Tony Xu - A Human and Math Problem - [Founder’s Field Guide, EP. 28]

My guest today is Tony Xu, co-founder, and CEO of DoorDash. Tony started DoorDash 7.5 years ago, and today it is one of the largest food delivery and logistics platforms globally, with operations in the US, Canada, Australia. In our conversation, we discuss the initial problem that DoorDash set out to solve, DoorDash’s counterintuitive approach to building their product, and the surprising benefits of capital constraints in DoorDash’s early days. DoorDash’s business model is an equal parts logistics nightmare and a human coordination problem, or as Tony puts it, a human and math problem. After talking to Tony, I feel his personality makes him a great candidate to solve those problems. For the full show notes, transcript, and links to mentioned content, check out the episode page here. ----- This episode is brought to you by Vanta. Vanta has built software that makes it easier to get and maintain your SOC 2 report at a fraction of the typical cost. Founder’s Field Guide listeners can redeem a $1k off coupon at vanta.com/patrick. ----- This episode is brought to you by LinkedIn Jobs. With LinkedIn, you get access to an active community of professionals with more than 722 million members worldwide. LinkedIn is the easiest place in the world to post a job and message qualified candidates. Getting started is easier than ever, and now you can do this all from your mobile device. When your business is ready to make that next hire, find the right person with LinkedIn Jobs. And now, you can post a job for free. Just visit linkedin.com/fieldguide to post a job for free. Terms and conditions apply. ----- Founder's Field Guide is a property of Colossus, Inc. For more episodes of Founder's Field Guide, visit joincolossus.com/episodes. Stay up to date on all our podcasts by signing up to Colossus Weekly, our quick dive every Sunday highlighting the top business and investing concepts from our podcasts and the best of what we read that week. Sign up here. Follow us on Twitter: @patrick_oshag | @JoinColossus Show Notes [00:03:07] - [First question] - The catalytic moment that inspired Tony to build DoorDash [00:06:33] - Network density and its role in scalability [00:09:15] - Solving a math and sales problem simultaneously [00:11:33] - Becoming aware of multiple problem sets for restaurant owners [00:14:34] - Designing a team to quantify friction and turn it into useable data [00:16:49] - How to approach difficult problems and compartmentalize them in a practical way [00:18:38] - Accounting for high-level variance in long-term goals [00:21:54] - Factoring in time horizons for variance and planning for the future [00:23:08] - Slowing down an action process to minimize a high-consequence decision [00:26:48] - Generating demand after completing their software infrastructure [00:28:15] - Strategic choices between network health and unit economics [00:30:32] - Why DoorDash struggled to raise financing in their seed-stage [00:32:23] - Solving a small fraction of a larger logistics problem [00:35:00] - Differences between local commerce and distribution centers [00:36:44] - Creating the DashPass subscription service and how premium experiences influence user behavior [00:38:34] - Advantages and potential diminishing returns in servant leadership [00:40:41] - His broader philosophy on leadership and how it’s changed over the years [00:44:55] - Creating an incentive-aligned culture of ownership at scale [00:47:02] - The most surprising lessons learned about company building [00:50:38] - Unexpected positive outcomes that may emerge in other areas due to DoorDash’s growth in the coming years [00:53:44] - Will DoorDash’s presence change behavioral patterns in where people choose to live or start their business? [00:55:36] - What has him most excited for the future [00:56:55] - Kindest thing anyone has ever done for him

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Patrick O'Shaughnessy?

  I know firsthand how complex the tech stack is for asset management firms. And seemingly every new tool and data source makes the problem even worse, adding more complexity, more headcount and more risk. RidgeLine offers a better way forward. One unified platform that automates away the complexity across portfolio accounting, reconciliation, reporting, trading, compliance and more, all at scale. RidgeLine is revolutionizing investment management, helping ambitious firms scale faster, operate smarter and stay ahead of the curve. See what RidgeLine can unlock for your firm. Schedule a demo at ridgeline.ai. This episode is brought to you by Vanta. Does your startup need a SOC 2 report to close big deals? Or do you already have a SOC 2 report and want to make it easier to maintain? Vanta has built software that makes it easier to both get and renew your SOC 2.

• 0:45Patrick O'Shaughnessy?

  With Vanta's continuous monitoring solution, you avoid hosting auditors on site and taking screenshots to prove that you're compliant, so you can focus on building your business. Vanta partners with audit firms who file your SOC 2 report directly inside of Vanta at a fraction of the normal cost. Hundreds of companies, including more than 100 Y Combinator businesses, are leveraging Vanta's today to streamline compliance and focus on building their businesses. Founder's Field Guide listeners can redeem a $1,000 off coupon at vanta.com/patrick. That's vanta.com/patrick. This episode is brought to you by LinkedIn Jobs. 2021 means new opportunities to grow your business. If part of your strategy is adding new members to your team, LinkedIn Jobs finds the right person quickly. To make things better, your first job post is free.

• 1:29Patrick O'Shaughnessy?

  With LinkedIn, you get access to an active community of professionals with more than 722 million members worldwide. LinkedIn is the easiest place in the world to post a job and message qualified candidates. Getting started is easier than ever, and now you can do all this from your mobile device. That's how LinkedIn Jobs can help you hire the right person faster. When your business is ready to make that next hire, find the right person with LinkedIn Jobs. And now you can post a job for free. Just visit linkedin.com/fieldguide. Again, that's linkedin.com/fieldguide to post a job for free. Terms and conditions apply. Hello and welcome, everyone. I'm Patrick O'Shaughnessy and this is Founder's Field Guide. Founder's Field Guide is a series of conversations with founders, CEOs, and operators building great businesses. I believe we are all builders in our own way, and this

• 2:14Patrick O'Shaughnessy?

  series is dedicated to stories and lessons from builders of all types. Founder's Field Guide is part of the Colossus family of podcasts, and you can access all of our podcasts, including edited transcripts, show notes, and resources to keep learning at joincolossus.com.

• 2:32Speaker 01

  Patrick O'Shaughnessy is the CEO of O'Shaughnessy Asset Management. All opinions expressed by Patrick and podcast guests are solely their own opinions and do not reflect the opinion of O'Shaughnessy Asset Management. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of O'Shaughnessy Asset Management may maintain positions in the securities discussed in this podcast.

• 2:58Patrick O'Shaughnessy?

  My guest today is Tony Shu, co-founder and CEO of DoorDash. Tony started DoorDash seven and a half years ago, and today it's one of the largest food delivery and logistics platforms globally with operations in the US, Canada, and Australia. In our conversation, we discussed the initial problem that DoorDash set out to solve, DoorDash's counterintuitive approach to building their product, and the surprising benefits of capital constraints in DoorDash's early days. DoorDash's business model is equal parts logistics nightmare and human coordination problem, or as Tony put it, "a human and a math problem." After talking to Tony, I feel his personality makes him a great candidate to solve all of these interesting issues. Please enjoy my conversation with Tony Shu.

• 3:39Patrick O'Shaughnessy?

