Max Levchin, Founder & CEO @ Affirm:The Biggest Surprise Scaling to $18.7BN Market Cap

Max, I am so excited for this. It has been a long time, so thank you so much for joining me today.

Thank you for having me.

I would love to start with something you said before. When it comes to hiring brilliant people, you said, “Brilliant people have extreme personalities more often than not.” When I thought about that, I wondered: Do you want them on your team as a result of those extreme personalities?

You mostly do. Everything is a range, so you inevitably find yourself asking the question: How extreme is extreme, and what will it cost me if this person blows up or goes into some sort of behavioral tailspin? But most of the time, truly brilliant people are worth the quirks, and it is a hallmark of a good manager to know how to bring out the best in people, even if they are extreme.

When we think about extreme personalities, they are often dislikable. They are not that amiable and they are very opinionated. How do we think about that being a pro versus a con?

I think opinion is exceptionally valuable. The pinnacle of teamwork achievement is to have really strong opinions, not hold them back, and yet manage not to profoundly offend the person or teammate you are judging or providing feedback to, so that they choose to opt out, whatever that means for the problem.

You want someone who can walk into a room and say, “That is a pile of garbage, and I love all of you. You are amazing people, and I am proud to be part of this team, but what we have built here together is not going to ship because it is terrible. Now we are all going to regroup and rethink and build something better.” That is one way of trying to couch the blow, but if you do not give the blow, you are going to ship garbage.

I do that, and people get offended. They get emotional and upset, and I get told that I am too direct. Help me. What do I do?

Empathy is important. When you are judging someone’s work, there are 2 possibilities. It is bad because they did their best and circumstances intervened, so they can do better, but it was not good enough and they need to know that. The moment is about telling them, “This is not your best work, but I know what you can do. This is an embarrassment, but it is not an embarrassment of your ability. It is an embarrassment of the time you had or the effort you were able to put in, and that is okay. We can fix it.”

Sometimes you look at the work and say, “This person cannot do better. Full stop.” Do not worry about being direct. The conversation you are going to have with them very soon is, “I do not think you have a place on this team anymore.” That is a much harder conversation, but you have to have it too.

You always said to me—and I know you said this years ago—“When there is doubt, there is no doubt.” I have always loved that and always think of it. On the extreme element, where does extreme good transition to extreme bad? Have you had that before, and what have your lessons been?

Any form of narcissism, selfishness, or self-orientedness is probably the most important trigger for turning an extreme personality from good to bad. Said in the opposite way, you can be as great at throwing tantrums or as direct in your criticism as you want to be, if you maintain a good degree of humility. You have to remember that the second it becomes about you being better than everyone else, everyone else will start hating you.

If you are their boss, they are going to hate you quietly, and that is the worst possible situation. What you really want is to constantly remember that you are just a part of this team. When you are criticizing someone’s work or telling them that it has to be redone, you are always redoing the work together. Maybe you are not personally going to move the pieces, but you are going to have to be involved, because they are not going to be happy with a second round of, “This is still garbage.”

I think there are many ways in which extreme personalities flip into being toxic, so you have to constantly watch out for it. But it is the price you pay. Brilliant people are often over the top, and that is part of the ingredients that make them brilliant.

When brilliant people flip into being toxic, is there a redeeming route back, or is that generally a one-way road?

I believe in second chances in general in life, except in matters of integrity. The hard red line for me is always, once I cannot trust you, I can never trust you again. From that point, in my worldview, there is no walking back.

That is not true with kids. Kids lie because they are exploring what to do, but even with my own kids, I have had very difficult conversations where I said, “Look, this is something that, in my professional and personal world of adults, is the hardest thing to come back from. The fact that you just told me something that is not true is really hard for me to hear.”

You are not going to talk to a teammate like you are talking to a 13-year-old. You are going to say, “Look, I cannot trust you ever again. Get out.” Other than that, I think there is a way to walk back from a lot of situations. No matter how crazy or toxic a situation could be, the judgment you have to make as a leader is: Is the person profoundly toxic, or did the situation turn toxic?

Situations come and go. People do not really change that much.

You said earlier, “Has the person done all they can to execute on the task and achieve what they want?” Where I struggle is that we have differences in expectations around what one should and can do. If you are a head of sales, there is always more you can do. You can send another gift, click “thoughtful” on a comment, or provide an additional resource. I will do everything to fucking win, and people on teams do not always think as bluntly or obsessively as that. How do you think about that alignment between what is rational to expect?

My brilliant wife frequently reminds me that, more often than not, it is about incentives. The reason you are so obsessed with fucking winning is because it is your company, your project, and your passion. You have been doing this since you were 14 years old, and the only goal is to do more of it, bigger and better.

You do not need motivation. Your incentives are aligned. You were essentially born with a microphone in your mouth, trying to get bigger. There is nothing to be ashamed of in drive and grit. They are the other side of the coin, and they are rare. People who have them often end up being entrepreneurs or doing something resembling entrepreneurship, because it is such an unstoppable force.

If you do not have it, it is something you cannot really teach, or at least it is very difficult to teach. On the other hand, there are plenty of other people whom you not only need but require in order to achieve your goals. No one is an island in pursuit of their entrepreneurial ambition, and you have to figure out what motivates them.

In many cases, their motivations are not at all relatable to you. You decided at some point that you were willing to work Saturdays and Sundays to succeed, and that was it. Some people go to the pub and others go skiing. Everybody has their thing, and you decided, “No, I am just going to work.”

Having that conversation with someone who has just joined your team—“We are all going to work Saturday and Sunday because I want to succeed”—is not going to do anything for them. What you need to do is figure out what they are looking for and what their goal is.

