The Exchanges

Q kind of you. As I was saying, I did not start quite like this. Um, I would love to start though on a little bit of like the re-founding moment because Platter's had many iterations over, you know, the years. And in the last two to three years, it would seem like there's been like a re-founding, latest chapter. Can you just take me to that and how that's evolved?

A For those that, that, that aren't aware, um, in twenty-twenty, uh, we signed paperwork to sell the company to Visa. Intervening period, uh, the, the, the regulators in the U.S. had investigated it to say that, uh, to kind of ask the question, was Visa a monopolist, and was this a kind of monopoly-creating transaction? And the amazing opportunity we had within the business was we'd grown quite massively from twenty-twenty to twenty-twenty-one. Uh, COVID had, had been a huge tailwind for us, and we had the opportunity to actually, uh, walk away at the end of that deal, and, and so, uh, elected to do so. Um, but no, we started working on the new product expansions before that acquisition.

AI assessment note: “we had the opportunity to actually, uh, walk away at the end of that deal”

Q In terms of the problems that they're pursuing, we chatted before, and you said to me about a brilliant Style of problem that you like called the grinder problem. But, um, I just wanted to ask, what is a grinder problem, and why do you find them attractive to do as a company?

A The early history of Plaid was, um, we needed to integrate with all the banks. And these integrations were not like, you know, flip a switch, magical integrations on. It's like a lot of work, uh, to actually set up integration with the bank. Uh, in the early days we had to actually build, uh, screen scrapers, uh, for the banks because the banks didn't have APIs, um, and like every screen scraper was different. And we had this realization, um, you know, not very far into the company. If we're able just to grind out all the integrations, we're gonna be better than anyone else. We will build better technology, yes, for these integrations, but also no one else is gonna do this thing that we're doing. Like, no one else is as crazy as we are to be willing to, like, go grind out, you know, 12,000 bank integrations and then, like, make them all scale and, like, build all the software that you need to, like, self-heal the integrations. It was just a mess. Um, and anytime we find a problem like that, I'm like, yeah, that's it. Like, I get excited about it. And everyone inside the company, like, even though we talk about this concept, people inside the company are often like, ah, it's gonna be so hard, it's gonna be, like, you know, it's gonna be difficult, there's like a long path, it might not work, so on and so forth. And I just believe that sometimes if you're willing to grind it out,…

AI assessment note: “if you're willing to grind it out, you're gonna have a unique product”

Q In terms of the problems that they're pursuing, we chatted before, and you said to me about a brilliant Style of problem that you like called the grinder problem. But, um, I just wanted to ask, what is a grinder problem, and why do you find them attractive to do as a company?

A The early history of Plaid was, um, we needed to integrate with all the banks. And these integrations were not like, you know, flip a switch, magical integrations on. It's like a lot of work, uh, to actually set up integration with the bank. Uh, in the early days we had to actually build, uh, screen scrapers, uh, for the banks because the banks didn't have APIs, um, and like every screen scraper was different. And we had this realization, um, you know, not very far into the company. If we're able just to grind out all the integrations, we're gonna be better than anyone else. We will build better technology, yes, for these integrations, but also no one else is gonna do this thing that we're doing. Like, no one else is as crazy as we are to be willing to, like, go grind out, you know, 12,000 bank integrations and then, like, make them all scale and, like, build all the software that you need to, like, self-heal the integrations. It was just a mess. Um, and anytime we find a problem like that, I'm like, yeah, that's it. Like, I get excited about it. And everyone inside the company, like, even though we talk about this concept, people inside the company are often like, ah, it's gonna be so hard, it's gonna be, like, you know, it's gonna be difficult, there's like a long path, it might not work, so on and so forth. And I just believe that sometimes if you're willing to grind it out,…

AI assessment note: “if you're willing to grind it out, you're gonna have a unique product”

Q In terms of the problems that they're pursuing, we chatted before, and you said to me about a brilliant Style of problem that you like called the grinder problem. But, um, I just wanted to ask, what is a grinder problem, and why do you find them attractive to do as a company?

