Transcript

[Harry] Jag, I am so excited for this show today. I've heard so many great things from many different people with different stories, but especially the main man, David Vélez. So thank you so much for joining me today, my friend.

[Jag] I'm so excited to be here, Harry. I'm a huge fan of the show, as you know.

[Harry] That is so kind of you. But when I spoke to David, he asked the first most important question. You went from a cushy job at Meta, working in the Valley, to working for a Brazilian fintech that no one really knew about at the time. Why did you do it? He asked.

[Jag] David is an extraordinarily persuasive guy, so that's the first thing. The first conversation I had with David, when he asked me towards the end, would you ever consider moving to Sao Paulo, I said, David, the obvious answer is no, and this is a crazy request. But he's a very persuasive guy, because when I gave him that answer, he said, there are people working here who are far more definitive than you are, so we should keep chatting. Doug Leone played a part in persuading me, but the larger reason is I had spent the previous almost 15 years working at Google, at Quantcast, at Facebook, basically in media tech and ad tech as it was the center of what was happening in Silicon Valley, and had come to realize that I had a ringside seat to the great disruption story of the previous 15 years, and that was amazing, intellectually, incredibly interesting. I had developed a toolkit that I was applying to serve what is effectively the richest billion people on the planet. The US, Europe, Japan, Australia, New Zealand, and what David and NuBank offered was the chance to take that same toolkit and apply it to, roughly speaking, the bottom half of the global income pyramid. The second reason, and the thing that became clear to me as I started talking with David and the NuBank team, was media tech and ad tech had played itself out, and I had been on both sides. I had worked for Google and Facebook, I had worked for a challenger startup, but the game had consolidated. Fintech is, even today, three years later, still very much in its early innings, and we still have this 20-year game to play itself out. Being in at the ground floor in the beginning of an industry that's just starting to fundamentally change, I firmly believe the next set of bits to be disrupted after the media industry, the obvious next set of bits, is reais, dollars, pesos, pounds, and the technology's ready and it's time to go, so it was just too big an opportunity to pass up.

[Harry] There's so many things for me to unpack there. You mentioned the word toolkit multiple times and utilizing that toolkit. I think there's this common question in product, which is, to what extent is it art or science? Incredibly difficult question, which is an utterly shit question, really, but I thought I would ask it anyway. Is product more art or science, and what would you say it would be if you had to rank it 70-30, 80-20?

[Jag] Yeah, I've heard you ask this question of a number of product leaders over the last year or so, from Gustav to Kevin Weil and others, and I've liked their answers. I'll give you my take. I have tried and we have tried at NuBank, especially over the last three years, to make it as scientific as we possibly can, and we pushed the envelope and I think there's still further to go. I think you can make it 90% science. The thing about product management is it's a pretty new discipline. When an engineer shows up at a tech company or a salesperson shows up at a tech company, they know what the job is. When I showed up at Google in 2006, no one really had a definition for what a product manager was. You see where the bathroom was and then they dropped you in a team and said, go. Over the last 15 years, and you've had Marty Kagan and others on, it's become gradually more scientific. I think we can make it, and we're on the cusp of making it, I think we're quite there, but we're on the cusp of making it 90% science. What is the science of understanding product market fit? How do you make it metrics-driven more than gut feel? Having said that, I think the last 10% and the most important 10% and the 10% that makes all of the difference is still art. Once you get the science, you get to the start line of having a shot at winning the race, but it's the 10% of art that is how you win that race.

[Harry] What is in the art? If we're in the science and we're determining product market fit, all of the metricized elements that we now have with the data tools that we have, what is in the 10% of art still?

[Jag] I think the heart of it, and it's the hardest part to make scientific, is the empathy with the customer. What is that specific thing that we're going to focus on that creates the magic for them that really taps into a deeply unmet need? The thing is, and the pitfall is, you can do customer research. You can go ask customers, but I tell the product team at NuBank all the time, we are not in the business of taking dictation from our customers because if we were, we would just send the engineer instead of us with no need for the middleman. We have to listen to the customer, better yet observe the customer to interpret what their unmet need is and what the hierarchy of those unmet needs are and where the magic is. Then you apply business filters and everything else to see if there's a profit to be had. That empathy and that intuition for where the real unmet need is something that is very hard to make scientific and is something that you can be called with the official title of product manager or you can be a guy named David Velez who came up with a magical idea 10 years ago and just had that intuition for where the pain was that was sharp and that would create this visceral reaction from the Brazilian consumer and that is hard to replicate.

[Harry] What makes a good question? I think the majority of customer development actually sucks because people don't ask good questions. What's a good question?