If their goal is to go skiing 3 days a week instead of 2, you may have made a hiring mistake. But if their goal is to feel great about their work, the investment you have to make is not beating them up about the quality of their work or telling them to work harder. It is saying, “I know you are motivated by quality. Let me spend all available time getting to a place where you look at every pixel you put down or every sound you record and you are falling out of your chair about how good it is. It cannot get any better. I know that is what motivates you, and that is what we are going to do together.”

That is what team leadership is all about. The difficult part of leading teams is deciding that something is irredeemable, because you want people to have second, third, and fourth chances. At some point, though, you have to say, “You are great, but you are a hiring mistake, and I am so sorry.”

At what stage are you forced to hire B-team players?

I find it absolutely absurd that we think we can continuously hire only A players. A players, by definition, are exceptions. You can do that for the first 50, maybe 100 people—I still do not think you can do it with 100—but when you are the size of a firm, it is impossible to have only A players.

When are you forced to hire a B team, and how do you think about that?

The story about knowing the letter grade of every person you hire in your company, whether it is 2,000 people like Affirm or I do not know how big you are, is just silly. People are coached, trained, and fine-tuned to be exceptional interviewees. Somebody comes in and tells you their life story, and they have probably watched an interview you conducted with someone who is a professional interviewer and spelled out what it looks like when you are hiring an A player.

Someone is going to say, “An extreme personality is very important. Max Levchin said so. How do I make myself more extreme?” I do not think you can really judge someone’s overall A or B grade at hiring time. You can assess someone’s coding ability, but a really good assessment test requires a take-home assignment.

You can implement quicksort on camera very quickly, but so can a C player these days. You have to give people a lot of time, and when you give people a lot of time, they will inevitably stray from the rules of whatever the test is. You can filter out C players and definitely filter out D players. Differentiating between B and A is what happens once you have brought them in.

You might say, “Oh my God, that was a brilliant hire,” but more often than not, you say, “I thought they were going to be a great hire. They are a good hire. I am happy we have them.” The question then is: What do you do to get the absolute best work from this person?

The Steve Jobs quote, “A players hire A players, B players hire C players,” is profoundly true. There should be an asterisk explaining why it is true, and the reason is fear. A players love working with other A players because they know the demands of the job, and the team dynamics make them better.

You want people who challenge you and tell you, “I know you are an A player, but that was not good. Let us redo this together,” or, “You go redo it.” A players just want to accelerate away from the pack.

B players frequently say, “Look, I am a worker bee. I know I am not an A player, and that is okay. I am just going to grind out good code, art, sound, or whatever.” But some of them are fearful and say, “If I have too many B players around me, let alone A players, I will look terrible. It will be obvious that I am not that good. There are people here who are a lot better than me.”

The way to fix this is for them to become a manager and hire a bunch of people who are not very good—much worse than they are—so they look like the tallest mushroom. That is the real danger of B players. Having people who are perfectly willing to do good work day in and day out is not the worst thing in the world.

There are plenty of things that A players will turn their noses up at because they are not challenging enough or accelerating them out of the A-player pack. B players are very good for that, and a lot of B players, over time, grind their way to becoming A players. You are investing in their future, and they know it, so that is also positive.

But B players who are afraid of being found out are the ones you have to watch out for in terms of the actions they take.

I love the statement you made: “We take calculated risks. Do the calculating.” When you review the journey, what was the biggest calculated risk you took, and, whether it succeeded or failed, what did you learn from it?

When I started the company, I sat down with a bunch of bankers and said, “I am going to start a company where I am going to do away with this imbalance. I am going to eliminate fine print so consumers do not have to read fine print to find out what the business model is,” along with several other ideas like that.

Several of them, including some very well-known people, basically laughed at me and said, “You are insane. This is not just the way you make money in this industry; it is the only way you make money. This industry has been competed down to the point where the only money you have is in the form of things that the customer does not understand, because everything else has been competed away.”

I said, “I do not think so.” I did not have a full model in Excel, in my head, or on paper proving that I was right. This was not a calculated risk. But I do not think it is worth running a company that fits neatly into the mold of “screw the customer and make some money.” I wanted to run the one that says, “Never screw the customer. Let us see if we can make some money.”

At this point, we are public, our financials are visible, and we do very well financially, so it was totally possible. This was a non-calculated risk at the very beginning of the mission. I simply decided that we were going to do it. The calculus of the first bet was not a calculus at all. It was, “I think it is going to work, but I am okay finding out that I am wrong.”

The second part completely worked out. If you look at the quality up and down the stack of Affirm, it is really exceptional, in part because of this.

What calculated, highly calculated risk did not work out, where your process was right but the outcome was wrong?

If you did a good job calculating, you have the right hedges, so even if it does not work, it works. That is the stock answer and the PR answer.

Well, I mean, it is also factual. If you have done a lot of calculating and it completely backfired, is everything calculable?

No, absolutely not. There are many things that you stare at into the void and ask yourself, “Will it work out?” The honest answer is, “I have no idea.” You try to minimize those, but they exist.

What strategic decision did not work out, and what did you learn?

One of the other things I must have said years ago is that I do not really believe in regret. I think it is a wasted emotion, so I have a hard time recalling profound moments of, “I should not have done that.”

Do you not think that is where lessons derive from?

I do. We have a whole culture of postmortems at Affirm, where we record every decision we make. I will give you a very tactical example.

The word “postmortem” and the culture around it come from engineering. I am an engineer by training and by past, so it is easy for me to reason in those terms. Years ago, we decided to very rationally scale our infrastructure because we expected a giant launch with a crazy amount of transactions. It was going to be 10 times bigger than what we were at the time.

