Invest Like the Best with Patrick O'Shaughnessy - Dave Girouard - Making Better Loans - [Founder’s Field Guide, EP. 52]

Starting point is 00:00:00 This week's episode of Founders Field Guide is sponsored by Clavio. Want to deliver marketing moments that last a lifetime, Clavio is the ultimate marketing platform for e-commerce. With targeted segmentation, email automation, SMS marketing, and more, Clavio helps you create your ideal customer experience and keep your customers coming back. See why brands like Living Proof, Solo Stove, and Nomad trust Clavio to grow their business. Get a free trial at clavio.com slash founders. That's K-L-A-V-I-O-O.

Starting point is 00:00:30 com slash founders. This episode is brought to you by Verset. Verset designs, builds, and scales digital platforms for some of the world's most ambitious companies like TD Bank, Getty Images, and American Express. If you require a high-performance team to tackle a hard or ambitious problem, then Verset is the firm to call. Listeners to the show can get free access to their private internal repository containing the most interesting essays, memos, and reading the Verset team has discovered across the internet over the past 10 years. To check it out, visit verset.com slash Patrick. That's V-E-R-S-E-T-T-T-com slash Patrick.

Starting point is 00:01:07 Hello and welcome, everyone. I'm Patrick O'Shaughnessy, and this is Founders Field Guide. Founders 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 series is dedicated to stories and lessons from builders of all types. Founders Field Guide is part of the colossus family of podcasts, and you can access all of our podcasts,

Starting point is 00:01:27 including edited transcripts, show notes, and resources to keep learning at join colossus.com. 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'Shaunacy Asset Management. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of O'Shaunosite Asset Management may maintain positions and the securities discussed in this podcast. My guest today is Dave Girard, co-founder and CEO of Upstart, a lending platform that leverages AI to make loans more accessible and affordable. Dave started in Silicon Valley as a product manager at Apple and later spent eight years

Starting point is 00:02:14 at Google, where he built their suite of cloud apps. In our discussion, we cover the lessons Dave has learned about building speed into a habit, the intricacies of training an AI model to predict the future, and what it was like to start a fintech business as an outsider. We also discussed the past, present, and future of lending, why Dave is a intricacies. Dave and his team have no plans to build a super app and the differences between public and private market investors from a founder's perspective. One of the tropes you'll hear these days is that lending has become a customer acquisition tool for fintechs. But as Dave explains, the market and opportunity

Starting point is 00:02:45 set in lending itself should not be underestimated. Please enjoy this great conversation with Dave Gerard. So Dave, I think the best place to begin this conversation is with just a really fun, interesting topic that you've written about before we get into the great journey of Upstart and everything you've learn building that company. The topic to begin with is this idea of speed as a habit in business. And this is very popular notion in technology, you know, the move fast and break things, probably the most famous example of a phrase that exemplifies this idea. But I think you've really thought about the various levels of building speed into a habit and why that's good and maybe there is ways in which it isn't good or can be dangerous. And I'd love to just begin with this

Starting point is 00:03:28 concept. I think it's really fun and one that a lot of people could apply to their own lives. walk us through what you've learned here and why you think it's interesting. Yeah, thanks, Patrick. So I only occasionally contribute something to the, I don't know, the blogosphere or whatever you call it. Sometimes I just keep thinking about something until at some point I feel like I have something useful to contribute. So several years back, I really started thinking about how some companies just move much more quickly than others. And I sort of kept thinking about why that is and how that is and what the implications of that were. I had been at Google for a lot of years.

Starting point is 00:04:01 And I thought of Google, at least at the time, as a company that even at significant scale, was able to move quickly. So speed as a habit really came down to trying to break down, what is it about how some companies do things that create speed? And can you habitually do that? And that's where the title comes from. But basically it came down to the idea that most of what we do in business is, number one, making decisions, and number two, acting on those decisions.

Starting point is 00:04:28 And then really as a sort of third thing, a lot of times you have to sort of get other there's other companies, other people on board with those decisions. And if you could do all three of those things faster habitually, you're going to just have more trips to the plate, more swings, whatever you want to describe it as. And that was the idea that went down to like how a company makes decisions. Some of the ideas in there have come out a little differently in terms of like one-way doors, two-way doors, the Bezos concept of decisions and whether they're revocable. I was just really into the effort of how you make decisions and which ones should take a long time or short time and just how you get things done. My belief was, particularly for founders,

Starting point is 00:05:06 but really for any type of executive, is you can begin to reinforce behaviors that will make your company move more quickly. And that was the heart of it. It was really thinking about, you know, maybe one of the most foundational ideas is the most important thing about a decision is to decide up front how long you're going to take to make that decision and who needs to be involved in it, which sounds like an obvious thing, but most people don't really think that way. And if you do that up front, you'll find you may not always make the right decision, but you make an expedient decision, and that's almost always better. So if there's room for improvement or faster speed at three levels, making the decisions,

Starting point is 00:05:42 and maybe there's even one before that, which is like knowing which decisions to spend any time on, but making the decisions once you know you have to make one, executing on them, and sort of convincing or getting others that matter outside the business to move at the same speed, I'd love to just pick apart each of those, I'll call it four stages, and just hear what you've learned. So I guess the first one is, how do you know that something is deserving of a decision in the first place? This seems like a triaging that has to happen before we get to the other three. What have you learned just in selecting your decisions in the first place? As a CEO of a company, your goal is not to make more decisions than you need to. You want to

Starting point is 00:06:17 build a company capable of operating where important things are done far below you in the org. And so that's the first thing is like, if I'm involved in every decision the company's making, And that's a bad sign. I should be breaking ties or helping to work some of the most important decisions that I have a unique perspective on. So that's the first thing. And then beyond that, it really comes down to like, we're all making decisions all day long. We just have to discern which ones are worth two minutes and which ones are worth two days and who should have input. And the trick as a leader, I think generally is to give people a voice, but don't try to create consensus because consensus is almost impossible if the decisions are even worth thinking about.

Starting point is 00:06:58 There's that rhyming phrase. I think it's to look around at the parks in all your cities, you'll see no statues to committees. So group decision making doesn't exactly work well. So if the first thing is making sure you're not the bottleneck and the second is sort of identifying how much time you're going to spend on a given decision, what have you learned about the execution layer of all this? Once you've made a decision, how to speed, then translate, into what the teams actually do. Another way of thinking about this is like, what produces the opposite?

Starting point is 00:07:27 What slows things down in your experience to an unacceptable degree? Execution often starts with meetings and discussions as well. As much as we all hate the concept of meetings and people love to bash meetings. It frankly is how a lot of things get done. But in meetings, you have to organize a plan and execute a plan. And oftentimes, there's not enough consideration to time. How many times I've sat in meetings where somebody would say, okay, here's the next two things or three things we're going to do. Thanks, let's go. And no one ever said,

Starting point is 00:07:56 when are we going to be done with those? How fast can we get that done so we can move on to the next thing? And having a habit of asking, when will this be done? Why can't it be done sooner? And not of everything. I think one of the important things you really have to know is what is on the critical path here. There's no sense us all worrying about something that, frankly, doesn't need to be done for two or three weeks or two months. you have to really just understand the nature of the project where the critical path is and just almost always challenge the when, challenge whether we can't do this sooner. When I was at Google, Larry Page was just adamant about this. Every time somebody would say, we've come up with an incredibly aggressive plan.