  So Tony, where to begin? Such an incredible business story since 2012 when you started DoorDash. Let's just start at the beginning. I love the founding insights or aha moments early on in a company's history that make you know you're going to chase after something. What is the first of those moments that you experienced in the DoorDash journey?

• 3:57Speaker 02

  Our founding store didn't really start with delivery. It didn't even start with food. It really started with a desire to help local business owners. People like my mom. I grew up in a really classic immigrant family story where my family came to this country from China, came to the US from China when I was five. My dad was getting his graduate degree and my mom was working three jobs a day to put food on the table. She worked for all local businesses, three of them, 12 hours a day for pretty much 12 years. I was able to moonlight as a dishwasher alongside her inside of a local Chinese restaurant. For me, that's kind of how I got motivated in the first instance to help brick and mortar businesses. And for my co-founders, there were similar stories and pursuits. And that's really what got the co-founding team together.

• 4:42Speaker 02

  I think for us, the founding motivation really started with a customer segment before we actually identified any problem. And then in 2013, as the team got together and we started speaking with business owners, whether they were people like my mom and sold food or others that sold flowers or furniture or other retail items, we discovered that actually there are a lot of problems to solve. None of them are ready to compete in today's digital or convenience economy because everything was thought of in the framework and paradigm of their four walls. If you think about customer acquisition, it was about where were they located? Or what phone number in the phone book or on Google they were? If you thought about customer service, it happened inside the store there and then.

• 5:26Speaker 02

  If you got the wrong item made inside of a restaurant or you couldn't find something inside of a retail store, all of that was solved by waiters, waitresses, clerks. Everything was done inside the four walls. But we all knew that business was happening more and more, not always inside those four walls. Everything needed to be reinvented. But look, for a small group of four people working outside of my apartment, there was no ability to solve all of those problems. We kind of had to pick and choose. So we chose the customer segment, these brick and mortar businesses. We wanted to transform them, went through a journey of identifying different problems. We actually touched upon a couple of different problems before we ultimately landed on the problem of delivery.

• 6:08Speaker 02

  Really around delivery, what happened was we spoke with a local macaroon store in Palo Alto. The store manager Chloe showed us a book of orders that had a bunch of customer information, even their payment details, but she had to turn all of them away because they were all delivery orders and she was a one-person shop. She couldn't fulfill the orders as obviously she was doing a bunch of other things, just running the store. And so that was kind of the first moment in which we discovered, wait a minute, this seems bizarre. This is 2013. Something as simple as delivery should exist, yet outside of New York City, really nowhere did it exist, at least in America. That was really how we started.

• 6:52Speaker 02

  It really started with a founding motivation, very authentic to each of the founders backgrounds and journeys through life. And then we identified the problem in talking with physical business owners.

• 7:06Patrick O'Shaughnessy?

  I'm absolutely obsessed with this modern class of businesses that allows a creative focus for entrepreneurs and takes away the logistics or sometimes even business focus of those same entrepreneurs. It seems like you unlock a certain amount of and quality of creativity when you enable smaller businesses to do the ancillary stuff that they can't figure out themselves. And it seems like early on in that journey, network density is really important. So I know the business started as PaloAltoDelivery.com. So obviously that's a pretty narrow focus. What did you learn early and across the journey about the importance of network density when building something like this?

• 7:48Speaker 02

  Well, certainly we didn't think too much about scalability in the early days, especially naming the service or the product. But look, I mean, when it comes to something like last mile delivery, one of the most important things is creating node density. Last about population density. I think this was one of the early insights we discovered, which was that we were fulfilling deliveries actually faster in places where the population density was much lower, like a Palo Alto or other types of suburbs versus say a San Francisco, which is a much more dense city by population.

• 8:23Speaker 02

  But when you start getting into delivery, you start realizing, wow, there's actually a lot going on. You have to think about where all the stores are located. You have to think about how easy it is to park. You have to think about which doorway you might want to enter through. Because sometimes there are alleys and sometimes there are front entrances. Sometimes one is easier to get into versus the other. You have to think about where are all of the stores relative to where are all of the customers? You know, most cities in America actually follow a very nice hub and spoke model. There's a main street in a downtown area and there tends to be homes nearby or offices nearby that surround it. Now, a place like New York City, that's basically the ultimate node density where you have those main streets in every neighborhood.

• 9:07Speaker 02

  Whereas for the rest of America, it's a bit more spread out. And so the early discovery for us was really thinking through what does node density look like? Because the nodes in our system, you have stores, you have consumers. You also have where dashers, the drivers on our platform, need to be positioned. It was about creating the most node density, less so about the population density, such that we can construct the most efficient routes, such that we can create the most comprehensive and accurate catalog of what was happening in the physical world, so that we can actually fulfill all of the activities necessary.

• 9:48Patrick O'Shaughnessy?

  It sounds like both a data and math problem on the one hand, and a sales problem on the other hand. You've got to convince three different kinds of constituents in a market, dashers, consumers and merchants, all that this is a good thing to be doing. How did you think about solving a math and a sales problem side by side?

• 10:07Speaker 02

  Yeah, we've always had a saying in the company since day one that DoorDash really is a marriage between a math problem and a human problem. We actually started by solving the human side. When I back up and actually when we were just starting the company, we spoke to a lot of our professors, whether it's at Stanford or Berkeley where we studied, around actually the math behind this. Really, this type of problem have been solved before. Because I think a lot of people think of the traveling salesman problem or the multi-vehicle routing problem. But in those instances, a lot of the demand and supply can be forecasted with very great accuracy or some distribution of what's to come. In our system, well, you can't control when people get hungry or when they want to order food.

• 10:50Speaker 02

  You can't force dashers to accept any order. You can't require a merchant to make things faster or slower, or have more things in inventory or less. All of those things are unknowns in our system. This was, from a math perspective, really an unsolved problem. We actually started, therefore, by trying to understand the human problem. Because in our business, it's really not the bits that we have to understand. We almost have to first collect all of the information about the atoms first, so we can create structured information about the physical world, and then apply some math. So for us, that meant doing deliveries. All of the co-founders, all of the employees in the company,

• 11:33Speaker 02

  for the first two years in the company's life, did deliveries pretty much every single day. We would code, talk to customers, restaurant owners and dashers. We would do deliveries in between, pretty much every day, customer service included. And that was really how we learned what were the human activities? Who were these people? What motivated them to do what they did as part of our ecosystem? And then, how do we actually understand their behavior such that we can create information to solve their problems, such that we can solve problems for them?

• 12:06Patrick O'Shaughnessy?

  What's an example of a problem set that you encountered, maybe even you, when you were doing the deliveries early on for a restaurant owner? You mentioned the unfilled delivery orders of the macaroon maker, but give us a sense for some of these other problems that were manifest.

• 12:20Speaker 02

  There were a lot of problems. I think one of the early things, and this came from growing up in a restaurant. I knew that restaurants could always turn food faster than they could turn tables. No one's right now, during the pandemic, eating inside a restaurant. But soon, you and I will have lunch and dinner or a meal together somewhere. Physically, that restaurant owner, they're not going to kick us out of our conversation or our meal together, but they can certainly make more food in the excess capacity they have in that kitchen. From a business perspective, we always have the hypothesis, and this was verified as we spoke with all the restaurants, that one of the problems they had was, how do I make more sales out of my fixed costs? If you think of a restaurant more as a manufacturing plant, they have a lot of fixed costs.

