What Google Learned From Its Quest to Build the Perfect Team
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 aresearcher for two professors at Harvard, which was interesting but lonely. Maybe a bigcorporation would be a better fit. Or perhaps a fast-growing start-up. All she knew forcertain was that she wanted to find a job that was more social. ‘‘I wanted to be part of acommunity, part of something people were building together,’’ she told me. She thoughtabout various opportunities — Internet companies, a Ph.D. program — but nothingseemed exactly right. So in 2009, she chose the path that allowed her to put off making adecision: She applied to business schools and was accepted by the Yale School ofManagement.
When Rozovsky arrived on campus, she was assigned to a study group carefullyengineered by the school to foster tight bonds. Study groups have become a rite ofpassage at M.B.A. programs, a way for students to practice working in teams and areflection of the increasing demand for employees who can adroitly navigate groupdynamics. A worker today might start the morning by collaborating with a team ofengineers, then send emails to colleagues marketing a new brand, then jump on aconference call planning an entirely different product line, while also juggling teammeetings with accounting and the party-planning committee. To prepare students for thatcomplex world, business schools around the country have revised their curriculums toemphasize team-focused learning.
Every day, between classes or after dinner, Rozovsky and her four teammates gathered todiscuss homework assignments, compare spreadsheets and strategize for exams.Everyone was smart and curious, and they had a lot in common: They had gone to similarcolleges and had worked at analogous firms. These shared experiences, Rozovsky hoped,would make it easy for them to work well together. But it didn’t turn out that way. ‘‘Thereare lots of people who say some of their best business-school friends come from theirstudy groups,’’ Rozovsky told me. ‘‘It wasn’t like that for me.’’
Instead, Rozovsky’s study group was a source of stress. ‘‘I always felt like I had to provemyself,’’ she said. The team’s dynamics could put her on edge. When the group met,teammates sometimes jockeyed for the leadership position or criticized one another’sideas. There were conflicts over who was in charge and who got to represent the group inclass. ‘‘People would try to show authority by speaking louder or talking over each other,’’Rozovsky told me. ‘‘I always felt like I had to be careful not to make mistakes around
them.’’
So Rozovsky started looking for other groups she could join. A classmate mentioned thatsome students were putting together teams for ‘‘case competitions,’’ contests in whichparticipants proposed solutions to real-world business problems that were evaluated byjudges, who awarded trophies and cash. The competitions were voluntary, but the workwasn’t all that different from what Rozovsky did with her study group: conducting lots ofresearch and financial analyses, writing reports and giving presentations. The members ofher case-competition team had a variety of professional experiences: Army officer,researcher at a think tank, director of a health-education nonprofit organization andconsultant to a refugee program. Despite their disparate backgrounds, however, everyoneclicked. They emailed one another dumb jokes and usually spent the first 10 minutes ofeach meeting chatting. When it came time to brainstorm, ‘‘we had lots of crazy ideas,’’Rozovsky said.
One of her favorite competitions asked teams to come up with a new business to replacea student-run snack store on Yale’s campus. Rozovsky proposed a nap room and sellingearplugs and eyeshades to make money. Someone else suggested filling the space withold video games. There were ideas about clothing swaps. Most of the proposals wereimpractical, but ‘‘we all felt like we could say anything to each other,’’ Rozovsky told me.‘‘No one worried that the rest of the team was judging them.’’ Eventually, the team settledon a plan for a microgym with a handful of exercise classes and a few weight machines.They won the competition. (The microgym — with two stationary bicycles and threetreadmills — still exists.)
Rozovsky’s study group dissolved in her second semester (it was up to the studentswhether they wanted to continue). Her case team, however, stuck together for the twoyears she was at Yale.
It always struck Rozovsky as odd that her experiences with the two groups weredissimilar. Each was composed of people who were bright and outgoing. When she talkedone on one with members of her study group, the exchanges were friendly and warm. Itwas only when they gathered as a team that things became fraught. By contrast, hercase-competition team was always fun and easygoing. In some ways, the team’smembers got along better as a group than as individual friends.
‘‘I couldn’t figure out why things had turned out so different,’’ Rozovsky told me. ‘‘It didn’tseem like it had to happen that way.’’
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 corporatecampuses and within university laboratories, psychologists, sociologists and statisticiansare devoting themselves to studying everything from team composition to email patternsin order to figure out how to make employees into faster, better and more productiveversions of themselves. ‘‘We’re living through a golden age of understanding personalproductivity,’’ says Marshall Van Alstyne, a professor at Boston University who studieshow people share information. ‘‘All of a sudden, we can pick apart the small choices thatall of us make, decisions most of us don’t even notice, and figure out why some peopleare so much more effective than everyone else.’’