A The early history of Plaid was, um, we needed to integrate with all the banks. And these integrations were not like, you know, flip a switch, magical integrations on. It's like a lot of work, uh, to actually set up integration with the bank. Uh, in the early days we had to actually build, uh, screen scrapers, uh, for the banks because the banks didn't have APIs, um, and like every screen scraper was different. And we had this realization, um, you know, not very far into the company. If we're able just to grind out all the integrations, we're gonna be better than anyone else. We will build better technology, yes, for these integrations, but also no one else is gonna do this thing that we're doing. Like, no one else is as crazy as we are to be willing to, like, go grind out, you know, 12,000 bank integrations and then, like, make them all scale and, like, build all the software that you need to, like, self-heal the integrations. It was just a mess. Um, and anytime we find a problem like that, I'm like, yeah, that's it. Like, I get excited about it. And everyone inside the company, like, even though we talk about this concept, people inside the company are often like, ah, it's gonna be so hard, it's gonna be, like, you know, it's gonna be difficult, there's like a long path, it might not work, so on and so forth. And I just believe that sometimes if you're willing to grind it out,…

AI assessment note: “if you're willing to grind it out, you're gonna have a unique product”

Q spikes. I haven't actually heard that before. I do want to ask you, because kind of given the trajectory of the company, and as we said, the incredible people that you work with earlier, in terms of the requirement to hire, it's incredibly steep So how do you manage the tension between keeping that bar so high, whilst also sustaining the steep requirement on growth curve of additional team members?

A It's incredibly hard to balance the need to hire people with a very high hiring bar. We're fortunate to be in a business where our revenue doesn't scale directly with the number of people that we have, meaning we don't have to go out and hire thousands or tens of thousands of people in order to scale a great deal as a company. And so we really do look for high leverage in the people that we're hiring. We think about, is this person going to come in and be able to truly increase the quality of the team or drive a meaningful piece of work forward that's really going to make a difference? And so we're able to, when we need to, make the decision to not hire someone if they're not at the appropriate level. And we're fortunate in our industry to have that luxury. I will say that probably makes us hire just a little bit too slowly. And that's a frustration, but a frustration that we're willing to take on in order to keep the talent pool really amazing at Platt.

AI assessment note: “revenue doesn't scale directly with the number of people that we have”

Q If you find great talent, how, what's been your biggest lessons on how to do outreach well?

A One of the things that someone told me really early in my career is that, um, the best thing that you can do is go learn from other people that are willing to talk to you. So I got really good early on at just sending cold emails to people saying like, hey, I love this thing that you do. I'm super, super impressed by it. Will you, will you talk to me on the phone for 15 minutes and tell me about it? And people were shockingly willing to do that early in my career. And it applies the same way for recruiting. Um, so, so by the way, exactly what I just said you can do just to get people that you kind of want to recruit but aren't recruitable. Um, but in general, just saying like, hey, you know, I'm Zach. I started this company called Plaid. I've been super impressed by, you know, whatever you're impressed by for that person. Um, and just say, hey, we're hiring for this job. Are you willing to talk to me about it? Um, and you know, they ignore you a bunch of times. They say no a bunch of times, but some people say yes. If you do it enough, then it starts to work. And like that, the same thing you can do over email, you do it over the phone, you get introductions, but like it's the same process of like, hey, I'm genuinely interested in you and like, I have an opportunity that you might be interested in. Like, can we have a conversation?

AI assessment note: “I've been super impressed by, you know, whatever you're impressed by for that person.”

Q I'm so enjoying this, dude. I totally with you on The importance of the grind and the value that comes from it. When does grinding not necessarily lead to better or more value?

A Well, I guess it would be situation specific, but, you know, if your customers don't care, ah, then it's hard. Um, you know, I'll give the plat example. Um, in the US, the banking system is incredibly fragmented. Um, there are tons and tons and tons of different banks. In Canada, it's less fragmented. There's still a decent long tail. In the UK, there's a decent long tail, and in Europe, there's, there's quite a long tail. In certain markets, there's like four banks. Um, there's just not an opportunity for a grinder problem there in terms of, like, building make integrations. That's not the hard thing. There are many other hard things about building financial infrastructure in that market, and, like, we will find those things and build in those markets over time. Um, but we kind of, like, look for those areas where, like, you know, the repeated work and the scale of the repeated work is, is, is really important.

AI assessment note: “there's just not an opportunity for a grinder problem there in terms of”

Q investing, game on. You do whatever you want. It's your money. Go, go for it. But if you're raising money from other people and then spending time investing their money, to me, I'm like, is that not a distraction to the core and to the investors that you've had invest in Plaid? And actually that's, that's not how it should be. How would you respond to people who believe that?

