The Founder’s Delusion

Give a founder a successful exit and a microphone, and something curious happens. The story of the company becomes the story of how to build a company. Every hiring decision starts to look deliberate. Every setback becomes a necessary lesson. Even the mistakes acquire a certain charm. They were the mistakes you had to make on the road to winning.

Nobody asks the obvious question: How does the founder know which of those decisions actually caused the success?

They know what they did. They know what happened afterward. Those are not the same as knowing why it happened. Yet startup culture treats the difference as a technicality. We put the winners onstage, hand them a marker, and ask them to explain the rules.

The problem isn't that successful founders have nothing to teach us. Many have a great deal to teach us. The problem is that a successful outcome is not, by itself, proof of the explanation offered for it. And when an explanation becomes advice, a bad one can send the next generation of founders off to make very expensive mistakes.

In 2021, researchers George Lifchits, Ashton Anderson, Daniel Goldstein, Jake Hofman, and Duncan Watts set out to measure just how easily success stories can bend judgment. They showed people examples of successful startups and asked them to bet on the prospects of a new company. One group saw examples emphasizing founders who had graduated from college. Another saw examples emphasizing founders who had dropped out. A third group saw no examples at all.¹

The underlying companies were the same. What changed was which successful co-founder the researchers called attention to.

Among participants shown the graduates, 87 percent bet on a graduate founder. Among those shown the dropouts, only 32 percent made that same bet. Without examples, 47 percent chose the graduate. A 55-percentage-point swing caused by a small, carefully selected set of true facts.

Here's the remarkable part. Participants had to acknowledge that the examples were cherry-picked. They knew. And they still made their bets, often confidently, then supplied plausible explanations for why graduates or dropouts were better suited to start companies.

The researchers hadn't needed to lie. They simply chose which truths to display. The audience wrote the theory itself.

That is the danger of a founder story. It can be completely accurate as a record of events and thoroughly unreliable as an explanation of what caused them. A biography is not a controlled experiment.

Psychologists have a name for part of this: self-serving attribution. People tend to credit themselves for good outcomes and look outward for explanations when things go badly. In a meta-analysis covering 266 studies, Amy Mezulis and her colleagues found a substantial self-serving attribution bias across the samples they examined, although its size varied by group.²

For a founder, the temptation is easy to understand. If a company wins, years of exhaustion suddenly have a shape. The late nights mattered. The risky hire was brave. The product decision everyone questioned turns out to have been visionary. If the company loses, the exact same behavior may become stubbornness, recklessness, or poor judgment. We read the character of the decision backward from the result.

There is a second distortion, and it's harder to see. We only hear detailed origin stories from companies that survived long enough to have an audience. Consider two startups that ignored customer research, launched early, and spent aggressively to grab a market. If one becomes a billion-dollar company, its founder may describe those decisions as fearless speed. If the other dies, the same decisions become cautionary tales about impatience. The behavior did not change. The outcome did.

This is survivorship bias: examining the winners while overlooking the businesses that behaved similarly and lost. It doesn't mean the winners did nothing right. It means we cannot learn what made them different by looking only at them.

This is where the easy version of the argument falls apart. If we claim that successful founders always mistake luck for genius, the research doesn't support us.

Benjamin Lough and Alyssa McFadden examined interviews from the podcast How I Built This with Guy Raz. In 183 episodes, founders had answered a direct question about whether their success came from luck or from skill and hard work. On average, their answers landed close to an even split. Some talked about their parents, their country of birth, or an opportunity they happened to recognize. Others stressed preparation and persistence. A number credited the people they hired.³

These founders were not, as a group, insisting that they had personally engineered every fortunate event. And a podcast interview cannot measure the actual share of success attributable to chance. What the study can tell us is that successful founders do not all tell the same self-congratulatory story.

But even the founder who says, 'I was lucky,' may not know where the luck entered. Was it the market opening six months earlier than expected? A competitor getting distracted? A supplier calling back? One large customer arriving at precisely the right moment? Luck isn't a mysterious substance sprinkled over a business. It is a name we give to events whose timing and causes were partly outside our control.

A candid founder can recognize uncertainty and still be wrong about which decision mattered most. That is the real delusion worth investigating.

There is a danger in pushing the argument too far. If every success story is partly accidental, it becomes fashionable to pretend no one is particularly good at building companies. That is just as lazy as worshipping the founder.

