of project confidence and imagine your way out of this thing. And you imagine a solution, which is, okay, well, we're in New York right now not a lot of people want to stay in their apartments in New York, they're traveling upstate in San Francisco where I was not a lot of people wanted to be in San Francisco they went to Napa and so slowly but surely it didn't happen in March, it didn't happen in April but in May we found resilience in the business model where people started to travel again, just in their own backyards, in their own country and that allowed the the company to come out of that.

1:19:01And the thing that he did last was to lay off employees. He knew that the employees were going to have a harder time finding another job during that time, during the pandemic. And so that was the, you just had these principles on how he's going to go solve this problem. Nobody would have told him to do that in that order. That's an incredible example. Thank you. And then on the flip side, this is the craziest thing. So because, you know, Airbnb was the poster child of the number one IPO possibility for 2020, and they had seen their revenue go down 80%. And then they came out of that and then went on to be the number one IPO of that year in December.

1:19:46And DoorDash, you know, sort of, I had the opposite experience of DoorDash, which was, boy, you know, they were also trying to go public that year. But instead of being beaten down by the pandemic, they really accelerated during the pandemic. And also there, Tony Hsu had a lot of principles behind what he wanted to get accomplished. And it was like anything else during the beginning of the pandemic, it was very, very scary. Because around the world, there are some places that were completely shut down, including restaurants. And if you shut restaurants down because of his experience having worked in a restaurant with his mom, by his mom's side, washing dishes with his mom, he knew that restaurants wouldn't survive.

1:20:34They have less than 30 days of cash. And, you know, that's less than 30 days of cash when they actually are running at full speed. so now they're they have less they have less cash the the restaurant is is uh shut down for a period of time people didn't know whether they can order food and it would be okay as soon as we figured out that the restaurant shouldn't be shut down and the restaurants can remain open for delivery and take out he went to full force and making sure that the business could help merchants bridge the gap. You know, this is not something that other companies couldn't have done.

1:21:16But one of the things I really respect about Tony and the team is they decided like these merchants are going to be hurting and we need to help them. And so they proactively went out and uploaded menus from different merchants onto DoorDash and said, hey, I know you guys are hurting. The only thing you can do is delivery. We're here to help you. We've uploaded the menu, your menu. We found your menu, whether it's on the internet or we've got a copy of your menu, we uploaded it. It's ready to go. Just let us know if we can turn it on for you. And that allowed DoorDash to grow very, very quickly during the pandemic.

1:21:56I thought that DoorDash was a well-run company in 2019. And it was a process by which they grew very fast. They probably grew too quickly. Their unit economics wasn't great. They had a few hard fundraising rounds for the C and D. They fixed the unit economics. So in 2019, it was a walled oil machine. And I saw this walled oil machine push to another level in 2020 during the pandemic when they really did help restaurants when they were in a world of hurt. So the pandemic is a crucible moment for every business. It was a crucible moment for every business. What differentiated the founders in Sequoia from the ones who took advantage of that opportunity and sort of came out of it stronger and the ones who didn't?

1:22:58I think we were, you know, I hate to say this and it's baby tripe, but going through the pandemic, we all came out stronger. It was a shared experience. Sometimes when you go through crucible moments and crises, you don't have the shared experience. But to answer your question more succinctly, I think there are people who knew that the crisis was an opportunity, and then there are those who just thought of it as a crisis. and I think it was Andy Grove that says that good companies get better and they're defined by a crisis whereas there are a lot of companies that are destroyed by a crisis and many great companies were defined and they changed the way they did things in a way that they found solutions that you would not have imagined when the crisis didn't happen.

1:23:59And I think the founders that are, they realize they're in a crucible moment. They realize they need to do something differently. And they stay calm. They stop. They look around and they figure out, they look around 360 degrees and say, that's the direction I'm going. even though the whole world is like revolving around them and it's a complete whirlwind. And they imagine why they should go into that direction. And in most cases, they had to find a different solution, a different product, a different way of operating, and then scale that into that direction. Those are the ones that come out really, really successful.

1:24:45form you know in airbnb's case they had to read they had to change cross-border travel to local travel they changed to long-term stays they changed to having experiences in door dash's case they simply had to sort of do things that they've never done before there wasn't a new pro it wasn't a product change but they had to imagine a different solution they're not going to wait until merchants call them and say, negotiate a deal. They said, look, we've uploaded your menu onto the website, into our app. These are our standard terms. Do you want to go or not? That wasn't a product that they had designed before the pandemic.

