I was speaking at a conference yesterday when someone asked me about consensus. If a team is making an important decision, they asked, surely getting everyone to agree is a good thing?
It can be, with some big caveats. The problem is we often confuse two different jobs a group has to do: figuring out what is true, and deciding what to do about it. Consensus is useful for the latte, but is surprisingly unreliable for the first part of that process.
We’re a deeply social species, and much of what we know comes from other people. We don’t independently verify whether every food is safe or whether every social convention makes sense. We learn by watching the people around us.
Models of cultural evolution suggest that this tendency can be adaptive. When individual learning is costly, what researchers call “conformist transmission” (adopting the behavior of the majority) makes sense as it points us toward behaviors that have already proved successful.
There’s also a coordination benefit. If ten people are trying to cross a river, it matters less whether they have identified the theoretically optimal crossing point, than whether they agree to cross in the same place.
But this mechanism that’s useful for basic social learning and coordination can become dangerous when we use it to answer more complex questions, because the desire to agree can narrow the search too early.
Research comparing decision-making methods found that groups using devil’s advocacy or dialectical inquiry made better decisions than groups told to seek consensus. And in one study, groups encouraged to challenge one another actually ended up with greater consensus and acceptance of the final decision than groups pushed toward agreement from the start.
The problem isn’t seeking consensus itself, it’s seeking consensus too early.
So what should you do if you’re faced with a complex question, or navigating a project with lots of uncertainties? Instead of seeking consensus, aim to foster collective curiosity.
Instead of asking, “Can we agree on this?”, the group asks, “What don’t we know yet?” Switching from consensus to curiosity changes our relationship with uncertainty, especially when it’s paired with experimentation.
If two people suggest competing solutions, you don’t need another hour of debate. Ask what evidence would help decide between them, then design a tiny experiment to gather that evidence. Show an unfinished prototype to five customers. Try a new process with three teams. Change one part of the onboarding experience for one week.
This is the distinction I find useful:
- Collective curiosity increases the amount of information a group can use;
- Experimentation increases the quality of the evidence the group can use; then only…
- Consensus increases the group’s ability to coordinate action based on that evidence.
All three are important, but we shouldn’t use them in the wrong order, and consensus should never be a substitute for the messy process of experimenting to figure out our next steps.