Is overconfidence killing your business?

Probably. Here’s how to start fixing it

There is a dark side to what we talked about last week. And among those who study how the brain works, what I’m about to tell you isn’t very controversial. It’s not even debatable anymore, really:

You are overconfident about your knowledge and abilities, especially when it comes to predicting the future.

Don’t feel bad though. Unless you’ve gone through specific, lengthy, repeated training, you’re no different from any of your friends, family or co-workers. Overconfidence is just hard-wired into us by our deep evolutionary past. But in the modern world, our natural overconfidence ends up being a big problem. And it keeps killing promising businesses:

  • Killer 1: Project scheduling is a nightmare when estimates are consistently optimistic. Just ask Bent Flyvbjerg, who literally wrote the book (many in fact) on mega-project cost overruns. Those rosy estimates of costs, time and revenue mean you end up taking on unprofitable projects that eventually sink you.
  • Killer 2: Depending on the country and the study, around 70–90% of people believe they’re better-than-average drivers, even though obviously only half of them could be. Even worse, 36% of drivers felt that they were better-than-average drivers even while texting during driving! What fraction of your engineers think they’re better-than-average? Salespeople? What about your managers? It’s hard to instill a culture of continuous improvement if almost everyone thinks they’re doing better-than-most already.
  • Killer 3: Perversely, the Dunning-Kruger effect shows us that people who are bad at something often believe they’re actually good at it. Their lack of knowledge and ability prevents them from realizing that they lack knowledge and ability. How’s your organization’s employee training? Do new hires understand what’s expected of them, and do they get training on the skills they’re expected to develop and improve?
  • Killer 4: Culturally, we reward confidence and the appearance of success. “Fake it till you make it” combines with “It’s not a lie if you believe it” in a heady stew of self-delusion. We’ve all seen dysfunctional office politics where this phenomenon gets rewarded. It leads to a vicious cycle of cynicism, and the best employees end up leaving to go somewhere less dysfunctional. You’re left with just the employees that couldn’t find a better job elsewhere.

Well-run organizations are aware of these issues. They train their people in the self-awareness, self-examination, and self-correction skills necessary to mitigate the effects, and they create a culture where expressions of overconfidence are safely and naturally corrected over time. In particular, painfully sometimes, upper management starts to understand that the image of the decisive, never-wrong leader went out along with suspenders and gold cufflinks a while ago. Uncertainty isn’t weakness, it’s realism.

Having taught companies these skills, we know it’s a process that takes time and effort. But even so, there are simple techniques you can start implementing today that can start you on the right path.

Train your ability to estimate probabilities

If I ask you to give me a probability that your project is going to be late, you might say “20% tops”. Ok, let’s put the letters A, B, C, D, and E in a hat. You’re about to grab a random letter from the hat, but before you do I predict you’ll pick, say, B. Then you grab a letter, and sure enough it’s B. Think of your intuitive level of surprise. You wouldn’t be shocked, but you’d be at least a little bit surprised at this 1 in 5 chance that I correctly predicted your random letter. Here’s the question:

What would surprise you more, me correctly picking the letter in the hat, or your project coming in late?

On paper, you should be equally surprised. But I bet that in most situations, the late project would actually surprise you less. The probability of lateness is higher than 20%! The letters-in-a-hat analogy has helped you calibrate your confidence level.

Systematically track your estimates

It’s easy to forget a bad estimate if no one bothers to write it down. Hindsight is 20/20, and we’ll always find a way to tell ourselves (and others) that we actually predicted what happened, even if we actually didn’t.

Tracking estimates, probabilities, and similar predictions is the simplest, most fundamental way we can get better at making them accurately. It doesn’t take much: start a simple spreadsheet today to keep track of your and your co-workers’ estimates and beliefs about the future.

Bet on it

You had to know this was coming. Very few activities are better at making us come up with real, honest estimates than putting your money where your mouth is. Or as Nassim Taleb puts it, Skin In The Game.

You don’t have to bet your house or anything. Usually the social effect of losing $20 is more than enough to make you reconsider those brash statements about how much revenue that new customer is going to generate. Bet with odds. If someone thinks something is 80% to occur, make them bet putting up 4:1 odds. If you think they’re wrong, give them a little incentive, make it 3:1. If they’re really and truly 80% confident, they should be happy to take that 3:1 bet. If they’re hesitant, you know that 80% isn’t worth the paper it’s written on.


As I said at the beginning, none of this stuff is controversial or particularly new. And yet I see few companies that have instituted a robust culture of making accurate estimates and good self-evaluations, or a feedback mechanism that improves those things over time. For those that do, accurately seeing and predicting the future gives you an enormous competitive advantage. Let your competitors kill themselves with overconfidence. Yours shall be the world, and everything in it.

Helping organizations make their best decisions

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store