Starting point is 00:08:34 We thought this would take six months, but we think we can get this done in two months. And Larry would say, well, why can't you get it done in a week? No matter what you brought to him, he could one up you and just challenge you. And there's an art to it. You don't want to just be a pain in the butt and just pushing on things when people are already working hard. But there is an art to it. What would that reveal? So if Larry's asking, why not a week instead of two months, why is that a valuable question?

Starting point is 00:08:59 What's on the other side of that question that's valuable? Often there's hidden assumptions in that two months. There's something going on there that they assumed this can't happen until that is done. This project can't start until that has done. Larry or whomever might want to challenge that assumption. Why can't we do these things in parallel? I don't actually think we need to wait. And that simple thing of, why are we waiting for X to start Y?

Starting point is 00:09:22 I feel like it is just a repeated record in my history. I've heard it and said it a thousand times where people think serial when they need to at least try to think parallel. In the final stage of getting other companies with whom you're collaborating or selling to or whatever to move at your same pace, that seems like maybe the most challenging. Any interesting lessons there of getting people aligned? quicker. Yeah, I mean, the art of persuasion really can be challenging. This is part of what you're doing is negotiating. Part of it is just asking. It could be a vendor. It could be, you're just trying to move

Starting point is 00:09:55 the will of somebody else toward it's something you need. And I just find often you appeal to them as the rational, logical person that you are where they just can't do anything but agree with you. I mean, it sounds funny to say that. If they understand where you're coming from, what you need to do and why if they were you, they would be asking for the exact same thing you are. Do this place where they almost can't deny that you are a rational, reasonable, and pretty darn nice person, and they want to be helpful to you. And I think that's often, that's a little bit opposite, maybe the classic Steve Jobs, where he just berates vendors or things.

Starting point is 00:10:30 He's like, we know, we need this ship this week. But I think there's only so much you can get away with, and Apple can get away, or Steve Jobs could get away with things that others probably can. that I think there are ways to sort of move people in the direction you need them. Prior to Upstart, you ran large chunks of Google. What other lessons, apart from Larry's great question there, do you most take from your experience there that you've brought with you to Upstart as valuable ways of running the company or ways of thinking about product or what have you?

Starting point is 00:10:57 There's a lot to it. I would just say in many ways, we've exported a lot of what I, or we as a whole bunch of ex-Gougal people at Upstart have learned there and kind of like adjusted it to our world. and Google didn't do everything right. So there was a lot of where we're picking and choosing of what we like. But if there's any singular lesson, I think it came down to having the quality of people around you is the whole game. When I started at Upstart, beginning of 2004, it was before they were public. I walked in and just sat down and I was like, wow, these people work at Google.

Starting point is 00:11:28 This is so cool. And it was before 50,000 people worked at Google. And when I'd be in rooms and hearing the discussions and trying to contribute, I was just like, this is a smart. The brain density in this company was intense. It just struck me always that it's the thing you hear elsewhere. You know, one exceptional software engineer really can be worth 10x what an average software engineer is. And you can take this to an extreme where it is dysfunctional.

Starting point is 00:11:52 But I would just generally say, we've been successful at Upstart, if anything, not because we had the right game plan and not because we always made the right decisions, but ultimately because we had the right people around the table that helped us get eventually to the right answer. And I think your hiring practices, your people development practices, your mission, all that contribute to who you can bring in the company and keeping them in the company. And that is really central to success. One of my favorite things when talking to people like you that are both founder and CEO of a company and have been doing it for a while with success is to, yes, one, tell the story of the business, which we'll do here in a minute, but also use it as an excuse to talk about, like,

Starting point is 00:12:30 major fields and what building a business in those fields like artificial intelligence, We might talk a little bit about crypto, hard to avoid if you're talking about FinTech these days, and really pick it apart at like the major things driving change in business today. Before we do that, I want to set the stage for what Upstart is and how it got going. And maybe the first timeline marker I'm interested in is when did it first become like a twinkle in your eye? Like what was sort of like the founding insight or the first moment that you knew you wanted to do something in this space? And how did you get to there? The end of my time at Google, I had been there about eight years.

Starting point is 00:13:04 and I started what became Google Cloud. And I just kind of got this thing where I had built something fairly interesting. It was still a very small part of Google, but it was interesting and compelling. And at some point, you start to question, like, maybe anybody could have showed up at Google and built this. The company had so much mojo in those days that you sort of question whether it was you or whether it was just Google. And so I had a little bit of that. I want to prove I can do something beyond Google. I actually had somebody who worked for me, who mistakenly copied me,

Starting point is 00:13:34 on an email to somebody outside the company. And it was one of those moments. And the email said, he's done pretty well here at Google, but I don't think he'd survive on the outside. And of course, he felt terrible. And there was a rationale for the email, this and that. But it's just one of those little moments that stuck with me.

Starting point is 00:13:51 Like, really? I think I could do okay. It was that combined with just a lot of ideation around things I thought might be interesting. And Upstart became one of them. It was the idea of providing access to young people so they could do more interesting things, access to capital. In the original version of Upstart, you would actually not get a loan, but you would get something called an income share agreement,

Starting point is 00:14:13 where you would pay it back as a fraction of your future earnings. And income share agreements, which, but the only thing we really contributed to that in the long run is the name because we actually created that name way back in 2012. But in any case, that was the notion. And my co-founder had built a model to effectively predict what somebody would earn over their life, time based on a whole bunch of factors. And that became the foundation of upstart. So that was

Starting point is 00:14:39 pre-pivot, but that was sort of the founding notion. And for me, it was like, look, I had done well at Google. I had been there eight years. And I probably could have strapped in and gone another five or ten years and done well. But I just felt like it wasn't the last cool thing I was going to do with my life. I wanted to have that experience of starting a company. I had at least a couple different ideas I was compelled to investigate. And so I went through this several month period, still at Google, investigating, and then ultimately pulled the rip cord. What was it like in the early days approaching, in this case, a finance function lending or capital sourcing or whatever you want to call it? As an outsider, you didn't come up necessarily through the ranks of finance,

Starting point is 00:15:20 so you're kind of approaching this problem laterally or horizontally. How did you investigate? like what were you curious about? What were you learning? What seemed crazy? Like, I'm just always interested when people come at like a very traditional, there's no function as old as lending. It's been around forever. So what did it feel like to kind of play with that concept in the early days? Well, I think coming from a completely different industry with no background in something can either be an enormous asset or liability. I suppose it depends on the person. I guess I knew what I didn't know. I spent a couple of months just driving around in the city, meeting people, putting coins into meters, having lots of coffee, and just learning and asking about this thing