• 13:02Speaker 02

  Rent, obviously, labor, I think sometimes is misunderstood, it's actually largely fixed. Some semi-variability when it comes to seasonality and things like this, but it's mostly fixed. So the only variable part is really the food, some of the packaging costs. And so they were always trying to think, how do I make more stuff out of my plant? And make the most out of my interactions with customers? So that was certainly one of the things we learned. I think other things that we learned was really just how there are different customers within inside of a restaurant. You have the owners, which I think a lot of people think about, and I think a lot about their founding stories as a fellow founder. But there is the store general manager, who's literally running the store day in, day out. There is the staff, that staff is divided into two.

• 13:46Speaker 02

  You have the back of the house where I used to work as a dishwasher. I wasn't qualified to be a chef or a helper in the back. But then there's also the front of the house, people who are serving the customers or people who are managing maybe some of the takeout operations. And there's even a whole set that you don't even see, that doesn't even work inside the restaurant. But they work with the restaurant, such as the accounting team or the team that designs the technology, payment systems, kitchen display systems, bookkeeping systems, et cetera. There are a lot of customers inside of a restaurant. The other thing you learn from a logistics point of view when it comes to restaurants is you notice that there's variability in the operations of a restaurant.

• 14:26Speaker 02

  Sometimes a restaurant is able to manage a different amount of orders per unit time, depending on which staff is in that restaurant. Other times, restaurants have dual kitchen capacities, where they have second make lines. And you think about, you know, how fast that is. Sometimes restaurants have alleyways in which they may make for better operations for something like delivery. Sometimes restaurants are located on the sixth floor of a shopping mall. In which case you have to think about, well, how do you actually get someone to go from that restaurant to meet a driver for delivery? So there are a lot of also logistical challenges as you do the deliveries that you notice about a restaurant.

• 15:07Patrick O'Shaughnessy?

  So that's the human component. You're in these places, learning the constituents and the stakeholders, they're all different. How do you then build a team that can map that into data and on to basically make it legible to software for you to then work with? What did that early team look like? Just, I'm curious, the intensity of the math problem? I think you read math dissertations for fun. So this is a perfect fodder for you. Just talk us through that math problem early on and ongoing.

• 15:31Speaker 02

  I used to read math dissertations for fun, but with a couple of kids running around in the household, it's been a little bit trickier. In the beginning, really, the team was two parts. You can think of it as engineering and operations. All the engineers did deliveries, too. It wasn't like it was just the folks not writing code that were doing deliveries. Everyone did deliveries together, really to learn some of these problems. How do you, for example, decompose a delivery into five to 10 steps? And then within five to 10 steps, how do you decompose that into its lowest level of detail? That was the early team. We always believed you had to get both right. This wasn't a business where, sometimes in places like Silicon Valley, when we observed some of the consumer technology companies

• 16:16Speaker 02

  that have come in some of the earlier decades, there was kind of like one critical focus. Some companies were known as product-driven companies. Others were known as engineering-driven companies. Maybe some were known as even sales-driven companies. At DoorDash, we kind of have to be great at all of these things. You can't say, "We write great code," but we don't pick up the phone if somebody were to call us about an issue, or that we don't actually meet a restaurateur where they are in terms of looking at their business, looking at their P and L, and helping them actually grow sales and profits. It's a combination of these things. And so, early on, it was engineers. And then, I think, on the operations side, we called them Rhodes Scholars meeting Navy SEALs. It's people

• 16:59Speaker 02

  who are Swiss Army knives, who, I think, demanded excellence of themselves in some field. It didn't have to be professional. Prior to DoorDash, almost nobody had any experience in logistics or restaurants. They really came at the problem as a result from first principles of mapping all of the activities, creating structured information, and then actually doing an appliance and technology with it.

• 17:22Patrick O'Shaughnessy?

  But when we last talked, we had this fascinating mini-topic that I think is an interesting sidebar right now, which is when faced with huge potential markets and really fascinating potential problems, that sometimes it makes sense strategically to just break those problems down into something much simpler. I think the context of our discussion was around level five autonomous vehicles. Can you talk us through this kind of mental model for how to approach, yes, an exciting big problem, but make it practical?

• 17:48Speaker 02

  We have a saying internally at the company, the dream big, but to start small. And that's because I think it's really important to know where you're going. That's the dream big part. But it's also really important to get going. And it's really hard to get going on a big problem when you're an under-resourced team, or maybe you don't even yet have the capabilities or the knowledge of what problems you're going to encounter. I mean, that was the story of DoorDash. When we started, the goal was to transform every brick and mortar business, building them a marketplace to give them demand, and also to build them a platform in which we can give them tools to create their own digital businesses. But that was way too big of a problem statement for the four of us to get started on. And

• 18:31Speaker 02

  so we narrowed that problem down to delivery. We narrowed delivery down to restaurants only, not all types of stores. And then we narrowed that down to one locale called Palo Alto, which is where we started. Even within Palo Alto, we narrowed that down to one street on Palo Alto, which at that time was University Avenue. There were a couple of main streets with restaurants, but at the time, we focused on University Avenue. And so the point being that the vision was always to work on a much bigger problem, but taking the first step and which first step to take, I think, is really important because it does have some path dependence, depending on which position you took.

• 19:10Patrick O'Shaughnessy?

  How do you think about variance in all of this? If you're thinking about the future and you're dreaming big, some dreams have higher or lower variance. How does that play into how you think about running a business and planning in general?

• 19:25Speaker 02

  I'm still trying to develop our own mental models on it. I'm constantly learning about different models to handle variance. The first thing that we tend to do when we look at problems that have high variance is we try to reduce the scope of the problem. Usually variance is something that builds on multiple risks. If you talk about something like climate change, it's not one thing. I wish it were because we could solve it much more quickly. Or even, you know, our current COVID pandemic. It's not just one thing about understanding the disease, but also everything from vaccine distribution and masks and containment. And there's a lot of things that add to variance.

• 20:04Speaker 02

  The first thing that we try to do when it comes to variance is we try to simplify the problem and reduce the variance in some sense. And maybe in the early days, we don't know what that means. We just have to almost look at the major components. Take, for example, when we started DoorDash, there was huge variance on understanding whether or not you can actually build a sustainable business, facilitating all of these moving parts of consumers, merchants and dashers, for instance. Okay, so what did we do? What we did was for each audience, we kind of took a hypothesis, what we thought it would take to almost manufacture an outcome that we wanted.