We were preparing at breakneck pace and scaled up every part of our backend infrastructure, then our frontend infrastructure, and everything else. On the day of the launch, it fell apart massively because we overloaded our backend with our frontend. We were all dressed up with nowhere to go, and the entire database and data layer of Affirm was in shambles for several hours while we cleaned it up.

It was a self-inflicted wound. We did not know how big the launch was going to be, so we assumed it would be super huge. It was very modestly sized at launch and turned out to be a disaster.

What did you learn from that?

“Do the calculating.” This was not an unforeseeable error. Someone should have sat down and said, “We are scaling the front tier faster than we are scaling the back tier. At some point, given the architecture of the system, we will have what is called an oversubscription of frontend connections to the backend.”

The backend will start choking, because even if the frontend does nothing, simply holding those connections is expensive. If you have a 10-times mismatch, you are going to bring down the database. That is what happened. Someone should have done that work, and that would have been a calculated risk that would not have backfired. This time, it did. It was a long time ago.

You mentioned the postmortem culture. People always say you learn so much from your failures. Do you learn more from your successes or from your failures?

Failures, for sure. Analyzing success is boring. It is, “We did well. It is good. I am tired. I am going to go to sleep.” That is the natural reaction.

Could you not learn the same thing from success? We just do not analyze success because it was successful. We say, “Move on.” If we analyzed why we were successful, we might see that we were thoughtful about something like customer acquisition cost over time.

The problem is that with successes, the texture of each individual decision is much harder to detect. If you look at the world as a funnel analysis, which is a really good way of thinking about a lot of things, you start over here by becoming aware of something, then start thinking about it, considering it, deciding, and finally succeeding or failing.

There is a real blur between the first 5 steps. This is true of web apps, checkout funnels, advertising, and all kinds of other things. If everything has gone pretty well, there is a lot of effort that has to go into saying, “That step to the next step of the funnel had a conversion rate of 95%, and the next one was 95%, and the next one was 95%. Which one did the best?” I do not know. They all seem like 95%. That is pretty good.

You should ask, “Could it have been 96? Why did we do better?” You are inevitably asking the question in terms of failure again. You are not saying, “How were we so good and smart that 95% was really high?” You are asking whether we should have done 94% and whether that would have been enough. Those questions are boring and pointless to ask.

I really believe in asking what we should have done better, even if it is the most successful thing ever. It is nice to take a victory lap and high-five the team, but ultimately the question has to be about what we could have done better.

I have a friend who grew up in a very managed parenting environment. When he showed up home from school with a 98% result on a test, his parents would ask, “What happened to the other 2%?” I am not sure that is a great way to have a childhood, but it is a great way to run a company.

If I am a founder listening to this and thinking, “I want to have a postmortem-centric culture, but I do not really know how to do that,” what advice would you give me on how to implement one today?

For everything, create a document space. Every time we have a blip—whatever the definition, whether it is in our systems, processes, or models—a separate conversation is carved off in Slack, and a separate conversation is documented in a Google Doc, Notion, or whatever tool we are using for that topic.

It is segregated. It is very difficult to peel away the details of event 1 if you are commingling them with event 2. The first and most important thing is to make sure you have separation of data streams.

If you get good at postmortem culture, you will gather a lot of data, and a lot of it will be irrelevant. People will say, “This also happened at the same time. I see it in the logs. Something just went down.” Inevitably, you will discover that it is unrelated and happened because of something else.

You want your data to be plentiful. You want as much of it as possible segregated into a separate pipeline of raw content as soon as possible. You want to preserve that information in a special place, and you want someone responsible for writing it up and distilling it from pages and pages of raw data into essentially a white paper.

Then you want the entire team to comment on it. The most important thing you can do throughout the process is completely separate any form of apologizing or taking responsibility. You cannot say, “We should do better next time.” You have to be clinical about it: This happened, and it was not good. Here is why.

You have to give the people who feel directly responsible for the flop the freedom to comment on what happened. You need to encourage them to acknowledge that they made a mistake, but without any sense of, “I need to make myself look better,” or, “I need to cast a certain light on the decisions I made.” Get rid of all emotion and describe exactly what happened.

Do you have a structured time for these postmortems? Is it, “We have a weekly postmortem on what happened last week”? How do you make sure it is not just a nice idea that you then fail to implement?

You have to give people responsibility for doing them. You need a directly responsible individual for authoring and presenting the postmortem. That is very important. You also need to make sure their manager knows that their list of responsibilities has now been enhanced by this postmortem, and that they are on the hook for it.

We typically say that a week, or maybe 2 weeks, is the most you are allowed to take to generate a real postmortem. Then there is a weekly review within the teams that are conducting the postmortem. They go through it and say, “This happened, and the postmortem has been prepared.”

Someone even more senior will then say, “That is a reasonable explanation. I buy it,” or, “This is not quite deep enough. We need a better analysis of what went wrong.” They might say, “I do not think this is accurate,” or, “You described what happened, but you did not explain it.”

That is where you need your A players in the room. They have the experience, intuition, and past knowledge to say, “That is a deep enough postmortem,” or, “You just described what happened.”

What was the most controversial postmortem, where opinions about the causes that led to the outcome were highly differentiated?

These days, the ones I see are very matter-of-fact. It is never about the individual; it is always about the events. It always comes down to this: Something happened, and a choice was made either by a program or by a human.

If it was a program, it was a bug or a poorly calculated outcome. If it was a human, it can be a judgment error. Someone might think, “I will flip the switch, and the lights will go on,” but it turns out to disable power for the entire city. That is a conversation we have had here.