1:25:29Often you need to find a completely different solution to, I mean, they seem so simple after the fact, but at that moment is completely unclear that that was the right decision. But that's why you need to stay calm. That's why you need to look around. You need to stop, look around, and find the right direction that you want to go in, and then accelerate. I think that's really an interesting approach, right? Where you sort of stop, you don't panic, you evaluate what's going on as rationally as you can in the moment, despite everybody wanting things from you and all this pressure that you probably have from investors, but you also put on yourself to take care of your employees, to take care of your customers.

1:26:11take care of your family, and everybody's worried about the same thing. So you're operating in a very amplified environment to begin with. Yeah. And the Beth's founders are doing this day-to-day, but here the timeframes are much, much shorter and much, much more compressed, and people's voices are much louder. Just recently, Tony said that it's about many good decisions about sequencing. We all know what we need to do, but sequencing it right is is actually what matters um and the sequence of how you do something often determines whether you're remembered or not remembered in in the example that we talked about for airbnb there was a sequence of like what is easiest what is the what is the values that i have for the company and let's go through the sequence that most matches the values so So Brian wanted to take care of the employees the most.

1:27:11And so the layoff was the last thing that he did. He tried to do everything else before the layoff. In Tony's case, the sequence was to make sure, because he had a heart for merchants, the sequence was to take care of the merchants. So there wasn't... It's just when you have these well-understood values, it makes the sequencing a lot easier and often people forget they write down their values for for a good reason because when you don't have a playbook anymore all you have left is the essence of what the company is which is the values of the company often that's the essence of the founder the values of that founder and how they want to operate but you forget that because it's noisy Lots of people, lots of different parties wanting your time and your attention.

1:28:06Work from home has become a somewhat contentious topic, but I really want to hear your thoughts on the pros and cons and how you see it.

1:28:20Pros and cons. I think it's very, very hard to build an early stage company remote. remote and the reason i say that is unless you you've worked with each other before i i view i view there to be a well of trust that is created when you work together in this you know the reason why people who have worked together for a long period of time can like anticipate each other's movements and what they're going to do next is because they've been around each other for a long time and as you know that most of our communication is not verbal, there's body language, there's how you sit, how you react. There's a lot of communications that we miss through work from home.

1:29:06And you just can't read it as well on the screen versus in person. Because there's only your, you know, on screen, in Zoom, you only have your head, there's the rest of your body that you don't see. And so I think it's quite important for early stage teams to work together. I mean, they don't have an office. They find the coffee shop to work together at. Why do you do that? You can all work from home, but why do you get together a coffee shop to work together? It's noisy. There are a lot of people around. There's a reason why you want to do some of these things. And I think, you know, does it mean that every single day I'm more productive in the office?

1:29:45No. There are days when I'm writing a memo. It's probably better if I sit in the office or I'm at home and just cranking. It's also very, very hard to build culture if you're all separate and apart from each other. There's no ritual. Culture is the values, but how do you express those values? They're your behaviors. They're the rituals. They're the narratives that we tell each other. And those things are just a lot harder to do over Zoom. so i think i think all companies should go back to being in the office a certain a certain number of days a week it doesn't have to be five i think that we can't put the genie back in the bottle on remote work but at the same time you know we thought the world was going to be different when we were all online and it turns out we actually enjoy seeing each other in person we enjoy having dinner together i don't enjoy having dinner over zoom with someone i enjoy it when we're we're breaking bread right off right next to each other do you believe that you can please both sides and that that's sort of the hybrid model where people are we're going to work at the office two or three days a week and then you can work from home two or three days a week do you believe in that being effective or do you believe you have to pick a predominant one and focus, because your operational backend, your company culture, your cadence, they're all going to be determined by what you pick.