Yet many of today’s most valuable firms have come to realize that analyzing andimproving individual workers — a practice known as ‘‘employee performanceoptimization’’ — isn’t enough. As commerce becomes increasingly global and complex,the bulk of modern work is more and more team-based. One study, published in TheHarvard Business Review last month, found that ‘‘the time spent by managers andemployees in collaborative activities has ballooned by 50 percent or more’’ over the lasttwo decades and that, at many companies, more than three-quarters of an employee’sday is spent communicating with colleagues.
In Silicon Valley, software engineers are encouraged to work together, in part becausestudies show that groups tend to innovate faster, see mistakes more quickly and findbetter solutions to problems. Studies also show that people working in teams tend toachieve better results and report higher job satisfaction. In a 2015 study, executives saidthat profitability increases when workers are persuaded to collaborate more. Withincompanies and conglomerates, as well as in government agencies and schools, teams arenow the fundamental unit of organization. If a company wants to outstrip its competitors,it needs to influence not only how people work but also how they work together.
Five years ago, Google — one of the most public proselytizers of how studying workerscan transform productivity — became focused on building the perfect team. In the lastdecade, the tech giant has spent untold millions of dollars measuring nearly every aspectof its employees’ lives. Google’s People Operations department has scrutinizedeverything from how frequently particular people eat together (the most productiveemployees tend to build larger networks by rotating dining companions) to which traitsthe best managers share (unsurprisingly, good communication and avoidingmicromanaging is critical; more shocking, this was news to many Google managers).
The company’s top executives long believed that building the best teams meantcombining the best people. They embraced other bits of conventional wisdom as well, like
‘‘It’s better to put introverts together,’’ said Abeer Dubey, a manager in Google’s PeopleAnalytics division, or ‘‘Teams are more effective when everyone is friends away fromwork.’’ But, Dubey went on, ‘‘it turned out no one had really studied which of those weretrue.’’
In 2012, the company embarked on an initiative — code-named Project Aristotle — tostudy hundreds of Google’s teams and figure out why some stumbled while otherssoared. Dubey, a leader of the project, gathered some of the company’s best statisticians,organizational psychologists, sociologists and engineers. He also needed researchers.Rozovsky, by then, had decided that what she wanted to do with her life was studypeople’s habits and tendencies. After graduating from Yale, she was hired by Google andwas soon assigned to Project Aristotle.
Project Aristotle’s researchers began by reviewing a half-century of academic studieslooking at how teams worked. Were the best teams made up of people with similarinterests? Or did it matter more whether everyone was motivated by the same kinds ofrewards? Based on those studies, the researchers scrutinized the composition of groupsinside Google: How often did teammates socialize outside the office? Did they have thesame hobbies? Were their educational backgrounds similar? Was it better for allteammates to be outgoing or for all of them to be shy? They drew diagrams showingwhich teams had overlapping memberships and which groups had exceeded theirdepartments’ goals. They studied how long teams stuck together and if gender balanceseemed to have an impact on a team’s success.
No matter how researchers arranged the data, though, it was almost impossible to findpatterns — or any evidence that the composition of a team made any difference. ‘‘Welooked at 180 teams from all over the company,’’ Dubey said. ‘‘We had lots of data, butthere was nothing showing that a mix of specific personality types or skills orbackgrounds made any difference. The ‘who’ part of the equation didn’t seem to matter.’’
Some groups that were ranked among Google’s most effective teams, for instance, werecomposed of friends who socialized outside work. Others were made up of people whowere basically strangers away from the conference room. Some groups sought strongmanagers. Others preferred a less hierarchical structure. Most confounding of all, twoteams might have nearly identical makeups, with overlapping memberships, but radicallydifferent levels of effectiveness. ‘‘At Google, we’re good at finding patterns,’’ Dubey said.‘‘There weren’t strong patterns here.’’
As they struggled to figure out what made a team successful, Rozovsky and her
colleagues kept coming across research by psychologists and sociologists that focusedon what are known as ‘‘group norms.’’ Norms are the traditions, behavioral standards andunwritten rules that govern how we function when we gather: One team may come to aconsensus that avoiding disagreement is more valuable than debate; another team mightdevelop a culture that encourages vigorous arguments and spurns groupthink. Norms canbe unspoken or openly acknowledged, but their influence is often profound. Teammembers may behave in certain ways as individuals — they may chafe against authorityor prefer working independently — but when they gather, the group’s norms typicallyoverride individual proclivities and encourage deference to the team.