It is never controversial. It is in the past, and we need to do better. You add a big guard to the switch that says, “Before you turn off the lights for the entire city, please read this warning.”

Speaking of the decisions that lead to the postmortem, you said before, “Make reversible decisions fast.” It makes me think of the one-way and two-way doors.

Honestly, I think everything is really a one-way door. There are very few things that are two-way doors.

How do you respond to that?

I do not agree. Things that change your reputation, change relationships with partners, or do that sort of thing are basically irreversible once you have done them.

When you look at the transience of news and reputations, you can look at Chamath on All-In talking about the Uyghurs and saying that it was below his line. A week later, he was number 1 on Spotify. Chamath is Chamath, and honestly, it has not impacted his reputation one bit.

There are probably some people who would disagree with you on that topic. I do not know exactly who they are, but I think we can guess.

I agree that superficial statements—even explosive or incendiary ones—fall off the news cycle faster than you would expect them to. I think people’s reputations recover, get managed, and get manipulated, and maybe it is true for companies too. But I think rational people have long memories and do not forget. They are not easily persuaded by being in Spotify’s top 10 or by whatever is on the front page of The Wall Street Journal today.

The decisions I care about much more are architectural decisions. You can decide to invest a year in rebuilding your systems in some new fashion, and if that is a bad decision, you are going to look back and say, “It was a terrible call, but we are too far into it. We cannot just uproot it. Too many people would be upset.”

Or you agree to a deal that is economically nonviable and find yourself staring at a giant black hole. You want to get out of it, but you signed a contract and cannot. Maybe it is not that much money, so you do not care, but sometimes it can be a lot.

I believe in one-way doors, and part of postmortem culture is identifying one-way doors that you thought were two-way doors. You say, “We should not have done that, because it turned out not to be something we could easily come back from.”

I agree that a lot of decisions that seem like one-way doors are not. That is probably a very fair statement. Is there a decision you thought was really reversible—“Make it fast”—that turned out not to be so reversible?

I think for every 10 decisions I thought were irreversible, only 1 or 2 were actually irreversible.

You have also said something that I love: “Care about how we make things. Mind the quality of the invisible parts.” I would love for you to expand on that.

That is another reference to Steve Jobs. He was obviously such a brilliant product person, and I aspire to be a good product person myself.

There is a repeated story about how they signed the insides of the motherboard on, I think, the first Mac. The team put their signatures on it, and the narrative was that if you build something beautiful and amazing, you do not mind—in fact, you take pride in—signing the insides, because if anyone ever looks, they will be blown away by how beautifully the whole thing is designed. That is where you want your artist’s signature.

There is a natural and healthy tension between shipping fast and building things beautifully. Building things beautifully requires design, time, and consideration of what the next version will look like, so you do not have to rebuild from scratch.

There is no hard-and-fast rule for knowing what matters so much that you want to do a great job. You will always have a team cleaning up, refactoring code, or dealing with technical debt. That is a property of every startup that has been a startup for a decade. You undoubtedly have someone refactoring part of your code because it has grown enough moss that you need to deal with it.

But there are parts that are obviously important and valuable. It could be as simple as a basic UI decision or as complicated as a machine-learning model that deals with credit underwriting, which is one of our crown-jewel achievements. We have some very advanced ones.

Those things matter. You cannot let them be anything but the absolute best you can possibly produce. When somebody tells me, “We are doing X, and it is going to be a month late,” my natural reaction is to jump out of my chair and say, “How can it possibly be this late? How dare you be so late with this thing? Why could we not ship it yesterday?”

But I absolutely make exceptions for things I know are going to be crucial for months and years to come. That is where you have to mind the quality of the invisible parts. A lot of the time, these are in the backend, whether it is credit, scalability, infrastructure, or something else.

It does not make it any easier to hear someone say that a project has slipped by a month because something has to be redone, but it is important, and you try to cope.

How do you feel about the concept of the lean startup and the MVP? Ship when it is ready. Get the first thing you can into customers’ hands. Does that lead to a generation of garbage software, or is it actually a great way to get customer feedback?

There are definitely things you want in customers’ hands as quickly as possible. We have a prototyping team at Affirm that is trusted and empowered to ship things that are not beautiful or production quality, simply to get customer feedback.

It is really important to find out whether some crazy idea we have has any odds of success once it is out there. You get data, and if the data tells you that it is great, you do not just say, “Cool, let us leave it out there and move on to the next one.” We pull down the feature or product and say, “Now we are going to do it right. We are going to make it scalable, make it beautiful, and make sure it stands the test of time.”

The idea is to find out whether the idea was really important. But if you believe that ideas speak for themselves so much that the product can be garbage, you will end up with a collection of garbage and a generation of garbage software.

You have to know when the moment comes to say, “That really does have traction. Someone will build a better version of this if I do not do it myself.” That is what you have to do.

Do you think speed is the most important determinant of success in startups?

It is one of the most important. It is very rare that I can think of more than a handful of companies that achieved huge success and huge scale by just rolling along.

I am sure I do not know the inside story, and what seems to me like rolling along was probably not rolling along. They were probably feeling like it was breakneck pace on the inside; they just did not show it very well.

Some founders and CEOs have the duck-foot property. They look composed above the water, but their feet are moving at a crazy pace below the water. I am not one of those. You can always tell if I am trying to go faster from the feet up.

But there are people who are capable of maintaining composure, and maybe their teams are composed too.

It is interesting that you said you are a writing culture. Do you favor short, pithy 1-pagers over novels or live rants? I personally like a live rant, being born with a microphone in my hand, as you kindly suggested.