A So that's a good framing and I appreciate the space. Um, but Going back to, like, one of the themes that we've had throughout this conversation, like, you have to examine the specific rationale and reason. Um, so, for me, as Plaid got bigger, I got sucked later in later stage. I was spending time with the gigantic banks, I was spending time with the gigantic companies, and I found that Plaid was, like, in some sense losing some of the founder mentality. It was, like, losing some of, like, the fine-grained attention that I needed to pay to the early stage market. And so, in, call it, like, 2016, I started angel investing, just because I wanted, like, a reason to go talk to these companies. And I did a bunch of it. Like, you know, I had reasons to go talk to founders, had reasons to engage with them deeply. Um, it kept me fresh. It built a great network. I learned things that I brought into how we, we developed products at Plaid. Um, I met amazing people who we ended up hiring, and it's created this, like, really virtuous cycle. Angel investing, actually, I think, is very distracting to founders. The reason for it is, um, it's personally on you. People want to talk to you all the time. You don't have a system for it. You, you haven't built infrastructure. You have no leverage. Um, and so for me, um, a good friend and I, uh, sat down and she said, hey, I'm gonna go be a full-time …

AI assessment note: “I learned things that I brought into how we, we developed products at Plaid.”

Q How did you think about resource allocation between the products? Were you like, hey, split investment across all three. One's a much bigger opportunity. Let's put 80% here in 10 and 10. How did you think about that? Is it? Well, first off, actually, that's a shit question. Just, do you believe that the best CEOs are the best resource allocators?

A In the long term, yes. It kind of depends on the horizon and scale of the business. Like, in the short term, like, if you're, if you're a founder, and you're amazing at, like, you know, finding product market fit. Like, actually, no, that's, that's the best thing. Just go, go spend your time with product market fit. Um, at the mid stages is about, like, can you build an effective company? In the long term, like, when you look at the CEOs that are, like, truly the best, like, 50 year public, I don't know, maybe not 50 year, no one's, no one's in C for quite that long, but, like, you know, 10, 20 year public CEOs are the best resource allocators.

AI assessment note: “In the long term, yes. It kind of depends on the horizon and scale”

Q Sorry, an atomic team. What's an atomic team? Like a generalist team?

A Yeah, like a product manager, whatever the smallest version of a sufficient team is. Um, so it's a product manager, an engineer, maybe a few engineers, maybe a designer if you need it, maybe a data scientist if you need it, depending on the type of product, and you send them off into the ether. Um, and as a founder, your job is like, you know, you need to give them, um, Sufficient, ah, you know, food to continue to exist, like sufficient resources and attention that they can continue to exist. Sufficient protection that, that they don't get killed by what I call the DDoS attack, where if we have a, ah, like a new product that we're working on, we tell, if we tell the whole company, we have a thousand people now, um, you know, 950 people are gonna reach out to the team saying, I'm super excited, how can I help? 950 messages saying, I'm super excited, how can I help? And actually, like, DDoS's the system and, you know, they're, they're thus not able to move fast. Um, uh, with this Atomic team, you give them, um, sufficient, uh, uh, resources, uh, you give them, like, sufficient protection, uh, which oftentimes means, like, kind of ignoring them or hiding them, uh, in some sense from the rest of the company. Um, and then, uh, yeah, you help them figure out how to get to their, like, at least in our case, uh, to get to, to their first design partner. So, uh, they need, like, one or…

AI assessment note: “whatever the smallest version of a sufficient team is. Um, so it's a product manager”

Q If you find great talent, how, what's been your biggest lessons on how to do outreach well?

A One of the things that someone told me really early in my career is that, um, the best thing that you can do is go learn from other people that are willing to talk to you. So I got really good early on at just sending cold emails to people saying like, hey, I love this thing that you do. I'm super, super impressed by it. Will you, will you talk to me on the phone for 15 minutes and tell me about it? And people were shockingly willing to do that early in my career. And it applies the same way for recruiting. Um, so, so by the way, exactly what I just said you can do just to get people that you kind of want to recruit but aren't recruitable. Um, but in general, just saying like, hey, you know, I'm Zach. I started this company called Plaid. I've been super impressed by, you know, whatever you're impressed by for that person. Um, and just say, hey, we're hiring for this job. Are you willing to talk to me about it? Um, and you know, they ignore you a bunch of times. They say no a bunch of times, but some people say yes. If you do it enough, then it starts to work. And like that, the same thing you can do over email, you do it over the phone, you get introductions, but like it's the same process of like, hey, I'm genuinely interested in you and like, I have an opportunity that you might be interested in. Like, can we have a conversation?

