How to design teams to improve almost any decision-making process

Not all teams are the same.
Not all teams are the same.
Image: Reuters/Vivek Prakash
We may earn a commission from links on this page.

By now, the science is well established: groups of diverse individuals make better decisions. But the workplace statistics are painfully clear: diversifying can be a slow process.

While we wait, there are other ways to engineer some of the benefits of diversity, by manipulating the structure and processes of workplace teams.

Mirta Galesic, a psychologist and professor of human social dynamics, says team structures can almost always be optimized to produce better decisions. At an event on the science of diversity organized by the Santa Fe Institute in March, she explained that while some conditions are fixed—the size of a corporate board can’t be manipulated from one meeting to the next, for example—others can be tinkered with, like how the team makes decisions or how frequently it shares information. Like the sliders on a stereo equalizer, the variables can he adjusted up and down, depending on the needs of the team, and the nature of the decisions it’s trying to make.

If you’re trying to optimize the output of your team, here are some important factors to keep in mind.

Team size

In a famous 1906 experiment at a British county fair, statistician Francis Galton surveyed the nearly 800 entrants in a contest to guess the weight of an ox. While many were wildly off, their average guess of 1,197 lbs. was almost identical to the ox’s true weight of 1,198 lbs. The statistical lesson is clear: for some problems, the more inputs, the better.

If your group is trying to answer a relatively simple question that has one definitive answer, having a large number of people around the table can be helpful. But when teams are asked to make more complicated decisions, simply adding more members won’t help, Galesic says.

To figure out the optimal size of a team when trying to tackle a mix of simple and complex questions, it’s important to know the expertise of the people in the room. If they are very good on most tasks, but occasionally make errors, then—statistically—the right number of people is often between three and fifteen, she says. After that, the returns will diminish.

“The accuracy of the group’s majority will increase quickly on easy tasks and decrease slowly on difficult tasks,” Galesic says. After a certain point “the larger the group size, the accuracy can only fall. You’ll reach a peak at some point and there’s no more return to adding experts.”

What if you don’t have an expert group? Does adding more people to the team help? In a word, no. According to Condorcet’s jury theorem, arrived at by the Marquis de Condorcet in 1785, asking multiple people who are likely to be wrong doesn’t improve the odds of them being right. In those scenarios, it’s better to reduce your decision makers to one person, who may stumble into the right answer.

Team rules

What’s the best way for a group to finalize its decision? By majority rule? Simple plurality rule? By relying on the most senior or expert member? As with picking the optimal size of the team, it helps to know what kind of problem you’re solving, Galesic says.

If it’s a group of experts (that is, a group where the members are, on average, going to be right more than half the time), majority rule is best. If most people are likely to be wrong, though, it’s better to rely on whomever is most likely to get it right.

Choosing the best decision makers also might be shaped by how information is shared in a group. Does the team meet just once a month, like at a board meeting, or are they in continual contact on a messaging platform like Slack? If they’re talking constantly, they may arrive at an answer quickly, but their answer may be inferior to one that could be arrived at with more time.

If a team is in constant communication, Gelasic recommends slowing down the decision-making process and adopting a majority rule, which forces more deliberation and discussion. If the team meets infrequently, then it makes sense to leave the decision-making to one person, because the need to win a majority could slow the process to a crawl.

“There is a sweet spot between the speed of the network and the speed of the decision rule,” she said.

The worst system combines a slow decision-making process with the slow transmission of information—a body that meets rarely, and reaches all decisions by majority vote. Consider, for example, a university’s faculty senate.

Team pressure

The amount of pressure felt by a team also can shape its decision making. Fear can be a powerful motivator and in the face of external threats, smaller differences fall by the wayside and groups can cohere around a central idea.

If a team has simple tasks that need to be accomplished quickly, threats can help speed the process. The threat can be literal (docking pay or suspending privileges) or arbitrary, like a deadline. For tasks where creativity isn’t required, diversity of thought can actually be a detriment, while external pressure can help harness the ruthless qualities of homogeneity.

But with more complex tasks, outside pressure can backfire, Galesic says. Teams under threat are more likely to fall victim to “groupthink,” a term invented by social psychologist Irving Janis to describe what happens when individuals succumb to the pressure of a group. Groupthink can flourish when team members share the same background, when they don’t have access to outside opinions, and when there aren’t clear rules for making decisions. Some of history’s most famous examples of flawed groupthinking—the US’s blundering into Vietnam, the rush to declare war in Iraq—came from groups under real or perceived threat.

Removing the pressure of threats like deadlines can give teams room to breathe, and perhaps come to a better solution than they would have under the gun.