Large teams of scientific researchers have long worked together to tackle complex projects. But in a surprising new study published in the journal Nature, researchers found that small teams actually do more innovative work than large teams do.
In the study, the researchers analyzed 65 million papers, patents, and tech products from 1954 until 2014, using three main databases as sources: the Web of Science, the US Patent and Trademark Office, and GitHub. Working backward, they found that the smaller teams of scientists consistently found completely new ideas, while larger teams more often added information to existing ideas.
“Large consortiums are still important drivers of progress,” writes Benedict Carey, reporting on the findings in the New York Times, “But they are best suited to confirming or consolidating novel findings, rather than generating them.” Carey points out that while there is room for both large teams and small teams in science, the two sizes serve different purposes, and for more innovative research, the study proves small teams are the better bet.
Though the researchers focused on scientific work in particular, the same idea can apply to other fields as well — and some of the most successful companies function best this way. Google, for instance, runs on the small-team model, as the company separates their employees into tightly focused teams who are focused on specific projects on a micro level. At Amazon’s headquarters, CEO Jeff Bezos enforces a “two-pizza” rule, where employees are discouraged from holding meetings that would require more than two pizzas to feed the entire team, in an effort to improve productivity and keep people from having to participate in an endless series of meetings that aren’t immediately relevant to their work.
“This [study] is the frontier of open questions that show how this finding can apply to work settings, business, and start-ups,” James Evans, Ph.D., one of the study’s researchers from the University of Chicago, tells Thrive. Evans says by studying the structures of teams in different settings, we can find common threads that point to similar results. “Many things we observe outside science and technology line up with these findings,” he says. “There are other correlates of small teams that may be associated with risk taking and disruptive success.”
When it comes to finding the magic number for team size, Evans says each workplace has to find the amount that works best for them, but confirms that small ones do seem to be best for innovation (and he tells the Times that a group of three would be likely to produce less disruptive work than a team of two, and so forth, with that relationship holding steady no matter the size of the teams being compared). That said, it may be important to have varied team sizes within your organization, since not every group needs to be coming up with deeply innovative concepts. “Because small and large play such distinctive roles in science and technology, the system is much better off with a diversity of small and large team sizes rather than the whole population somewhere in the middle,” Evans explains. So when in doubt, variety is key. “An ecology of small and large teams does better than a population with just one,” Evans adds. “It’s a multi-equilibrium solution.”