Starting point is 00:16:01 because this concept, and a lot of people would educate me, but they would also probably tell me they weren't that sure this idea makes a lot of sense, etc. So it was just like an experience, but I think ultimately, if you know what you don't know, you are a fast student and you can challenge a lot of the assumptions, but not be blind to the way the industry works. It can work to your favor to come completely from the upside, not to bring assumptions. When I found at Upstart, I was 45 years old. So I had plenty of times around the block, if you will, just not in that particular industry. And that was helpful. I knew we needed to hire a lawyer very quickly. We weren't going to make much progress without having legal expertise in the company. So just a lot of

Starting point is 00:16:44 things like that, that I think in the end were very helpful. What does it feel like in the early days in terms of the opportunity to build an improved product. So you're coming into a space that's well established. There's lots of ways you can get a loan for lots of different reasons. And presumably that means you need to offer something, if not radically different, just very different from incumbent solutions. You referenced the pivot. Maybe now's the time to talk about the early evolution of the product. But how did you figure that problem out? Because it seems like a really hard one to try to improve on something that's available in lots of other forms. Financial services, along with maybe healthcare are, you know, two of the most difficult industries to penetrate. And I'd just say for the

Starting point is 00:17:24 first 20 years of the internet, they were largely avoided by most entrepreneurs who'd rather go to less regulated industries. Certainly in the last 10 years, that's changed pretty radically. And when we got into the industry and said, hey, we can create a better loan product by using very sophisticated math and a lot of data, there were large players in the industry doing online lending, but they weren't really focused on the problem we were focused on. They were more just bringing the product online, making it available, making it a little more efficient because you could do it on the internet or on your phone, et cetera. We really wanted to look right at the heart of lending, which is the credit decision,

Starting point is 00:18:02 who gets a loan and for what price. And our basic assumption was if you could have a much more accurate model identifying who's likely to payback a loan and who isn't, some very good things would happen. And that's sort of proven itself up. At one level, it seems very obvious. If you know a lot more and you have more sophisticated risk models, some good things are going to happen. But for some reason, just enormous disbelief that it's possible to build a better mousetrap in lending. It's just viewed as the ultimate commodity subject to macroeconomic conditions, et cetera. So we just had a lot to overcome in terms of convincing the world. You could build a better mousetrap in lending. Can you describe credit score? And like, I realize I've never really thought about credit score before. What drives it? Why is it suboptimal as the sort of predictor of why someone will pay or won't pay, repay a loan? And then I really want to hear the full story of building an AI function like this or a data model function like this into a core product or business because I suspect we'll see this applied in a million other industries over the next couple decades. And so hearing sort of the blood, sweat, and tears that goes into that process in some detail will be great. But first, maybe credit score is something everyone's familiar with. What is the credit score? Why? Is it interesting or not? Where does it fall down?

Starting point is 00:19:17 It may be interesting to start just before the credit score. So the credit score was invented about 30 years ago, 1989-ish. And before that, I mean, there was nothing. If you were going to try to get a loan, you would sit down across the table from somebody, a loan officer at a bank, and they'd ask you a bunch of questions that have a bunch of rules that want to know where you work and what you paid for your mortgage, et cetera. And that obviously was a very bespoke process with all sorts of problems. problems of fairness, problems of accuracy, of performance. And when Fair Isaacs came out with the FICO score in

Starting point is 00:19:52 1989, suddenly there was this universal three-digit number that gave a sense of how credit-worthy you are. So at the time, it was a radical leap forward for any particular type of bank or lender trying to make a credit decision to actually have a number that means something based on your prior use of credit. And in a weird way, it became a crutch for the industry. 30 years later, it's still is the centerpiece of how credit decisions for consumers are made, whether that's a credit card, a mortgage. The government uses it to decide which mortgages can be sold to the government sponsored entities, Fannie Mae and Freddie Mac. And it just became encrusted into how the world of credit works. But when you think about it for a moment, a three-digit number is never going to

Starting point is 00:20:38 capture all the subtlety of a person and whether they would pay back a loan and when and what size loan and what type alone. So there's just a myriad of things that it naturally can't be encapsulated in a three-digit number. So we had a huge palette of opportunity to improve on that. In any data model exercise is kind of some common elements, and I'd love to understand them through the lens of upstart and what you were trying to build at the start and how that's evolved ever since. So if we think about the three-digit number as like some sort of predictor of whether or not people are good borrowers or not. And we know that there are certain, like, features or inputs that go into that credit score. I don't know them all, but there's some number of them.

Starting point is 00:21:17 Effectively, you need to replace that. So you need different features. You might need some sort of different outcome that you're targeting. And then obviously you need the algorithm in between those two things. So talk us through how you approach that problem. Like, what does it like to build a data model or an AI model? And what are the discrete steps? It comes down to data and algorithms. And what that really means is you go from a three-digit number to maybe 12 different variables or maybe 20 or 40 or 60. And the thing about AI or machine learning in general is the more data you have, and we like to think about rows and columns, the columns are like the things we know about every person applying for a loan. The rows are kind of like the repayment data of all the loans we've ever done over time. Every month, there's 40 or 50,000 repayments or delinquencies that happen.

Starting point is 00:22:05 And that's just a bit of training data. Now, what you have is machine learning algorithms that are interpreting that training data and then predicting what's going to happen to the next applicant. But there's a weird interaction between them where if you don't have enough training data, you can only use a fairly simplistic algorithm. Using more sophisticated algorithms requires more training data. So there's this codependent relationship on the two. that means, as we've grown over, been doing this now six or seven years, the amount of data we

Starting point is 00:22:37 collect about an individual has grown, the sort of rows of performance data have grown, and then we're constantly changing in and out the engine, swapping it in for more sophisticated models that create a more accurate model. And what our model very simply is trying to do is predict the cash flow for every month of a loan, which sounds simple, but actually it's quite different than anything that's been done in history, which is if you took out a three-year loan, there's 36 monthly payments, what's the chance that Patrick is going to either miss one of those 36 months in which one is it going to be or that he's going to actually pay back that loan early during one of those months? So time-based cash flows is the essence of what our core part

Starting point is 00:23:17 of our models predict. And that leads to a pretty dramatically more accurate system. What did you find early that was surprising in terms of data that mattered that maybe others weren't considering or things that everyone thought mattered but didn't at all? Or were just some early findings on the sort of information that could help you make these valuable predictions. Well, when we started early, based on where we had pivoted from, we were very interested in young borrowers post-college, people with little to no credit history, what in the industry they would call a thin file. So we kind of said, well, look, there's a lot of other data you could use to understand more about this person. For example, the level of education obtained or what they

Starting point is 00:23:56 studied in school because these things are predictive of, if not income, stability of income over time. So it seems self-obvious. And my co-frounder at the time, who was only 20 when we started the company, he would go out to a few websites and apply for loans. And he would get rejected. And I was like, wow, this guy went to Yale. He had a perfect SAT score. He was earning a six-figure income. He had zero debt. He was getting rejected all around because he didn't have three years of credit. And I was like, wow, I think one website gave them a 24% interest rate on a $10,000 loan. And I was like, wow. So it became obvious that there's more things you can know about somebody to inform the right price of a loan. And education was the thing we were known for in the early days.

