top of page
  • Writer's pictureDipyaman Sanyal

If Data Driven Decision Making was Easy, No One in the World Would Smoke

Think about a Founder-CEO who has started a unicorn. From nothing they have built something – and something worth a billion dollars. At that point it is widely accepted that the CEO is incredibly smart, has put together the right team, has taken valuable decisions and is the darling of the media. People ask for their signatures and the CEO pontificates on all matters on global television, even matters that are not their core area of expertise. At that point if the CEO has a ‘pet-project’ which is quite obviously not worthwhile, why would she/he/they listen to me or you? Why would they listen to the data which is screaming at them that this is a stupid move?

It is said that back in the Roman Empire there were slaves called Aurigas, whose job was to crown the leader during the Roman Triumphs, and while they put on the laurel, they were to whisper in the ears of the Caesar the words “Memento Mori” which means, “Remember You Are Mortal”. Alas, we do not have such job profiles anymore! Successful leaders believe in an omniscience till they make devastating mistakes on their ‘pet-projects’ (a subject studied extensively in the corporate governance literature).

When I was teaching my first corporate training program, it was to a group of very senior engineers who take decisions on multi-million- (or even billion) dollar projects for the government of India. I was to teach them data driven decision making! I was 32 and the average age of the learners was north of 50. I had a wonderful PPT prepared and was ready to teach but the evening before the class I was introspecting about why should this group listen to me? What do I bring to the table to convince them that the decisions they take can be improved upon based on the quantitative techniques that I was preaching about?

So, I made some changes, and this was arguably the first instance when a corporate analytics training started with a quick primer from psychology. Looking at the Heuristics and Biases program (Daniel Kahneman, Amos Tversky, Richard Thaler, their co-authors and the large group of behavioral economists were my source) I began with how the human mind can take wonderful decisions very fast but often because of this speedy decision making, we tend to get biased in our planning. It clicked and since then I have used this start to dozens of trainings for senior managers or CXOs.

Like smoking, power is addictive. It makes you do things which otherwise you would never dream of doing. Think of the Stanford Prison Experiment, where student subjects in the role of prison guards acted obnoxiously with other student subjects who were in the role of prisoners. Power, also, corrupts. An easy introductory paper on the subject from MFRK De Vries (1991) titled, “WHATEVER HAPPENED TO THE PHILOSOPHER-KING? THE LEADER'S ADDICTION TO POWER”, begins with the abstract from which I quote, “… the existence of a number of conscious and unconscious forces which make leaders reluctant to relinquish power. It is suggested that a major reason for the unwillingness to let go is the transferential effects of leadership. Both mirroring and idealizing transference reactions are discussed. In addition, there is the ageing factor which may contribute to the addictiveness of power, since ageing can evoke a strong need for compensatory strivings. The talionic principle - meaning in this case the conscious or unconscious fear that the loss of power will be followed by some form of retaliation for previous acts - may be another factor contributing to power's addictiveness. In this context the role of envy is emphasized.”

So, the leader’s addiction to power is a complex process which arises from a variety of factors, many of which are not conscious even to the smartest of leaders. If data driven decision making was easy, would anyone smoke? Addiction to nicotine or to power are not easy things to give up. And now think of data driven policy. While the data on smoking is clear and the outcomes are fairly near term, the effect of policy is felt after decades or even centuries. Think of climate change. Think of education policy. Think of labor mobility.

Maybe it is time for leaders – from managers and business unit heads and CEOs to presidents and prime ministers – to hire (or rent! we are available :) a new generation of Aurigas. The data scientists/analytics officers, whose role, along with solving problems, will be to remind the chiefs that you are mortal and your ideas maybe, just maybe, faulty, because the data says so. At least let us put it up for debate.


bottom of page