This stone has a soup in it
Startups that go badly often start like the stone soup story.
Like that guy who was charged with fraud because his AI shopping app didn’t have any AI in it.
It’s the stone soup story all over, because of course no startup has amazing tech on day one. In the spirit of “fake it till you make it”, founders tend to skimp on the expensive features until they have a sense for product-market fit. So instead of “real AI” they may start with building a simple interface—the stone!—and hire humans behind the scenes to do the work that AI is supposed to handle.
This manual work is meant to be temporary until the dev team catches up on real AI stuff and the soup is ready to be served. But then somehow the marketing material gets confused with what investors are told, and this slippery slope ends with the Department of Justice knocking on the door.
And all the same time the story could have gone totally well: the dev team delivers the real AI stuff, everybody makes a ton of money and the founders are celebrated for their ingenuity. They write books about building product on a shoestring budget. People eat delicious soup.
Anyways, this is a long way of saying that we should only judge whether a decision is good or bad before the outcome is known. Solving problems with the best available technology is generally a good idea, lying about said technology is a bad idea, no matter who gets caught.
How bad is a bad idea
I’m sure everyone can agree though that gambling with company money is a bad idea, regardless of whether someone wins or loses. Yet, the day Frederick Smith took FedEx’s last $5,000 to Las Vegas in 1974, he probably wasn’t thinking about such niceties as the “ethical implications” of gambling with employee paychecks. It was an act of survival. The company was bleeding $1 million a month, fuel costs were skyrocketing, and no investor would touch them.
When Smith returned with $27,000 in blackjack winnings—just enough to keep the company’s airplanes in the air for another week—he became a part of start-up folklore. Today, the FedEx story is often celebrated in business schools as an example of entrepreneurial grit and determination.
But really, what if he had lost?
If I may say so myself
As Jonathan Baron and John Hershey explain in their 1988 paper Outcome Bias in Decision Evaluation, people may hold themselves responsible for both good and bad luck, becoming smug in their success or self-reproachful in their failure.
White-collar criminals can do just enough mental gymnastics to avoid ever having to see themselves as dishonest people, which is just as well—part of a CEO’s job is to be good at building self-serving narratives.
A chief executive’s true performance is quite difficult to measure. Unlike say, a sales representative who can point to concrete business won or an engineer who ships specific features, a CEO’s impact is often based on their reactions to external factors. Some of their most consequential decisions—like entering a new market or shutting down investments in a product line—might not reveal their true value for years. And in the absence of clear metrics, an executive will be measured by the strength of their story.
This kind of selective accountability is quite useful when writing quarterly reports or public statements. Corporate success becomes less about measurable outcomes and more about constructing compelling leadership narratives that satisfy our need for simple, hero-driven explanations.
All things being equal, outcome bias will drive public perception, so if you’re a CEO who must do something bad—at least try to succeed at it.
