top of page
  • Thomas Thurston

Others may be average, just not you

Updated: Jul 16, 2023

I’ve spent my career studying one basic (but controversial) question: how do people predict if a business will survive or fail? What methods do they use and how accurate are they?  What can we learn about such predictions and, perhaps most importantly, how can we improve?


One of the first things I learned when starting down this path was that people tend to be pretty optimistic – at least about themselves. We like to cast ourselves in the best possible light.  We think we’re special. This is especially the case with innovators.  When it comes to launching a new business, the statistical odds of success can be downright horrific. We all know the odds, yet sill we march forward in confidence that our wits will defy probability. Innovators proceed in an on-going mission: to explore strange new worlds, to seek out new life forms and new civilizations, to boldly go where no man has gone before… or something like that.


Throughout this research we’ve administered experiments to thousands of innovators. We sit them down and ask them to look at business plans. Next we ask them to predict which businesses will survive and which will fail. This lets us test their methodologies and accuracies. The results have been consistent: everyone’s random. Most groups are between 42% – 48% accurate (close to 50/50) given binary predictive models (survive/fail), regardless of experience, education, industry, nationality or other factors. From the glass tower CEO to the valet who parks their car, nobody’s better than random. Ouch!


However if you ask people what they think about their predictive abilities you get starkly different results. In a recent survey we found people to be around 5 times more likely to believe they were better than average at predicting business survival or failure as opposed to worse than average.






Not only were people more prone to believe they were better than average, but those with more education were disproportionately likely to hold this belief. In other words, nobody's accuracy is better than random but people with more education are more inclined to think they're better off.



Before we're too hard on people for being delusional about their business predictions, it's worth noting that we all tend to be a bit over-optimistic about ourselves. It isn't just a business phenomenon. For example, we recently asked people to rate their driving abilities and found them to be 7 times more likely to believe they were better than average drivers as opposed to worse than average.



This goes a long way towards explaining back-seat drivers. We're also 7 times more likely to think we're better looking than average as opposed to worse.




If you thought that was bad, it pales in comparison to how good a kisser most of us think we are. People were a whopping 8 times more likely to believe they were better than average kissers.




However unlike business predictions, perceived kissing abilities didn't necessarily increase as a function of education. It seems college helps many people feel good about their smooches (perhaps they get in a little extra 1:1 learning after class) but those with no degree seem to do just fine.



We'd all like to think we're better than average and, of course, some people are. But when it comes to innovation, predictive abilities are on par with a random guess. You may as well flip a coin. While a little gloomy (especially if you're paying back B-school student loans) this is an important thing to keep in mind. People don't have any idea what businesses will survive or fail. For some this is a call to portfolio investing (make lots of bets hoping at least a few will win), but for me it's a call to science. We must change the way we look at innovation, study it and apply it because the current tools aren't proving any better than randomness. Think about it. If your GPS never took you where you wanted to go, how long would you continue to use it? It's time to admit the problems so we might have a chance to improve.


Of course, if you don’t like my suggestion, rest assured. I’m sure your predictions really are better than random. Your driving is superb. You’re extremely good looking and your kisses move mountains. Others may be average, just not you.

Comments


bottom of page