In 2013, Alistair Shepherd asked everyone in a business-school pitch competition to complete a survey with questions inspired by online-dating sites. It asked things like “Do you like horror movies?” and “Do spelling mistakes annoy you?”
Using only the survey results—and knowing nothing about the business proposals—Mr. Shepherd predicted how well the eight teams would collaborate internally and how they would ultimately fare.
He ended up ranking all eight correctly. What made for a great team? Those in which people had the most tolerance for their teammates’ perspectives—and those in which people had the greatest diversity in personalities. Introverts and extroverts, for example, or improvisers and organizers.
Mr. Shepherd’s experiment represents an attempt to get beyond the usual approach to workplace chemistry. We often have an intuitive understanding about what qualities make people click or clash in the workplace, but don’t follow up on that gut feeling in a systematic way. We don’t always make sure our assessments of people’s personalities are actually correct, and we don’t analyze what happens when we throw together people with certain traits.
Now a number of organizations are trying to get more rigorous about the process. They’re picking employees’ brains to get a deeper understanding of their personalities and motives, and then analyzing team performance to see what type of person works best with whom. In some cases, they’re measuring very subtle personality signals—such as how often different people speak to each other, and even the body language that they use—using wearable tracking devices.
One of the biggest efforts to quantify what makes teams work together comes at Alphabet Inc.’s Google. Its People Operations department crunches data to answer questions such as whether a team’s productivity was correlated to how often its members socialized outside the office (not necessarily). Google also found that the best teams created a culture of “psychological safety,” meaning team members could share thoughts, ideas and concerns without fear of ridicule or punishment. The analytics team also looks at issues not directly related to teamwork, such as the optimal length for paid maternity leave. (Longer leaves generally result in lower attrition rates for new mothers.)
The survey says…
For companies that can’t muster an in-house research department, a handful of consulting firms have sprung up to combine existing and original research about what makes for optimal teams. Then they send clients reports and reminders about how to work better.
In some cases, companies are using surveys from Saberr, the firm Mr. Shepherd founded, to help vet job applicants. The applicants take a survey, and Saberr compares the results with those of existing employees; applicants get a score ranging from 0 to 100 to show how compatible they are with the team they’re vying to work for. Among the questions asked: How important it is to take risks or have fun at work?
Saberr and other consultants also use those strategies to figure out how best to organize existing teams. Mr. Shepherd says one client, a call center with 300 employees that was divided into teams of three or four people, reorganized some of the teams as a result of Saberr’s analysis.
As part of the reorganizing, “we put together people who were motivated by the same thing,” Mr. Shepherd says. “Are they motivated by safety and security, or are they trying to take risks? Are they seeking power?” Mr. Shepherd says there were no right or wrong answers, but it was most important that members of the group were motivated by the same values. “If you have different goals, you could be pulling in opposite directions,” he says.
Saberr has been testing a new artificial-intelligence-based service in which employees exchange messages with a chatbot, a computer program that acts like an executive coach. The chatbot asks questions about the effectiveness of office meetings and about whether employees feel that they have well-defined roles and responsibilities. Employees can choose to answer anonymously or not, and Saberr sends the feedback to the entire work team, along with suggestions on how to improve productivity.
The nonprofit Human Dynamics Lab at the Massachusetts Institute of Technology also uses surveys to get a picture of employee traits at its two dozen or so clients. Alex Pentland, who runs the lab, says his group then supplements survey data by measuring things such as how often workers communicate by email and phone.
The group also looks at how people interact in person by using a wearable device like a name badge, which can log when an employee speaks, as well as tone of voice and body language. (It doesn’t record what people actually say, though.) Every week, the lab analyzes the data with artificial intelligence and produces a chart that shows who’s talking to whom and how they’re talking with each other.
Simply seeing the data encourages employees to adapt their behavior, such as trying to boost engagement among the more silent members. “Do the sales people actually talk to the design team?” Mr. Pentland says. “If they don’t, then you’ve got a problem.”
In studying workplace teams that have tangible markers of success, such as sales teams, researchers have found that the most successful groups share some characteristics, Mr. Pentland says. They are:
1) Each member of the team is engaged. That means everyone talks and listens in roughly equal measure and that they talk with everyone, not just a manager.
2) There are a diversity of ideas, and everyone is willing to consider new ideas. This can be measured through surveys, as well as the tone of voice people use.
3) Everyone is setting goals for a project.
“You need everyone exploring slightly different things, but going in the same direction,” Mr. Pentland says.
Meanwhile, the use of tracking devices makes some potential clients hesitate. Some clients try the name-badge devices for only a week, hoping they can retain what they learned from five days of analysis. “People don’t always like this,” Mr. Pentland says. “They feel like they’re being spied on sometimes.”
**This article originally appeared in the Wall Street Journal written by Stu Woo.**