1:31:16And if you pick in the middle, you're probably making trade-offs on both sides and not getting maybe the advantages because you don't want to make trade-offs. I think it's best if you take a stance, take a side. Hybrid, I would cripple with the fact that hybrid doesn't work at all. I think you can be in person, you can be remote, you can be hybrid you just need the systems around it to to make up for the fact that you're going to operate in this way every multinational corporation when they have multiple locations was doing some form of hybrid right you have some people who are in this office some people in this office they're connected through tele you know telecommunications um and some some people on the road and like any company that had what gets a certain skill the salespeople in the field they're not actually at headquarters and you are having an all hands those people call in but when you have an important all hands you call everybody in you have an important off-site you call everybody in and there are just certain things that should be done in person there's certain things that are okay to do over Zoom, over telecommunications, over email.

1:32:35I think that there's a different medium to do different jobs, depending on the objective is, and you should understand what the drawbacks and the advantages of each one. At Sequoia, I mean, we take a lot of first meetings over Zoom. and during the pandemic when we were only doing meetings over zoom you just never got the right texture for the company the cadence of the how the company operated and yeah we might still take first meetings over zoom but we're going to meet the team you know at some point in person did you analyze your investing during that period where you're meeting more over zoom versus more in person when it comes as approaching the investment decision?

1:33:22Yeah, we can make those assessments, but the hard part is the environment is also different, right? So 2021 valuations were nutty. Was it the issue of making those investment decisions over Zoom or was it the environment that drove valuations higher so therefore the returns are going to be more difficult, more challenged when valuations are higher? Why do you play when the valuations are higher? when sort of like the game is loaded against you in a way? Like why choose to keep playing? Why not back away and then come back when the odds are in your favor? Well, historically, valuations have gone up.

1:34:01But if they keep going up, then you... How does that work? I think our LPs pay us to make money during good times and bad times. And here's the thing that is most interesting to me. you can decide to shut down but then you won't get the texture of those companies being developed and sometimes it's you gotta keep meeting companies so you don't shut things down you can decide to invest fewer dollars when things are hot and you can decide to invest more dollars when things are cold but the market is also smart so when things are hot it's usually because something new is happening and things are called.

1:34:48Maybe it's just not a good time to invest too. If you could identify when it's irrationally exuberant, yes, you decide not to invest in the irrationally exuberant situations. But in 1999, which was one of the, if you go back in history and look at megatrends in investing, You know, I'm 52 years old right now. When I was in junior high school, one of the things I did as a business was to build PCs, clone PCs, because the IBM PC XT, AT, they were really expensive. and and so there is one that was one mega megatrend which is the pc revolution then in then there was the internet revolution and then more recently the cloud revolution the mobile revolution some right now we're in the ai revolution and the internet revolution when i first started working professionally after college was the ad lake exchange and then we started investing in 1999 That's when VentureFrog started.

1:35:55And if you decided not to invest in that time, you would have missed out on Google, Salesforce, PayPal, Zappos, OpenTable. There are a lot of great companies that were still founded during that time. And paying market prices for those investments would have been just fine. And you can wait until... It would have been better than just fine. It would have been better than just fine. But you could wait until 2001, and maybe you could invest in lower valuations. But often, in the public markets, you can wait because those companies are available when the multiples are low. In private companies, just because the valuations have come down doesn't mean the company that you want to invest in is looking to raise a round.

1:36:42Maybe, you know, in 1999, you wanted to invest in Google. In 2001, they happened to be public. You could invest in Google, but if they were private, maybe you wouldn't be able to. Yeah. John Bragg said this thing in my recent interview with him that I had never fully appreciated, I don't think, which was he had a reputation for overpaying for all of the land he was acquiring. And I said, why do you do that? He's like, well, if we're easy to deal with, people want to deal with us. And he's like, these assets only go for sale once. And he's like, so maybe it takes 12 years instead of 10 to get our payback, but it doesn't really matter because there's no opportunity 10 years from now to buy it again.

1:37:22Yeah, I think that's an example for us. It's the same is true in investing in private companies. It's still much better to pick the right company and slightly overpay than to invest in something that is cheap. you brought up ai as sort of like the next revolution where are we going with ai where do you see that and obviously the longer we go out on the horizon the harder it is to predict so maybe let's start with like what do you see as the next 12 months in ai and where do you think we we sort of like how do you envision sort of 12 months to five years it's funny because i think in the history of technology changes, and I think Bill Gates said this, it's easy.