What does a writing culture truly mean to you, Max, and why do you think it is so important?

It feeds back into that notion of postmortems. We have a postmortem repository where people can review every major or minor thing we have done going back to the beginning of Affirm and learn from it.

It is really valuable to have a library of case studies internal to the company saying, “That was a screw-up. The great outage of whatever year—what happened there?” We managed to overload the backend with the frontend, and so on.

It is valuable to have these things not as oral tradition but written down. To do that well, you have to have a writing culture where people are comfortable typing up the random words that come into their minds and saying, “This makes sense in my head, but I do not know whether it will make sense to you.”

You need some baseline quality requirement for writing. The most important thing about technical and business writing is simplicity. If you are reading a 12-page document describing a feature, you are probably going to miss details. If it is a feature meant to be confined to 25 by 25 pixels, the description should be 2 or 3 pages.

You should not have many pages describing the impact on the world that you expect to have, how it relates to world hunger, and so on. Pithiness is important. Live rants are good as motivation sometimes, and they can align the team around a story and a strategy that the story represents.

But in my experience—I host an all-hands meeting at least once every 2 weeks, and often more frequently—you end up saying 12 things and people remember 2. The shorter the better. Being short and pithy is much more valuable, even when you are speaking.

In writing, people simply run out of time. We have distractions and notifications popping up. If you have a 12-page document to read, you think, “Oh, man.” Some people are very good at focusing, but I think the majority of people are less focused than they would like to be and less focused than they think they are.

Are you back in the office now?

It is quite hybrid, and it really depends on the geography. We have offices worldwide now, which is pretty cool. By offices, I mean places, because some places do not have an office.

World headquarters, notionally, is still San Francisco, so that is where I will be today. I think it will be about half full, maybe a little more than half full. Our other East Coast office in New York will have standing room only. There will not be enough seats or cubby holes to hide in because people will be there in such quantities.

Everything in between is represented too. It is not true that we are fully back.

Would you like to be fully back if you could? Every CEO I speak to either says, “I wish we were remote,” “I wish we were fully back,” or, “We are fully back.”

I think there is a ton of value in being fully back. Again, like everything else, there is texture to this value. If you have a C-level team, one way of pushing them into the B level is to make sure they all sit together and are surrounded by A players who constantly tell them, “Go better. Do more.”

If you have a great team, some people are actually more productive when they can cut out all distractions. They can hide themselves in their spare bedroom and write code for 12 hours, design a product for 12 hours, or do whatever it is.

I care a lot more about output and productivity than I care about where you are physically. The thing I am completely non-compromising about is that there has to be enough time together.

The notion of everybody being back in the office, even if all of your meetings are on Zoom because you are talking to our team in Madrid from San Francisco, is hard to argue for. It is not a productive use of commute time. If you live in the East Bay, ride BART into the office, get into a cubby hole, and then Zoom for 12, 5, or 6 hours, it is hard to make the argument that you should have been on that train in the first place.

The argument that you have never met your team but are just as productive in person, while only seeing them on Zoom, is stupid. It is simply not true. You form relationships, understand each other’s decision-making styles, break bread together, and form camaraderie. That has to happen. It is entirely non-optional.

Anyone saying, “I can be fully remote and these people never have to meet in person,” is kidding themselves.

When was morale lowest in the company, and what did you learn about leadership from that period?

We had to do a layoff about 2 years ago, and that was the lowest point for me and, I think, everyone around me. Those are never easy. I have had to do more than 1 in my life, and every time it does not get any easier. It is always brutal.

How did you learn anything about messaging from the first time to the second time?

The very first time I had to do it was certainly not at Affirm. Every time I have done it, I have said, “I hope I never have to do this again.”

The first time, I did not know what I was doing, and I was terrified of owning the responsibility that I had screwed up. I was the CEO every single time I had to do a layoff, and the first time you feel like everyone is going to blame you. It is your fault. You want to run and hide. It is terrible, and you cannot wait for the moment and the day to be over.

You just want not to be there. A good friend of mine, who at the time was an executive at that company, took me aside and said, “This sucks for all of us. You are the leader. You have to be out there helping people pack their boxes. You have to be part of the goodbye.”

“This is not a death sentence. It is a professional event. It sucks no less than any other major trauma, but you can play this from the comfort of your office or in the middle of the floor where everyone is crying. Go be with the people. It will feel better in the end, and they will feel better in the end.”

That is what I did. It was somewhere between cathartic and therapeutic, and it was a really important lesson to learn. You cannot hide from the grief that a layoff creates.

The other thing I learned over the years, from watching my friends have to do it and certainly from having to do it myself a couple of times, is that empathy is really important. Being there with people in person, helping them pack their bags, and giving them a big hug is important.

You will also find that if you have built a great culture, the blow is much softer than you think it will be. This is not something you can fix in the moment, but if the culture of the company is great, people understand that you tried every possible strategic or tactical idea to avoid going through this.

This decision was the last decision in a long list of decisions you could have made. People are inevitably more understanding and less hard on you than you are on yourself.

One of the most amazing and emotional events for me during the Affirm layoff a couple of years ago was that every single person I talked to said, “I get it. This happened. We grew too fast and over-hired. I understand. I even understand why me and not the next person. I hope there is an opportunity to come back. I really love this place, and I want to come back.”

Is over-hiring a calculated risk?

Yes, it certainly is. Actually, that might be a good answer in retrospect to your question about calculated risks that did not pan out.