AI assessment note: “just sending cold emails to people saying like, hey, I love this thing”

Q I'm so enjoying this, dude. I totally with you on The importance of the grind and the value that comes from it. When does grinding not necessarily lead to better or more value?

A Well, I guess it would be situation specific, but, you know, if your customers don't care, ah, then it's hard. Um, you know, I'll give the plat example. Um, in the US, the banking system is incredibly fragmented. Um, there are tons and tons and tons of different banks. In Canada, it's less fragmented. There's still a decent long tail. In the UK, there's a decent long tail, and in Europe, there's, there's quite a long tail. In certain markets, there's like four banks. Um, there's just not an opportunity for a grinder problem there in terms of, like, building make integrations. That's not the hard thing. There are many other hard things about building financial infrastructure in that market, and, like, we will find those things and build in those markets over time. Um, but we kind of, like, look for those areas where, like, you know, the repeated work and the scale of the repeated work is, is, is really important.

AI assessment note: “There's just not an opportunity for a grinder problem there”

Q How did you think about resource allocation between the products? Were you like, hey, split investment across all three. One's a much bigger opportunity. Let's put 80% here in 10 and 10. How did you think about that? Is it? Well, first off, actually, that's a shit question. Just, do you believe that the best CEOs are the best resource allocators?

A In the long term, yes. It kind of depends on the horizon and scale of the business. Like, in the short term, like, if you're, if you're a founder, and you're amazing at, like, you know, finding product market fit. Like, actually, no, that's, that's the best thing. Just go, go spend your time with product market fit. Um, at the mid stages is about, like, can you build an effective company? In the long term, like, when you look at the CEOs that are, like, truly the best, like, 50 year public, I don't know, maybe not 50 year, no one's, no one's in C for quite that long, but, like, you know, 10, 20 year public CEOs are the best resource allocators.

AI assessment note: “In the long term, yes. It kind of depends on the horizon and scale”

Q If you find great talent, how, what's been your biggest lessons on how to do outreach well?

A One of the things that someone told me really early in my career is that, um, the best thing that you can do is go learn from other people that are willing to talk to you. So I got really good early on at just sending cold emails to people saying like, hey, I love this thing that you do. I'm super, super impressed by it. Will you, will you talk to me on the phone for 15 minutes and tell me about it? And people were shockingly willing to do that early in my career. And it applies the same way for recruiting. Um, so, so by the way, exactly what I just said you can do just to get people that you kind of want to recruit but aren't recruitable. Um, but in general, just saying like, hey, you know, I'm Zach. I started this company called Plaid. I've been super impressed by, you know, whatever you're impressed by for that person. Um, and just say, hey, we're hiring for this job. Are you willing to talk to me about it? Um, and you know, they ignore you a bunch of times. They say no a bunch of times, but some people say yes. If you do it enough, then it starts to work. And like that, the same thing you can do over email, you do it over the phone, you get introductions, but like it's the same process of like, hey, I'm genuinely interested in you and like, I have an opportunity that you might be interested in. Like, can we have a conversation?

AI assessment note: “And it applies the same way for recruiting.”

Q Not at all, but I want to kick off with a little bit of context, Zach. So tell me, how did you make the move from consultant at Bain to founding one of the hottest companies in fintech with Plaid, and what was that aha moment for you?

A You know, it's funny. Founding Plaid in the early days was a series of accidents and a focus on one specific problem. But a series of accidents that actually took us there. I left being kind of along with my co-founder in end of 20 12. And we had this thesis that we could build some interesting products and financial services. We were trying to solve this problem that financial stress, stress about money is the number one stress on most consumers in America. And we wanted to help them live better and easier financial lives. Started out just focused on building a set of consumer focused products, tools to help consumers better manage their money. Realized that there was a bit of infrastructure missing, meaning the connectivity between a bank account and any application. It was so incredibly difficult to build a product on top of banking data that we decided to build infrastructure instead. And a large part of what we did was just be at the right place at the right time. Maniacally focused on this, this very specific set of problems around the infrastructure behind fintech. We could never have guessed that the fintech market would grow as fast as it did. And it's been really just quite a journey of trying to hang on and scale as fast as our customers need us to.