Starting point is 00:24:43 And it was because we did start focused on younger people. But there's so many other things, the industry that you work in the company, you work for, just a myriad of other things. And that's the idea is we're trying to create a 360-degree view. view of the individual. But we aren't judgmental about the data in any way. We really are letting the software interpret performance. It's not that upstart thinks, this school is better than that school or that, this job's better than that job or anything else. It's really the software interpreting and learning based on the performance of the loans. How do you think about bias in a model, like things getting in there, which show up as quantitatively predictive, but for some reason or another,

Starting point is 00:25:22 you clearly don't want those as aspects of the model? The whole motion of AI bias far beyond the industry we're in more generally is something we've been in the center of since we started. We went to the appropriate regulator, the Consumer Financial Protection Bureau. Before we even launched as a company, a lot of people thought we were a little naive doing this, but we said, here's what we're trying to do and here's what we think we can achieve with the idea that more accurate model is actually going to improve credit outcomes for everybody, for every demographic. And over time, we worked on a process where we actually test every single applicant for bias. We provide that data to the CFPB, the preeminent regulator for

Starting point is 00:26:02 consumer protection, every quarter on behalf of all of our banks that work in our system. So I guess maybe that's a long-winded way of saying, if you're worried about AI bias, the right answer is test all outcomes rigorously. We are set up such that if somehow our model moved in a way that it was biased against any particular demographic, we could revert. we could go backwards and avoid that. But fortunately, what we've seen is our algorithms actually increase, improve credit outcomes for every single demographic we can name. Race, gender, ethnic origin, et cetera, age. And that's really powerful. And regulators care about that. Higher approval rates, lower interest rates for every demographic. And that's a powerful combination.

Starting point is 00:26:48 So we haven't talked at all about interest rates, a critical variable in all of this for a lot of reasons. Obviously, all things equal, a borrower wants a lower interest rate. Competing on a lower rate is a really good way probably to win business or acquire a customer in lending. Also, you need to charge a rate that's high enough for you to have a good business. Teach me everything you've learned about interest rates and what you found surprising and sort of how it figures into your business model. Yeah, so we're kind of a technology player in the middle. We market to consumers. They enter a funnel. We essentially refer them to one of our bank partners. that fund the loans. So we're not a lender ourselves. We're not a bank. We sit in the middle as the

Starting point is 00:27:27 technology provider, but first and foremost, rates matter a lot. Our whole business tends to grow when our funnel gets more efficient. That means for every person who requests a rate might want a loan, how many convert into loans. In our primary way we grow, as the models get smarter, we can improve more people with lower interest rates, and we also can remove as much of the friction as possible for someone who says, yes, I want that loan. Today, about 70% of our loans, there's no human intervention at all. It's literally lights out processing,

Starting point is 00:28:01 which, by the way, it was very much inspired by Google, to go back to Google again. When I was interviewing at Google in 2003, the recruiter said to me, you can't come in an interview this week because they're all at Lake Tahoe skiing. And I was like, wow, the entire company went to Lake Tahoe skiing.

Starting point is 00:28:17 She said, but the funny thing is, they're still making like $7 or $8 million a day in revenue. And I thought, wow, that's a great business model. We've always wanted to have the type of business that can generate a lot of revenue isn't very human capital intensive. So now maybe you could describe the actual business model itself, what literally is happening today in 2021 and who the stakeholders are. So you mentioned that you're a middle layer between the consumer and the banks. You're not doing the actual lending off your own balance sheet. I'd be curious why. Why not? Just describe maybe like a revenue event. So what is it that causes the cash register to ring for Upstart? And what's the

Starting point is 00:28:56 timeline around that ringing? We are kind of a two-sided business where we market to consumers on one side. We spend a lot of money, tens of million dollars every month to bring consumers to our platform, running them through our models and then see which bank offers them the best loan. And we refer them to that bank. Almost all of our revenue, something in the range of 97% of our revenue, is fees that banks pay us. It can be number one. for referring somebody to them, a referral fee. Number two, what we would call platform fee, which means the cost of originating the loan on their behalf. And then we service loans for them as well in their name. Those three fees make up almost the entirety of our revenue. The reason behind that

Starting point is 00:29:37 is the first few years, we really just worked with one small bank as we refined in this process. And then as the industry was developing, it became clear that most of the participants were going to actually pursue bank charters and become banks. And that was a viable path. Not when we wanted to pursue, we just didn't see ourselves fitting the model of a bank. So our view was generally, look, we can be a partner to banks. We can bring our technology. We can leverage their balance sheets and their desire to have a better consumer offering. That's really what we've done. We now have 25 plus banks on our platform. They are the ones actually originating funding the loans. They actually get to have all the knobs and the dials of the types of loans they want, who they want to

Starting point is 00:30:18 approve what their return targets are. So it's not a blind system at all. It's actually a system where they have a lot of dials and controls and what we like to call nerd knobs of exactly type of credit they want. So they've got a control over it, but it's our technology underneath. So if I think about it from the bank's perspective, then I want to come back to the customer journey or feel. But if I think about it from the bank's perspective, what's valuable is you're effectively probably reducing my loss ratio. If I work with you, I'm going to have better prediction of who's going to repay or not repay or prepay or whatever, all those things you described. And because I have that intelligence,

Starting point is 00:30:54 I can probably charge a more competitive price. I'm sure it's oversimplified. But generally, from the bank's perspective, is that sort of the value proposition? Yeah. Think of it as knobs. You can turn either aggressively or non-aggressively. If you want to just keep your approval rates the same, we can knock out sometimes in the range of 75% of your losses, which is pretty awesome. But that's kind of the most conservative way to turn the dials. Another way to turn them is to say we can improve two or even three times as many people and keep your loss rate constant from what it is today. Banks will land somewhere in the middle. But the ultimate value prop is you can have a more inclusive lending program. You can improve more of your customers or new potential customers

Starting point is 00:31:36 at lower rates, and you can be more profitable. So what's not to like about that? And so if we skip to the other side and think about this from the perspective of a customer, I love the story about your 20-year-old co-founder, unable to get a loan. Who is the average customer today? And what does the journey feel like to them? Like from the time they hear that Upstart exists as a company through to getting a loan, what is sort of the felt experience?

Starting point is 00:31:58 The median age of our borrowers is probably the late 20s. It still trends toward younger relative to the American demographic. We serve everybody. Clearly, it's broadened. And personal loans is our core product. We're just expanded into refinancing auto loans. So we're kind of expanding the types of loans we support. But it still trends towards young people online. They tend to be relatively highly educated. They're not entirely. And the thing I think that shocks people about us is they come to upstart, they can get a rate in a few seconds. And then if they like it, most of them are approved literally in the moment, no documents to upload, no phone calls to have or anything of that nature. So people are

Starting point is 00:32:39 sometimes just shocked that it's actually that easy to get a 10 or 15, $20,000 loan. But that's part of the magic is just the beauty of lending is there's not that much to it. There's price and there's experience. The price is obvious. The experience, it just means making it easier, removing all the friction and reducing the time and the weight and all that. And those are the two pillars of our value prop as a company. And fortunately, they're very intrinsically tied to artificial intelligence and what that technology can do. Is it fair to say that if I think about the compounding asset that you are trying to foster or build it upstart, is this core data model and data set that you're just every year going to get better than the next relevant competitor at knowing something

Starting point is 00:33:22 about the consumer in question here. And that with that knowledge, you become almost like a platform. Like, if you've got that insight on a customer, that insight can be used right now to make loans, but potentially for lots of other financial services, too. So do I have that right? And if it is right, like, how do you think about intentionally compounding that value? Because I assume that's, you know, where a lot of your competitive advantage comes from. They are centralized artificial intelligence model. So every consumer that gets a loan, every bank that works with us is both contributing to and benefiting from these centralized AI model. So we do get this advantage because the model with more data, more experience is the best one.

