In some ways it’s remarkable the Saturday Night Live team gelled at all. Michaels, it turned out, had chosen everyone precisely because of their disparate tastes. Zweibel was a specialist in borscht belt one-liners. Michael O’Donoghue wrote dark, bitter satires about such topics as the assassination of JFK. (When a distraught secretary told O’Donoghue that Elvis had died, he replied, “Smart career move.”) Tom Schiller aspired to direct art films. And everyone could become scathing critics when their sensibilities clashed. “Great, Garrett,” O’Donoghue once said when he read a script the actor had spent weeks writing. Then he dropped it into a trash can. “Real good.”
“Comedy writers carry a lot of anger,” said Schiller. “We were vicious to each other. If you thought something was funny and no one else did, it could be brutal.”
So why, given all the tensions and infighting, did the Saturday Night Live creators become such an effective, productive team? The answer isn’t that they spent so much time together, or that the show’s norms put the needs of the group above individual egos.
Rather, the SNL team clicked because, surprisingly, they all felt safe enough around one another to keep pitching new jokes and ideas. The writers and actors worked amid norms that made everyone feel like they could take risks and be honest with one another, even as they were shooting down ideas, undermining one another, and competing for airtime.
“You know that saying, ‘There’s no I in TEAM’?” Michaels told me. “My goal was the opposite of that. All I wanted were a bunch of I’s. I wanted everyone to hear each other, but no one to disappear into the group.”
That’s how psychological safety emerged.
Imagine you have been invited to join one of two teams.
Team A is composed of eight men and two women, all of whom are exceptionally smart and successful. When you watch a video of them working together, you see articulate professionals who take turns speaking and are polite and courteous. At some point, when a question arises, one person—clearly an expert on the topic—speaks at length while everyone else listens. No one interrupts. When another person veers off topic, a colleague gently reminds him of the agenda and steers the conversation back on track. The team is efficient. The meeting ends exactly when scheduled.
Team B is different. It’s evenly divided among men and women, some of whom are successful executives, while others are middle managers with little in the way of professional achievements. On a video, you see teammates jumping in and out of a discussion haphazardly. Some ramble at length; others are curt. They interrupt one another so much, it’s sometimes hard to follow the conversation. When a team member abruptly changes the topic or loses sight of their point, the rest of the group follows him off the agenda. At the end of the meeting, the meeting doesn’t actually end: Everyone sits around and gossips.
Which group would you rather join?
Before you decide, imagine you are given one additional piece of information. When both teams first formed, each member was asked to complete what’s known as the “Reading the Mind in the Eyes” test. They were each shown thirty-six photos of people’s eyes and asked to choose which word, among four offered, best described the emotion that person was feeling.*2
This test, you are told, measures people’s empathy. The members of Team A picked the right emotion, on average, 49 percent of the time. Team B: 58 percent.
Does that change your mind?
In 2008, a group of psychologists from Carnegie Mellon and MIT wondered if they could figure out which kinds of teams were clearly superior. “As research, management, and many other kinds of tasks are increasingly accomplished by groups—both those working face-to-face and ‘virtually’—it is becoming even more important to understand the determinants of group performance,” the researchers wrote in the journal Science in 2010. “Over the last century, psychologists made significant progress in defining and systematically measuring intelligence in individuals. We have used the statistical approach they developed for individual intelligence to systematically measure the intelligence of groups.”
Put differently, the researchers wanted to know if there is a collective intelligence that emerges within a team that is distinct from the smarts of any single member.
To accomplish this, the researchers recruited 699 people, divided them into 152 teams, and gave each group a series of assignments that required different kinds of cooperation. Most teams began by spending ten minutes brainstorming possible uses for a brick and received a point for each unique idea. Then they were asked to plan a shopping trip as if they were housemates sharing a single car: Each teammate was given a different list of groceries to buy and a map showing prices at various stores. The only way to maximize the team’s score was for each person to sacrifice one item they really wanted in exchange for something that pleased the entire group. Then the teams were told to arrive at a ruling on a disciplinary case in which a college basketball player allegedly bribed his teacher. Some teammates represented the interests of the faculty; others were standins for the athletics department. Points were awarded for reaching a verdict that maximized each group’s concerns.
Each of these tasks required full team participation; each demanded different kinds of collaboration. As the researchers observed groups going about the tasks, they saw various dynamics emerge. Some teams came up with dozens of clever uses for the brick, arrived at a verdict that made everyone happy, and easily divvied up the shopping trip. Others kept describing the same uses for the brick in different words; came to verdicts that left some participants feeling alienated; and managed to buy only ice cream and Froot Loops because no one was willing to compromise. What was interesting was that teams that did well on one assignment also seemed to do well on the others. Conversely, teams that failed at one thing seemed to fail at everything.
Some might have hypothesized that the “good teams” were successful because their members were smarter—that group intelligence might be nothing more than the intelligence of the individuals making up the team. But the researchers had tested participants’ IQs beforehand and found that individual intelligence didn’t correlate with team performance. Putting ten smart people in a room didn’t mean they solved problems more intelligently—in fact, those smart people were often outperformed by groups consisting of people who had scored lower on intellect tests, but who still seemed smarter as a group.
Others might have argued that the good teams had more decisive leaders. But the research showed that wasn’t right, either.
The researchers eventually concluded that the good teams had succeeded not because of innate qualities of team members, but because of how they treated one another. Put differently, the most successful teams had norms that caused everyone to mesh particularly well.
“We find converging evidence of a general collective intelligence factor that explains a group’s performance on a wide variety of tasks,” the researchers wrote in their Science article. “This kind of collective intelligence is a property of the group itself, not just the individuals in it.” It was the norms, not the people, that made teams so smart. The right norms could raise the collective intelligence of mediocre thinkers. The wrong norms could hobble a group made up of people who, on their own, were all exceptionally bright.
But when the researchers reviewed videos of the good teams’ interactions, they noticed that not all norms looked alike. “It was striking how different some of them behaved,” said Anita Woolley, the study’s lead author. “Some teams had a bunch of smart people who figured out how to break up work evenly. Other groups had pretty average members but came up with ways to take advantage of everyone’s relative strengths. Some groups had one strong leader. Others were more fluid, and everyone took a leadership role.”
There were, however, two behaviors that all the good teams shared.