[Jag] It's about exploring the customer's problem rather than the solution and even when you're exploring the problem, you don't ask about the problem directly. You ask about the customer's life around whatever area you're interested in and customers don't even realize when they are employing workarounds, when they are making do, when they think everything is fine but if you somehow showed up with something completely different, you can change their lives. So I think where we tend to fall short in the industry today even still is we spend a lot of time trying to pitch the solution, ask about the solution. We spend not nearly enough time talking about the problem and when we are talking about the problem, we're mostly asking about it rather than asking around it and observing around it because you are just synthesizing that.

[Harry] Do you worry that by asking around it, you can lead them to a conclusion? Jag, don't you find it difficult when you have to wait three days for your salary to come through? Yes, I guess so, yes, but do you know what I mean? You're leading the witness.

[Jag] Yeah, for my money and one of the hardest things, especially product managers who are earlier in their careers to wrap their head around is we should never fall in love with our belief or our hypothesis and most of us can't help but become the lawyer, the advocate for the idea and what is really important is to have a hypothesis, to be really clear on your hypothesis, not to go in with a vague set of, I want to ask you about this general area of your life, but to have a couple of sharp hypotheses you're out to test, but to be entirely agnostic, not in love with any of those, to approach it more as a judge than a lawyer. So you've got to have a sharp hypothesis, but also not fall in love with the hypothesis. It runs completely contrary to human nature to come up with an idea and keep your distance from it at the same time and that's I think what makes it hard.

[Harry] How do you do it? You spend time on it, with time you have commitment and with commitment you have love. Fuck me, this is a dating show, Jag, but when you've sacrificed other things to dedicate time and commitment and energy and passion to it, it's human nature. How do you detach yourself from it?

[Jag] The one step which is very hard but is not even enough is to at least fall in love with the problem, not the specific solution. David, just to take that NuBank example, was in love with the potential of unleashing the Latin American consumer who was being really badly served. He wasn't in love with I need to build a credit card for because of X, Y, Z reason or business case and so first you start with at least being in love with the problem and then when you're in love with the problem, the solution can come out in any one of a dozen or two dozen or a hundred different ways and you're just keeping your ear to the ground to see how can I unleash this bigger idea, this bigger opportunity. That's the key. Whereas falling in love with a specific idea before you've proven it just means you're going to end up leading the witness and you're going to hear what you want to hear and that's always a trap.

[Harry] Jag, before we dive into the meat of the show, I do just have to ask you, you spent several years at Facebook as a director of product management. It's a pretty heavy role at a pretty important time for the company. What are one or two takeaways for you that really impacted how you think about product today?

[Jag] Facebook has its share of pitfalls in how it operates but it has a few things it does extremely well which in at least a couple of cases dovetailed with things that I really needed to learn, especially at time. One of them and this sort of builds actually ironically of what we were just talking about is I had a tendency to A, enjoy the arguments and B, get entrenched in my position. Not with the customer but internally and one of the things that Facebook does incredibly well especially at the senior levels is I have a strong point of view as director A and you have a strong point of view as director B and we need to collaborate and Facebook is built for it. You have to collaborate across business areas to make something big happen and Facebook was incredibly good at saying and I had to learn this from a number of people but they were incredibly good at simply saying, you think X, I think Y. By the end of this week, the latest, we are going to either agree, disagree and know Y and escalate and the speed of escalation, it wasn't I think X, you think Y, let's come back in a week, discuss it some more and then in a week after that, we are going to escalate it. It was like I think X, you think Y, by the end of day, we are going to agree if we can come to a consensus or escalate and that speed of escalation I think ties to Facebook's historical always impressive speed of execution and that was probably the single biggest thing I learned there which has been so valuable for me since.

[Harry] I was speaking to a $50 billion CEO today who is a friend of mine and he said at the end of the day when you get to a size like ours and bluntly like NuBank is adding that for me to use that and you are fighting off eventual mediocrity and you are fighting off eventual slowness. My question to you is when you think about Facebook operating in that execution speed, how do you retain that speed and agility when you are NuBank today with such large teams and very real processes and bluntly heavy corporate structures?

[Jag] We have a CEO who is a founder. He fights the natural creep of corporate process including ones that I try to introduce at times and I think that's a helpful counterforce but I think there are two things and again, these things are over the last 10 years getting more and more codified amongst the really great companies. My wife used to work at Amazon and I heard in the early days about two pizza box teams. At Google, we used to talk about teams should be like a family. You should be able to feed it around the table. We at NuBank, we do a couple of things religiously. We try to keep teams very small actually. You might have many of them focusing on many things as the company grows but we try very hard to keep the size of teams small. That's a really important thing and the second thing which we just talked about is when you get stuck, don't stay stuck, escalate. Someone will make a call, a bad decision which can be iterated on is better than no decision at all because you start learning from the market from customers and then very few decisions are one-way doors. So escalate and decide and keep your team small and at NuBank, there was a religion which is sometimes hard as a senior exec, there was a religion around autonomy and we try to give our general managers and even folks who are more junior than that a lot of autonomy to make decisions about where they're taking their product or their business.