Starting point is 00:34:03 Now, the other part is we have to teach it new tricks. We started with unsecured lending, which is a very simple form of credit. Now we've moved into auto lending, which is a secured form of credit, right? You have the consumer to underwrite, but you all of a sudden have this asset that's back in the loan, which in this case is the car. So now we've sort of taught the model new tricks. And instead of being backed by an automobile, it might be a home. It might be the cash flows of a small business. It might be a piece of heavy equipment. So our idea is generally centralized models that are learning as quickly as possible and then taking side steps to learn new tricks. We're suddenly almost any type of lending in the world and potentially domains beyond lending, we have significant

Starting point is 00:34:45 advantages in. And I think that's the heart of what we're building is an AI model that can properly price, almost any flavor of credit, just gets a little bit better at doing that every month, every week, every month. And that's an enormous opportunity. Not easy to navigate, a lot of things you could do wrong, but I think with strong execution, it's almost unlimited addressable opportunity there. What has been the hardest hurdle that you've had to clear around this concept of building this core AI model? So in the entire upstart journey, what episode that was useful or required to move the ball forward in this area, would you least want to go back and do again? What was the most challenging? I think the hardest part has been convincing the world that

Starting point is 00:35:31 AI can be fair, and in fact, more fair than the credit score system it's eventually replacing. There's almost this assumption where if it's a FICO score, if that's not fair, well, that's just what it is. We all use FICO scores. But the Fed put out this study that 30% of black Americans have FICO scores in the lowest decile, the lowest 10% of Americans. So that itself says, wow, is this really the right way we're going to allocate credit? There's so much sort of fear and loathing about AI and technology in general that to win that mind share and convince people, AI can actually be an equalizer is a really big challenge. We invest in enormous amounts of resources in doing that every day right now.

Starting point is 00:36:17 My co-founder, probably talking to the CFPB multiple times a week, and not just with the regulator. We're talking about lawmakers. We're talking about consumer advocate groups of different types. It's just a constant journey. I feel like ultimately, to go back to your question, yeah, I certainly would have loved to not have invested all that time and energy. But at the same time, I think we've built a great moat.

Starting point is 00:36:40 I mean, I think ultimately our business and our AI is very defensible in terms of its very pro-consumer. It is helping people get into the banking system that we're not otherwise in the banking system, and that's ultimately our goal. Say a bit more about the entire fintech ecosystem in which you've operated and how you've seen it change from the start of Upstart through to today. So if you take us back then, what did it feel like? How many companies were there? You know, was it impressive? And how much has that changed through to today? You're operating in one specific area of fintech and there's lots of applications that are being built. But what does it felt like to sort of be in that ecosystem to you? I feel like for a long time, we operated almost

Starting point is 00:37:20 invisibly because we were in a sector that prior to say 2016, the idea of online lending, there were some companies that went public and then suddenly they didn't do so well. So the particular area we were in got out of favor. But at the same time, you had people like Stripe and Square who were just suddenly getting all the attention and raising insane amounts of money as private companies. And so FinTech as a sector overall just really started to take off. I think if you sort of fast forward to today, you have a bit of like a bifurcation of types of companies. There's a whole move toward the super app, the one app to rule them all. Revolutes trying to do that in Europe. And you have several companies. Robin Hood seems to be headed in that way. You have

Starting point is 00:38:04 Square. You have Sofi. So there's all these efforts to do that. And we've been sort of, I just say, the contrarian model to that, very, very focused on lending. It's been a little hard to convince the world that, no, there's so much value in lending. I mean, it is the center beam of revenue and profits in financial services. So there's absolutely nothing wrong with pouring all your energy into creating a transformative lending product. In some sense, we've had to fight an uphill battle because there's just fascination with notion of super apps or other types of experiences. What does that overall pie look like? So if you say lending and the associate revenue is sort of the centerpiece of financial

Starting point is 00:38:43 services revenue overall, if we had just a big pie chart in front of us on financial services revenue, how is that allocated today? It's hard to say where you draw the bounds on financial services, just to include insurance companies, things of that nature. So it's a little hard to say. The thing I will say that is important is if you looked at the net interest income earned in lending in the U.S., it is comparable to all profits in the technology industry. Now, these aren't exactly apples to apples. You're comparing net interest income to a bottom line

Starting point is 00:39:11 profit number. But the scale of it is so vast that people fail to appreciate how big the concept of lending is. But of course, it's why banks exist. Banks were formed as an entity to take deposits and make loans. And unsurprisingly, it's still the main thing they do. And I'd further say, if you don't want to do that, you probably don't want to be a bank. So it's just so vast that from our point of view, if you can actually make a significant improvement there, not 10 basis points, but 100 percentage points, the opportunity there is so vast. And that's why we're focused on it. I keep going back to Google. But it reminds me in the early days of Google, search was considered a commodity and all the sort of portals at the day were adding all these other features and just

Starting point is 00:39:54 thought of search as a checkbox. And then Google showed out and said, no, actually, that's where all the money is. And suddenly, this company that was maniacally focused at that time on in search suddenly just won the day. We think there's something to that. I mean, payments are obviously enormously important. There's so many giant companies being created in the payments sector. But lending is really where most of the profits are made. So it seems natural to us that there's great opportunity there. Why are there more profits in lending than payments? Well, I'd say because there's more risk in unknowns. I mean, payments in some sense tend to be somewhat commoditized in the sense of it's a volume game. And if you look at the business models of

Starting point is 00:40:38 Square or Stripe, it's a very commoditized business, which of course means you want to add on other services around it. But it's so vast, of course, that you can build obviously giant, very valuable companies that are centered on payments. But lending is what just feels like this giant bespoke scary thing. I remember before a couple of years ago, Jamie Diamond said something like, He was just sensing that the economy was in a precarious place. This was a year or so before COVID. It wasn't really core related. But he said something to the effect of, I'd be pretty happy if our loan officers are playing

Starting point is 00:41:08 a lot of golf right now instead of making loans. It just struck me is that is the state of lending, which is he's looking at his crystal ball, he's nervous, and he wants to stop making loans. Then we clearly have a big opportunity to make some improvements here. So now we have to talk because of your focus on lending about just lending the function itself and the ways that loans are made, like what are the major types of loans by volume or by type, and how that might change in the future. The whole by now, pay later thing has become a hugely popular concept, even though it's kind of an old concept. How would you teach a little high-level

Starting point is 00:41:43 class on the types of loans that exist today? And then I'd also like to talk about how that might change in the future and what services like yours might unlock in terms of new types of loans and why that's good. Yeah, maybe I'll just speak to what the big categories are and then how they're shifting. So I think they are in motion right now. Starting from what's known to be the smallest is what we do, personal lending, $100 billion or so annually. COVID has disrupted a lot, but it's not the largest category. It was for many years pre-COVID growing faster than any other category because a simple loan you can get in a few minutes that you can use for almost anything, pay for a wedding, pay off your credit cards, what have you, relocate to a new city. It's just as high consumer utility, but most banks had not

Starting point is 00:42:27 offered them historically because they couldn't do it economically. But it's a small fast-growing sector. Moving from there, I mean, something like auto, we're moving into five or six times, 700 billion in originations. Most people who have cars have car loans or leases. It's a much, much larger category that's been around a long time. You have things like home equity and things of that nature. But the big one in consumer lending, of course, is mortgages. It's another order of magnitude larger. And it has all sorts of history and disruptions to it, as we all know, and credit cards and student loans are in there as well in the sort of unsecured lending. It's a vast category.