1:38:08We almost always overestimate what things are going to happen in a year and we underestimate what's happened in 10 years. Yes, it's harder to imagine 10 years out, but in the business I'm in at Sequoia, we are paid to make investments and to hold them long-term, and we do have to sort of imagine 10 years out. In the next year, these foundation models are getting so good that you can see certain things being automated that we used to do that we're not going to do as much. But I think the next step after that is if you think about, lots of people talk about customer service and how we can automate some of the customer service emails or calls and things like that.

1:38:57Okay, let's say we do that. What's next? well, then you should reimagine the customer experience. And so AI will help us automate some of these things, and there'll be massive cost savings. But AI should also allow us to reinvent the customer experience. And that's why we're very excited about a company like Sierra. They're not just going to help you automate customer service tickets. They're going to help you reimagine what customer service looks like. Just similarly to what happened in the internet. And the internet happened. We could book travel on the internet. We didn't have to call a travel agent.

1:39:36It changed the experience. We could search. That changed the experience. It broadened. I had to imagine where I wanted to go and call someone to get me a techie or a hotel. Now I can search. So my imagination could be broadened. And that just changed my experience of travel. And, you know, the airlines didn't allow you to change tickets once you bought them. Because that was in some backend system. They didn't want to open up that system. Now we can change flights on the website. Well, the same will be true with AI. Like, start with automation, start with new experiences, and the experience will completely change over time.

1:40:18And it always takes a little longer. I mean, I remember we talked about autonomous vehicles probably 10 years ago. and today we have autonomous vehicles working in a few cities with Waymo. People imagine that that would be solved very, very quickly. But where are we going? I think we're going in a world where many of the problems that we have in the past will get solved through automation. Many of the things that, many new problems will get solved through AI because they're going to be as good as we are in doing analysis and being creative and things like that. And so I think it will be a very, very productive megatrend.

1:41:00One way to explore this topic is to ask you what you're looking to avoid with AI investments. And I'm thinking, you know, maybe something that comes to my mind is like it's a wrapper over chat GPT or something. What are you trying to avoid when it comes to making AI investments? So the wrapper is a good example of something. it's just a different user interface on top of a foundation model. But I think we're trying to make sure that something persists. And what if it persists? It has its unique distribution. It's embedded into your workflow. It has true ROI for a long period of time and it changes the experience.

1:41:42Other things I'd say we try to avoid is things that are roadkill along the way of a company like OpenAI or Google or Amazon or Meta. So small refinements into what the model doesn't do well today, that's just going to be roadkill along the way. And then I think we're trying to avoid things that sound good, but they're not good businesses. You know, back to, you know, 1999, there were a lot of things that had lots of usage that not had no revenue. And so right now there's a different example where there's a lot of test revenue, but the churn is very high because people don't stick. So there are a lot of things where you see very, very fast adoption, but the churn rate is very, very poor.

1:42:36So we're trying to avoid those kind of companies. Do you think Google and Microsoft, How do you think you compete with them on a foundational model basis? A, they can scale into enterprise overnight. B, they can spend$100 billion on GPs. How do you think about that? Are we going to end up in a world where there's only a few foundational models and we're all going to plug into them? To be honest, I don't know. But I'll tell you this. In the history of technology investing that I've been involved with, If you're afraid of the incumbents, you should just punch out. But in 1999, you could have been afraid for Microsoft because they have lots of money and lots of servers, and yet they didn't win search.

1:43:24Someone else won search. Microsoft won the browser war for a hot socket. It was Netscape against Internet Explorer. I don't know many people who use Internet Explorer today. most people I know use Chrome. So this is not a static situation. I think you just have to sort of have an open mind on why you will win. And often the reason why large companies don't win is because they have their own internal issues. That slowed them down. So Microsoft during the internet, it was the antitrust situation. And that slowed them down. They could no longer bundle IE into the operating system, and so they had to break all those things apart.

1:44:15Google has its own regulatory challenges, and so does Microsoft. So I don't presume that they're all just going to win because they have a lot more money and they have a lot more capital and they have a lot more people. And besides, if you thought that, then I would be out of a job. So how important do you see what Facebook's doing then, where they're spending$50 to$100 billion, but they're making it open source? I think open source is a very, very important aspect of what has happened in technology. And you have these religious fights between closed and open, and you have religious fights between Apple and Microsoft, or Apple and IBM.

