Class14b

Return to home page.

What Google Learned About Teams?

like most 25-year-olds, Julia Rozovsky wasn’t sure what she wanted to do with her life. She had worked at a consulting firm, but it wasn’t a good match. Then she became a researcher for two professors at Harvard, which was interesting but lonely. Maybe a big corporation would be a better fit. Or perhaps a fast-growing start-up. she told me. She thought about various opportunities — Internet companies, a Ph.D. program — but nothing seemed exactly right. So in 2009, she chose the path that allowed her to put off making a decision: She applied to business schools and was accepted by the Yale School of Management.

When Rozovsky arrived on campus, she was assigned to a study group carefully engineered by the school to foster tight bonds. Study groups have become a rite of passage at M.B.A. programs, a way for students to practice working in teams and a reflection of the increasing demand for employees who can adroitly navigate group dynamics.

Every day, between classes or after dinner, Rozovsky and her four teammates gathered to discuss homework assignments, compare spreadsheets and strategize for exams. Everyone was smart and curious, and they had a lot in common: They had gone to similar colleges and had worked at analogous firms. These shared experiences.

Instead, Rozovsky’s study group was a source of stress. ‘‘I always felt like I had to prove myself,’’ she said. The team’s dynamics could put her on edge.

So Rozovsky started looking for other groups she could join. A classmate mentioned that some students were putting together teams for ‘‘case competitions,’’ contests in which participants proposed solutions to real-world business problems that were evaluated by judges, who awarded trophies and cash.

One of her favorite competitions asked teams to come up with a new business to replace a student-run snack store on Yale’s campus. Rozovsky proposed a nap room and selling earplugs and eyeshades to make money. Someone else suggested filling the space with old video games. There were ideas about clothing swaps.

Rozovsky’s study group dissolved in her second semester (it was up to the students whether they wanted to continue). Her case team, however, stuck together for the two years she was at Yale.

It always struck Rozovsky as odd that her experiences with the two groups were dissimilar. Each was composed of people who were bright and outgoing. When she talked one on one with members of her study group, the exchanges were friendly and warm. It was only when they gathered as a team that things became fraught.

our data-saturated age enables us to examine our work habits and office quirks with a scrutiny that our cubicle-bound forebears could only dream of. Today, on corporate campuses and within university laboratories, psychologists, sociologists and statisticians are devoting themselves to studying everything from team composition to email patterns in order to figure out how to make employees into faster, better and more productive versions of themselves.

In Silicon Valley, software engineers are encouraged to work together, in part because studies show that groups tend to innovate faster, see mistakes more quickly and find better solutions to problems. Studies also show that people working in teams tend to achieve better results and report higher job satisfaction.

Five years ago, Google — one of the most public proselytizers of how studying workers can transform productivity — became focused on building the perfect team. In the last decade, the tech giant has spent untold millions of dollars measuring nearly every aspect of its employees’ lives.

The company’s top executives long believed that building the best teams meant combining the best people. They embraced other bits of conventional wisdom as well.

In 2012, the company embarked on an initiative — code-named Project Aristotle — to study hundreds of Google’s teams and figure out why some stumbled while others soared.

project Aristotle’s researchers began by reviewing a half-century of academic studies looking at how teams worked. Were the best teams made up of people with similar interests? Or did it matter more whether everyone was motivated by the same kinds of rewards? Based on those studies, the researchers scrutinized the composition of groups inside Google: How often did teammates socialize outside the office? Did they have the same hobbies? Were their educational backgrounds similar? Was it better for all teammates to be outgoing or for all of them to be shy? They drew diagrams showing which teams had overlapping memberships and which groups had exceeded their departments’ goals. They studied how long teams stuck together and if gender balance seemed to have an impact on a team’s success.

No matter how researchers arranged the data, though, it was almost impossible to find patterns — or any evidence that the composition of a team made any difference.

Some groups that were ranked among Google’s most effective teams, for instance, were composed of friends who socialized outside work. Others were made up of people who were basically strangers away from the conference room. Some groups sought strong managers. Others preferred a less hierarchical structure.

As they struggled to figure out what made a team successful, Rozovsky and her colleagues kept coming across research by psychologists and sociologists that focused on what are known as ‘‘group norms.’’ Norms are the traditions, behavioral standards and unwritten rules that govern how we function when we gather: One team may come to a consensus that avoiding disagreement is more valuable than debate; another team might develop a culture that encourages vigorous arguments and spurns groupthink.