Starting point is 00:43:04 I would say a couple of things that are happening that are interesting. One is that unsecured lending is taking a lot of market share away from secured lending. People are using unsecured loans to redo their kitchen. Instead of trying to get a home equity loan, which can take you two months, I'll get a $10,000 loan in a matter of a day. And I may pay a little bit more interest, but nothing compared to the two months I'd wait to get a home equity loan done. They're also taking market share away from, as you mentioned, credit cards.

Starting point is 00:43:34 Buy Now, Pay Later is really a small installment loan. That suddenly, for various reasons, is appealing to consumers. And so you can think of unsecured loans that are just underwriting the individual, no collateral, as gobbling market share because of convenience and because of increasing ability to price them properly and to issue them economically. So I'd just say those are some of the trends out there. Consumers, of course, have zero tolerance for waiting, zero tolerance for like requesting documents and conversations and this and that. So if you can solve for that, a couple percentage points of the interest rate, one way the other doesn't mean all that much. Do you think that this whole buy-now pay later concept is a major business threat to some of the really powerful incumbent like payment rails type companies given how popular it's become? It seems to me like almost anything online these days.

Starting point is 00:44:24 You can rather than pay it up front, you get like a zero percent APR. So I'm curious why that's the case. And paying installments over time, which can, I understand, circumvent visa and MasterCard and companies like this. How do you think about this as just like a competitive new field of gravity for, payment slash lending. I would say my perspective on it is changing by the day as I kind of learn more about what these companies are doing. But sometimes I think of it is, is it a threat to like banks lending through credit cards or otherwise? I certainly think it can be. I think, first of all, the strong network effect of Visa and MasterCard has been around a long time and just unbreakable, frankly, the toll

Starting point is 00:45:04 they take is significant and there's just been nothing because in order to do so, you need to have both presence with merchants and presence with consumers. And it's just been almost impossible for any entity to accomplish that. But the buy now pay later players, particularly when you think about one of the largest of them after pay, being in the hands of Square now, since Square acquired them, suddenly you have a two-sided network that actually can really put some stresses onto that Visa MasterCard model. I think there's a lot there. It is not so much about lending in that case as it is about just payments in sort of creating a payment structure that consumers like that does not go through the Visa MasterCard rails and potentially very threatening to them. Now, maybe one of the

Starting point is 00:45:50 open questions is, are merchants actually paying more for these services than they're actually paying for Visa MasterCard and how will that play out? I think there's some open questions about how the industry evolves, whether there's pricing power, commoditization, et cetera. But I think for sure Visa MasterCard have reasons to be concerned. So if we go back to your role in this whole ecosystem and this unfolding story and you're sort of just reliably every year going to have a more predictive, more useful platform data model that others can build on top of can lend using your bank partners, for example, how do you think about five years out or 10 years out, whatever the appropriate distance in the future for a radically different future that your company enables? Like if things are the most different in terms of how people borrow money or whatever 10 years from now, and you've been a major part of that story, how do you think that plays out? From the consumer perspective, at a moment's notice, you have the very best offer of credit

Starting point is 00:46:47 you have for any particular circumstance you find yourself in. Whether you're considering a purchase, a small purchase or a large purchase or up to a home, there is no process. It is just an entity associated with you. It knows you and you just get bits. I always picture it as these numbers just floating over your head, which is exactly what you can get at any moment. There's no process to go through. And it's adjusting to the world.

Starting point is 00:47:10 It's adjusting to the reality of your financial life. And that to me is from the consumer perspective, where it has to get to? It's just maybe self-obvious that this is not that far off. And I think generally one of the things my co-founder also says is 90% of the interest paid in this world is entirely unnecessary. It's because of the obtuseness of risk models. When you think about from the other side, from a lender, just the price of credit, there's so much opportunity to just dramatically reduce the amount of interest paid to borrow money with just simpler, better, smarter models and less friction, which translates into cost.

Starting point is 00:47:50 So we do think of it from the consumer perspective. We think of it from the lender perspective. And it's just such a crazy, inefficient industry. Sometimes we think about Renaissance technology coming up with these trading algorithms, trading commodities, trading stocks or bonds, et cetera, trying to shave off three basis points. And we kind of think it's laughable because we're working at something that has two orders of magnitude, more inefficiency in it. So it's that much of an opportunity. Just touch a little bit more on the inefficiencies that still exist today. Certainly understand them when you got started six, seven years ago. But what today stands out as the single most or a couple most inefficient parts of this. process that you guys can trip away at? The center most piece of it is that people with low FICO scores are not necessarily bad credits. There's just so many strange reasons why somebody ends up with a 580 or a 620 FICO score and the vast majority of banks are out there will just say, hey, if you're below 680, we just don't take that risk. And it is literally, as we discussed,

Starting point is 00:48:51 a 30-year-old notion that a three-digit number can represent you. So we've basically said, look, everyone operates on a continuum of risk. And we can very particularly measure that risk. And that opportunity is so vast because half a more of Americans are completely left out of the system, meaning a bank would not offer them a loan. Or if they offer them a credit card, it would be a terrible rate. And the strange part is, even people with very high FICO scores, you might have a 750 or an 800 and think you're getting a great deal. But guess what? You're actually subsidizing the losses elsewhere in the system. So you're actually paying too much as well. So when there's inefficiencies that are that rampant, there's just a lot of room to fix.

Starting point is 00:49:33 And we don't have to be perfect, by the way. No system is perfect. No model is perfect. We just have to be better. And we are, without question, a lot better than the status quo. When we zoom out a little bit away from the actual job being done, but more just generally into building this business, back to our opening idea of lots of decisions made efficiently and with some speed, what have been the most important decisions that you've made in terms

Starting point is 00:49:55 of deciding what kind of company it's going to be, how you're going to operate. Like, what are the big heavyweight or freight decisions that you've made that you're proud of? I kind of started with the notion all by myself, and then I happened into my two co-founders, Paul and Anna. I mean, we couldn't be more different. It turns out I'm older than Paul's parents. I discovered that in my first week knowing him, I discovered I'm older than his parents. And Anna was a lot younger than me, and we soon to have her first child.