You decide that you want every incremental engineer to build more features while fundraising is good, and then suddenly it is not. What do you do? There is nothing you can do other than say, “I am sorry, but some of you cannot have a seat anywhere.”

The theory was that the features would pay off and return cash, but they did not. Or it was, “We will have a continuous cash tap from fundraising,” and that did not happen either.

I do not think it is ever that simple. The analysis is typically about how fast something will grow, how well you can have the market protect your cash flow if it does not, and how profitable something will be. It is roughly what you said, but the calculus is a little more complicated. You are typically estimating more than a few variables.

One of my favorite internal lines at Affirm is that it is a constant multivariate-analysis problem. It is very rare that the question is simply, “Is that number 5 or 10? Is it going to be 500?” Those are easy to estimate.

It is more like, “We have 9 variables, they are all nonlinearly distributed, some multiply with others, and we do not really know how some of them relate to the first group. We are going to have to estimate all of them and predict the outcome based on that.”

It is a hard problem to solve, not just for Affirm but for any company.

You mentioned growth. Can I ask a blunt one? Is Klarna growing faster than Affirm, and how do you think about that in terms of Klarna’s U.S. expansion?

I think they announced that they grew 32% last quarter. I think we announced that we grew 35% in the last few quarters. Their numbers are not public, but they are about to go public, so it will be a little easier for everyone to track them.

Based on estimates from the last several quarters, we have been taking share in the market against all competitors, not just Klarna. We offer more services, whereas the majority of our competitors specialize in paying for something very specific.

We made the decision from day 1 to provide every possible service, from a 45-day, super-short-term loan all the way out to 4 years. Very few people do everything. It should be a natural consequence of our ability to serve all those needs that we take a disproportionate share.

If entrance into the market is pretty well done, is the size of the market the issue? Is it a matter of not grabbing the opportunity from your side? It is a big task to go after the U.S. market from Sweden, of all places. How do you think about that? Is it, “We dropped the ball,” or is it, “It is a massive market”?

You are asking me to analyze their strategy. I think that is a postmortem they will have to write for themselves. I am not sure I am qualified, because I do not have all the information.

I also think it is very healthy for startup CEOs, of which I am one, to spend very little time on other people’s strategies. Competitive analysis is valuable at the product and sales-strategy levels. Asking, “What do they do that we could do better?” is a really good question.

Asking, “What is their strategy?” is like asking what the weather will be 12 weeks from now. You can say, “They have been really smart in this part of their go-to-market strategy,” or, “That partnership was clever.” We do a lot of that.

We tear down all our competitors’ products and sales efforts—everything we can get our hands on—and learn from it, because it is free. Someone is making mistakes for us, and we should learn from those too.

Is regulation a friend of buy now, pay later, or an inhibitor to growth?

We are always on the record as being pro-thoughtful regulation. I think you should not be paying any late fees, so I was very happy about the late-fee cap. It is a good development. People should not build their profit models around simply raising late fees.

Generally speaking, I am pro-thoughtful regulation. There is always the risk of regulators deciding, “I understand this. It is black and white. Do this and do that,” and having a chilling effect on the industry. Products that should have existed can become nonviable.

Historically, I think we have mostly stayed on the right side of these decisions, so regulation has generally been positive. But the best thing you can do as a company, certainly one of our size, is engage directly with regulators and speak very clearly: “This is a good idea. This is a bad idea. We would like to offer you some feedback.”

They are human too. They make mistakes, and you want to engage with them.

Do you like being a public-company CEO?

Some days, yes. Other days, no. I think it changes. One day you are thinking, “Yes,” and another day you are thinking, “No.”

The days when it is yes are when it is not very different from being a private-company CEO. Being public generally gives you access to more capital and exposes you to really smart investors with whom you can engage.

There are investors who will only invest in public equities, and they are brilliant people. They have seen it all before and could write books about what a great public-company CEO looks like. You want to spend half an hour with them and say, “Teach me how to be as good as this person.”

If you are a private-company CEO, they might take your meeting, but they are probably busy. If you are a public-company CEO, they might own your stock, and then they want to talk to you because they want to assess you as much as they want to help you.

That is a cheat code for accessing some of the best brains out there. The days when you are thinking, “Oh, man,” are when you have to read a filing that the company has to make, because it is your responsibility to understand everything you are putting out into the public. You cannot avoid that.

Those are generally small prices to pay, but you do have added responsibilities that you did not expect when you were getting a computer-science degree.

Is your mental health tied to the market cap of the company? It is so difficult to detach yourself from the valuation of your company. What have been your biggest lessons on that?

The short answer is that I am sure it is, but I am also sure I put an enormous amount of effort into making sure it is not. It is a balance you are constantly trying to strike.

What is harder is when you think you are doing a fantastic job and can look at the internal numbers and the external numbers and say, “I am absolutely crushing it.” Sometimes it is tempting to think that the market should know by now and give you high marks, which means the stock should go green, green, green.

More often than not, you have a temporal mismatch. You are doing great, but the market thinks you are not. If you are a startup or a relatively high-volatility stock, you end up thinking, “I did great. Why am I being punished?”

The short answer is that it is the market. You do not get to tell it how to react to good news or bad news. There is a great quote attributed to Warren Buffett, although it actually came from his mentor, Benjamin Graham: “In the short term, the market is a voting machine; in the long term, it is a weighing machine.” I take great comfort in saying that to myself whenever I think the market has misjudged me.

I have also gone to great lengths to make sure I do not accidentally stumble upon my ticker. On my Apple desktop, if I click in the top-right corner and the stocks drop-down appears, I cannot see my stock ticker. I do not accidentally discover that today is a red day or a green day.