AI assessment note: “It was so incredibly difficult to build a product on top of banking data”

Q Not at all, but I want to kick off with a little bit of context, Zach. So tell me, how did you make the move from consultant at Bain to founding one of the hottest companies in fintech with Plaid, and what was that aha moment for you?

A You know, it's funny. Founding Plaid in the early days was a series of accidents and a focus on one specific problem. But a series of accidents that actually took us there. I left being kind of along with my co-founder in end of 20 12. And we had this thesis that we could build some interesting products and financial services. We were trying to solve this problem that financial stress, stress about money is the number one stress on most consumers in America. And we wanted to help them live better and easier financial lives. Started out just focused on building a set of consumer focused products, tools to help consumers better manage their money. Realized that there was a bit of infrastructure missing, meaning the connectivity between a bank account and any application. It was so incredibly difficult to build a product on top of banking data that we decided to build infrastructure instead. And a large part of what we did was just be at the right place at the right time. Maniacally focused on this, this very specific set of problems around the infrastructure behind fintech. We could never have guessed that the fintech market would grow as fast as it did. And it's been really just quite a journey of trying to hang on and scale as fast as our customers need us to.

AI assessment note: “Realized that there was a bit of infrastructure missing, meaning the connectivity between a bank account”

Q investment. When I was chatting to Rick at NEA, he said one element that you've done incredibly well at is keeping a really high bar when it comes to hiring, which you touched on earlier. So I'd love to dive on this. What have been some of your biggest lessons on really attracting that A-star talent? And how have you thought about kind of constructing the optimal recruitment process, Zach?

A So if you know the answer to what the optimal recruitment process is, please tell me. It's a constant game of figuring out how to get better, figuring out what new tactics and processes And at every stage of scale, it changes massively. For us, I think the best way to get great people is to prioritize recruiting very clearly. Myself and my co-founder have spent an inordinate amount of time recruiting our team, being sure that we have really great people coming in the door, iterating our hiring processes, figuring out our questions, creating the rubric for how we evaluate people, testing our values, so on and so forth. In the earliest days, we knew that we as a company would only grow if we hired great people. And so almost the history of the company, I've probably spent about a third of my time recruiting people. And that involves everything from personally sourcing, going out and finding the best candidates that are out there, to selling people, convincing them to get into the hiring process, evaluating, of course, and then closing at the end of it. We have a couple principles that we think quite a lot about. One of them that I like is hiring for spikes. So what that means is we look for the areas where someone is truly world class, and we prefer to hire people that have a few elements in which they're totally amazing, where they can really lean into their strengths, than hiri…

AI assessment note: “One of them that I like is hiring for spikes.”

Q No, I absolutely love business history books. Add in economic psychology is two of my more nerdy passion points. Tell me though, Zach, what do you know now that you wish you'd known at the beginning of your time with Plaid?

A I think the biggest lesson is persistence. We have run into so many problems. There's been, in the early days, so many times that the company almost died, or we had to deal with some big, crazy issue. One of the things that one of my friends told me was just to say, hey, don't let the company die. You will become a much larger company if it doesn't die. And for a long time, I had this saying, when someone would say, hey, how's Plaid going? I say, well, we're not dead yet. We're still growing. It's a joke in the way that I say it. I think persistence in being willing to push through all of those things that are really hard at the early stages of companies, that's something that I wish I'd been told earlier.

AI assessment note: “I think the biggest lesson is persistence.”

Q Not at all, but I want to kick off with a little bit of context, Zach. So tell me, how did you make the move from consultant at Bain to founding one of the hottest companies in fintech with Plaid, and what was that aha moment for you?

A You know, it's funny. Founding Plaid in the early days was a series of accidents and a focus on one specific problem. But a series of accidents that actually took us there. I left being kind of along with my co-founder in end of 20 12. And we had this thesis that we could build some interesting products and financial services. We were trying to solve this problem that financial stress, stress about money is the number one stress on most consumers in America. And we wanted to help them live better and easier financial lives. Started out just focused on building a set of consumer focused products, tools to help consumers better manage their money. Realized that there was a bit of infrastructure missing, meaning the connectivity between a bank account and any application. It was so incredibly difficult to build a product on top of banking data that we decided to build infrastructure instead. And a large part of what we did was just be at the right place at the right time. Maniacally focused on this, this very specific set of problems around the infrastructure behind fintech. We could never have guessed that the fintech market would grow as fast as it did. And it's been really just quite a journey of trying to hang on and scale as fast as our customers need us to.