Research on venture-backed entrepreneurs finds that past success predicts future success. In a study published in the Journal of Financial Economics, Paul Gompers, Anna Kovner, Josh Lerner, and David Scharfstein reported that founders whose previous company had gone public had about a 30 percent chance of succeeding in their next venture, compared with about 21 percent for first-time founders, under the study's definition of success.⁴

That is a meaningful advantage. It is also a long way from certainty. Roughly seven out of ten previously successful founders in that analysis did not repeat the same kind of success.

Another study, following 10,541 founder-startup pairs, found associations between founder personality characteristics and outcomes at different stages, from raising capital to a public offering or acquisition. The findings matter. So does the word associations. A founder characteristic that correlates with an outcome is not a universal instruction manual.⁵

The Gompers team found something especially revealing: entrepreneurs who entered favorable industries at favorable times tended to repeat that pattern. There appears to be skill in recognizing the right market at the right moment. But that's a different skill from the founder's ability to describe why one ad campaign, one feature, or one management style made the company succeed.

The useful question isn't 'Skill or luck?' It's 'Which skills travel, and which advantages belonged to that particular company at that particular time?'

Now imagine the successful founder sitting across from someone building a very different company. The veteran says, 'We never bothered with market research. We built what we believed in and customers followed.' That might be a faithful account of their experience. It might even be good advice in the narrow conditions they faced. But a new founder who copies the behavior is making a much bigger bet: that the conditions are similar enough for the result to repeat.

Usually, the missing information is not what the winner did. It's what happened to all the people who did the same thing. How many businesses skipped research and failed? How many took their time and succeeded? How many were selling into a market that had already started buying, before their first pitch deck existed?

The successful founder can describe the route they traveled. To know whether it's a good route for someone else, you need to see the other roads, including the ones that end in a ditch.

That problem gets worse when founders become investors, advisers, or public authorities on entrepreneurship. A memorable story travels easily. A comparison group does not. Soon a decision that worked once has become an industry commandment, repeated by people who were nowhere near the original company.

The consequences become tangible when a founder's theory of success enters the marketing budget. 'We grew because we moved fast.' 'Our customers don't respond to research.' 'Advertising is how you create a market.' These may be hypotheses worth testing. They are terrible substitutes for tests.

Consider Baroo, a pet-care startup founded by Lindsay Hyde. The company offered services for residents of apartment buildings, and early enthusiasm helped produce high growth expectations. In a Harvard Business School case discussion, Hyde and professor Tom Eisenmann described how an early, misleadingly positive signal from investors contributed to an unsustainable growth strategy. The business later became a case study in how early signals can send a startup in the wrong direction.⁶

The lesson is not that early adopters are useless or investors are always wrong. It's that approval is not demand, demand is not profitable demand, and the preferences of a first cluster of customers may tell you very little about the next ten thousand.

Eisenmann calls one version of this a false positive: a signal that looks like validation but isn't strong enough to support the leap a founder wants to make. In his research on startup failure, he also describes false starts, when teams rush to build and test products before adequately understanding the customers' problem.⁷

A marketing team can make either mistake much more expensive. Once the founder declares that the story has been proven, the job shifts from finding out whether the premise holds to producing numbers that make it look true. Paid acquisition goes up. The dashboard gets prettier. More people arrive, but they don't stay, or their lifetime value never covers the cost of finding them. Yet each new campaign becomes a reason to spend again because the original explanation must, somehow, be right.

This is where a good marketer has to be willing to annoy the room. A click is not a customer. A customer is not necessarily a profitable customer. A few glowing interviews are not representative evidence. And doubling the budget does not turn an untested assumption into a proven one.

Marketing should be one of the places where a founder's story gets challenged by behavior. Which buyers actually return? Which segment converts without an expensive incentive? What changes when the message changes, while the audience and offer stay comparable? What would sales look like if the company stopped buying its own momentum?

The answers may still favor the founder. Good. Now the belief has earned its place.

A founder's account becomes more useful when it includes the things business magazines usually cut: the reasonable decisions that failed, the assumptions that survived too long, the person who disagreed and was right, the customer evidence that forced an unpleasant change. Just as important, it should include a clear admission of what cannot be known.

No founder gets to rerun the same company in the same market, with the same team, the same customers, and the same economic conditions, changing only one decision. That alternate company does not exist. So an honest success story can identify plausible causes and useful patterns, but it cannot establish every cause with confidence.

For aspiring founders, that changes how you listen. Don't ask only what the successful person did. Ask what they tried that didn't work. Ask what evidence changed their mind. Ask whether they have seen the same approach succeed in a different company. Ask what conditions made the decision sensible then, and which of those conditions are missing now.

And when you encounter a beautiful origin story, make yourself ask the least glamorous question in entrepreneurship: Who did the same thing and lost?