Failure is as unfashionable as ever. And that’s a problem.

Failure has had a cheerful rebrand. Yet, despite the warm words about failing fast and failing well, humans remain deeply uncomfortable with it. Failure avoidance infiltrates most systems. Take education. In the 1960s, general systems scientist George Land evaluated the creative potential of school children over time. Some 98% of five-year-olds performed at genius level. Yet only 30% of 10-year-olds achieved such status. By age 15, the proportion of geniuses stood at 12%. And adults? Sorry, folks, only 2%. Why? 

One explanation is that we systematically reward children for getting the right answer and punish them for getting the wrong one. Furthermore – as multiple psychologists have demonstrated – well-intentioned parents try to protect their children from any conceivable risk, cutting off the experimental elements of their development. 

Such featherbedding is deadly for innovation and creativity. Many popular and important innovations – from cultural phenomena such the James Bond movie series, to lifechanging biotech, such as the mRNA vaccines – were ridiculed when first proposed. People encouraged to conform and to avoid failure are unlikely to risk suggesting a controversial new idea. 

Yet hiding failures, or being unaware of them, creates systematic errors in learning. Learning is fundamentally a process of considering what we expect to happen (the theory), then comparing it with what did happen (the reality). When the unexpected occurs, that’s a signal to update our expectations. Let’s examine a few ways in which mismanaging failure can hobble learning. 

Survivor bias

This is a huge issue in any study that purports to discover the secret of success by studying successful people or organizations. A far greater number are likely to have disappeared along the way. Without being able to compare the successes and the failures, we have no way of drawing sensible conclusions about whether a particular practice was more likely to lead to a good (or bad) outcome. Many pages have been written about the success of bet-the-company entrepreneurs. Accounts of bad bets are rarer. With such information it’s easy to conclude that risk-taking is good, without understanding when it works and when it doesn’t. 

Missing data points 

Scientists understand that a well-designed experiment in which a hypothesis was not supported remains valuable. In business, however, ventures or projects that fail are buried and become undiscussable. The result is that no-one learns from the original failed effort and are prone to repeating the same mistakes. 

Management by exception

Management by exception is popular, the idea being that issues need not be escalated to the next level unless something goes wrong. This has two problems. First, the manager to whom problems are being escalated lacks any idea of the total number of problems that are coming up. (Is this one in 10? One in 100?) Second, the practice cuts off the creativity of those doing the escalating, as they worry that if they fail to escalate, they can be blamed if their preferred course of action disappoints.

Limited goals 

The easiest way to hit performance goals is to ensure they are unchallenging. There are several risks here. The leader of a business growing at 5% might be seen as doing a great job, unless you realize that comparable rivals are growing at 15%. A manager routinely delivering results on an existing situation doesn’t know (nor do their superiors) how effective they could be in a novel one. 


It needn’t be this way. Postmortems on failed projects can assist learning. Keeping a record of things we tried with interesting results, even if they weren’t the expected ones, is helpful. And making failures discussable – turning them into humorous stories – can go a long way to correcting business’s endemic learning disability.