AI assessment note: “Realized that there was a bit of infrastructure missing... decided to build infrastructure instead.”

Q investment. When I was chatting to Rick at NEA, he said one element that you've done incredibly well at is keeping a really high bar when it comes to hiring, which you touched on earlier. So I'd love to dive on this. What have been some of your biggest lessons on really attracting that A-star talent? And how have you thought about kind of constructing the optimal recruitment process, Zach?

A So if you know the answer to what the optimal recruitment process is, please tell me. It's a constant game of figuring out how to get better, figuring out what new tactics and processes And at every stage of scale, it changes massively. For us, I think the best way to get great people is to prioritize recruiting very clearly. Myself and my co-founder have spent an inordinate amount of time recruiting our team, being sure that we have really great people coming in the door, iterating our hiring processes, figuring out our questions, creating the rubric for how we evaluate people, testing our values, so on and so forth. In the earliest days, we knew that we as a company would only grow if we hired great people. And so almost the history of the company, I've probably spent about a third of my time recruiting people. And that involves everything from personally sourcing, going out and finding the best candidates that are out there, to selling people, convincing them to get into the hiring process, evaluating, of course, and then closing at the end of it. We have a couple principles that we think quite a lot about. One of them that I like is hiring for spikes. So what that means is we look for the areas where someone is truly world class, and we prefer to hire people that have a few elements in which they're totally amazing, where they can really lean into their strengths, than hiri…

AI assessment note: “best way to get great people is to prioritize recruiting very clearly.”

Q Sorry, an atomic team. What's an atomic team? Like a generalist team?

A Yeah, like a product manager, whatever the smallest version of a sufficient team is. Um, so it's a product manager, an engineer, maybe a few engineers, maybe a designer if you need it, maybe a data scientist if you need it, depending on the type of product, and you send them off into the ether. Um, and as a founder, your job is like, you know, you need to give them, um, Sufficient, ah, you know, food to continue to exist, like sufficient resources and attention that they can continue to exist. Sufficient protection that, that they don't get killed by what I call the DDoS attack, where if we have a, ah, like a new product that we're working on, we tell, if we tell the whole company, we have a thousand people now, um, you know, 950 people are gonna reach out to the team saying, I'm super excited, how can I help? 950 messages saying, I'm super excited, how can I help? And actually, like, DDoS's the system and, you know, they're, they're thus not able to move fast. Um, uh, with this Atomic team, you give them, um, sufficient, uh, uh, resources, uh, you give them, like, sufficient protection, uh, which oftentimes means, like, kind of ignoring them or hiding them, uh, in some sense from the rest of the company. Um, and then, uh, yeah, you help them figure out how to get to their, like, at least in our case, uh, to get to, to their first design partner. So, uh, they need, like, one or…

AI assessment note: “whatever the smallest version of a sufficient team is”

Q No, I absolutely love business history books. Add in economic psychology is two of my more nerdy passion points. Tell me though, Zach, what do you know now that you wish you'd known at the beginning of your time with Plaid?

A I think the biggest lesson is persistence. We have run into so many problems. There's been, in the early days, so many times that the company almost died, or we had to deal with some big, crazy issue. One of the things that one of my friends told me was just to say, hey, don't let the company die. You will become a much larger company if it doesn't die. And for a long time, I had this saying, when someone would say, hey, how's Plaid going? I say, well, we're not dead yet. We're still growing. It's a joke in the way that I say it. I think persistence in being willing to push through all of those things that are really hard at the early stages of companies, that's something that I wish I'd been told earlier.

AI assessment note: “I think the biggest lesson is persistence.”

Q I'm so enjoying this, dude. I totally with you on The importance of the grind and the value that comes from it. When does grinding not necessarily lead to better or more value?

A Well, I guess it would be situation specific, but, you know, if your customers don't care, ah, then it's hard. Um, you know, I'll give the plat example. Um, in the US, the banking system is incredibly fragmented. Um, there are tons and tons and tons of different banks. In Canada, it's less fragmented. There's still a decent long tail. In the UK, there's a decent long tail, and in Europe, there's, there's quite a long tail. In certain markets, there's like four banks. Um, there's just not an opportunity for a grinder problem there in terms of, like, building make integrations. That's not the hard thing. There are many other hard things about building financial infrastructure in that market, and, like, we will find those things and build in those markets over time. Um, but we kind of, like, look for those areas where, like, you know, the repeated work and the scale of the repeated work is, is, is really important.

