Studies in collective intelligence have played an important part in our initiative to make teams work. Collective intelligence (CI) is what makes groups better or worse than the sum of the individuals in the group.

It turns out that the IQs of the people in the team only weakly correlate to the team's CI. What makes far more difference is the level of empathy in the group, how much communication there is, and how distributed that communication is.

Bees have brains the size of a grain of sand, yet they can collectively make complex decisions like finding the ideal location for a new hive.

For example, bees have brains the size of a grain of sand, yet they can collectively make complex decisions like finding the ideal location for a new hive. The secret lies in how they work together. The hive is far more intelligent than sum of the individuals.

Teams that have little communication or where communication goes through one or two people have lower CI than teams where everyone communicates freely with everyone else.

Airline flight crews know this well. They pay careful attention to the science of how groups work most effectively together, for obvious reasons. A crucial aspect of flight crews is Crew Resource Management (CRM). Good CRM has everyone knowing what their task is and focusing on that but at the same time there is constant communication. No one is "the boss" or "the expert". Everyone speaks up and contributes, paying attention to their own possible blind spots.

Poor CRM, where the captain acts as the boss, where the others are intimidated or afraid to speak up, or where the captain constantly criticises the less senior members, has actually crashed passenger planes all on its own, even when there was nothing wrong with the plane. Conversely, when flight crews have saved a plane that seemed impossible to save, it comes down to excellent CRM.

We discovered, by accident, that having the team swarm on a feature amplifies the performance of the team dramatically. When my team started we were very small and had to hire in half the team. To make sure that we at least got the most important features done while we filled out the team, we decided to have the whole team swarm on the most important feature. We were surprised at how fast that went and at how few bugs came up in QA. So we continued doing that.

Two quarters later another team tried it as well. They were expecting to have around 300 bugs come up in QA. They flew through the feature and had only around 30 bugs in QA. The difference was so dramatic that they also adopted swarming.

Apart from a buzz of task-related communication, swarming also means that tasks are highly interdependent and that there is high task visibility. These are the two main factors that reduce social loafing and free-loading, which helps explain why new team members are surprised at how motivated they feel.

We also want to be able to measure this. What exactly is going on? Why does it work?

Answering these dynamics is what I and some other leads have been trying to figure out. Using psychometric principles, we've developed an Impact 360 to measure the real-world impact of each person on the team. This tells us what is working, why, and how we can get better.