[Harry] Do you inherently trust people? I know that sounds weird but you said about autonomy, there's two ways which is full trust from day one and it's yours to be lost or you gain trust over time. I inherently don't trust people which is why I will die alone with my cat but do you trust people from the start or is it to be gained over time?

[Jag] I tend to default to trusting people from the start. We had a mantra at Facebook and speaking of lessons from there about assuming good intent, it was built into the cultural training there. I think assuming good intent, trusting people. I don't trust people at the recruiting stage. When I'm recruiting someone, my default answer is they're not good enough to join the company but once they've proven that and we've decided to extend an offer, I'm going to assume that you are NuBank quality. We try to keep a very high bar and then I'm going to try to give you as much room to run and show me what you can do. Look, if you make mistakes along the way, we'll correct them. We all do and if you make too many mistakes, then we have to have a more difficult conversation but I think the right starting point is to give people that trust and give them the room to run.

[Harry] You mentioned there the pre-hiring process. We have about 700,000 founders that listen so no worries that you're about to impact many thousands of teams but I do have to ask about the hiring process. I know we're mixing the schedule up but fuck it, let's do it. How do you structure your hiring process today? What's the right way to structure a hiring process for a product team?

[Jag] Yeah, I really struggle with this. I don't know that I have a good answer and if anything, I've lost confidence over time and over the years and over the last decade rather than gained it.

[Harry] Why? What has led you to lose confidence?

[Jag] I've made some mistakes in hiring, some great understanding that first of all, understanding that credentials matter less than I thought, that someone who had gone to the right school, the right program, the right grades, the right company and any combination of all of those things isn't necessarily right for your situation at your time, one. Two, a gradual understanding and I know you're a football fan and I'm a huge football fan. The team matters. I think we've seen that in the world of football since Pep came along 15 years ago. The team matters and so no matter how many Galacticals, how many All-Stars you've got on the team, it doesn't mean that the whole is going to be greater than the sum of the parts and so recruiting someone who is going to fit the need, fit the team, fit the problem and you don't have the normal signposts of credentials necessarily as a definitive guarantee. After consulting, I joined Google and Google is famous for the fact that Larry literally looked at every hiring packet including resume. I what? 12 years after graduating college, I had to provide my SAT scores to Google. I had to go find them. I had to provide my college transcript which everyone did. It was a very credential-driven place which was accounted for a lot of the talent density there but also led to blind spots. Just take another example. Google, I was one of the last product managers hired into Google who did not have a computer science degree and because of that computer science requirement to be a product manager, people like Kevin Systrom weren't allowed to be product managers at Google so he went off and founded a little company called Instagram and there were a number of those types of people who went on to found great companies. The thing that I used to think was very scientific has become more in terms of recruiting and I used to feel a lot of confidence and clarity about it, feels more and more like art to me and I think that means I just don't understand it well.

[Harry] With the signals of clarity being less definitive in terms of quality, school, grades, prior companies, you name it, then there's something else. What questions do you ask to try and unveil my character? And imagine I'm an angel investment of yours and you're advising me I'm a CEO. I don't know what to ask. What questions should I ask to determine quality of product?

[Jag] There are two flavors of questions I tend to ask these days. One is a set of questions around drive at Facebook. It was an interview stream called leadership and drive but the thing I really look for is drive because there are a lot of really smart people who don't actually have that inherent drive for greatness, who will walk through fire, break through walls to achieve greatness, especially in a startup type environment where the odds are stacked against you. You're supposed to fail. The odds are that you will fail. Drive is more important than IQ points. There's a set of questions around what's the hardest thing you've done. Tell me about a time when you're in a situation where you didn't know what the answer was. I took a class in leadership. It was supposed to be a leadership class in graduate school and it basically was inherently designed, the professor was a psychologist, and it was inherently designed to be a Lord of the Flies exercise over the course of the semester when you had Israeli fighter pilots and BBC journalists and these amazing variety of people. For example, one class the professor walks in, puts his briefcase down, sits down and doesn't say a word for 45 minutes and you just see what 100 type A Harvard people will do and it's just this amazing psychological experiment. But this drive to take command of a situation, but also to collaborate and bring people in, it's not drive for power's sake, it's drive for getting something done, is a really important thing. The core thing I took away from that class after spending three quarters of it not understanding why in the hell I was in it, I was just like, what are we doing here? He had a phrase near the end, which really resonated with me. I use it almost every week is, are you able to stomach your own incompetence? Are you able to enter into a situation where you do not know what the answer is and dive in anyway? Which I think is the essence of building something new. So there's a set of questions I ask around that. The other one is clarity. It's pure coincidence that you had Kevin Systrom on recently, but I'm not generally in the habit at least for the last 10, 15 years of looking for advice from 20 somethings. So Kevin Systrom back before Instagram was famous, certainly before it was bought by Facebook, I heard Kevin at some conference say, we may not be right, but at least we are clear. And I love that because I think that's the other thing I'm looking for someone. And so I would ask questions simply, you used to work at this company or you founded this company 10, 12, 15 years ago. It may not have been a successful company, it doesn't really matter. Tell me about the company. And I just want to know how clearly you can explain to me as someone who knows nothing about your business so that I can understand if it seems to make any sense. It may not make sense. I don't have the technical capability to understand if your startup made sense, but did it make sense or your product or whatever you're working on. But I want to know if you can tell me clearly if you spend five minutes telling me and I'm zoned out and I don't understand what you said anyway, which by the way is three quarters of most answers. And so I'm looking for clarity and I'm looking for drive. And I still don't know if I have the right questions to get to those things, but that's what I'm reaching for.

