The problem of management in two graphs

Inside every team, there are two wolves...

Managing teams is a race, and these two graphs (from the same paper), show you the two sides: slack & collective intelligence.

The left-hand side shows how, as team size grows (the little numbers 1, 2, 4, 8 indicate team size), per-person productivity drops (the drop in mean team person-hours). People on larger teams work less than people on smaller teams - they start to slack off, allocating their work to less productive tasks and generally working less urgently. And this experiment doesn’t even fully take into account the other problem with larger teams. Increased coordination costs and team-based process loss (all the social reasons why teams have issues) further decrease the value of every added team member.

This is the curse of all large teams, and why Amazon has the two pizza rule, limiting the size of teams to one that can be fed with two pizzas. Larger teams become slower and less efficient, so limiting team size is one solution. And you shouldn’t use teams at all when solving simple problems: "the recommendation would be to ask a group to solve the problem when the problem is complex but to ask independent problem solvers when it is simple."

But I also think it is a bit defeatest to just give up on large teams, and the graph to the right shows why. It suggests, that as team size increases, you can also increase the collective intelligence of the team. Larger teams mean more people with different talents and abilities, and more chances to split up work. It also means Linus’s Law (named after the inventor of Linux) is more likely to be at play: “with enough eyes, all bugs are shallow.” To someone, any problem is trivial, you just have to find the right person. A larger team has more talent to draw on.

Being a good manager is all about increasing collective intelligence faster than slack. And it matters: teams in the top 15% of collective intelligence perform 18% better than if the same team had average collective intelligence. Understanding how to maximize collective intelligence is an emerging area of research. Some early findings:

• Psyschological safety establishes a baseline for team participation. Psychological safety is a research-backed way of making sure everyone on the team feels comfortable taking risks, which means team members will be willing to speak out when they have something to add. At studies at Google it was the key predictor of innovative team success. This is a great guide to how to instill psychological safety on teams.

• The social intelligence of your team matters. Collective intelligence depends on factors beyond just the raw IQ of team members, and is limited by the emotional intelligence of the least sensitive team member. But adding people with high social skills to any team greatly improves team collective intelligence. Team composition is really important!

• But it may be less important than process. Teams that have a good process are more likely to have high collective intelligence. Process lets people coordinate without getting in each others way.

• Turn off your Zoom video? Video chats undermine the collective intelligence of groups, but in a surprising way. It turns out it is the video, and not the chat, that is the problem. Teams with video on during calls end up syncing up less, and have less even turn-taking during meetings.

• Alternate individual and group work. Individuals are more creative working alone But groups can combine ideas to improve them. So alternating between group and solo work can maximize collective intelligence.