1:44:57But let's go back to and. The world is built on and. And open source in the internet days, there were lots of things that were built on open source, and some of it was closed source and some of it was open source. And I think the same will be true with AI. There are going to be things where you just want to host it on a closed source model because you know it has the breadth of functionality. And there are things where you can only make it work if you use your specific data, train it very specifically for you, and you're going to use open source models for that. And just like Google built a great business that is pretty close source on open source software, like Linux, I think the world of AI is going to be in hand as well.

1:45:51There'll be close source and there'll be open source. I want you to imagine for you're in charge of, let's say, Canada and the US and maybe the UK. And I tell you that we are optimizing to be the best in the world at AI. What are the policies that you think about? How do you approach this problem? And how do you encourage us to be the world leaders? I'm maybe overly optimistic about human nature. And so I believe that you want to have as little regulation as possible in the early days. Let people experiment. Let people try things. And is it all going to be good? No. But if we had over-regulated the internet, it would not have grown to the scale that it is today.

1:46:45And do we have problems on the internet? Absolutely. And so I believe in sort of having regulation that is more open than not, and then slowly close off things that you know is not good or the behavior is not good. Regulators often worry about the fact that if you wait too long, you won't be able to regulate it. I don't know if I believe that. And then on the flip side is, it's always been, I don't know of a time that has ever worked to fight against technology and to overregulate technology. You could have overregulated the fact that the scribes were going to no longer have jobs because we came with the printing press.

1:47:28You can go all the way back in time like that. Or figuring out how to have a plow so that you need fewer people to sow seeds because the plow could do the work of three or four people all the way up to a tractor. How have we done trying to regulate or over-regulate or control things that sort of make human beings much more productive? That's what we're talking about with technology, and that's what we're talking about with AI. and then you know if you want to start applying regulation there's there's maybe a slightly different framework that i would use which is you want you know as an example with the combustion engine nobody regulated the engine we regulated what engines can go on the streets what can go into a plane.

1:48:19You regulated how it was applied. But in a lab, and people were, you know, sort of doing research and trying to make the combustion engine even more powerful, being more fuel efficient, you don't regulate that. You regulate it when it gets in the hands of a consumer. So that was like removing barriers almost. But like, what about amplifying? What about the push? How do we encourage more? How would you embrace this as a country? The market is pretty good at pushing. When they see opportunity, more and more capital goes into it, more people go into school. In almost every single technology revolution I've been part of, there are too few people who understand that technology.

1:49:12technology. So when there was the PC industry, there are some people who understand mainframes, very few people understand PCs. More and more people, more and more investment went into PCs, more and more people bought PCs, more people started using PCs, more people started building applications for PCs. And so over time, there was just a lot more people and a lot more investment in it. People went to school to sort of write software specifically for the PC as opposed to framework, the mainframe. That happened in the internet. Most people didn't know how to build a website. More and more people learned how to program in HTML and Python and Linux and learn how to use Linux to build.

1:49:53Today, that happened with cloud and mobile. So today, we have probably too few people who really, really understand the technology and its application. But you have so many people going into it today because they see the future is going to be powered by AI. And so wait a year or two. I bet you there's plenty of push today. That's why the valuations are high for AI companies. And wait a year or two. There's going to be a lot of things that are going to be built. In fact, the hot take is that many things will be overbuilt. So maybe we were building too many GPUs for a year or two years out. Maybe there are too many data centers that have GPUs a year or two years out.

1:50:39And maybe those things will get cheap. And when they get cheap, we'll find new applications when it gets really, really cheap. That would be sort of the history of America. Actually, it's interesting that you brought up don't fight technology. There's a saying in investing, don't fight the Fed. It's sort of the same sort of. The Fed is way more powerful in many respects. Than technology? No. Well, I think the Fed can constrain or can loosen money supply in a very dramatic way. So it applies throughout the whole economy as opposed to just about technology. Historically, businesses have cycles. We tend to overbuild when we're optimistic.

1:51:21And then, you know, you go through that phase you just talked about. Out of the large technology companies today, who do you think is best positioned? That's hard. Well, if you want to use Don't Fight the Tape, when I started in my technology career, Microsoft was the largest technology company, and today they're the largest company. And so they seem to have figured out through their ups and downs how to stay relevant and stay on top.

1:51:57i think google are google's having some challenges with their own regulatory issues and apple don't ever count apple out because back to they weren't the first to come up with the pc they weren't first to come up with the mobile phone they weren't there's so many other mobile phones. They just wait until they can create a great customer experience. And then when they do, they pounce. They weren't the first to come up with an MP3 player. Or a phone. Or a phone. How do you think of NVIDIA?

