Max Levchin on Hiring, Postmortems and Leading Affirm

Affirm founder and CEO Max Levchin joins Harry Stebbings to discuss hiring, decision-making, postmortems and leading a public company. He also comments on consumer lending and competition with Klarna. His product and regulatory views reflect his direct commercial interest in Affirm; the interview does not independently validate comparative figures or establish whether any loan is suitable for a particular consumer.

Levchin says highly talented people can have strong personalities, and a manager must weigh their contribution against the cost of difficult behavior. He values strong opinions delivered with enough empathy to keep teammates engaged: criticize the work directly, but make clear that the team will improve it together.

When work is poor, he distinguishes someone who can improve under different circumstances from someone whose abilities do not fit the role. Stebbings says direct feedback can upset people; Levchin advises explaining the gap while acknowledging effort and potential. A decision that someone cannot succeed requires a separate and more difficult conversation.

He sees narcissism and self-interest as warning signs that a forceful personality has turned toxic. Humility matters because a leader who makes criticism about personal superiority can create resentment. He generally believes in second chances, but says loss of trust through dishonesty is his personal hard boundary.

Stebbings asks how a founder who works obsessively can align expectations with employees. Levchin says founders have distinctive incentives and must learn what motivates each teammate rather than assume everyone shares the founder's goals. Someone motivated by craft, for example, may respond to support in producing work they are proud of rather than demands to work weekends.

Asked when a company must hire 'B players,' Levchin challenges the idea that an interview can reliably assign a permanent letter grade. Practical assessments can filter obvious mismatches, he says, but the difference between a good and exceptional hire often becomes apparent after joining. His labels are managerial shorthand, not objective measures of people's worth.

He argues strong colleagues often seek other strong colleagues because they raise the standard, while insecure managers may hire weaker people to protect their own position. At the same time, he values dependable contributors and says some develop into exceptional performers. The danger he identifies is fear-driven hiring, not merely being less experienced or less visibly exceptional.

Levchin recalls founding Affirm with the aim of making the terms of consumer finance clearer and avoiding reliance on fine print. Bankers questioned whether that could make money. He says he did not have a complete financial model at the time and accepted the possibility of being wrong; this is his account of the founding thesis, not a comparison of all lenders' terms.

As an example of inadequate calculation, he recalls preparing for a launch expected to be much larger than existing traffic. The front-end capacity was increased faster than the backend, and connections overwhelmed the database for hours even though launch traffic was modest. He says the imbalance could have been anticipated through an architectural calculation.

Levchin prefers to study failures, though Stebbings argues success can also be analyzed. Levchin responds that even a successful funnel can be examined by asking which step could have improved. His emphasis is on learning what to do better while still acknowledging a team's achievement.

To build a postmortem culture, he recommends separating records for each incident, preserving raw information and assigning someone to distill it into a written analysis. The team should be able to describe mistakes without blame or performative apologies. He wants the document to explain not just what happened but why.

At Affirm, he says a directly responsible person normally prepares the postmortem within one or two weeks, and teams review the work regularly. Senior colleagues can challenge a shallow explanation or disputed facts. Levchin emphasizes recording decisions and improving safeguards rather than making the discussion a trial of an individual.

Stebbings questions whether decisions are truly reversible. Levchin says some choices affecting reputation, partner relationships, architecture or contractual commitments are difficult to undo, while others that feel permanent can be reversed. He sees postmortems as a way to recognize when a supposedly reversible decision was not.

He urges teams to care about the 'invisible parts' of products, invoking the idea of taking pride in the inside of a machine. There is tension between shipping quickly and building durable systems. He considers some backend capabilities, including Affirm's credit underwriting models, important enough to accept a delay for quality; his assessment of those models is a company claim.

Affirm has a prototyping team empowered to put imperfect ideas in front of customers, he says. If feedback supports an idea, the team should then rebuild it for quality and scale rather than leave the prototype as the final product. Speed is essential to startups in his view, but he does not think it excuses permanently poor work.

A writing culture supports the postmortem archive and keeps lessons from becoming oral tradition. Levchin prefers short, clear documents to long prose and says even spoken all-hands messages are more useful when concise. His recommendation is a communication practice, not a claim that every complex issue fits on one page.

Affirm's work arrangements are hybrid and vary by location, he says. Levchin cares about output and recognizes that some people focus better remotely, but insists teams need meaningful time together to form relationships and understand how colleagues make decisions. Stebbings notes that other CEOs favor a fuller return; Levchin distinguishes shared time from commuting merely to spend the day on video calls.

He recalls layoffs at Affirm as a low point for morale and describes an earlier lesson from another company: a leader should be present with affected people rather than hide from the consequences. He advocates empathy and taking responsibility. His account of conversations with laid-off employees is his own perspective, not evidence that everyone experienced the event the same way.

Asked whether overhiring is a calculated risk that failed, Levchin agrees it can be. Companies weigh the prospect that new hires will create valuable features against uncertainty in funding, cash flow and growth; those assumptions can be wrong. He describes company decisions as involving multiple interacting variables rather than a single forecast.

Stebbings asks whether Klarna is growing faster and how its US expansion affects Affirm. Levchin cites 32% growth for Klarna and 35% for Affirm, saying he believes Affirm has been gaining share; the interview does not specify a fully comparable metric, geography or period for the percentages. His share claim is a competitor's assessment, not an independently established market-share result.

Levchin says Affirm offers a range of loan durations and argues breadth can help the company serve more customers. He declines to infer Klarna's strategy without its internal information, preferring to study competitors' products and sales practices for lessons. Product availability, approval and terms depend on the specific transaction and are not established by this interview.

On buy-now-pay-later regulation, he says he favors thoughtful rules and opposes late fees, while warning that poorly designed restrictions could make useful products unviable. He describes engaging with regulators to provide feedback. This is the position of a lending company CEO, not a legal conclusion or proof that any product is cost-free.

He says being a public CEO gives access to capital and investors with deep experience, but also brings reporting responsibilities. He tries to detach daily decisions from the share price, even hiding the ticker from his desktop. He warns employees not to assume a high post-IPO stock price represents a settled long-term valuation.

In quick fire, Levchin says he likes learning quickly but sometimes misjudges how quickly he can master a field; he cites his ongoing effort to understand large language models. He expects systems around models to evolve and imagines multiple models debating answers. That is a technical possibility in his view, not a demonstrated guarantee of accuracy.

He dislikes an AI story created purely for the market and says companies should use genuinely useful tools. Public investors ultimately need to evaluate growth, risk and profitability, he argues. His distinction between a short-term voting market and long-term weighing market is an investing analogy, not a forecast of Affirm's stock.

If he could not fail, he says he would revisit difficult problems including food, water, shelter and energy. Asked which PayPal colleague he would start another company with, he chooses Peter Thiel because he knows him well under pressure, while praising Elon Musk too. He says both have an ability to encourage people around them to attempt harder problems.

Levchin admires Airbnb's 'belong anywhere' message and describes his affection for La Marzocco coffee machines. He discloses that the coffee company is an Affirm customer, while saying he liked the brand before the relationship. His comments are personal preferences and include a commercial connection.

He jokes about a future pickle company after a failed homemade batch grew mold and was discarded. He also reveals that Affirm shareholder letters contain references to The Big Lebowski, a favorite film shared by members of the management team. These anecdotes illustrate his interests and team culture.

Reflecting on family, Levchin says he grew up in an academic household where advanced degrees were a measure of achievement. Although his family encouraged a doctorate even after PayPal's IPO, he wants his children to be free to explore entrepreneurship without feeling the same academic obligation. He describes a personal choice, not a universal prescription for education.