It had been 7 hours, 21 minutes and 31 days, but I got a reply from Nobel Prize winner Daniel Kahneman. An angry reply: “I said no such thing as a % of unconscious decisions”.

I couldn’t be happier with the angry reply. I was angry too. After a long search for the truth about the numbers, I finally got the answer. Kahneman never said that 95% of our decisions were made unconsciously.

A falsehood promoted by many communication, business and behavioral design professionals in Europe, by Harvard Business School professors and more. An unaccounted number snowballing in (college) universities all over the world and even linked to the aura of Kahneman.

But wait, let’s rewind to the summer of 2020.

Online workshops in Amsterdam
For me, it all started with enthusiasm. With a keen interest in the field of behavioral design. With the guiding question: How can you nudge people in their decision making? Loss aversion, primacy effect and whatnot. The masterpiece of Daniel Kahneman, Thinking, Fast & Slow, opened mental doors.

I decided to follow a four-day online course in Amsterdam. The onset looked promising. Numbers with impact stunned my brain. “We make 17.000 decisions a day, 95% of our decisions are unconscious, in a podcast Kahneman even talked about 98%, …”.

Alas, let me stop you there. Not a single number above was checked and referenced with real research. Urban myths in the higher echelons of academia and consultancy? I hate to break it. I emailed the agency with my concerns, even contacted the often quoted Harvard professor Gerard Zaltman (who politely answered that it cannot and has not been measured).

A number hoax, and not the only one. What struck me most? To experience how hard it was to change the mind of behavioral design professionals. A field organized to change behavior. A few weeks later I got this mail from the agency:

After trying to debunk the myth for a couple of weeks, showing them the Zaltman response, they continued promoting the false numbers. Even summarising the book by Kahneman with 95% versus 5%. Djeezzz.

Kahneman to the rescue. His e-mail seemed to trigger the right action. Finally, the agency promised to stop promoting false numbers and to stop linking them with Kahneman.

One down, 7214 consultants to go?

Geert, aren’t you overreacting? I hear you think. Well, I might. But since this deconstruction of the numbers, I see false numbers everywhere. A rude awakening for a 37-year-old college professor and consultant.

This week I listened to a podcast of a well-respected copywriter and trainer. He tells us that stories are remembered 22 times more easily than other information. Wow. That triggered my freshly activated bullshit-radar for numbers.

Five minutes later the number already seemed debunked. Urban myth? I contacted the trainer, who referred to a book by Gaia Vince, so I contacted her for the academic source. And even found a whole list of stunning numbers in a review by Nature.

I am still waiting for a scientific source from Vince. The trainer then referred to ‘Actual Minds, Possible Worlds’ (1987) by Harvard professor Jerome Bruner as original source. I read it. No research to be found in these collected poetic-literary essays. Fact-checking, the place where fun goes to die.

Stockholm-syndrome for false figures

What concerns me the most. That even when the numbers are debunked, smart people seem to experience an enormous difficulty to let them go. Another consultant who used the Kahneman numbers in his traveling roadshow, just got mad, when confronted. Stockholm-syndrome, anyone?

It’s easy to laugh with Trump’s crazy miscarriage of numbers. Often the above-mentioned consultants have and will. But ‘the number bias’ is everywhere. I fell for it too… and will fall for it again. And you, you will fall for ‘the number bias’ too.

“Tu quoque fili me”

Let me end with a personal falsehood I promoted for more than a year. I told many students and companies that images are transmitted 60.000 times faster than words. I read it in a couple of management books and on uncountable websites like the ones from Forbes or Entrepreneur.

Again, alas, no research there if you go in fact-checking modus. The closest I came to finding sources was combining a study from MIT (13 milliseconds to process an image) with a study from Cambridge (200 milliseconds to process a word). If we do the math, it’s closer to 15.4 times instead of the ridiculous 60.000. And even this combined analysis is wobbly.

So. In these times of fake news, tame your inner Trump. If it’s too good to be true, it often is. Do not use numbers if you aren’t absolutely sure.

Stop the count.

Geert Van den Eijnden

Ps: Ever heard that 55% of meaning is transmitted by facial expressions, 33% by vocal cues and only 7% by words? Myth. (Yes, I fell for it too.)