Starting point is 00:50:22 So we just came from these, they're both immigrants. I'm not an immigrant. So just three people who happened to like the same idea came together. And almost a miracle that it worked. But one of the enormous strengths of upstart, first of all, is that we still have the three founders day-to-day active, as well as a lot of the exec team has been with us since the beginning. We survived a brutal pivot. I mean, this beautiful idea of an income share agreement, so perfect on the whiteboard,

Starting point is 00:50:47 came up against the cruel reality of the world, which essentially decided it was a niche product. So going through that pivot, keeping the entire company on board to do that, I think was a huge proof point for us. I don't think we ever were ready to quit. I think we just felt that we were a good team. We were going to figure this out. And suddenly we found our way into our market that there was an opportunity for us. So I think that was a big moment for us. I think when we decided we're not going to become a bank, we're going to serve banks, which was suddenly changing the nature of our product pretty dramatically to think about having one bank originating all the loans and suddenly having many, many banks from a technical perspective, it was suddenly a big redo. So a lot of challenges

Starting point is 00:51:33 like that. Decided to go from one office, decided to go from the Bay Area only to open up Columbus, Ohio, which was a very big deal for us at the time. So just things of that nature. But most importantly, is really having a team that's been around the block together a lot, no matter what hits us in the face, and there's something new that hits us in the face probably every week. It is really comforting to know you have a team that has already worked through a lot of things together. If I was going to build a business that relied on selling to banks, what should I know? Raise a lot of money because you're going to need a lot of time. I mean, selling into banks is sort of the technology selling cycle in its worst possible form because banks are amongst the most regulated and conservative industries. We

Starting point is 00:52:17 weren't selling a chat app to put on your website or some sort of new thing for your data center. We're selling something right at the core, an approach to lending. But we've built a team and we've said, we're going to do this right. We're going to prove it over time. We're going to work not just with banks, but with regulators. It's been a journey, but again, I think we're getting there and the adoption rate. And it goes back to cloud computing or anything else that's brand new. There's a technology adoption curve. And you have to both respect it and then move it along it as quickly as you can. Dan. You mentioned this idea that a great outcome is that you and the two co-founders are still running the business and a key part of the business. How does it feel interacting with CEOs of

Starting point is 00:52:58 maybe public companies or just big company CEOs that were not the founder versus the experience of being both the founder and CEO? Because one of the very popular investing tropes these days is that the best performing companies are often founder led. And the data is sort of mixed on this, but certainly, you know, a lot of the biggest companies in the world were built by the founder for a long time. Draw that contrast for us interacting with hired CEOs versus you and other founder CEOs. I have a lot of friends who are CEOs who are hired in by the boards to replace a founder or something of that nature. So I haven't experienced that side myself, but I have good friends who have. And I always kind of laugh that I have such advantages over them. I mean, inevitably,

Starting point is 00:53:39 if you're hired by the board, there's a lot of issues there, first of all. They didn't hire you for no reason. There's some problems to fix and some issues. Oftentimes, you do have a founder who's still very involved, just enough to be disruptive to you and making it hard for you to operate with authority. And also the board brought you in. So the board wants to know how you're doing. They want to see how they did on that hire and they're going to be very active. Whereas when you're the founder, I mean, a founder of a floundering company is necessarily a great place to be, of course. But once you've achieved some success and you're on the right path, you have a lot of authority. I mean, you can easily say when we pivoted, which again, we hadn't

Starting point is 00:54:16 really found success yet. It was very easy for the three founders to say, we're going in this new direction. End of story. I think just the advantages I have and the clarity with which I, along with my co-founders, can make decisions and change directions. Even in this COVID world, we thought we were going to go one way. And then three months later, we completely changed our mind. And I think if I had been a hired CEO, they would have been like, what the hell is he doing? Get your act straight. As the founder, I'm like, no, no, no, I know this company. It is me. And I think you just, have enormous advantages. So sometimes when I look at public companies, I do look through that lens. I think DoorDash, for example, founder-led has incredible advantages. Uber, a little less so.

Starting point is 00:54:57 I think Dara's an amazing CEO, but it's not the founder. And I think there is some dynamic there that is really important. Obviously, we cannot talk about lending and evolving financial services without talking a little bit about decentralized finance and the role that crypto might play in all of these functions. So if one of the cool things that crypto enables is this, incredible speed, always on, public wallets that you could sort of, speaking of rich data, you could read into and see what assets exist and that may proliferate over time. How do you operate in parallel to this crazy fervor going on in Defi? Do you view it as a threat?

Starting point is 00:55:32 Do you view it as an opportunity, a bit of both? How do you view it? Well, I'd say, first of all, the two most compelling areas, particularly in FinTech, certainly are crypto and AI. So we're in one of the two, which I would take us the first. That is a good thing. The attraction of crypto to a lot of the talent in the world is so strong today that we do feel like it can be a bit of a liability that we're not a crypto company today. And not just about recruiting talent, but the way the industry might just go. We are crypto believers.

Starting point is 00:56:04 I think we will certainly look and understand things over time. There is a relationship there. I mean, first of all, the way I think about it is crypto is recording the past immutably. and AI is about predicting the future. If you have an immutable, clear, unambiguous record of the past, you can be much better at predicting the future. So the two are actually very related, and I think better products can be built that actually combine the two of them. So there's a lot of opportunities.

Starting point is 00:56:33 Credit reporting agencies could go the way of crypto and distributed. There's just all sorts of notions of reputation in pseudonymity, if that's the right word, in the cloud that I think could be very powerful. I think creating products that cross borders, it's very hard to build financial products in lots of countries because the rules are different. The infrastructure is different. And crypto certainly at least holds up the possibility of doing things in a more global way. How do you think about other areas that most have your attention or even excitement in the fintech world writ large? So you have a privileged seat in this space. You sort of get to see what's going on probably before others do and make decisions as to

Starting point is 00:57:12 or not you're going to get involved or not. Are there any other big themes, either outside or even within crypto and AI, that you think demand people's attention if they're interested in fintech today? Because we're such an AI company, we look at the world through that lens. And I just say, when I feel like AI and the risk models, it enables unlock an incredibly different consumer experience, that's when we get excited. I mean, risk models sound like this wonky thing that the risk committee and a bank would care about. But in reality, a model that allows you to approve somebody in a moment for a loan that would otherwise have taken days, that's what it's all about. It is really creating these breakthrough experiences where the underlying problem to be solved

Starting point is 00:57:53 is risk-related. That, to me, is just exciting stuff. So there's certainly other areas. I mean, insurance is, of course, it's like the other side of the coin from lending. It's an entirely risk-based industry. There's some challenges there in terms of the nature of the business. models and insurance companies. But I think we just definitely look and we say, look, we're in a couple of sectors today. We're going to probably expand to some more. And it would be easy to just gobble market share, just do as much as we can here. But we feel like we have to build a company with ambition enough to plant flags everywhere. I mean, we have such an advantage and it comes down to can we take this outside the U.S. when and how? Can we take it to more or even all flavors of

Starting point is 00:58:34 lending when and how. So it comes down to companies just trying to have sufficient ambition with also some reality basis to all of it of what you can accomplish. I think risk in predicting the future may sound a little crazy, but that's the heart of what we do. AI is about predicting the future, and I think the opportunity for that to create mind-boggling experiences is unlimited. Is there anything else about the state of the art of applied AI that you think people would find interesting, abstracted away from what you're doing, just very generally, I saw this insane demand curve for GPU, basically renting GPUs to train artificial intelligence models. And it was doubling like every three or four months or something.