Project Aristotle’s researchers began searching through the data they had collected,looking for norms. They looked for instances when team members described a particularbehavior 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 constantlyand that team leaders reinforced that behavior by interrupting others themselves. Onother teams, leaders enforced conversational order, and when someone cut off ateammate, group members would politely ask everyone to wait his or her turn. Someteams celebrated birthdays and began each meeting with informal chitchat aboutweekend plans. Other groups got right to business and discouraged gossip. There wereteams that contained outsize personalities who hewed to their group’s sedate norms, andothers in which introverts came out of their shells as soon as meetings began.
After looking at over a hundred groups for more than a year, Project Aristotle researchersconcluded that understanding and influencing group norms were the keys to improvingGoogle’s teams. But Rozovsky, now a lead researcher, needed to figure out which normsmattered most. Google’s research had identified dozens of behaviors that seemedimportant, except that sometimes the norms of one effective team contrasted sharplywith those of another equally successful group. Was it better to let everyone speak asmuch as they wanted, or should strong leaders end meandering debates? Was it moreeffective for people to openly disagree with one another, or should conflicts be playeddown? The data didn’t offer clear verdicts. In fact, the data sometimes pointed in oppositedirections. The only thing worse than not finding a pattern is finding too many of them.Which norms, Rozovsky and her colleagues wondered, were the ones that successfulteams shared?
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 youwatch 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 todo. When someone makes a side comment, the speaker stops, reminds everyone of theagenda and pushes the meeting back on track. This team is efficient. There is no idlechitchat or long debates. The meeting ends as scheduled and disbands so everyone canget back to their desks.
Team B is different. It’s evenly divided between successful executives and middlemanagers with few professional accomplishments. Teammates jump in and out ofdiscussions. People interject and complete one another’s thoughts. When a team memberabruptly changes the topic, the rest of the group follows him off the agenda. At the end ofthe meeting, the meeting doesn’t actually end: Everyone sits around to gossip and talkabout their lives.
Which group would you rather join?
In 2008, a group of psychologists from Carnegie Mellon, M.I.T. and Union College beganto try to answer a question very much like this one. ‘‘Over the past century, psychologistsmade considerable progress in defining and systematically measuring intelligence inindividuals,’’ the researchers wrote in the journal Science in 2010. ‘‘We have used thestatistical approach they developed for individual intelligence to systematically measurethe intelligence of groups.’’ Put differently, the researchers wanted to know if there is acollective I."Q. that emerges within a team that is distinct from the smarts of any singlemember.
To accomplish this, the researchers recruited 699 people, divided them into small groupsand gave each a series of assignments that required different kinds of cooperation. Oneassignment, for instance, asked participants to brainstorm possible uses for a brick. Someteams came up with dozens of clever uses; others kept describing the same ideas indifferent words. Another had the groups plan a shopping trip and gave each teammate adifferent list of groceries. The only way to maximize the group’s score was for eachperson to sacrifice an item they really wanted for something the team needed. Somegroups easily divvied up the buying; others couldn’t fill their shopping carts because noone was willing to compromise.
What interested the researchers most, however, was that teams that did well on oneassignment usually did well on all the others. Conversely, teams that failed at one thingseemed to fail at everything. The researchers eventually concluded that whatdistinguished the ‘‘good’’ teams from the dysfunctional groups was how teammatestreated one another. The right norms, in other words, could raise a group’s collective
intelligence, whereas the wrong norms could hobble a team, even if, individually, all themembers were exceptionally bright.
But what was confusing was that not all the good teams appeared to behave in the sameways. ‘‘Some teams had a bunch of smart people who figured out how to break up workevenly,’’ said Anita Woolley, the study’s lead author. ‘‘Other groups had pretty averagemembers, but they 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 aleadership role.’’
As the researchers studied the groups, however, they noticed two behaviors that all thegood teams generally shared. First, on the good teams, members spoke in roughly thesame proportion, a phenomenon the researchers referred to as ‘‘equality in distribution ofconversational turn-taking.’’ On some teams, everyone spoke during each task; on others,leadership shifted among teammates from assignment to assignment. But in each case,by the end of the day, everyone had spoken roughly the same amount. ‘‘As long aseveryone got a chance to talk, the team did well,’’ Woolley said. ‘‘But if only one person ora small group spoke all the time, the collective intelligence declined.’’