• 20:45Speaker 02

  So for consumers, we thought that they would be willing to pay for convenience. We set a price on that. Merchants, that they would be willing to pay for incremental business. We set a price on that and that dashers would be willing to earn in this flexible way. And we set a price on that. We didn't know. It wasn't like there were rules of engagement here, but we tried to manufacture that outcome early on. The whole point is just to get going so that we can collect information about understanding which parts did we get wrong. It's actually to look for the disconfirming evidence. That's actually why we start with the hypothesis. And then what you can do is if you can reduce variance on each one of your hypotheses, then what I would argue is A, you can get going faster because you're solving simpler

• 21:30Speaker 02

  problems by reducing the scope of variance. And B, you start understanding the interplay. Because a lot of times, big problems where it becomes really hard to solve and sometimes why it requires, at least in our opinion, interdisciplinary kinds of ways that looking at them is because you have interplane dynamics. In our marketplace, for example, every order touches consumers, merchants and dashers. Okay, well, you of course have interplane dynamics and there are sometimes even externalities such as inclement weather. And all of these things do happen every day. Traffic happens every day. Sometimes staffing shortages happen every day at stores or inventory outages happen every day.

• 22:13Speaker 02

  These are all things that contribute to greater and greater variance. The point is to reduce the scope. Therefore, you can understand where the variance is being caused. But most importantly, it means you can get going.

• 22:26Patrick O'Shaughnessy?

  What about time? There's scope is one dimension. In complex systems, high variance situations, more time seems like could be a gift or a curse. How do you think about time as a variable in planning for where to go next in the business?

• 22:41Speaker 02

  Startups and technology, it almost feels like time is always against you. You can never do something right faster or sooner. And that's why, again, back to the point of getting going, I think, is one of the most important points. But with respect to something like time, especially when the variance leads to possibly irreversible outcomes or consequences or consequences that just have large magnitude, maybe not irreversible per se, but very high magnitude. That's when actually you almost have to put a mental block on the get going. And then you have to start taking more time upfront to think through the sequencing of what it is that you want to do. Because maybe you don't have to make a certain decision today.

• 23:26Speaker 02

  And by the way, if you made that decision today, it could really cost you, like I said, a very, very high consequence. When there is the time component, one of the things we try to think about is at which point do we need to make certain decisions?

• 23:41Patrick O'Shaughnessy?

  Is there a good example in the company's history of a high consequence decision that you intentionally slowed the action process down for that you think was the right decision to slow it down?

• 23:52Speaker 02

  One of the most important decisions, maybe very early, as early as month one of the company's life, was we actually spent very little time building the consumer product. Actually, our very first product as a consumer product was actually our Dasher app. Back to this notion of how do you actually build a system of last mile delivery where you can bring everything inside of a neighborhood to consumers in minutes, not hours or days? We actually decided to not work on maybe what seems to be most obvious, which is a consumer product in which you can find all of the things or buy all of the things from all the restaurants near you. Instead, we said, that can wait. We built a website so that you could actually order.

• 24:37Speaker 02

  In the very, very V1 of the company's life, the product was a phone number that rang Google Voice that rang the cell phones of the founders. And that's because the core thesis we were testing was if we delivered from places that never offered delivery before, would you care? And how would we actually deliver upon a good enough experience such that you care enough to come back? We focused instead, and really for probably a couple of years, when I look back on it, on spending, when I think about the allocation of time for our engineering team, 90-plus percent was dedicated in the earliest years to dashers and to merchants. And this is all building things that consumers could not see.

• 25:21Speaker 02

  A Dasher app that a consumer could not see, a logistics dispatch system that nobody could see, an exception handling system, point-of-sale integrations, kitchen display integrations on the merchant site, a merchant portal, a merchant tablet, and other order protocol systems such that they can actually receive orders and communicate to us back and forth. Those are things in hindsight were very expensive decisions. I mean, a couple of years of time, right? A very precious, small number of engineering hours that we actually had were dedicated instead to that. But that's because we thought the core hypothesis to test in this whole system was if we could bring it to you, would you actually care? It's not like delivery is this new thing.

• 26:05Speaker 02

  If you think about it from first principles, well, maybe the reason why delivery isn't outside of New York City is because no one cares. Consider the alternative where we spent two years instead building the consumer product. But maybe we didn't spend as much time on other components such as the dispatch system or the merchant products. And maybe we couldn't deliver from the places you wanted, or we couldn't deliver accurately or on time. If you think about the two paths in terms of the emphasis on which product approach we took, you might reach very different conclusions about the likelihood of success of this business, which over time created lots of value. I think you have to have some points of view.

• 26:45Speaker 02

  And so our point of view, it started with having a point of view that it was really about making sure that the product wasn't necessarily the UI of the consumer app. But much more so, can we bring you what you want in the time that you expected? And in the condition that you would expect. And if we can do that, then we could buy ourselves time to work on other things. And along the way, we're collecting information. It wasn't like it was just, okay, we must just stick with this hypothesis for two, three years. And then let's see what happens at the end of it. Along the way, we're collecting information every single day.

• 27:20Patrick O'Shaughnessy?

  It's a really interesting story because the classic view of a two-sided marketplace would be often that demand is much harder to get. If you aggregate the demand, suppliers will tend to show up. Suppliers like selling stuff, they like business, they like lead gen. But in this case, really you're working on behind the scenes harder work. Seems like it's clearly much harder to build everything you just described than a front end consumer UI. When you did get to the consumer side and started building that UI, what did you learn then about generating demand that belonged with this great new infrastructure you had built for dashers and for merchants?

• 27:59Speaker 02

  I think there were a couple of different inflection points. You know, the first was just the realization that consumers came for really the content that we were showing them. The amazing local Mediterranean hummus shop, the amazing cake store, your favorite fast food place, whatever it was, that was probably the first recognition that the content was what people were coming to look for, more so than any other consideration. I mean, obviously, consumers have to vote with their wallets on many dimensions. What can I get delivered? How fast can I get it delivered? How expensive or affordable their service is? In the case that something went wrong, how good was customer service?

• 28:41Speaker 02

  Obviously, they're judging us along many dimensions, but the very first one, we thought, was the content.

• 28:48Patrick O'Shaughnessy?

  How did you think early on, given that this is such a network-driven business? You're building tools to serve a big network that's complicated. And to be successful, obviously, you have to kind of own a liquid network. How do you match that in the early days of company building against what I'll call unit economics? The gross profit that DoorDash makes on a per-delivery basis, or a per-dasher, workday, or whatever the right unit is? It seems like those two things might be at odds and that you had to make strategic choices early on and maybe ongoing between network health and unit economics. How did you deliberate between those two things?

• 29:27Speaker 02

  I might slightly change the premise where I don't know if your unit economics in a marketplace and your network health have to disagree. I think what might disagree is your overall company profitability and maybe your network health. I think what's important in that process, though, is to make sure that your unit economics always work. And that was true for DoorDash. And one of the things that really helped with this was a constraint, which was our budget. We had very little money in the early days. DoorDash struggled to raise a seed round. And when we got started, so we had very constrained wallets. And as a result, I think we started noticing that something seems to be working because our bank accounts are not going down, even though we have a very small bank account.

• 30:11Speaker 02

  Part of that, I kind of say this tongue in cheek, but was this maniacal focus on always making sure that our unit economics for orders worked. In fact, when I look back over the seven and a half years that we've been building DoorDash, it was critical to do that because in the first five, five and a half years of the company's life, we were very undercapitalized compared to any of our peers, maybe one one hundredths of some of our peers. And that capital efficiency, I think, started, yes, because of the constrained budgets. It also started, though, because we figured out very early on, because we partially had to, that we had to make the unit economics work. So I think unit economics

• 30:53Speaker 02

  and network health, you can actually tie. And it's understanding the relationship of how many units you would need before you could achieve maybe overall company profitability. That's an investment decision.