AI assessment note: “if your customers don't care, ah, then it's hard.”

Q How did you think about resource allocation between the products? Were you like, hey, split investment across all three. One's a much bigger opportunity. Let's put 80% here in 10 and 10. How did you think about that? Is it? Well, first off, actually, that's a shit question. Just, do you believe that the best CEOs are the best resource allocators?

A In the long term, yes. It kind of depends on the horizon and scale of the business. Like, in the short term, like, if you're, if you're a founder, and you're amazing at, like, you know, finding product market fit. Like, actually, no, that's, that's the best thing. Just go, go spend your time with product market fit. Um, at the mid stages is about, like, can you build an effective company? In the long term, like, when you look at the CEOs that are, like, truly the best, like, 50 year public, I don't know, maybe not 50 year, no one's, no one's in C for quite that long, but, like, you know, 10, 20 year public CEOs are the best resource allocators.

AI assessment note: “In the long term, yes. It kind of depends on the horizon and scale”

Q And then you do, like, milestone-based financing, which is, like, hey, if we find sufficient product market fit, We'll open up new resources. How do you think about the next stage post-atomic team?

A Yeah, if you, if you try to over-processize that, then it doesn't really work. Like, a lot of people try to run, like, an internal VC style model, which is like, you know, we'll fund you for a million bucks for a year and we'll figure it out. Yeah, and you can do that, but it becomes over-processized, and then everyone starts to, like, try to hack the process as opposed to, like, really finding the, the unique insight. Um, I'd rather just, like, let people go run at a thing for a while, and, like, usually they will come, like, at least I've found, they will come back and say, hey, this is not gonna work. Or, like, hey, there's something very real here. They might come back in a week and say, like, there's something very real here, and, like, I, I have a customer that's, like, chomping at them to buy it. Like, I really need to fund it now. I don't want them to wait 18 months to come back and ask for more resourcing. Um, on the flip side, uh, you know, no one inside the company wants to go, like, off in a foray for, like, three years finding nothing.

AI assessment note: “I'd rather just, like, let people go run at a thing for a while”

Q investing, game on. You do whatever you want. It's your money. Go, go for it. But if you're raising money from other people and then spending time investing their money, to me, I'm like, is that not a distraction to the core and to the investors that you've had invest in Plaid? And actually that's, that's not how it should be. How would you respond to people who believe that?

A So that's a good framing and I appreciate the space. Um, but Going back to, like, one of the themes that we've had throughout this conversation, like, you have to examine the specific rationale and reason. Um, so, for me, as Plaid got bigger, I got sucked later in later stage. I was spending time with the gigantic banks, I was spending time with the gigantic companies, and I found that Plaid was, like, in some sense losing some of the founder mentality. It was, like, losing some of, like, the fine-grained attention that I needed to pay to the early stage market. And so, in, call it, like, 2016, I started angel investing, just because I wanted, like, a reason to go talk to these companies. And I did a bunch of it. Like, you know, I had reasons to go talk to founders, had reasons to engage with them deeply. Um, it kept me fresh. It built a great network. I learned things that I brought into how we, we developed products at Plaid. Um, I met amazing people who we ended up hiring, and it's created this, like, really virtuous cycle. Angel investing, actually, I think, is very distracting to founders. The reason for it is, um, it's personally on you. People want to talk to you all the time. You don't have a system for it. You, you haven't built infrastructure. You have no leverage. Um, and so for me, um, a good friend and I, uh, sat down and she said, hey, I'm gonna go be a full-time …

AI assessment note: “I learned things that I brought into how we, we developed products at Plaid.”

Q And then you do, like, milestone-based financing, which is, like, hey, if we find sufficient product market fit, We'll open up new resources. How do you think about the next stage post-atomic team?

A Yeah, if you, if you try to over-processize that, then it doesn't really work. Like, a lot of people try to run, like, an internal VC style model, which is like, you know, we'll fund you for a million bucks for a year and we'll figure it out. Yeah, and you can do that, but it becomes over-processized, and then everyone starts to, like, try to hack the process as opposed to, like, really finding the, the unique insight. Um, I'd rather just, like, let people go run at a thing for a while, and, like, usually they will come, like, at least I've found, they will come back and say, hey, this is not gonna work. Or, like, hey, there's something very real here. They might come back in a week and say, like, there's something very real here, and, like, I, I have a customer that's, like, chomping at them to buy it. Like, I really need to fund it now. I don't want them to wait 18 months to come back and ask for more resourcing. Um, on the flip side, uh, you know, no one inside the company wants to go, like, off in a foray for, like, three years finding nothing.