[Harry] You said clarity there and it takes me to something you said to me before the show, which was on the lack of importance of execution and the importance of strategic clarity. What did you mean by the lack of importance or execution being overrated and strategic clarity being underrated?

[Jag] I'm being a little contrarian. I spent a number of years as a strategy consultant. It's almost mantra at this point, I actually think it's fairly lazy mantra in Silicon Valley and in tech in general to say strategy is easy and execution is everything. And look, to be very clear, execution is massively important, but execution without strategic clarity is basically frenzy. It is not progress. It is not forward motion.

[Harry] I'm interrupting. I'm so sorry. Is it not about the amount of at-bats though? With every bit you learn, I always say the speed of execution is the single biggest driver in the companies that I see succeed that I've invested in because they have a structure and a mechanism where they execute, learn, change, execute again and again.

[Jag] Completely. I think there's one piece missing in the framework you just outlined. If you are executing and you don't know what you are executing towards, you cannot take the signal that's coming back at you and iterate. If you have a clear hypothesis of the case, a theory of the case as to what you're supposed to get back, what your strategy is trying to achieve, then all the data you're getting back becomes signal and you can react very clearly and definitively so your next iteration is in the right direction. It doesn't have to be perfect. It has to be clear so that you can take the signal back from the customer or the market and iterate in some clear direction. Otherwise you're off and running in service.

[Harry] And that's what would be an example of this, just so I understand it more clearly. You said they're actually strategically being clear about knowing where you're going or what you're wanting. What would be an example of that in product?

[Jag] I'll give you, let me give you an example, which is one we have wrestled with fairly recently at NuBank. If you've got a hypothesis for a product and you launch it out in the market and you're not careful about who you've launched it to or how you've positioned it, and you get a bunch of signal that says, Hey, as we did at NuBank, hey, the NPS on this product is actually quite mediocre. And you just look at your average NPS and says, eh, this product just doesn't seem to be resonating. You might iterate in any number of directions, kill the product, iterate the product in some incorrect direction. What we ended up finding was, hey, peel underneath the data, there is a set of customers, which was actually the set of customers we meant the product to be for, who love the product. They're actually a small minority of the people who we've given the product to right now. And there's a bunch of customers who don't like the product, but you know what? It turns out this product wasn't designed for them. It wasn't built for them. Of course they don't like it. There's no reason why they should like it. So that's in part a story about de-averaging, but it's also a part of the story about saying, and this is what strategy is. I am building this product for this specific set of customers, this specific group segment of customer to meet this deep unmet need. Do they like it or not? If those customers don't like it, then you start to ask a series of why questions. Why don't you like it? When is it about the product that isn't working? Is it, we executed it poorly? Is it fundamentally misconceived? How do we iterate? If it turns out, however, which is what we found quite recently, those customers love it, but we've been giving it to a bunch of customers for who it wasn't intended. And they don't like it. You've actually got confirmation that you're on to something. And the problem isn't to iterate the product. It's to get the product into the hands of the right customer. Completely different answers. And that's what I mean. If you do not have a clear hypothesis, a clearly articulated strategy of how you plan to win, you can execute with speed, with velocity, and never get there. And that happens all the time.

[Harry] Do you know what I find so hard about how you plan to win? I had this realization I was with Alex Schultz, the CMO of Facebook the other day, Meta, sorry, he did the rebrand. He'd kill me if he heard me say Facebook. And he taught me actually the most brilliant thing, which was when you think about how to win or setting goals, the truth is most of it is gameable. If you say, I want a thousand users, your growth and marketing team could spend on ads and get shit users that churn, but your goal was a thousand users. My point is, how do you align goals to strategic outcomes that move the needle for the business?