You have to think long term. If you catch yourself thinking short term—by short term, I mean, “Let us have an amazing quarter”—you are in trouble.

Do you see a dip in employee morale when the stock price goes down?

I do not, but I make it my business that every time I talk to our employees about the outcomes of the company, I take great pains not to talk about the stock price.

The only time I talked about the price of our stock to employees at length was when it was flying very high after the IPO, during the heyday of 2021. I repeatedly said, “Do not get accustomed to whatever numbers you are reading on the screen today. We will not know what this company is worth for years.”

“If you think it is too low, good for you. If you think it is too high, it may well be. Just do not look, and do not develop the conviction that this is correlated to some tangible metric.”

First, we have to become profitable. Then we have to show revenue growth rates, and then we have to show net-income growth. In the long term, that is what the weighing machine weighs. It is not weighing sentiment.

We are going to do a quick fire. I will say a short statement, and you give me your immediate thoughts. Does that sound okay?

Sure, but sometimes I will take a very long time. You will have to edit out the awkward silences.

That is the choice of podcasting. What do you believe that most people around you disbelieve?

I have this conviction that I can learn anything very quickly. About 8 out of 10 times, I am right, and 2 out of 10 times, I am completely wrong.

What have you recently been wrong on?

Two years ago, I decided that I was going to read the paper “Attention Is All You Need” and become an expert in large language models. I was going to read all the papers after that.

I am still in the process of getting anywhere near expert status. I try to keep up, but there is a lot of really interesting stuff happening in AI.

I had Marc Benioff on the show about 12 hours ago, and he said we are getting to the upper limit of LLM efficiency. Agree or disagree?

Agree, but I do not think we are going to stay with LLMs forever. I think we are going to expand the systems that use LLMs.

Reasoning models are already not pure LLMs. They have iterative processes within them. I think we will invent more really interesting things around that.

An obvious idea, which I know has been explored and will be explored more, is the notion of multiple competing LLMs essentially debating each other to arrive at the right answer. That is not a pure LLM model.

You can amplify human intelligence by having 3 smart people in a room instead of just 1 person.

As a public-company CEO, do you have to have an AI story today?

I generally dislike stories that have no substance. You have to ask yourself, “Are we taking full advantage of the best and most interesting tools available right now?” The most interesting tools available right now are AI.

If you look through your closet and say, “We are not really using this stuff, but we need a story for the market,” you are doing 2 things wrong. You are telling a story that is not true, and you are not using the most interesting tools available.

If you are using the most interesting tools, your investors might be happy if it tells a story. But if it does not, I do not think they should care that much. As you go public, a lot of this becomes: Are you growing at the right scale? Are you taking appropriate risks? Are you managing to profitability?

The storytelling hour of, “By the way, we have this shiny item that we are so excited about,” feeds the voting machine. I am sure some of it is important, but I spend a lot more time feeding the weighing machine.

Max, what would you do if you knew you could not fail?

I would do exactly the same thing I do now. Before Affirm was a thing, I made a list of difficult problems that I thought required solving. I decided that financial services, with money at the core, was a problem I was well suited to solve, given the experience I bring.

If I knew I could not fail at anything, I would revisit my list of hard problems. It inevitably included things like food, water, shelter, and energy, and I would solve them all.

If you could start a company with only 1 member of the PayPal Mafia again, who would it be?

Peter.

Why?

He is amazing and brilliant. More so than Elon? Elon is brilliant too. If I had a choice between starting a company with Peter or starting a company with Elon, it would be a very hard choice.

I think Peter is a more known quantity for me, which is why I answered the way I did. We started the company together. Elon’s company merged into ours to form the modern PayPal, so I know exactly what he is like under pressure and when he is stressed, trying to raise money and not succeeding for a time, and dealing with personal-life issues.

I have seen him go from his mid-20s to being older.

How has he changed over time?

When you observe someone closely for a long time, you do not notice the change. You see the episodes, but you do not know how the person is really changing.

We have all grown up and probably become more measured and slightly more careful—maybe slightly less crazy, or maybe slightly more crazy. I do not know. He has not changed very much.

The thing that makes Peter Peter is actually very similar to Elon in that sense. There is a really interesting echo between the two. They both have the amazing quality of bringing out the best effort in the people around them.

They somehow make you feel that you can and should do better work. Conversations with Peter always left me thinking, “The problem I thought was too hard to solve—I should go give it another try. I think I can do this. I think this is actually doable.”

He is genuinely excited about you solving a problem through the capacity that you have. That is an amazing thing to work with someone on, because they fuel your desire to stay late and do more work.

What is your favorite consumer brand, and why?

I love Airbnb. It is not only because the overall brand is amazing and it is obviously a super-successful company. Their tagline, at least the one they promoted for a while, was “Belong anywhere.”

I think that is such a profound, 2-word brand statement. You know exactly what it means and exactly what they stand for. It is evocative. You imagine yourself in all kinds of crazy places—in the Himalayas or the Amazon—and think, “That is an amazing story told with 2 words.”

That takes me back to my love of brevity. In the old world, my favorite brand is probably the one I look at every morning. I am a huge fan of coffee and bicycles—not at the same time, frequently, but sometimes at the same time.

Every morning, I roll out of bed, stumble downstairs, and make myself an espresso. My favorite espresso-machine brand is La Marzocco. I do not actually know that much about them—they are an Affirm customer, so I have to give that full disclosure—but they were a huge favorite of mine long before I convinced them to give Affirm a try.

It tells this amazing story of Italian manufacturing: someone with a mustache smelling the coffee fumes and saying, “This is a perfect espresso.” It is a story in a cup. You think it must go back hundreds and hundreds of years, to when they were making coffee in Venice with a mortar and pestle or something.