AI assessment note: “if you try to over-processize that, then it doesn't really work.”

Q Ok, so we get on the call, or we get in a process with this person. In terms of talent assessment, how have you changed your approach, mindset, appreciation towards talent assessment?

A So we have a couple of these, like, core philosophies, and one or two of them have evolved. One of them is higher for spikes. This one hasn't really changed. We look for people that are incredible at some things, even if they have, like, glaring weaknesses in other areas. So as long as we have a team that can balance that out, we would rather have the spikey person, because it turns out, you know, a well-rounded person that maybe doesn't have any really tall spikes, um, if you build a team of those, you'll have, like, you know, one certain, uh, you can imagine, like, a level, uh, or, like, kind of a, if you were to chart this out, you can imagine, like, they're, they're only reached a certain max on the vertical axis. Uh, but if you have spikes and you, like, make the spikes all kind of work together, then, like, the, the level that will reach on the vertical axis is far higher. Uh, so higher for spikes is a big one for us. It also means that we like people that are weird, that have unique experiences, that are different, that have a different background, or a different way of thinking about the world. So that's been a big one for us. Um, one that I've changed was, uh, there's this thing internally called the experience trap, where we basically thought that anyone that had been doing a certain thing for, like, more than 15 or 20 years, like, kind of wasn't a fit for Plaid. Um, …

AI assessment note: “one that I've changed was, uh, there's this thing internally called the experience trap”

Q Do you remember the moment you made it?

A Oh my gosh, yeah, yeah. It was, um, my leadership team and I, it was, uh, right, Um, ah, we were doing this off-site because we didn't want to be in the office. Um, ah, and we were sitting in my living room. I'd had them all there. We were, like, had this offer, and, and we were going through, like, the pros and cons of it. And, um, you know, we went around. We talked about what people were thinking, and then I kicked them all out, and I was like, look, I need to think. Like, I'm making this call, like, at the end of the day. Like, I've gotten all of your input, and I just need to think. And it was, like, me sitting in my living room. Um, ah, you know, my wife is out of town. I'm just kind of alone sitting there. Um, and ended up making what I think was a 51 49 call to say, like, it's the right answer to sell the business. Um, For a variety of reasons at the time. A year later, we had that same meeting, same group of people sitting in my living room, having the same, same style of conversation. And that was a, like, you know, ninety-nine-one decision. I was, like, pretty certain that the right answer was to walk away. I just wanted to make sure that everyone else was aligned. And what had changed in that innovating period is that our business had grown quite massively. We'd reached this escape velocity, um, and the concept of digital finance had really taken hold of something t…

AI assessment note: “It was, like, me sitting in my living room... ended up making what I think was a 51 49 call”

Q Sorry, an atomic team. What's an atomic team? Like a generalist team?

A Yeah, like a product manager, whatever the smallest version of a sufficient team is. Um, so it's a product manager, an engineer, maybe a few engineers, maybe a designer if you need it, maybe a data scientist if you need it, depending on the type of product, and you send them off into the ether. Um, and as a founder, your job is like, you know, you need to give them, um, Sufficient, ah, you know, food to continue to exist, like sufficient resources and attention that they can continue to exist. Sufficient protection that, that they don't get killed by what I call the DDoS attack, where if we have a, ah, like a new product that we're working on, we tell, if we tell the whole company, we have a thousand people now, um, you know, 950 people are gonna reach out to the team saying, I'm super excited, how can I help? 950 messages saying, I'm super excited, how can I help? And actually, like, DDoS's the system and, you know, they're, they're thus not able to move fast. Um, uh, with this Atomic team, you give them, um, sufficient, uh, uh, resources, uh, you give them, like, sufficient protection, uh, which oftentimes means, like, kind of ignoring them or hiding them, uh, in some sense from the rest of the company. Um, and then, uh, yeah, you help them figure out how to get to their, like, at least in our case, uh, to get to, to their first design partner. So, uh, they need, like, one or…

AI assessment note: “whatever the smallest version of a sufficient team is.”