[Jag] Yeah, that's what Facebook calls execution, is aligning your success metrics and agreeing the success metrics that actually align with winning. There are all sorts of very elegant product management frameworks about the stages of building and launching and having a product succeed. I tend to break it down into basically two stages. There is pre-product market fit and there's post-product market fit. And your job in each stage is completely different. And the problem companies often have, product people often have, frankly CEOs often have, is they think of building a product as I build it. I get it launched and then I start scaling it and it's successful. And there's actually the most important piece in the middle, which is I got to get this product that's just launched, that probably isn't quite right yet, to product market fit, to customer love, to use a more evocative phrase. And if I don't get customer love, scaling is like pushing the rock uphill and I can game it. I mean, any great growth person can run ads, can do any number of tricks of the trade, and you can get to the next round that can go in a good bull market for a long time. But if you have customer love, and this was the magic of early NuBank, your customers are your market and the snowball is rolling downhill instead of you trying to push the rock uphill. And so there are a couple of ways to answer your question. The first way, the most mechanical way to answer your question is if you don't have a customer love metric, I learned actually when I moved to Sao Paulo back in the beginning of 2020 from the NuBank product team about this technique that was starting to take off in Silicon Valley where I was living called the Sean Ellis, which is an attempt to bring science to product more fit that I've come to really love. You can use NPS scores, there's a number of technical mechanisms you can use, but we try to be pretty maniacal about customer love first before we even talk about significant scaling. Because yeah, I can scale the product. We've got 70 million, 75 million customers at NuBank. I can make any dead cat bounce. I can make any product I launch look great for the first year, but that doesn't matter. What matters is do the customers love it because then it'll explode out on its own accord. And that's what we really try to focus on.

[Harry] You mentioned dead cat bouncing there. You mentioned it to me earlier and I had no idea what you were talking about. And in the notes I was like, oh, I haven't heard this. What is the illusion of dead cats bouncing before I forget to ask this?

[Jag] So sure, from your world, from the world of investing, there are lots of dead companies that have a nice rally at some point on their way to zero and that can often trick investors. The way that tends to happen in massive and largely successful technology companies is the following arc. Company launches a product, great entrepreneurial energy, great customer focus. That product takes off. Maybe there's a second one of those that are really successful. Then if the company gets big, the founder becomes the CEO, cannot be as close to driving that stuff quite as tightly and a professional product team takes over and forgets the inherent, you got to get customer love before you can explode out. On top of it, you've got the CFO and the CEO saying our next wave of products has to be really big to matter. And so you start focusing on making early launches really big before they are genuinely loved and then all of the incentives can go haywire. And I've seen this at pretty much every company I worked with. It's one of the things I obsess about and I'm paranoid about it. NuBank as we start to get really big is making sure that we aren't bouncing dead cats, that we aren't launching products that customers don't actually like where we're milking the NuBank brand, where we're milking the hyper successful initial products we've got rather than adding to the brand and adding to the product portfolio in terms of value to customers. Just take Facebook as an example, you've got a billion plus daily active users. I guarantee you I can launch any product and make it look great for the first six months, the first year. And guess what? I can get promoted that way. Now what are my incentives? My incentives are to make every product look good even when it isn't actually working.

[Harry] So what are the metrics that reveal true customer love in that time period? If you have a smart practitioner who can make anything look good, how do you determine between the true good versus the fake good?

[Jag] The ultimate metric that I try to focus on is churn because if I have a leaky bucket in the early days, I can fill it up because I got a billion daily users I can pour into. But you will never fill that bucket if it's got a bunch of holes in the bottom. And so churn is the ultimate metric. And then you'd start looking for what are the leading metrics for churn. And that's where things like the Sean Ellis survey and even NPS, which we tend to use maniacally and pretty religiously. And if we cannot get a Sean Ellis survey score, that's actually above what Sean himself recommends well above, we're not focused on scaling it. We're focused on iterating it again for love, not for scale. Back to one of your first questions. That's our attempt to turn art into science. I also, we've started going back to having leaders of business product managers, just go call 10 customers who are using the product and just get a visceral sense and don't ask someone else in the customer experience team to call them. You call them because there's no other way. And by the way, bring your engineering lead in that call and just listen to what they have to say. That part is some of the art. Let's just listen if there's real enthusiasm or if they're trying to be polite to you. Human beings are polite by nature and that's great in life and not great for product managers.

[Harry] How do you feel about the segmentation or the category that you're in altering churn by pure category? And what I mean by that is if you are in productivity tools or CRM or whatever, CRM is actually a bad example, but productivity tools or consumer, churn is very high, very easy. Banks is like phone contracts. People don't change banks with as much track like ease and flexibility as they do pretty much every other product, I think maybe homes are being saved. Is churn really a good predictor of love for you?

[Jag] Yeah, I think it's a great point that you make that there are a lot of people who get trapped with companies and being trapped is different than being in a committed relationship.

[Harry] That's called marriage, Jag. You're going to get... No comment.