• 31:05Patrick O'Shaughnessy?

  You mentioned this interesting point about struggling to raise a seed round. I think it was my friend, Daniel Gross, who first asked the provocative question, can you name an iconic huge company that did not have trouble raising a seed round? Or where the price of the seed round was, in hindsight, clearly wrong? I mean, obviously, I guess that's definitionally true. But why do you think that was? This seems to me, with the benefit of hindsight, enormous market, the biggest possible market, and endless places to go. What do you think was behind the difficulty raising financing in the first couple of legs of the business?

• 31:39Speaker 02

  And DoorDash just had challenges along the way of raising capital. But I think usually what it is, is there's a lack of information to make decisions on. I think that's kind of the high level bit. As a result, you kind of have to have convictions on really hypotheses where there isn't a lot of data and you kind of have to make the bet before the data is there. And so when I think about what data was not there, maybe as investors were evaluating the possibility to invest in DoorDash, well, one, do people really want this? Or is this just a service for the wealthy or a certain segment of the population in places like New York City, for example? Second, can you even solve the logistics problem in any

• 32:21Speaker 02

  Sustainable, profitable way because again, when I told you when we started and spoke with some professors about this logistics problem from a technology perspective, I bet investors did similar homework and discovered that this is a problem where it hasn't really been solved before. And then I think third, just how would the markets actually grow? Because there probably will be a few of these networks or several of these networks, who's going to be able to compete in the most effective way? I think those are all probably questions that investors had, but you're going to have to ask them.

• 32:56Patrick O'Shaughnessy?

  As we fast forward through to today, what's fascinating about DoorDash is the percentage of the problem that's been solved. So we think about this like, oh, great. You did it. Successful public company, et cetera. My guess is you would argue that we're a relatively small percentage of the way into solving this bigger problem. How do you think about that? How far into this do you think we are? Not just DoorDash, but just generally speaking, legibility and logistics problem for physical stuff in the world. Is it still as early as it seems it might be?

• 33:25Speaker 02

  It depends on how you define the problem, but I think no matter the problem definition, and I'll talk about, I think nuances of the definition here, but the progress bar is very low. We're very, very, very early. I mean, even when you take something like just the very first core problem that we are solving at DoorDash of just bring all of the restaurants in your neighborhood to you, that industry is still single digit, maybe just clearing 10% of overall industry restaurant sales, even with some of the tailwinds from the pandemic. And you compare that to something like ticketing, which is now 40, 50% online or commerce,

• 34:05Speaker 02

  e-commerce is 20, 25% online. Bring all of the food to you is still in its earliest innings. And obviously food is a massive category, given how often we consume. When you start looking at the problem from bring all of local commerce online. That I'm not even sure we even show a blip on the progress bar right now. I mean, we still don't know information like, again, we're in a pandemic. But if you can think back or remember back a year ago, pre pandemic, or look forward post pandemic, we still don't know where the last parking space available is in a certain city. We don't know how fast people make food, how long it takes to make a salad versus baking a deep dish pizza. Where are the apples left in aisle six?

• 34:50Speaker 02

  These are problems that are really important to solve. I think one of the greatest business model transformations that is happening, and frankly, one of the biggest business opportunities that's occurring right now is all of local commerce is coming online. I mean, look at what happened during the pandemic. Every brick and mortar store, if there's any silver lining, participated in e-commerce and they actually took share for the first time in e-commerce because the collective sum was so massive. I think that should be reason for optimism. Why, when you do digitize and allow all of local commerce to be cataloged and brought online, then many, many new business opportunities will emerge.

• 35:32Patrick O'Shaughnessy?

  What do you think are the most important differences between the local commerce story and the massive distribution center, Amazon type commerce? What's important for us to know about how those worlds differ? Because obviously they're very different things and 20 minute delivery is possible in one, it's not in the other. How do you think about the important differences between those two?

• 35:56Speaker 02

  In an overly simplistic or reductionist view, it's a world of centralization versus decentralization. Because at the end of the day, the customers want the widest selection. They want the highest quality. They want the lowest prices. And that's going to be true in either world, whichever world. I think when you think about it, well, in the centralization world, you can have ownership over the catalog. You can put as many things into it as you want, or as few things you want into it. But you're right, because of the lack of proximity, it's farther away. Because maybe it's part of an ecosystem already, it isn't as inviting to local merchants. Now, when you look at the decentralization play, you can bring things certainly quicker and cheaper because of the proximity to your point, but maybe you can't bring everything.

• 36:41Speaker 02

  Because you have other things you use a city for, not just storage or inventory for commerce and content. I think the two worlds are starting to melt. Where I think that every store on the street recognizes they're really selling you two families of products. They're selling you experiences that we all crave, especially now, because we've been stuck at home for a year. And they're selling you convenience and they're trying to figure out how to do that. Do I do that out of my current square footage? Do I look for adjacions for footage? And I think that's something that is a very, very exciting evolution that will be coming.

• 37:17Patrick O'Shaughnessy?

  When it comes to something that is introduced to the business, like the monthly, I think it's $10 dash pass that consumers can buy that sort of unlocks, it's like an inner ring to the game. If you're a more frequent DoorDash user, I'm sure it makes sense, almost like Amazon Prime. For sure be a member of this thing. How did something like that get conceptualized? And what have you learned about the way that something like that changes user behavior? Versus just giving them something nice?

• 37:43Speaker 02

  That was a project that took almost a year in testing the Dash Pass. Because with subscription programs, I think to your point, lots of things change. Consumer engagement may change, your economics may change also. Because obviously consumers are paying for the benefits of usage. And really it becomes more of a fixed cost to that consumer where they want to use it as often as possible to kind of recoup the fixed cost of membership. And so those kinds of activities, you have to be very careful as you're building a business where you have variable logistics costs, but only fixed revenues coming in from something like a membership program. So that took a lot of testing. But the idea really came from watching our consumers. Because consumers tell you a lot, not certainly when you talk to them, of course.

• 38:28Speaker 02

  I do customer service every day, they say a lot. But they also tell you a lot based on what they do, what they search for, what they're buying, how they're buying, how often they're buying. And that teaches you a lot about perhaps how we can create new products that might delight them and serve them better. But really a product like Dash Pass was one that's a good example, actually. Another good example of one that has very high magnitude of possible outcomes. How you think about rolling out a product like that, in our opinion, is quite important. And what are the different inputs and levers to actually create a program that actually makes sense? And then how do you continuously deliver value from it thereafter?

• 39:07Patrick O'Shaughnessy?

