In Meltdown, Chris Clearfield and András Tilcsik explore how complexity causes failure in all kinds of modern systems — from social media to air travel — and reveal how we can prevent disasters in business and life. Here, they explain for Management Today how to start a meeting to avoid bad decisions.
In 2013, U.S. retail giant Target expanded into Canada and opened more than 120 stores across the country in less than nine months. One month before the launch, senior employees gathered at Target’s Canadian headquarters to discuss the timetable for the expansion.
Two people expressed concerns about the overly ambitious schedule and the company’s supply chain problems, but no one openly argued that Target should delay its launch. ‘Nobody wanted to be the one to say "this is a disaster",’ recalls a former employee. At the meetings over the next few months, as journalist Joe Castaldo put it, ‘everyone knew the launch was a disaster and the company had to stop opening stores so it could fix its operational problems, but no one actually said so.’
We have all been there: a team is about to make a bad decision, but no one in the meeting room speaks up. Or by the time someone does, it’s too late: there is already a consensus, and the decision will not change. Over the past five years, we studied dozens of business failures, big and small, and saw this kind of situation time and time again.
The good news is that even a small change in how we run meetings can make a big difference. That’s the conclusion from a simple but powerful experiment conducted in the 1970s. To study how teams make decisions, psychologist Matie Flowers created 40 teams and gave them a problem to solve—a crisis at a hypothetical organization, Eastern High School.
In the scenario, Eastern High is facing a number of problems. The local school system suffers from budget shortfalls, falling tax revenues, and conflicts with the teachers’ union. Though Eastern High has always been an elite public school, it now faces an influx of less-well-prepared students because of redistricting. And some teachers just can’t keep up. Ms Simpson, an aging algebra teacher, for example, can no longer maintain order in the classroom.
The superintendent puts together a team to deal with the crisis. The team includes four people: the superintendent himself, the principal of Eastern High, the school counselor, and a member of the school board who represents the parents.
Each person brings different pieces of information to the table. The superintendent, for example, knows that other principals in the system were asked to allow Ms Simpson to be transferred to their schools and refused. The principal knows that Ms Simpson had a mild stroke two years earlier and that other faculty members are very fond of her. The counselor knows that students usually get easy As from Ms. Simpson and that they don’t think she is a good maths teacher. And the school board member knows that parents are opposed to increased taxes.
In the experiment, Flowers assigned each participant to one of four roles (superintendent, principal, school counselor, or school board member), and each person received an information sheet describing six or seven facts that were relevant to the situation but were unavailable to other team members. The idea was to simulate a decision situation where different people have access to different information. The teams then got to work and tried to come up with solutions.
But there was a twist in the experiment: Flowers had randomly divided the participants playing the superintendent role into two groups: directive leaders and open leaders.
Directive leaders were trained to state their own suggested solution at the beginning of the discussion and to emphasize that the most important thing was for the team to agree on its decision. Open leaders, in contrast, did not state their own preferred solution until all other members had the opportunity to offer their ideas. And open leaders emphasized that the most important thing was to discuss all possible viewpoints.
The power of an open mind
Flowers tape-recorded the sessions, and two independent judges analyzed the recordings. They counted the number of different solutions that team members proposed, and the number of facts from the information sheets that came up during the discussion.
The results were clear: teams with open leaders came up with more solutions. They also shared more facts during their discussion. But the most striking pattern emerged when the judges separately counted the facts that had come up before and after the group agreed on a solution. Before reaching a decision, teams with open leaders shared almost twice as many facts as teams with directive leaders.
So an open leadership style didn’t just produce more solutions but also a better-informed discussion leading up to a decision. Under directive leaders, in contrast, nearly one-third of the shared facts came up only after consensus had been reached. At that point, of course, new facts made little difference and could only be used to justify a decision that had already been made.
Flowers’s study showed just how little it took to gather more facts and elicit more solutions. It didn’t take a born leader with exceptional charisma and skill. Open leaders in the experiment were picked randomly and were given only a brief training. Yet their teams consistently did better than the other groups.
A few words can make a big difference. You can start a meeting like this: ‘The most important thing, I think, is that we all agree on our decision. Now, here is what I think could be done.’ But to avoid making a bad decision, it’s much better to say: ‘The most important thing is that we air all possible viewpoints to reach a good decision. Now, what does each of you think should be done?’
Meltdown: Why our systems fail and what we can do about it, by Chris Clearfield and András Tilcsik is published by Atlantic Books.
Image credit: By anucha maneechote/Shutterstock