I do not think it is nearly that old. I think it is maybe 60 to 100 years old, but it is this timeless coffee awesomeness that I love. I use it every day, and every time I look at it, I think, “Yes. I love my coffee machine.”

If you could be CEO of any other company for a day, what company would you be CEO of?

I have all these unfulfilled dreams of running a much smaller company that makes widgets. I spend my entire life making things out of bits, and I would love to make things out of atoms.

One of my favorite executives came up with a brand name for a company that would sell pickles: Lethal Pickles. This comes from a story in the very early days of Affirm, when I tried to teach people how to make pickles because I love pickles too.

I managed to screw it up badly enough that they eventually turned toxic and grew mold. It was terrible. I was briefly trying to persuade people that it was fine and not actually a problem, and that they should eat them. I almost single-handedly wiped out the entire Affirm team.

Fortunately, no one ate any, and we threw them away. But the story of Max’s lethal pickles survives in the annals of Affirm’s postmortem history.

Someone said, “You should eventually fulfill your dream of making things out of atoms and molecules, and sell pickles online. Call it Lethal Pickles.”

One day, I will start Lethal Pickles.

I have the domain. We're going to name the show “Max's Lethal Pickles.”

That's great. See how it does.

Yeah, yeah, there you go. Final one for you: You get asked questions by employees, by investors, both private and public, by a lot of people every day, and by journalists. What question are you never asked?

It'd just be nice to be asked. It'd be interesting to be asked. So, if you are a careful student of my shareholder letters, of which I have now written on the order of—it cannot possibly be 16, but I think it's 16 or 15—you will see that every single one of them contains a reference to the movie The Big Lebowski. You have to look sometimes hard to find it, but no one's ever asked me, “What's with The Big Lebowski in your shareholder references?”

What is with The Big Lebowski? This comes as a great surprise. I'm a huge fan of movies. When I was a freshman in college, all I wanted to do was take computer science classes. I came into my university thinking, “I'm just going to take every possible computer science class and nothing else, because everything else is stupid and boring, and I just want to write code.”

I still love writing code. If I knew I couldn't fail at solving the world's problems, then I would just write code from that point on for fun. But it turned out that you actually had to take some required courses, so you couldn't just take computer science.

I took film studies as a “This will be easy; it'll be fun,” and I got to watch things that are actually gems of the movie-making world, things like North by Northwest and Hitchcock movies and all sorts of stuff that blows your mind. If you don't know about it, you wouldn't watch it, because these are old black-and-white movies. Why would you care? Then you watch them and you're like, “Oh, my God. These are stories that have to be required viewing.”

One of the movies I watched was Seven Samurai, and it became this obsession where I've seen it over 100 times. No one asks me about that one either, or they used to. Now it's sort of a known fact that I've watched it 100 times, and that's a separate podcast for some other time.

But the silliest movie, with yet a similarly profound cultural and, at least to me, intellectual impact on me, is The Big Lebowski. It's this comedy-tragedy-whodunit commentary on the 1980s United States. It has amazing dialogue. It's everything in 2 hours of cinema that, if you haven't watched it, you should watch. It is unbelievably funny. It's one of the funniest movies ever made.

Somehow, over the years, this is how you interview talent: One of the things you can do is figure out if you share cultural artifacts at the time of the interview. When I was interviewing a good half of my management team, I found out—maybe before or after I decided that they were the right ones—that they were huge fans of The Big Lebowski.

The team that writes the letter with me is this group of people who love The Big Lebowski, and we sort of made it our business to stick Lebowski quotes into our shareholder letters.

I have one final one for you that I know you won't have been asked, right? I don't think you'll have been asked, okay, but it's a kind of a meaningful one. What about the way that your parents brought you up have you deliberately decided not to apply to the way that you bring your children up?

I grew up in an academic family, and for them, the notion of advanced degrees was the ladder to climb. It was embarrassing that my mom never got a PhD to her mom, who was an exceptionally important influence in my life and my mom's life. So this notion of the way you achieve in the world, the goals to set, the ways to think about success and failure in life, is measured through academic success.

My grandmother has papers in the Harvard Library. She's long gone, but she was a Soviet scientist who managed to have her astronomy PhD thesis referenced in dozens of academic papers in the United States. So she achieved the academic peak early in life and stayed there. As she was very old and quite ill, she'd ask me, “What about the PhD track?”

I was just about to start my third company. I was like, “You know, I'm going to get a bachelor's, basically because of you. Otherwise, I would have long dropped out and moved to Palo Alto and started a company already.” It all worked out in the end, so I'm not complaining.

But they grew up in a world where entrepreneurship was not a factor. The Soviet Union was not a place where entrepreneurship meant anything in those days. I would love for my kids to at least explore entrepreneurship and decide that it's important. So telling them, “You must spend a decade on the school bench,” is counterintuitive to that.

I never told them, “Your great-grandmother, with her multiple PhDs, is just waiting for you to get your first advanced degree and keep going.” I felt very guilty about it. The day we took PayPal public, my mom asked me, “Will you finally move back to Chicago and get a PhD?” I was like, “But, but, but—this is not the response.”

Yet that is what I was brought up with, and it weighs on me. I don't want it weighing on my kids.

Have you ever been asked that question before?

I don't think so.

Good try. I like it. It's a really revealing question. Max, dude, it's so lovely to see you again. I so appreciate you doing this, opening up, being so brilliant. You are fantastic, and I just so appreciate the time.

Thank you. You're always fun to talk to. Good to see you again.