  You mentioned this idea of learning from customers. This has become maybe like the most popular trope. Be a customer obsessed company. Talk to them, watch them, learn from them. And when you start reading about DoorDash, you quickly stumble on this important idea of servant leadership, which I definitely want to talk about in a bit more detail. I think one of the components of servant leadership, you've already mentioned some of it. You do deliveries, you're talking to stakeholders in the ecosystem, not just once or twice, but a lot. And engineers are doing it, blah, blah, blah. Is there a diminishing return curve to that kind of discovery? What have you learned about that? I'm sure it's valuable to do a few times. Is it valuable to do 100 times? Or do you stop learning after relatively few experiences on the front lines?

• 39:51Speaker 02

  One of the things I've learned is that customer expectations always go up. And just like convenience is one that almost moves only in one direction. It tends to go in the direction of greater convenience or gravity tends to go in the direction of landing on the ground. I think customer expectations always go up. I think it's one of these things where you're not blindly trying to build for the customer. That's not the point. I think great products come from taking a point of view like DoorDash did seven and a half years ago, for example, of starting with the Dasher and the merchant product and then building our way towards the consumer. But you have to sanity check along the way.

• 40:32Speaker 02

  And the good news with customers, especially in a high frequency product like ours, you get that information very fast, especially if you want the negative signal. And as a result, we learned quite a bit. And it's not just learning to react necessarily to what new products we've shipped. It's also doing a bit of what you're doing, which is listening and learning about possible new things that maybe are still in exploration phase. So far, I don't know if we've ever actually measured, I guess, the return on, I suppose, time spent in this case. But because I believe that expectations always go up, it's one of the things that's most important for us to continue to do.

• 41:14Patrick O'Shaughnessy?

  What is your broader philosophy on leadership? And how much has that changed across the seven and a half years?

• 41:20Speaker 02

  It's changed a lot, I'm sure. In the beginning, I'm not sure if there was necessarily a philosophy. In fact, we waited as a company, as a founding team, probably for two and a half years, maybe, before we wrote down a set of leadership principles that described our culture. And that's because I think culture, it's eighty percent who you are and twenty percent who you aspire to be. And I think the same can be said about leadership. Because it's very, very hard, I think, sometimes just to read maybe a book or listen to a talk about leadership and then try to map yourself to that perspective. If that's not who you are, you're almost always fighting against human nature in some sense.

• 42:00Speaker 02

  Instead, maybe an easier way to, I guess, find your own voice or your own way of leading might be observing the person in the mirror and finding out what is your strength that you do already and then build a leadership set of principles around it. For example, around our culture, we talked earlier about customer obsession. But we didn't come up with the word customer obsession. Instead, we wrote down stories of what had already happened in the first two years of DoorDash's life. And then we derived the customer obsession principle, not competitor-focused idea, where, for example, in month three of the company's life, we had a massive outage on September 28th, 2013.

• 42:40Speaker 02

  I still remember it because of the fact that we were late on every order by at least forty-five minutes. And I remember it took the founding team maybe thirty seconds to make a pretty high magnitude decision, which was to refund every single order, which was a double-digit percentage of our bank account. You've got to remember, we couldn't raise our seed rent, so we were days away from cash out. And we also stayed up to bake cookies, delivered them by 5:00 AM the next morning to all of those customers. Or I think of stories where there were a lot of deliveries in the early days that actually made me nervous because they were going to the hospital. And I thought maybe something horrendously had gone wrong. But instead, actually, they were going specifically to the maternity ward.

• 43:22Speaker 02

  As moms didn't necessarily want, I guess, hospital food over and again after delivery. They wanted to get their favorite meal. And when we found out about some of these moms... We literally drove alongside the dashers with balloons, chocolates and other stuff, animals and brought them something special. Those are the things I thought about and we wrote down and those are the things that the team remembers that authentically describes customer obsession. We didn't necessarily write down the words customer obsession and then I guess try to fit stories against it. It was almost the other way around and that's how I think of leadership. For example, I like being in the details with our teams in our product reviews and our business reviews. And so one of our principles is to operate at the lowest level of detail.

• 44:05Speaker 02

  But in many ways, that came from something that I naturally did. Now, I think one of the things that's important as a company evolves, as it grows, is you have to check whether or not those things, those beliefs, principles that you authentically live up to every day actually matter to the company. Because companies evolve and they grow and you have to add things sometimes or subtract things sometimes because they may no longer fit that life stage of the company. For example, we added a principle in 2017 around making room at the table. And this was our commitment in 2017 as a three-year-old company to make sure early on that we would build both a management team as well as a company as diverse as our customer base.

• 44:50Speaker 02

  That came with commitments to recruiting, to other dimensions of the company. So that's an example where we didn't have it from the beginning. There were no stories that we could derive. Actually, it was the absence of stories, but it was something that we wanted to get better as part of that 20% that we wanted to add that was not in the building, that was not in the water, that we wanted to make in the water. I think that's how I think about leadership. It's evolving, but it starts with you and I think is very strong willingness to observe and know thyself, both the good things and the not so good things. Because I think that's how you can show up every day consistently.

• 45:28Patrick O'Shaughnessy?

  It's such a beautiful idea. It reminds me of Carlos Brito, who's the CEO of Anheuser-Busch InBev, talks about how you give equity, actual equity, as a reward for ownership, not to create ownership. Ownership is a thing that comes before equity. Your actual strengths and behavior come before the principles, not the other way around. It sounds so obvious when you say it out loud, but I think almost everyone does it the other way. Like, what do we want our principles to be? Like, what sounds good? What was in the book we read most recently, or something like this? And this seems to come from a position of strength in a much more clear and obvious way. That's quite wonderful. How do you do that at scale? You are you. DoorDash is a lot of people.

• 46:11Patrick O'Shaughnessy?

  How do you create that incentive-aligned culture of ownership as the company gets really big, just in the number of people working there?

• 46:22Speaker 02

  As a company starts, I guess, evolving, if you will, from a product into an organization, obviously, the CEO's job really becomes systems design. And when you think of systems design, you need objectives, you need constraints, you need mechanisms to kind of influence. One of the first things that's really important is you have to write things down. It's a mechanism that is true, not just for how you describe your culture or your leadership principles, but also it's how we do things at DoorDash. You know, every meeting starts with a document. This has been true for the last five or so years. Written communication scales much, much better than just conversation that you and I are having. Second, you have to think about who are you bringing into the building? Because it's a lot easier, I found, to get it right on the way in,

• 47:05Speaker 02

  in terms of matching with that value set or those sets of leadership principles, than to try to somehow force it on someone. Because, again, you don't want to fight human nature. You want to embrace human nature, and it's a lot easier to embrace these principles if you already are practicing them. So you look for them. And then the third is you want to find ways to reinforce them as well as to minimize the absence of the practice. But the reinforcement is really, really important and there's lots of ways in which you can do that.

• 47:35Patrick O'Shaughnessy?

  What has been the most surprising thing, aside from everything you just laid out, that you've learned about company building, just generally speaking, maybe relative to your early expectations?

• 47:44Speaker 02

  Well, the initial thing that comes to my mind is just how it's never ending. It's very different from maybe success in other genres that I've participated in before. For example, sports. I grew up playing basketball as a way to learn English. In basketball, there's a certain time clock.

• 48:05Patrick O'Shaughnessy?