1:52:37I think NVIDIA is a very great company for a variety of reasons. One of which is, it went from a graphics processing company to being the way we power all of AI. And they fought for so long about this idea of parallel processing. Well, the CPU worked a certain way, the GPU worked a different way, and the GPU is now shown to be more powerful. after we sort of got to the edge of Moore's Law, the GPU is winning out because they can process things much more efficiently. But I worry about... They've only had one leader in their whole cycle. I worry about what happens when Jensen decides not to be at the head of NVIDIA.

1:53:37Yeah, that is the question, I guess, right? I think that's the question a lot of people have. jensen's a remarkable sort of ceo jensen's a remarkable ceo and he does things very differently than anybody else we had him speak at our base camp which is our annual retreat for our founders and he just came out with so many great stories about nvidia about his personal life but also about his management style that is so different i remember being there and that was that was awesome Yeah. Sometimes, you've said in the past that sometimes the path to greater efficiency is doing things sequentially and sometimes it's doing them in parallel.

1:54:14Can you explain that? I think most people are trained to do things in serial they take a problem, they break it down into pieces and they do one piece at a time step number one, solve problem number one solve problem number two because we need the solution to problem number one to feed into problem number two and they eventually go down the path and a lot of, like if you're an individual contributor you just, you write code you just keep writing the code and you finish the code and you submit it but if you're managing a team you need to break down this piece of software into components so that you can give it to your team members you need to have all 10 of them working on it as opposed to you write it yourself even if you're a really good software writer you literally have to be a 10x engineer on top of your 10x engineer so you have to be a 100x engineer to do it in the same time frame that 10, 10X engineers can do it.

1:55:16And sometimes it's just more efficient to just break the problem down so that you can put them in parallel and then stick them all together. How do you think about companies and time span and specifically maybe with an angle towards positioning yourself to rapidly adopt to whatever the future brings versus trying to predict the future and running ahead of it? I think the consensus would tell you that it's very, very hard to predict the future. And your ability to predict the future is uneven, because in the world that you know, you may be able to be very good at predicting the future. but the world is not stationary and the world changes and in this new world trying to predict the future when you don't have many facts or many patterns that you can go against uh it's much better to have machine that does when to pounce than to try to predict what's going to happen go deeper on that a machine well your company is in some sense a machine And so you're trying to figure out when you should pounce on an opportunity.

1:56:28You're building, you're building, you're building, and you decide, I'm going to take this path now. And you're going to make your company go in this direction because you see the opportunity in this path. maybe make the example more concrete. Perhaps, you know, as an example, DoorDash started with food delivery, but specifically restaurant delivery, and they waited until they figured out an advantage that they could have in grocery. But the mechanism by which to deliver restaurant food and grocery food were pretty much the same. and figuring out when to enter the grocery market had a lot to do with whether the company was ready.

1:57:18The customers were ready. The market was ready. The grocers were ready. And you can predict that customers want this. before they want it or you can just wait until you see lots of customers bang on the door writing into DoorDash, hey, can you also get my groceries? And so it's a matter of what, do you want to be pushing the customer to do what you want or do you want the customer to pull you in the direction that they want to go? Is the biggest risk there that you try to tackle two problems at once? I'm assuming they knew they wanted to get into groceries. So part of the risk at the start is if we're focusing on restaurants and groceries, all the time we're spending on one of those two things?

1:58:06And not all of it, but a lot of it's coming at the expense of the other. Or do you see that as, no, it's not? It's a very sharp insight that you have. I think the one thing that you'll find is that there are lots of examples that I could use to support your conclusion there. Because before DoorDash happened, there were plenty of other solutions. solutions and most notably there was grubhub that was just the website and then they passed the order to the restaurant and the restaurant delivered and tony's observation was that's not deep enough of a solution and so why was it not deep enough well there are a lot of restaurants out there that don't have their own delivery staff yeah and why don't we just you're creating problems for them in a way exactly and so the only restaurants that would be on grubhub's website were all the companies that already had delivery staff available so what doordash did was to expand the market and by expanding the market they were very focused on that problem because they're like really you'll you'll deliver you'll take the orders i cook it and you'll deliver it for me i've never heard this before so you're just like educating the market for a period of time because you're just focused on that problem as opposed to um besides grub hum which didn't do that there was postmates originally postmates wasn't just in um restaurant delivery they were doing anything They would pick up from anywhere, and they would deliver it to anywhere.

