Research: When Small Teams Are Better Than Big Ones


Here is an excerpt from an article written by Dashun Wang and James A. Evans for Harvard Business Review and the HBR Blog Network. To read the complete article, check out the wealth of free resources, obtain subscription information, and receive HBR email alerts, please click here.

Michael Blann/Getty Images

*     *     *

The “discovery of the 21st century” was awarded the Nobel prize two years after it was made. In 2015, the Laser Interferometre Gravitational-Wave Observatory (LIGO) project finally detected what it was built to: gravitational waves—ripples in the fabric of space and time—caused by the collision of two black holes.

Among many things, the LIGO experiment is a testament to the power of teams in tackling the 21st century’s toughest challenges. Indeed, one of the most universal shifts in the innovation sector in recent years has been the growth of large teams in all areas of research and development, while solitary inventors, researchers, and small teams have all been on the decline.

This fundamental shift is critical for science and innovation policy, as it points to large teams as optimal engines for tomorrow’s largest advances.

But do large and small teams differ by type of innovation? That’s what we set out to test with our postdoctoral researcher Lingfei Wu. We examined millions of papers, patents, and software projects, summarizing our insights in a paper published in Nature. In short, we found that while large teams do indeed advance and develop science, small teams are critical for disrupting it—a finding with broad implications for science and innovation.

The Dominance of Large Teams

Decades of research on teams and collaborations has documented the growing dominance of larger teams over individuals and small teams in research, development, and creative tasks ranging from scientific studies to Broadway musicals.

Indeed, high-impact discoveries and inventions today rarely emerge from a solo scientist, but rather from complex networks of innovators working together in larger, more diverse, increasingly complex teams. This trend reflects an important conclusion that has become a simple prescription: when it comes to teaming, bigger is better.

Part of the reason that we need large teams is that some achievements simply aren’t feasible for smaller groups to pull off. For example, the Nobel-winning LIGO experiment involved firing two laser beams between two 4-kilometer tunnels housed in an ultrahigh vacuum, to detect a variation about a thousandth the diameter of a proton. It had by far the highest price tag of any project ever funded by the National Science Foundation. Hence, it is no surprise that the paper reporting the discovery listed more than 1,000 researchers.

Yet there are reasons to believe that larger teams are not optimized for discovery or invention. For example, large teams are more likely to have coordination and communication issues—getting everyone on board for an unconventional hypothesis or method, or changing direction to follow a new lead, will prove challenging. Large teams can also be risk-averse, as they demand an ongoing stream of success to “pay the bills.” As such, large teams—like large business organizations—tend to focus on sure bets with more established markets. By contrast, small teams—like small ventures–with more to gain and less to lose, are more likely to undertake new, untested opportunities.

This leads us to ask whether the narrative of relying only on large teams might be incomplete. Our research suggests that team size fundamentally dictates the nature of work a team is capable of producing, and smaller team size confers certain critical benefits that large teams don’t enjoy.

Large Teams Develop, Small Teams Disrupt

To examine the effects of team size, we analyzed over 65 million papers, patents, and software products that came out between 1954 and 2014.

We compared the work of large teams to that of smaller groups, with a “small” team defined as one having three or fewer members. We measured the disruptiveness of a work, using an established measure of disruption that assesses how much a given work destabilizes its field. This told us how the research eclipsed or made us rethink the prior “state of the art,” setting a valuable new direction for others to follow.

Our analyses uncovered a nearly universal pattern: whereas large teams tended to develop and further existing ideas and designs, their smaller counterparts tended to disrupt current ways of thinking with new ideas, inventions, and opportunities.

In other words, large teams excel at solving problems, but it is small teams that are more likely to come up with new problems for their more sizable counterparts to solve. Work by large teams tends to build on more recent, popular ideas, while small teams reach further into the past, finding inspiration in more obscure prior ideas and possibilities. Large teams, like large movie studios, more likely generate sequels than new narratives. We found that as team size grows from 1 to 50 members, the associated level of disruption drops precipitously.

Our results appeared remarkably robust against many tests and alternate explanations. For example, one could argue that certain types of people are more likely to work for smaller or larger teams, thus changing the outcomes associated with each. But when we compared the work of the same individual on a small team versus a large team, we found systematic differences in line with our results. We also found that team differences are not due to the different types of topics that large and small teams tend to study. This suggests it’s about team size rather than the efficient sorting of people and problems.

*     *     *

Here is a direct link to the complete article.

Dashun Wang is an Associate Professor of Management and Organizations at the Kellogg School of Management at Northwestern University and a core faculty at NICO, the Northwestern Institute on Complex Systems.

James A. Evans is a Professor of Sociology, Director of Knowledge Lab, and Founding Director of the Computational Social Science Program at the University of Chicago; and an external Professor at the Santa Fe Institute.


Posted in

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.