Second, the good teams all had high ‘‘average social sensitivity’’ — a fancy way of sayingthey were skilled at intuiting how others felt based on their tone of voice, theirexpressions and other nonverbal cues. One of the easiest ways to gauge social sensitivityis to show someone photos of people’s eyes and ask him or her to describe what thepeople are thinking or feeling — an exam known as the Reading the Mind in the Eyes test.People on the more successful teams in Woolley’s experiment scored above average onthe Reading the Mind in the Eyes test. They seemed to know when someone was feelingupset or left out. People on the ineffective teams, in contrast, scored below average. Theyseemed, as a group, to have less sensitivity toward their colleagues.
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 smartpeople, all optimized for peak individual efficiency. But the group’s norms discourageequal speaking; there are few exchanges of the kind of personal information that letsteammates pick up on what people are feeling or leaving unsaid. There’s a good chancethe members of Team A will continue to act like individuals once they come together, andthere’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 socializeinstead of remaining focused on the agenda. The team may seem inefficient to a casual
observer. But all the team members speak as much as they need to. They are sensitive toone another’s moods and share personal stories and emotions. While Team B might notcontain as many individual stars, the sum will be greater than its parts.
Within psychology, researchers sometimes colloquially refer to traits like ‘‘conversationalturn-taking’’ and ‘‘average social sensitivity’’ as aspects of what’s known as psychologicalsafety — a group culture that the Harvard Business School professor Amy Edmondsondefines as a ‘‘shared belief held by members of a team that the team is safe forinterpersonal risk-taking.’’ Psychological safety is ‘‘a sense of confidence that the teamwill not embarrass, reject or punish someone for speaking up,’’ Edmondson wrote in astudy published in 1999. ‘‘It describes a team climate characterized by interpersonal trustand mutual respect in which people are comfortable being themselves.’’
When Rozovsky and her Google colleagues encountered the concept of psychologicalsafety 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.’’ That team, researchers estimated, wasamong Google’s accomplished groups. By contrast, another engineer had told theresearchers that his ‘‘team leader has poor emotional control.’’ He added: ‘‘He panics oversmall issues and keeps trying to grab control. I would hate to be driving with him being inthe passenger seat, because he would keep trying to grab the steering wheel and crashthe car.’’ That team, researchers presumed, did not perform well.
Most of all, employees had talked about how various teams felt. ‘‘And that made a lot ofsense to me, maybe because of my experiences at Yale,’’ Rozovsky said. ‘‘I’d been onsome teams that left me feeling totally exhausted and others where I got so much energyfrom the group.’’ Rozovsky’s study group at Yale was draining because the norms — thefights over leadership, the tendency to critique — put her on guard. Whereas the norms ofher case-competition team — enthusiasm for one another’s ideas, joking around andhaving fun — allowed everyone to feel relaxed and energized.
For Project Aristotle, research on psychological safety pointed to particular norms that arevital to success. There were other behaviors that seemed important as well — like makingsure teams had clear goals and creating a culture of dependability. But Google’s dataindicated that psychological safety, more than anything else, was critical to making a teamwork.
‘‘We had to get people to establish psychologically safe environments,’’ Rozovsky told me.But it wasn’t clear how to do that. ‘‘People here are really busy,’’ she said. ‘‘We needed
clear guidelines.’’
However, establishing psychological safety is, by its very nature, somewhat messy anddifficult to implement. You can tell people to take turns during a conversation and to listento one another more. You can instruct employees to be sensitive to how their colleaguesfeel and to notice when someone seems upset. But the kinds of people who work atGoogle are often the ones who became software engineers because they wanted to avoidtalking about feelings in the first place.
Rozovsky and her colleagues had figured out which norms were most critical. Now theyhad to find a way to make communication and empathy — the building blocks of forgingreal connections — into an algorithm they could easily scale.
Illustration by James Graham
In late 2014, Rozovsky and her fellow Project Aristotle number-crunchers began sharingtheir findings with select groups of Google’s 51,000 employees. By then, they had beencollecting surveys, conducting interviews and analyzing statistics for almost three years.They hadn’t yet figured out how to make psychological safety easy, but they hoped thatpublicizing their research within Google would prompt employees to come up with someideas of their own.