1:59:52And they would even pick up something from, I don't know, Bloomingdale's or from Walmart. But in that example, there's just too many different things that you could do, and it wasn't focused. And even more broadly than not, there was TaskRabbit, where you can run errands. Well, one of the errands you can have them run, someone who's on TaskRabbit, is to go pick up food from any restaurant. So simplicity of the solution is the reason why I think you want to start with a very focused problem. It's much easier to tell people, this is what DoorDash does. And then over time, you expand that. You expand the remit because you've earned the trust of the customer to allow you to do more things.

2:00:40One of the problems when you were saying that is like, how did DoorDash find drivers in those early days? Yeah, it was hard because if you sort of believed that it was more profitable to transport humans than deliver food, then you would come to the conclusion that Uber and Lyft would get all the drivers and DoorDash would not get any of their drivers. And it turned out that, well, A, there's some people who prefer to deliver food or a package over delivering people. They don't want to have to have social interactions. And then there's a set of people with perfectly fine cars, but they're not perfectly fine cars to transport humans.

2:01:26And so early on, we thought it was going to be a problem, but it was less of a problem than we had originally imagined. one of the the problems that i anticipate or maybe from the outside looking in i can be completely wrong is when you sign up a grocery store customer they want to work with you in a particular way but it might not be a standardized way and then you might say yes because you want to sign up that customer maybe it's like albertson's or something but then you go to the next largest grocery store chain and they want to work with you in a different way how do you think through the standardization which is like, no, we work in this way versus creating all of these one-offs.

2:02:04And I see a lot of startups do this where they want to sign a customer so they'll violate sort of like their standard procedures and they'll create these one-off contracts and technical solutions and they're trying to get revenue so that they can still exist. How do you think about that trade-off? It's a very, I think it's a, I use the accordion example often, which is you pull the accordion too far and then you have to push it back in. At some level, if you are super rigid, you're not going to win your first few customers. So at the very beginning, the accordion is completely folded and you do pull the accordion and you allow things that you may not want to allow in that very early days.

2:02:54And then it gets super far stretched And you're like, oh shoot, I need to standardize a bunch of things. And then first, you're not going to go back on these customers or merchants. So you're going to go into the next inning with a lot more standardization. Or you break up your hundred customers into three personas or four personas and try to have four different standardizations. But trying to be overly standardized at the beginning is generally not a great idea. and the opposite of that is have a standardized product and then allow customization on the edges and I think a lot of software companies try to do that when build a platform look the platform is not changing you're buying the platform but on the edges we lack for customization and you allow you do it at the beginning you find other developers who are willing to do the professional services on the side and go from there.

2:03:58It's almost like you earn the right to do standardization when you're sort of trying to be the back end where you have to accommodate at the start and widen the accordion. And then as you get scale and network effects and more and more benefits to the merchant or partner that you're partnering with, you sort of can more standardize. Is that true? That is true. And I also just think it's very, there's a lot of hubris in thinking that you can standardize at the beginning when you don't even know whether that's the right solution for everyone to standardize on. And so bring some customers. They'll tell you your standardization doesn't work for, not only just because it doesn't work for them, it may be that it doesn't work for many people.

2:04:40And at least learn from that. What's the difference between working backwards and working forwards? So working backwards is this idea that Amazon popularized, which was you start with the vision of the world that you want to imagine, and you think about if everything goes right, what is the world that we live in, and what does this product become that you're trying to build? And at Sequoia, we call that the pre-parade. And if everything goes right with this company, what does it become? And then you work backwards from there. One year before that, what does it look like? Two years before that, what does it look like?

2:05:28And Amazon popularized by working backwards, meaning that go from the future all the way to today. And I love that concept in a sense, but I've also seen lots of founders struggle to take it from the future, which they have this great vision of the world, and really map it all the way back to today. They map it somewhere five years out, 10 years out, but they can't map it today. And I also tell founders, okay, that's the future that you're aiming for. Here are the realities of today. What does it look like one year ahead, two years ahead, three years ahead? That's working forwards. And if the two don't match, you're not on the same trajectory.

