Steven Levitt and Stephen Dubner, coauthors of “Freakonomics,” “Superfreakonomics” and “Think Like a Freak” became household names by seeing things differently. Speaking at the 2014 EY Strategic Growth Forum, Dubner discussed the importance of being willing to ask questions that other people won’t ask, even if it means going up against tradition and conventional wisdom.
“My personal view on business is that it’s too hard. It’s too hard to figure out what you’re supposed to do. No one is smart enough to know all of the answers,” Levitt says.
So, what you have to figure out is how you can harness data to be smarter than you, he says, and the best way to do that is through randomized experiments.
“How can you find data, or even better how do you experiment? How (can you) make decisions in such a way that two months later or six months later, the data can tell you, ‘Did I make the right decision or the wrong decision?’” Levitt says.
Distinguishing good data from bad
Data also is one of the best ways to determine what could happen in the future — another component that the most successful companies continually work on.
If you want to do your best at predicting the future you have to acknowledge what you can and can’t know, Dubner says. And if you’d like to try to solve problems, rather than thinking you know how the future will unfold — you need to try to get good data and try to distinguish good data from bad data.
Finally, you have to try to figure out the incentives that are in play to actually make people behave the way they do, rather than what they say or what you think they do, he says.
“We’re all making decisions every day based on some kind of data, but I have to tell you — self-declared data, survey data, is kind of the lowest form of data,” Dubner says. “Any time you’re making an important decision based on what people tell you they think will happen or even if they tell you the way they’ll behave, you’re bound to make a lot of mistakes.”
Drilling down to incentives
In order to examine how to predict how your customer might react, Dubner gave the example of a San Diego utility. It was trying to get people to use fans more than air conditioners.
When it surveyed its customers, people said the incentive they would most respond to was: Use less electricity now because it’s good for the planet later, he says.
So, in order to get real data on real behavior, he says the utility hung distinctive signs with four different incentives on people’s homes, and then tracked the actual electricity consumption.
Despite what they said on the phone, Dubner says in reality the incentive that most people responded to was: Use more fans because other people in your neighborhood are thinking about using more fans.
No one likes to be thought of as a herd thinker, Dubner says, but that’s the one that human beings actually seem to respond to more.
“It’s not about what I want to be true, it’s about what works,” he says.
“And we’re all trying to figure out what works; what actually solves problems,” Dubner says. “The way to do it is, again, don’t pretend you know the answer; find real data, not declared data, and figure out which incentives actually change human behavior, which is a lot harder to change than you think.”