Starting point is 00:59:17 Like, it was insane. And that's like nationwide. And that tells me that we're about to see an explosion of this concept of predictive model applied everywhere that it can be applied. Is there anything that you find interesting just about what's going on in the world of AI generally speaking that would surprise people? I always find that there's these new different applications than I ever would have imagined. And there was one I remember, this is a bit ago, but this might have been some Google research or something, but just the idea of building a memory chip, where you store

Starting point is 00:59:46 what information and how fast it can be retrieved. It's sort of the heart. I actually studied computer engineering. So it's this world I grew up in cash memories and first level cash, second level, all this kind of stuff. And you kind of think, well, what does that have to do with AI? It's fascinating to think you can actually use AI to predict where you should store a piece of information so that it will be optimally retrieved in ways that you just never would have thought of that as an AI problem, least I wouldn't have. Every time you sort of dig in, you kind of go, how could this be made better if you could predict what was going to happen next? That's a simple question to ask. And just find, as much as AI feels like autonomous driving or language translation or these obvious categories that

Starting point is 01:00:25 are large, in almost every application, if you could predict the future, you could make it better. That's maybe an obvious statement. And I just keep discovering more and more every day. I know it's not been a long time, but I'm very curious what you've learned or discovered about being a public market CEO versus a private market CEO. What is the most different? Did anything surprise you? The process of going public has become this huge issue. People are interested in in SPACs versus IPOs versus direct listing. Curious kind of your experience navigating that decision, yet another decision to talk about. And then just more generally, what it's felt like being public relative to being private. We went public in December 2020, so mid-COVID, I guess is maybe the way to describe it.

Starting point is 01:01:05 And we would have had COVID not happened, probably would have tried to go public earlier in 2020. But in short, the decision was, I don't think the private market ever really understood us or maybe just didn't like us. I don't know. Raising money as a private company was always hard. Our numbers were getting better. They looked really great. But as long as it had lending fintech tied to it, there was just so much skepticism. We finally said, I think the public markets appreciate growth and profits and these kind of things. So let's go public and we can prove it in the public markets over time that we actually are a unique and differentiated business. So that was sort of let us to go. But strangely, we were a traditional IPO. I did the whole thing right from this room I'm sitting in right now. I mean,

Starting point is 01:01:46 I didn't go anywhere. The entire thing was done in my spare bedroom on Zoom calls with the banks to prepare and the board and IPO itself and even post-IPO now. Of course, you have earnings calls and you have to do callbacks with analysts and investors after the calls and you go to these bank conferences and such. It's all on Zoom. So for me, strangely, it's like different Zoom calls. You know, it's all I can say. I think the world would be very different if it weren't for this kind of pandemic situation. It would feel very different. But I do invest a lot more time and need to invest a lot more time with analysts and investors, then it does not feel normal to me to do that. It feels like a little unusual just to keep repeating the story, telling the story, answering the same

Starting point is 01:02:30 questions over and over again. But I think it's just clearly part of the job to bring the market along and help them understand the company. What do you think that public market investors get or are focused on that private investors didn't? That's a kind of interesting comment that there's been a crossover of understanding as you've gone public versus private. What was that gap? the public investors aren't making long-term bets. They're buying something they could decide to sell the next day. So in some sense, the venture community is just looking for sort of like, I don't know, a notion that they saw before everybody else.

Starting point is 01:03:03 And we weren't the first in our sector. We weren't the first to say we were doing AI or this and that, difficulty in discerning real from fake. And so I just think it was hard for the private markets to understand whether we were really doing something different. Maybe it was also timing. We were early. We were not the largest in our industry. And once we decided to go public, I think we had the good fortune of the business really kicked in. The AI models mature. They were

Starting point is 01:03:27 just at this hockey stick like moment of how they worked. And that led to just a lot of growth and profitable growth, which in the fintech world is pretty unique. But I think the public market is much more willing to just look at that and understand it. And by the way, they can trade out of it later if they want. I didn't quite appreciate the difference between a public investor and a private investor in the sense of the public investors can change their mind at any time. Are there any other battle scars or major lessons from long-term company building that we kind of haven't talked about that you think are important to understanding the journey that you and your co-founders have been on? The part that we did struggle with and I'm sort of referencing in here was the sort of capital

Starting point is 01:04:06 raising part. At certain times the business was unclear which direction was going. It was pivoting, etc. But we never got through all our private rounds, except for our seed round, we never got more than a single viable lead term sheet. You hear these stories in Silicon Valley, people writing blog posts about how to deal with 10 different term sheets and evaluating which one you should take. And I just laugh at those. I'm like, I didn't have that problem. It's not that I go, wow, all these VCs were being really blind or stupid and not offering us term sheets. I tend to think, what did we do wrong? I'm like, we were not telling the story in a way that made sense. And I think maybe we just didn't focus enough. We were too inward looking in terms of

Starting point is 01:04:46 telling our story. We really needed to tell the story from industry perspective in a way that I think we would have been better understood. And that's the hard part. I always thought I was a natural company builder. I was not a natural fundraiser. So it was always like, oh, man, I got to do this again. So I had all the incentives in the world to become cash flow positive and to not need to necessarily go on the Sandhill circuit again. Well, Dave, this has been so much fun. I think this space is like quietly massive, the sheer size of it. I like the comparison of interest expense against all technology profits.

Starting point is 01:05:20 It's like such a stark way to think about this. It's a huge space. And I think what you're doing is reducing friction drastically. And we just know from the history of tech, when frictions go away, behaviors change in nonlinear ways and in really cool ways. And so it will be fun to see how Upstart continues to affect the ecosystem. I ask everybody on the podcast the same closing question. What is the kindest thing that anyone's ever done for you? So many things, but I sort of have to go back to my youth. I grew up in a family

Starting point is 01:05:47 with six kids, a very small house with one bathroom, and zero net worth as a family, paycheck to paycheck. And I have to go back to my dad. I mean, my dad, of course, worked 50 hours a week at his main job. And even when he had his two weeks off, he would go get another job as my family would drive to Hampton Beach, New Hampshire and go on vacation. My dad would literally work those extra two weeks. So I grew up in a family where I certainly didn't have a lot, but I never felt poor.

Starting point is 01:06:14 We weren't really poor. And it was just really through someone just committing themselves to making a good life for his family for our family. And so can go no further than that. Wonderful way to end it. Dave, thanks so much for your time. I've learned a ton. Can't wait to watch the business grow.

Starting point is 01:06:29 Thanks, Patrick. If you enjoyed this episode, check out join colossus.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.

Invest Like the Best with Patrick O'Shaughnessy - Dave Girouard - Making Better Loans - [Founder’s Field Guide, EP. 52]

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.