After Rozovsky gave one presentation, a trim, athletic man named Matt Sakaguchiapproached the Project Aristotle researchers. Sakaguchi had an unusual background for aGoogle employee. Twenty years earlier, he was a member of a SWAT team in WalnutCreek, Calif., but left to become an electronics salesman and eventually landed at Googleas a midlevel manager, where he has overseen teams of engineers who respond when thecompany’s websites or servers go down.
‘‘I might be the luckiest individual on earth,’’ Sakaguchi told me. ‘‘I’m not really anengineer. I didn’t study computers in college. Everyone who works for me is much smarterthan I am.’’ But he is talented at managing technical workers, and as a result, Sakaguchihas thrived at Google. He and his wife, a teacher, have a home in San Francisco and aweekend house in the Sonoma Valley wine country. ‘‘Most days, I feel like I’ve won thelottery,’’ he said.
Sakaguchi was particularly interested in Project Aristotle because the team he previouslyoversaw at Google hadn’t jelled particularly well. ‘‘There was one senior engineer whowould just talk and talk, and everyone was scared to disagree with him,’’ Sakaguchi said.‘‘The hardest part was that everyone liked this guy outside the group setting, but
whenever they got together as a team, something happenedthat made the culture go wrong.’’
Sakaguchi had recently become the manager of a new team,and he wanted to make sure things went better this time. Sohe asked researchers at Project Aristotle if they could help.They provided him with a survey to gauge the group’snorms.
When Sakaguchi asked his new team to participate, he wasgreeted with skepticism. ‘‘It seemed like a total waste oftime,’’ said Sean Laurent, an engineer. ‘‘But Matt was ournew boss, and he was really into this questionnaire, and sowe said, Sure, we’ll do it, whatever.’’
The team completed the survey, and a few weeks later,Sakaguchi received the results. He was surprised by whatthey revealed. He thought of the team as a strong unit. Butthe results indicated there were weaknesses: When asked torate whether the role of the team was clearly understood andwhether their work had impact, members of the team gavemiddling to poor scores. These responses troubledSakaguchi, because he hadn’t picked up on this discontent.He wanted everyone to feel fulfilled by their work. He askedthe team to gather, off site, to discuss the survey’s results.He began by asking everyone to share something personalabout themselves. He went first.
‘‘I think one of the things most people don’t know about me,’’he told the group, ‘‘is that I have Stage 4 cancer.’’ In 2001, hesaid, a doctor discovered a tumor in his kidney. By the timethe cancer was detected, it had spread to his spine. Fornearly half a decade, it had grown slowly as he underwent
treatment while working at Google. Recently, however, doctors had found a new,worrisome spot on a scan of his liver. That was far more serious, he explained.
No one knew what to say. The team had been working with Sakaguchi for 10 months.They all liked him, just as they all liked one another. No one suspected that he was dealingwith anything like this.
‘‘To have Matt stand there and tell us that he’s sick and he’s not going to get better and,you know, what that means,’’ Laurent said. ‘‘It was a really hard, really special moment.’’
After Sakaguchi spoke, another teammate stood and described some health issues of herown. Then another discussed a difficult breakup. Eventually, the team shifted its focus tothe survey. They found it easier to speak honestly about the things that had beenbothering them, their small frictions and everyday annoyances. They agreed to adoptsome new norms: From now on, Sakaguchi would make an extra effort to let the teammembers know how their work fit into Google’s larger mission; they agreed to try harderto notice when someone on the team was feeling excluded or down.
There was nothing in the survey that instructed Sakaguchi to share his illness with thegroup. There was nothing in Project Aristotle’s research that said that getting people toopen up about their struggles was critical to discussing a group’s norms. But toSakaguchi, it made sense that psychological safety and emotional conversations wererelated. The behaviors that create psychological safety — conversational turn-taking andempathy — are part of the same unwritten rules we often turn to, as individuals, when weneed to establish a bond. And those human bonds matter as much at work as anywhereelse. In fact, they sometimes matter more.
‘‘I think, until the off-site, I had separated things in my head into work life and life life,’’Laurent told me. ‘‘But the thing is, my work is my life. I spend the majority of my timeworking. Most of my friends I know through work. If I can’t be open and honest at work,then I’m not really living, am I?’’
What Project Aristotle has taught people within Google is that no one wants to put on a‘‘work face’’ when they get to the office. No one wants to leave part of their personalityand inner life at home. But to be fully present at work, to feel ‘‘psychologically safe,’’ wemust know that we can be free enough, sometimes, to share the things that scare uswithout fear of recriminations. We must be able to talk about what is messy or sad, tohave hard conversations with colleagues who are driving us crazy. We can’t be focusedjust on efficiency. Rather, when we start the morning by collaborating with a team ofengineers and then send emails to our marketing colleagues and then jump on aconference call, we want to know that those people really hear us. We want to know thatwork is more than just labor.