[Jag] No. What we have started to use for NuBank, and this is part of the journey for us over the last three years, but we have a definition of do we have a customer? Do we have an active customer? And do we have a primary, what we call a primary banking relationship? And we have been obsessed. Our board actually, the first board meeting I participated in back in, and it was March 2020, it was actually three days before pandemic locked us all down. And they pushed very hard on the question of you're growing like crazy, but are these people treating us as a primary banking relationship? And we got obsessed about that since that push. And we've been growing that metric pretty dramatically month over month for three years now. And so there is churn and we pay attention to that. And then there is, are you growing the share of your customers who are primary banking relationship and are we driving, and this is something that Facebook and other social networks use, are we driving our intensity of engagement, daily active users over monthly active users, monthly active users over all customers? That intensity of engagement combined with some of these customer love metrics, you start to cobble together a picture.

[Harry] Can I ask, how do you define primary banking relationship? The big thing in the UK and Europe is do they shift salary inputs from traditional bank to that new bank? Is that the same for you? How do you define that?

[Jag] Altered for the realities of how that payroll deposit works in Brazil and in Latin America, that's basically what we're getting towards is are they moving the bulk of their incoming money into NuBank before they start spending and saving and doing the rest of their financial lives?

[Harry] Can I ask what percent of your users is primary first bank is new bank?

[Jag] Yeah, by our metrics, and we just announced this as part of our quarterly update, we are well north of 40% of all customers, not of actives, but of all customers, and again, that's part of us holding ourselves to a high bar. Three years ago, that metric was about 10%, and so we've grown at 4x even as we've grown the customer base 3.5x in the last three years.

[Harry] Do you want usage? Sorry, I'm just so interested now. Do you want usage? Because on a bank, maybe they have the card, you don't want them to log into the app. They probably log in when there's a problem or when they aren't sure about something. How do you feel about usage when maybe it doesn't correlate to the love?

[Jag] We are keyed first on customer love. The first value of new bank when David, Chris, and Ed founded the company, it's one of the reasons, frankly, I joined back to your earlier question, was I visited Sao Paulo and in the lobby was this very odd statement, value number one. We want our customers to love us, interesting word, fanatically. That's the word we use. We want our customers to love us fanatically, and that is exactly how I think about it. We are keyed off on that from day one at new bank, value one at new bank. We believe in the context of what we are trying to do in our strategy, that intensity of engagement with our app is actually a leading indicator of that customer love, and we've proven that to ourselves. You're right that in concept, someone could never use the app or almost never use the app and still love the product because they're using the credit card or what have you. But for us, what we found is intensity of usage is a lead indicator for customer love. Ultimately, what we care about is customer love, but we track both because we believe one's a leading indicator.

[Harry] Can I ask, speaking of the product, you have so many different ways you can take product with new products, with features, how do you decide what to do next? How do you prioritise product decisions, Jag?

[Jag] Again, it's aligned with what we think allows us to become a leading player aligned with the mission of the company. We want to transform services, including financial services, across Latin America. We didn't say Brazil, we said Latin America. So launching in Mexico and Colombia three years ago was, I wouldn't say a no-brainer, but it was an obvious mountain we needed to climb. We started three, four years ago, as essentially a Brazilian credit card company. David was very clear that we were on a journey from a Brazilian credit card company to being a full solution Latin American financial services company. That's a mouthful, but it's clear. That means you have to have investments. You have to offer insurance. You have to do lending. Latin America consumers tend to consolidate their financial lives with one primary institution. We had made a lot of headway in deconstructing financial services in Brazil, and the reconsolidation with the traction we had was the next play. We've been climbing two mountains simultaneously over the last three or four years, one proving that the model extended beyond Brazil, which was a question three years ago, and second proving that we were more than a credit card company, which was also a big question three years ago.

[Harry] Why was it a question that you were more than Brazil? The cultures are relatively similar. It's not, you know, Southeast Asia compared to the flat town, which is obviously bigger. They are relatively aligned in terms of cultures.

[Jag] To me, it was an obvious next move, and the timing felt roughly right. But this was a question investors were asking. Are there unique aspects of Brazil that make this a one-off? Can you actually do it beyond the country? And there are some technical specifics. The working capital cycle for credit cards in Brazil is very favorable for a credit card company for what we're trying to do, which is almost, compared to anywhere else in the world, Mexico and Colombia are much more typical. So there were some technical reasons why, but we firmly believe that Mexicans and Colombians were roughly as poorly served as Brazilians were, and that the formula we had come up with, both starting with the culture and leading into the products, was something that we could extend. But we had to prove it. We had to prove it, which gets back to execution, go full circle.

[Harry] You're a leader of products. You have resource allocation at the front of your mind. You mentioned all those different new products that you have now, but that you had to build and build out over time. How do you think about the right ratio of allocating between core product, innovation and scaling and growth?