2:06:14And so I believe in having founders work both backwards and forwards. Often when you work backwards and you think you have a great plan and then you mis-execute and you don't know that you're not executing, it's because you haven't worked forwards on what the year one, two, and three look like to be on that trajectory. What are the two or three mental models you keep coming back to most when you're making a decision?

2:06:45I'm going to ask you that question. this is your interview nobody's here to listen to me i want to know what your three two or three mental models that you use i think the one that i learned from amazon is whether this is a type one or type two decision and you know if it's a type one decision i've learned from brian chesky to run out the clock because you can't go back as a type two decision and you you know you just move the move forward make a call and of course correct along the way type one being irreversible on type 2, yeah. So if it's irreversible, run out the clock, meaning get all the information, make sure that that's a high-confidence decision.

2:07:27Type 2, if you can always change it, make a fast decision and keep iterating. That's one model. The other mental model I think about is

2:07:42working backwards, working forwards, it's also top-down versus bottom-up is a similar sort of idea. Some problems are better solved thinking top-down. Some problems are more easily solved just working bottoms up. Another one generally is to break up hard problems into smaller and smaller bite-sized problems and solve each of those individually. And just the way I live my life, this is not a mental model. It's a lot about consistent compounding and just build just every single day doing the little things that sort of improve myself or oneself or the company that I'm working with. Talk to me how you think through the breaking problems thing, because you mentioned bringing it into smaller problems, solving those, and then you get the solution.

2:08:37How do you think of that versus the optimal solution for that particular sub-problem might not lead to the optimal solution for the bigger problem? hmm yeah so i think there is local optimization and global optimization you're talking about that's a much better way to word it if you want to solve a global optimization problem you need to pop up and if you're trying to solve a local optimization problem you want to go down and so i think the problem first and foremost has to be at the right altitude and then you can break the problem down from there is there an example that comes to mind for your thinking?

2:09:22If you're trying to climb a mountain and the top-down problem is, what's the path? So even before you climb the mountain, you're going to plot the path before you get onto the mountain. But once you're on the mountain and you have a path, you're just going one step in front of the other on that path. And if you find that the path is not right, hopefully you've had some backup paths and backup plans that you can go one foot in front of the other. Path A doesn't work, while path B might be a little better because I plotted the alternative plot to path B. So that's an example of like planning top-down on your path and your alternatives and doing a bunch of scenario analysis.

2:10:09And then when you're executing on a day-to-day period, you just literally just go execute. And I don't see that any differently than a company. You have a plan for the year. You have some scenario planning around it. And so the more scenario plans that you have, the more easily you can course correct. You know you're supposed to be on path A. You decide that you're still going on path A. You're slightly off. Let's go back on path A. Path A turns out not to be the right thing. We're going to pick path B because that's the second alternative. We're going to go right one foot in front of the other, plotting against Path B.

2:10:49We always end these interviews with the same question, which is, what is success for you? I know that this question was coming. And to me, the success is just the process and having a good process. because if you have a process that's consistent to who you are and your values and you live that every single day and you get up every day and follow through on that process, you're already successful. Lots of people talk about success is not the destination, it's the journey. I think it's just the process before the journey. Which direction you're going to go? Why did you decide to go in that direction?

2:11:37how fast you want to go in that direction before you can go on that journey. Hopefully you've planned things out and you have had a process to sort of plan that out. And if you did that right, you know you're going to be successful.

2:11:58Thanks for listening and learning with us. For a complete list of episodes, show notes, transcripts, and more, go to fs.blog slash podcast, or just Google The Knowledge Project. Recently, I've started to record my reflections and thoughts about the interview after the interview. I sit down, highlight the key moments that stood out for me, and I also talk about other connections to episodes and sort of what's got me pondering that I maybe haven't quite figured out. This is available to supporting members of The Knowledge Project. You can go to fs.blog slash membership, check out the show notes for a link, and you can sign up today.

2:12:37And my reflections will just be available in your private podcast feed. You'll also skip all the ads at the front of the episode. The Farnham Street blog is also where you can learn more about my new book, Clear Thinking, turning ordinary moments into extraordinary results. It's a transformative guide that hands you the tools to master your fate, sharpen your decision making, and set yourself up for unparalleled success. Learn more at fs.blog.com. Until next time.