  Game ends.

• 48:07Speaker 02

  There's a game over. There's a who scored more points during that duration? I think this is true in terms of the pedagogy in which so many of us were educated, where there's some definition of quote unquote goodness. You get an A in a class, you attend a certain school, you major in a certain field, secure some internship or some job somewhere or some graduate degree. And there's this belief that something ends, that there is this commencement and then there is the graduation. Company building, I feel like it's almost very unnatural because you have to be very long-term oriented. Because ideally, the best success is the company never ends. There is no definition

• 48:49Speaker 02

  of start-stop. I think that's probably the very first thing that it's taking me a while to realize. And also to therefore, to work with everybody inside the building at DoorDash to appreciate the journey that we're on, not some start-stop-final-destination kind of thing. I think the second adds to one of the comments we talked about earlier about playing your strengths, that it's so much easier, company building or appreciating the journey, if you will, if you can play to your superpower and learn other things along the way. That makes the journey a lot more fun. And when the journey is with other people, it makes it a lot more impactful and rewarding, no matter the score.

• 49:33Speaker 02

  Or the outcome and that chasing excellence certainly beats chasing the scoreboard. It's ironic because actually if you chase excellence, maybe the scoreboard will actually be something in your favor. But I don't think of that. I more think about how are we going to do 1% better today? How are we going to invent the next thing to delight our audiences? The final thing I would say is just, I've learned quite a bit around company building around what are some of these attributes of excellence? What are the attributes of chasing excellence and how that's been true? No matter what position in the company you play, the best people tend to have the following attributes. One, they can go high-low and they can really go to the lowest level of detail.

• 50:16Speaker 02

  Two, they have a very strong bias for action. They have that get-going movement indoctrinated inside their bones. Three, they're great recruiters because they recognize that it takes a team to win. Four, they try to intentionally get 1% better every day. There's this humility that no matter what standard or benchmark they've set for themselves or cleared, that there's always something that we can do better. Five, they have great followership. That's actually my definition of leadership. It's not how many people report into you or how many people listen to you. It's how many people want to follow you. And in fact, we have many individual contributors at the company who have great followership and they're immense leaders in their own right.

• 51:00Speaker 02

  And then finally, they can hold two opposing ideas at the same time. Their ability to constantly want to disconfirm their own bias.

• 51:08Patrick O'Shaughnessy?

  I absolutely love that list. When you step back and think about the problem space that we've spent a lot of time talking about, it has this other feature that I'd like to sort of close with, which is this idea of emergence. So in other, especially network-based technology businesses, what you often find are unexpected and unpredictable, interesting, positive outcomes that result from you reducing friction for merchants. I'd love to hear you riff on that a little bit. You talked about dreaming big earlier. What is the big dream now as you look forward for DoorDash's? Next decade, next seven years after the first seven and a half. How has that dream evolved and what sort of emergence do you think that your activities might create in the world?

• 51:53Speaker 02

  I guess a few points. The first thing I would say is we're still so early in just organizing this city. If you think about our goal of growing and empowering local economies, which produce the vast majority of GDP in this country and globally. We're so early. Now, the network grows in its value and its creation of value as we move from category one to category two. As we start thinking about building out a platform for these businesses so that they own the customer relationships, that they have all of the same products we have from customer acquisition to customer service, so that they can replicate the same business building activities that we do, so that they can be successful.

• 52:37Speaker 02

  And when I think about that network and what that can become in the next X years, I think it'll create more and more business opportunities. I mean, there's no reason why people can't. In fact, some of this is happening, you know, start a restaurant from their own kitchens or to open a retail store from their living rooms. We saw some of this during the pandemic. I think there is no stopping creativity. But the distribution of access, the distribution of knowledge, the distribution of capital, the distribution of some of these enabling functions, if you will, that is limiting and that is right now a constraint. And I think that if we can provide that, I think the activity within just right in front of you,

• 53:21Speaker 02

  the city that has been housing you and gives sustenance to you and your family will continue to be very, very successful. It's actually, I mean, if you looked at some of the statistics, the percentage contribution of GDP and job growth from the physical businesses on the streets have remained relatively the same in this country through a lot of turmoil in the last 80 years that it's been measured through ups and downs in the economy, obviously changes with technology adoption. But regardless of the decade, that percentage is still the majority for every new revolution that comes. That's what gives me. Optimism that the cities are very resilient and it's a complicated set of things that make that true.

• 54:01Speaker 02

  Our job is to help these cities create the next and enable the next set of business model transformations and capabilities, such that more creation can continue, such that these cities will continue being the biggest producers of jobs and GDP creation.

• 54:17Patrick O'Shaughnessy?

  Do you think that your presence in the world will change the nature of where people live? Because hearing you talk about this, I mean, it basically sounds like you're creating a cohesive operating system for businesses where local movement of goods, if that describes your business or the one that you want to build, DoorDash would be the place you go build on top of that. It is the platform on top of what you build in the same way that if you want to sell stuff online, maybe your Shopify is the place where you go. That seems like it could really change dynamics of how people exist, where they live, where they start businesses. Do you think that that's too much of a stretch? Or that will kind of stick to patterns of living that we've seen in the last 50 years?

• 54:56Speaker 02

  I've always believed that even though things take time, people everywhere are more similar than we are different. There's nothing special, you know, what I've learned in the last seven and a half years, I love New York City, I love being there and I can't wait to get back. There was nothing special in hindsight about the New Yorker who can get delivery. What was special was the access to the convenience. It's not like where I grew up in the Midwest, that my friends and family there didn't want convenience or that they didn't eat three times a day or that they also had little kids and they weren't able to cook every meal and things like that. And they were also just looking for time-saving solutions. They were, they just didn't have access to the same products and services that somebody maybe living in New York City did. And so absolutely, I think that's what fosters progress.

• 55:41Speaker 02

  And that's not just true about DoorDash and local commerce and the opportunity for local commerce. I mean, that's true for entrepreneurship. I mean, we're seeing this globally now that you don't have to be based in maybe the Bay Area or particular part of the world to start a company. In fact, there are great companies being started now in many places around the world because of some of the democratization of knowledge, of talent, of capital and of company building practices. And I think that should also be true for local commerce.

• 56:10Patrick O'Shaughnessy?

  What has you most excited about the future, just period?

• 56:13Speaker 02

  The pandemic was another great exhibit of just showing the best of humanity. I know that there were bad times and moments that we saw, in fact, during this pandemic. But we also saw some of the best things. We saw dashers, for example, create cards to make grandparents feel better because they no longer could see grandchildren by taking photos of some of their grandchildren and actually putting them on cards. I mean, who would think of these things? And the point I'm trying to make is I'm very optimistic just about human progress. And this is not just about commerce. I mean, climate change is one of the greatest problems that we, as a collective humanity, will have to solve.

• 56:56Speaker 02

  There's immense progress in a lot of areas. Now, I know we have to do it fast. There's immense progress. Look at some of what's happening with biology. My background was in math, studying mathematical biology, not just in identifying when diseases will actually come, so much in preventative health and just better understanding of the body, as well as better understanding of how to produce treatments. I mean, look at some of the speed in which some of the vaccines have been created during this pandemic. I'm very, very optimistic about humanity.