[Jag] Yeah, a bit of art with the rule of thumb. I grew up at Google. We used to talk about 70-20 then there. It's a poor heuristic, but it's something roughly tend to use. We want to make sure we don't lose sight of continuing to invest to improve in our core products and a core handful of products now. We put a disproportionate effort on making sure every pixel is perfect or as perfect as we can make it and improving. We want to make sure we have small teams going back to something we talked about earlier who can do big things. Our insurance or investments or lending team when they started was a dozen people. We're like, go change the world. Go come up with something fundamentally different, not incrementally better because incrementally better isn't going to get anybody to notice. We're going to obsess about you dozen people, come up with fundamentally better. Only when you get to fundamentally better will you get the next round of investment, meaning people. When you launch, make sure there's customer love because only when you get customer love are you going to get more investment at that stage. We try to intentionally limit resourcing with a small group of hyper-creative, hyper-driven people who think they can change the world. That gets back to the autonomy and the drive and the clarity that we talked about earlier.

[Harry] How often do you do product reviews and who's invited to them and any best learnings from what makes a great one great?

[Jag] We do at least a half dozen weekly.

[Harry] Half dozen weekly?

[Jag] Weekly, yeah. Because I'm doing product reviews on one small element of a product. We'll obsess about the details of this aspect of the flow. There are some areas at any one time where I'm personally obsessed by that either they are the next new thing or something might be struggling or whatever it is and I'm in a bi-weekly cycle with team. And then there is just a general slot where we have two or three open slots and we're like, sign up and people will come with, hey, we're thinking about XYZ problem. Making one great is hard because you're balancing two things. You're trying to be inclusive and particularly in a Zoom world, you can include a hundred people in your product review and a very inclusive media works against the second thing you're optimizing for, which is this is a safe space for us to brainstorm together. And frankly, I don't think we're succeeding at balancing those two things anywhere near optimally if I'm honest.

[Harry] Which one would you lean towards?

[Jag] I would like to lean towards having the safe space because a product team should feel safe to come in and not do a dog and pony show for senior management and tell us how everything is great, but tell us where the warts are, what things aren't working, what things they're struggling with and have ideas about what they're thinking about, some of which will be bad ideas. It should be a learning environment. I read once an article about Pixar and how they do it. And there are elements of them we've tried to incorporate or I don't think fully successfully. And what they tried to do is they'll have every director of a Pixar movie be in a room and they will give the current director of the current big movie bet brutal feedback, ruthless feedback, but then it's on the current director to decide which feedback to take. So they get everyone's opinion and then they're very clear about who the decision maker and how do you balance that safe space with that inclusivity? How do you balance senior participation with the autonomy for the person who's actually on the ground making the decision?

[Harry] That's hard. You said that decision. Before we move into a quick fire, David asked me to ask this one. When you review the decisions you've made, what's the product decision that you think is best that you've made at NuBank and what product decision do you think is the worst product decision you've made at NuBank?

[Jag] David's got you doing his work for him.

[Harry] Literally.

[Jag] One of the decisions I think that we did really well and we did it essentially in the first handful of months of the pandemic was we had, first for context, we were a credit card company. It was doing very well. We had about it over the course of the previous year. We had launched our bank account product and it had unleashed this wave of growth. We almost, I think, tripled the number of customers. We had an engagement challenge. Those customers who were getting the account but weren't yet getting the credit card. We were struggling to get them fully engaged. We did two things that I think were very powerful in driving the engagement and the primary relationship in going after. We were pretty decisive about making a shift from being a savings account only. We had a very specific and appealing value prop there to being a transactional daily bank account. That was basically by incorporating every transactional mode there was from debit card, which is the most old school thing you can imagine, which was actually massively successful, to central bank sponsored mobile real-time payment system in Brazil called PIX, which has taken off perhaps other than India, arguably other than India is easily the most successful real-time mobile payment system in the world is in Brazil. We bet big on that and we sweat every pixel of that flow for signup because customers were going to decide who was going to get those for their PIX. We made it as simple to sign up and as simple to use as we possibly could. We drove the team crazy. That was one bet. The second big bet is we launched a secured card product in Brazil, which said, hey, we can't yet sponsor you for credit because you haven't yet proven the credit worthiness that would allow us to do so in an economically sound way, but we can help you get on that first ladder of opportunity by proving that you will pay off your credit card against the security deposit, which is a common product from the US, but hadn't been done in Brazil. That allowed us to start offering credit and get people on their first rung of the ladder of opportunity and credit in the country, and that's been massively successful. Those two things together have been key drivers of our growth, so we're really, really proud of that. In terms of the product, we've struggled for a while to get exactly right, and it gets down to what we talked about earlier about clarity is our suite of products for the upmarket for the high-income segment in Brazil has not had, and I have not cracked myself, the degree of clarity of what makes that offering fundamentally different for that target customer set as we could have, and so that's something that you fail every day, and to be clear, when we say upmarket for us does not include, Harry, you're not in our group. We're not talking about private banking. We're talking about mass awful. We have a play, but it's not a place that we've cracked.

[Harry] Why do you think it hasn't cracked yet?