Which isn’t to say that a team needs an ailing manager to come together. Any group canbecome Team B. Sakaguchi’s experiences underscore a core lesson of Google’s researchinto teamwork: By adopting the data-driven approach of Silicon Valley, Project Aristotle
has encouraged emotional conversations and discussions of norms among people whomight otherwise be uncomfortable talking about how they feel. ‘‘Googlers love data,’’Sakaguchi told me. But it’s not only Google that loves numbers, or Silicon Valley that shiesaway from emotional conversations. Most workplaces do. ‘‘By putting things like empathyand sensitivity into charts and data reports, it makes them easier to talk about,’’Sakaguchi told me. ‘‘It’s easier to talk about our feelings when we can point to a number.’’
Sakaguchi knows that the spread of his cancer means he may not have much time left.His wife has asked him why he doesn’t quit Google. At some point, he probably will. Butright now, helping his team succeed ‘‘is the most meaningful work I’ve ever done,’’ he toldme. He encourages the group to think about the way work and life mesh. Part of that, hesays, is recognizing how fulfilling work can be. Project Aristotle ‘‘proves how much a greatteam matters,’’ he said. ‘‘Why would I walk away from that? Why wouldn’t I spend timewith people who care about me?’’
The technology industry is not just one of the fastest growing parts of our economy; it isalso increasingly the world’s dominant commercial culture. And at the core of SiliconValley are certain self-mythologies and dictums: Everything is different now, data reignssupreme, today’s winners deserve to triumph because they are cleareyed enough todiscard yesterday’s conventional wisdoms and search out the disruptive and the new.
The paradox, of course, is that Google’s intense data collection and number crunchinghave led it to the same conclusions that good managers have always known. In the bestteams, members listen to one another and show sensitivity to feelings and needs.
The fact that these insights aren’t wholly original doesn’t mean Google’s contributionsaren’t valuable. In fact, in some ways, the ‘‘employee performance optimization’’movement has given us a method for talking about our insecurities, fears and aspirationsin more constructive ways. It also has given us the tools to quickly teach lessons thatonce took managers decades to absorb. Google, in other words, in its race to build theperfect team, has perhaps unintentionally demonstrated the usefulness of imperfectionand done what Silicon Valley does best: figure out how to create psychological safetyfaster, better and in more productive ways.
‘‘Just having data that proves to people that these things are worth paying attention tosometimes is the most important step in getting them to actually pay attention,’’ Rozovskytold me. ‘‘Don’t underestimate the power of giving people a common platform andoperating language.’’
Project Aristotle is a reminder that when companies try to optimize everything, it’s
sometimes easy to forget that success is often built on experiences — like emotionalinteractions and complicated conversations and discussions of who we want to be andhow our teammates make us feel — that can’t really be optimized. Rozovsky herself wasreminded of this midway through her work with the Project Aristotle team. ‘‘We were in ameeting where I made a mistake,’’ Rozovsky told me. She sent out a note afterwardexplaining how she was going to remedy the problem. ‘‘I got an email back from a teammember that said, ‘Ouch,’"’’ she recalled. ‘‘It was like a punch to the gut. I was alreadyupset about making this mistake, and this note totally played on my insecurities.’’
If this had happened earlier in Rozovsky’s life — if it had occurred while she was at Yale,for instance, in her study group — she probably wouldn’t have known how to deal withthose feelings. The email wasn’t a big enough affront to justify a response. But all thesame, it really bothered her. It was something she felt she needed to address.
And thanks to Project Aristotle, she now had a vocabulary for explaining to herself whatshe was feeling and why it was important. She had graphs and charts telling her that sheshouldn’t just let it go. And so she typed a quick response: ‘‘Nothing like a good ‘Ouch!’ todestroy psych safety in the morning.’’ Her teammate replied: ‘‘Just testing yourresilience.’’
‘‘That could have been the wrong thing to say to someone else, but he knew it wasexactly what I needed to hear,’’ Rozovsky said. ‘‘With one 30-second interaction, wedefused the tension.’’ She wanted to be listened to. She wanted her teammate to besensitive to what she was feeling. ‘‘And I had research telling me that it was O.K. to followmy gut,’’ she said. ‘‘So that’s what I did. The data helped me feel safe enough to do what Ithought was right.’’