Project Aristotle’s researchers began searching through the data they had collected, looking for norms. They looked for instances when team members described a particular behavior as an ‘‘unwritten rule’’ or when they explained certain things as part of the ‘‘team’s culture.’’ Some groups said that teammates interrupted one another constantly and that team leaders reinforced that behavior by interrupting others themselves.

After looking at over a hundred groups for more than a year, Project Aristotle researchers concluded that understanding and influencing group norms were the keys to improving Google’s teams. But Rozovsky, now a lead researcher, needed to figure out which norms mattered most. Google’s research had identified dozens of behaviors that seemed important, except that sometimes the norms of one effective team contrasted sharply with those of another equally successful group.

Imagine you have been invited to join one of two groups.

Team A is composed of people who are all exceptionally smart and successful. When you watch a video of this group working, you see professionals who wait until a topic arises in which they are expert, and then they speak at length, explaining what the group ought to do. When someone makes a side comment, the speaker stops, reminds everyone of the agenda and pushes the meeting back on track.

Team B is different. It’s evenly divided between successful executives and middle managers with few professional accomplishments. Teammates jump in and out of discussions. People interject and complete one another’s thoughts. When a team member abruptly changes the topic, the rest of the group follows him off the agenda.

In 2008, a group of psychologists from Carnegie Mellon, M.I.T. and Union College began to try to answer a question very much like this one. ‘‘Over the past century, psychologists made considerable progress in defining and systematically measuring intelligence in individuals,’’ the researchers wrote in the journal Science in 2010. ‘‘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 I. Q. that emerges within a team that is distinct from the smarts of any single member.

What interested the researchers most, however, was that teams that did well on one assignment usually did well on all the others. Conversely, teams that failed at one thing seemed to fail at everything. The researchers eventually concluded that what distinguished the ‘‘good’’ teams from the dysfunctional groups was how teammates treated one another.

But what was confusing was that not all the good teams appeared to behave in the same ways. ‘‘Some teams had a bunch of smart people who figured out how to break up work evenly,’’ said Anita Woolley, the study’s lead author. ‘‘Other groups had pretty average members, but they came up with ways to take advantage of everyone’s relative strengths. Some groups had one strong leader.

As the researchers studied the groups, however, they noticed two behaviors that all the good teams generally shared. First, on the good teams, members spoke in roughly the same proportion, a phenomenon the researchers referred to as ‘‘equality in distribution of conversational turn-taking.’’Second, the good teams all had high ‘‘average social sensitivity’’ — a fancy way of saying they were skilled at intuiting how others felt based on their tone of voice, their expressions and other nonverbal cues. One of the easiest ways to gauge social sensitivity is to show someone photos of people’s eyes and ask him or her to describe what the people are thinking or feeling — an exam known as the Reading the Mind in the Eyes test.

In other words, if you are given a choice between the serious-minded Team A or the free-flowing Team B, you should probably opt for Team B. Team A may be filled with smart people, all optimized for peak individual efficiency.

There’s a good chance the members of Team A will continue to act like individuals once they come together, and there’s little to suggest that, as a group, they will become more collectively intelligent. In contrast, on Team B, people may speak over one another, go on tangents and socialize instead of remaining focused on the agenda.

Within psychology, researchers sometimes colloquially refer to traits like ‘‘conversational turn-taking’’ and ‘‘average social sensitivity’’ as aspects of what’s known as psychological safety — a group culture that the Harvard Business School professor Amy Edmondson defines as a ‘‘shared belief held by members of a team that the team is safe for interpersonal risk-taking.’’

When Rozovsky and her Google colleagues encountered the concept of psychological safety in academic papers, it was as if everything suddenly fell into place. One engineer, for instance, had told researchers that his team leader was ‘‘direct and straightforward, which creates a safe space for you to take risks.’’

Most of all, employees had talked about how various teams felt. ‘‘And that made a lot of sense to me, maybe because of my experiences at Yale,’’ Rozovsky said. ‘‘I’d been on some teams that left me feeling totally exhausted and others where I got so much energy from the group.’’

For Project Aristotle, research on psychological safety pointed to particular norms that are vital to success. There were other behaviors that seemed important as well — like making sure teams had clear goals and creating a culture of dependability. But Google’s data indicated that psychological safety, more than anything else, was critical to making a team work.

However, establishing psychological safety is, by its very nature, somewhat messy and difficult to implement. You can tell people to take turns during a conversation and to listen to one another more.

Rozovsky and her colleagues had figured out which norms were most critical. Now they had to find a way to make communication and empathy — the building blocks of forging real connections — into an algorithm they could easily scale.