• 57:27Patrick O'Shaughnessy?

  I love it. What an incredible place to turn to my closing question. I ask everyone the same thing. What is the kindest thing that anyone's ever done for you?

• 57:36Speaker 02

  It's a tough one. There's so many times in my life where there's such asymmetry, I suppose, where someone is literally, unconditionally, giving me something that may never see any payback. Brings me back to the story and the sacrifice that my parents made when they came to this country, also that my grandmother made. I was raised by my grandmother for the first 12 years of my life. I didn't really see my parents that often, mainly because they were just trying to create sustenance so that I can actually never have to defer my dreams. Like dad worked 40 hours a week while getting his graduate school degree. My mom worked three jobs a day, as I mentioned, for 12 years and very, very thankful to all of the sacrifices that they made. I'm a relatively speaking new parent these days.

• 58:21Speaker 02

  When you just think about... Keeping my kids alive, forget about like creating opportunities for them. I think about just the selfless nature that that entails. And then when I add on some of the circumstances that my parents and grandma had to overcome. And frankly, the dream that they had to defer, that's the kindest thing that someone has done for me.

• 58:42Patrick O'Shaughnessy?

  God, what an amazing answer. And it makes me think about when I had my kids going to my parents soon thereafter, just being like, holy shit. Thank you.

• 58:50Speaker 02

  Hundred percent.

• 58:52Patrick O'Shaughnessy?

  This is not. I've just absolutely loved our conversation, Tony. I'm so appreciative of the time and all the insight. Love the business that you have built and are building and all the just awesome lessons that you've shared with us. This episode was brought to you by Vanta. In this five part mini series, I sit down with Vanta CEO Christina Kachopo to hear about the origins of Vanta, how Vanta is automating security and when companies should look to partner with Vanta. In this week's episode, Christina and I discuss how Vanta approaches SOC two compliance with a modern software mentality and what it's like to partner with Vanta. So talk through how you began to approach this problem as a software company yourself.

• 59:32Patrick O'Shaughnessy?

  So I love companies that go from some arcane point in time, human based process to something almost more streaming and more software based. So just talk us through kind of the initial product and how it attacked this need for the market in a unique way.

• 59:45Speaker 03

  So the first thing we noticed was most companies didn't actually, or the security practices across companies should actually be really similar. Even if someone's, I don't know, a developer tool, developer productivity and someone else building an email client, they should both be keeping customer data secure. Both companies employees should, for example, have two factor on their email accounts. So I think the first thing we did, looking at this as software people is say, why is it so bespoke? We actually have a good understanding of security best practices. Let's just use those. Second part was then just looking at, okay, an auditor is going to come in. I'm going to think of them as a smart and skeptical person and I need to prove to them, you know, that I do everything I do. And so then it was just a game of how can we produce

• 1:00:28Speaker 03

  that data? And this was where the technical piece came in of I can pull a screenshot from AWS and tell them my databases are encrypted. That's what we used to do. I can also write a few lines of code and pull information from AWS every 10 minutes to see if the databases are encrypted. So just sat down, thought about what we would need to prove these practices were in place and then wrote a bunch of code against that.

• 1:00:51Patrick O'Shaughnessy?

  So let's say I'm a CTO, which is I'm assuming kind of the point of interaction between Vanta and another software firm. What does it feel like for the CTO to engage with you for the first time? And how does the timeline look different? So what are the steps to getting up and running with Vanta? And if one of my goals, we're going to talk beyond SOC 2 in a second, but let's say my initial goal is SOC 2 compliance. What kind of timeframe and workload am I talking about?

• 1:01:13Speaker 03

  We make setup really easy on purpose. We sort of know that we're working with startup CTOs that have approximately a million other things to do. So getting set up is just getting a handful of API connections so we can get the data we're looking for. And then, once we have that data, we could make what we call a gap assessment, just a list of all the things you need to do and a dashboard that you or someone on your team could work off of. Then it's just a game of finish off all the tasks, get your dashboard green, and then you're ready for your audit period. So it takes differing amounts of time, depending on how the company wants to prioritize it. But we've had founders do it in a weekend, literally. It probably wasn't the most fun weekend, but definitely possible.

• 1:01:53Patrick O'Shaughnessy?

  Better than a year and a half.

• 1:01:54Speaker 03

  Exactly.

• 1:01:55Patrick O'Shaughnessy?

  Once this initial on-ramp is complete, and let's say they're even through their first audit, which is made simpler by Vanta, what does the ongoing relationship look like with a customer? And talk about like the difference between a point in time audit and sort of a, an ongoing integration with Vanta and why that's helpful.

• 1:02:11Speaker 03

  It's a twofold answer here. So one, these auditors come from the accounting background, so they understand recurring revenue. And so these audits are annual events. And so you need to redo them once a year regardless. This isn't just the software people came and tried to put a subscription model on top. This is kind of a subscription model much before. But the other piece is Vanta is continuously monitoring your startup security in the background. And we think that's both important in general. Helps you as a founder know you're you're in good shape. You're not doing anything that will make you embarrassed later. And when we think about just what Vanta is trying to do in the world, we really want to move everyone to this sense of continuous monitoring and continuously verifying a company's security, not just relying on an audit every March, but really having this holistic, continuous picture.

• 1:02:56Patrick O'Shaughnessy?

  So there's almost like a, let's say this official angle. That's the SOC two stamp of approval, which often buyers require of you to to work with you. And you need that. But there's also this, it's almost like an insurance product that the most damaging thing would be if something really bad went wrong. And the odds of some security problem going wrong are reduced with this continuous monitoring. Is that sort of a fair summary of the twin benefits?

• 1:03:19Speaker 03

  Yeah, it is. And we talk about the the security pieces. Sort of just making sure you've locked all the doors and windows of your house. The idea is that is this baseline, again, you're not being a doofus. You're doing the reasonable things. No one's going to walk through the front door and, you know, take your TV. There's certainly other security products on the market that maybe are more intense, but much more narrow. And so the Vanta idea is actually, security is just, you're only as good as your weakest link. And so let's just kind of raise everything up a bit until you're you have that confidence.

• 1:03:49Patrick O'Shaughnessy?

  Maybe just give us an overview. What are the major pieces of a security pie? If I'm thinking about what security means for a software business, you've mentioned a couple of terms here and there. What are the major doors and windows, I guess, is what I'm asking, that that need to be locked down and closed?

• 1:04:03Speaker 03

  So there's a lot. So for a software company, key piece is, in fact, your cloud infrastructure. Really broadly, did you set up AWS or whatever you think? Did you set it up reasonably? And setting it up reasonably breaks down into about 50 things, but did you do that? You have the employees at your company and the accounts they have access to. So what tools and why and what permissions? And hey, does everyone need access to all of the customer data historically? You know, probably not, especially as you grow. Employee laptops end up being important because, again, they have saved passwords. They're just the key, literally and figuratively, to so much customer data. Those are kind of the big ones we think about.

• 1:04:39Patrick O'Shaughnessy?

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