[Jag] Part of what I tell myself is you're a Brit, so I'll come to you with a Churchill quote. Victory is going from defeat to defeat without loss of enthusiasm. The biggest single reason is we need clarity, and we need a degree of focus on what every product that I've ever been involved with that's succeeded has had one core characteristic. It has been fundamentally different, not incrementally better because the only way you break through the noise is you've got something fundamentally different for a very specific person in mind, and that person loves it so much that they tell their friends, and that's how we have our cost of acquisition so low at NuBank. That's the only way you break through. You cannot take on an entrenched player head on. That's a recipe for failure. It doesn't matter how much money you throw at the problem, and so we've got to come up with something that's just fundamentally different with a clear customer in mind, and that's what we're working.

[Harry] My friend, I want to move into a quick fire, so I say a short statement, you give me your immediate thought. Does that sound okay?

[Jag] That was great.

[Harry] I'm thrilled that we didn't send this out too far ahead of time because otherwise this would be a total deviation from it. Tell me, when do you listen to users versus when to ignore them?

[Jag] You always listen to users, which doesn't mean you follow exactly what they say.

[Harry] I've been fascinated about this one for a while. Have Brazilian banks responded? Have the incumbent Brazilian banks responded to NuBank, or have they just conceded that they've lost that market and that segment?

[Jag] I think the very best incumbent banks have started to respond, and one of the things I'm proudest of, I'm sure David even more, is we have forced the bar to be raised for customers across the board, and so that is something we face, is the bar against which we compete is rising all the time, quite rapidly over the last two or three years through the pandemic.

[Harry] Why do so few fintechs never master credit?

[Jag] I think credit is a very hard thing to crack. It is only partially technology based, and one of the, I think, most masterful decisions that the early NuBank team did was they decided to start on the hard thing first, rather than get going and figure out, we'll crack it later, and therefore it is built into the DNA of the company at a very deep level, and it is a great counterbalance or complement to the tech orientation that we've got as well, and so that combination is a very rare thing to have and a very hard thing to do.

[Harry] What's the biggest mistake founders make when hiring product teams?

[Jag] That's a hard one. I would say that the biggest problem and mistake founders make is they go for the credential of the person who succeeded at big company XYZ, which is a very different mindset than succeeding in a highly ambiguous pre-success mode of a startup.

[Harry] Tell me, what one piece of advice would you give to a product leader starting a new role today?

[Jag] Be incredibly clear about what hill you're going to take. Someone just told me yesterday, actually someone on our board, on the NuBank board, told me priorities is not a plural, and I like that. Be very clear on what is your priority and what you absolutely need to get right and be very focused only on that.

[Harry] What makes David so special?

[Jag] There are any number of things, from the courage as a Colombian to enter the Brazilian market not speaking Portuguese and have the audacity to think you'll take on some of the most profitable banks in the world, but the thing that I've observed that I would most call out from working closely with him is his complete openness to changing his mind. He holds strong opinions weakly in a way that is beyond the sound bite or the cliche of that phrase. He will make a strong argument in one direction and two weeks later, he will tell you why he was completely wrong with no ego and a complete willingness to pivot and it is very hard at times and the more junior you are, the harder it is to keep oriented because there's such an instinct to please the CEO and the founder, but he's genuine about it and he wants to hear the argument and he's genuinely processing the argument, not listening to you to rebut. I learn from that every day and I try to remind myself every day, it's not important to win the argument, it's important for us collectively to be right.

[Harry] The final one, what recent company product strategy have you been most impressed by where you thought, huh, that's smart, I like that.

[Jag] You had Gustav on, I don't know exactly when, but fairly recently.

[Harry] Yeah, true, but once again.

[Jag] The first product I did in Silicon Valley when I joined Google was Google audio and just a failed product in the end, but what one of the privileges of that role back in things, 2000, late 2006, early 2007 was I got to play with the desktop demo version of Spotify back in the day and I played a few Bob Marley tunes and thought this is pretty cool. To see the evolution of that company through mobile, through subscriptions, more recently through podcasts, I now probably spend three quarters of my Spotify listening time, which is at least doubled through the pandemic, listening to podcasts more than listening to music. It's been masterful. He's done the whole series. I know a podcast series on how they did it and how they thought about it and the way they have balanced this keen focus on the customer need with this deep, profoundly deep and Gustav embodies this technical, deep technical understanding of the technologies involved is it's just been masterful. So I'm in awe of what they've achieved.

[Harry] I call him the Daniel Craig of product, body of an Adonis and the voice is just so soft and silky. Listen, Jag, I've absolutely loved this. Thank you so much for being so brilliant and you've been an incredible guest.

[Jag] I've enjoyed it tremendously. Thank you so much for the chance to chat.

[Harry] I just want to say Jag was absolutely fantastic. Just such an awesome and wonderful person, such great energy interviewing him. If you want to see more from us, of course you can on 20VC.com or you can head over to YouTube and search for 20VC.