4 ingredients of successful teams

Steve Kelman is professor of public management at Harvard University’s Kennedy School of Government and former administrator of the Office of Federal Procurement Policy.

Richard Hackman is a professor of social and organizational psychology at Harvard University and one of the leading academic researchers on teams. He is most associated with the view that many discussions of how to get good performance from teams inappropriately focus on processes, such as the interpersonal dynamics on the team, when instead they should focus more on structure, such as team mission and composition or even whether the activity is appropriate for a team in the first place. Also, compared to many scholars who study the topic, he is unusually interested in teams in government contexts.

His latest book, “Collaborative Intelligence: Using Teams to Solve Hard Problems,” focuses on teams in the intelligence community, but 95 percent of what he writes could apply to any team involved with analyzing and/or making decisions based on data.

Much of a team’s fate is written before its first meeting, based on various decisions made about the team’s structure, Hackman said. To avoid problems down the road, he offers several recommendations.

  • Don’t create a team in the first place without a good reason. Teams are best for situations in which the work is helped by interdependent contributions rather than the efforts of smart or creative lone wolves.
  • Specify a compelling, motivating purpose for the team. Hackman said this is a job for the organization’s leaders, not the team’s members. Those leaders should not abdicate goal-setting to the team, but they should give the group autonomy in deciding how to reach the goal.
  • Pay attention to team composition. It is important for those establishing a team to consider the mix of skills and expertise the team needs to do its job. Hackman also said those who establish teams should take the time to find out, based on past performance, whether candidates for the team play well with others. There is also strong evidence that it takes time for teams to develop the ability to work together and know who has expertise in which areas. Indeed, Hackman cites the amazing statistic that 73 percent of airline “incidents” noted by the National Transportation Safety Board took place on a crew’s first day of flying together. Furthermore, there are significant costs to bringing new members up to speed on the activities of an established group — one of the reasons for the old adage that adding staff late in a software development process drags the project further behind schedule.
  • Focus on promoting productive, task-oriented information sharing and deliberation. Feel-good “team-building” exercises have generally been shown to have little impact on performance.

Hackman said intelligence community officials are increasingly interested in various kinds of prediction markets and crowdsourcing as alternatives to teams. Like teams, those approaches make use of information that comes from a variety of sources. But unlike teams, they require no deliberation or discussion of different opinions, only an averaging to get the crowd’s overall prediction. And they create no opportunities for peer coaching, which can improve the performance of individuals on a team.

If teams don’t function well, crowdsourcing is likely better than teams. But Hackman clearly believes that in many cases, we are better off aiming for well-functioning teams so that we can get a greater diversity of opinion rather than just an average and so that we can capitalize on the benefits of interpersonal learning.

About the Author

Kelman is professor of public management at Harvard University’s Kennedy School of Government and former administrator of the Office of Federal Procurement Policy. Connect with him on Twitter: @kelmansteve


  • Comment
    Pilot Class. The author and Barbie Flowers are first row third and second from right, respectively.

    How VA is disrupting tech delivery

    A former Digital Service specialist at the Department of Veterans Affairs explains efforts to transition government from a legacy "project" approach to a more user-centered "product" method.

  • Cloud
    cloud migration

    DHS cloud push comes with complications

    A pressing data center closure schedule and an ensuing scramble to move applications means that some Homeland Security components might need more than one hop to get to the cloud.

  • Comment
    Blue Signage and logo of the U.S. Department of Veterans Affairs

    Doing digital differently at VA

    The Department of Veterans Affairs CIO explains why digital transformation is not optional.

Stay Connected


Sign up for our newsletter.

I agree to this site's Privacy Policy.