One Two Four All - a technique for getting insights from groups

Categories: product-development techniques groupwork

I’ve started working my way through The Surprising Power of Liberating Structures - a hand book of techniques for collaborative work.

So far I’ve tried one technique from the book - one two four all.

The idea is super simple and is an alternative to open brainstorming or post-it note sessions. Before describing the technique with a few comments, I’ll just point out one of the weaknesses of a group work activity like a retrospective.

We have been doing a lot of retrospectives over the last year, and we usually get the team to all write up thoughts against a set of questions Then we do a cluster analysis on the combined post-it notes. Overall this is a great technique, but it has a couple of potential down sides. It doesn’t scale well once you get past about ten people. This technique also gives more weight to the ideas of people who write fast, and can churn out a lot of post-it notes.

“One two four all” is an elegant mechanism that avoids both of those problems.

The way it works is that you pose one or two questions to the group. The group is given one minute where everyone thinks on their own about their answer. Then they pair up in twos and are given two minutes to discuss. Then groups of four are created and are given four minutes to discuss. Finally each group of four reports back on the best idea that they feel they have discussed.

This allows ideas to filter up from people who might otherwise remain quiet, it scales horizontally, and is really easy to implement.

When we ran it a few weeks ago the first minute of silence seems awkward to me. I was doing the exercise with quite a high level stakeholders in the organisation. After that though there was great and engaged conversation and we really came up with a nice diverse set of thoughts on the questions posed. I feel given the room we had and the number of people involved (about 25), that we got a lot more out of this process than if we had tried to run a retrospective.

As I was setting out the process I got a few questions from the group; Should we do it once for one of the questions and again for the second question? Do they need to split directly into fours at the end, some of the tables had five people and it looked like that could work. We decided to have the groups work on both questions at the same time, rather than serially, and I think that worked well. Given this is a technique to enable high quality conversations I didn’t see the need to hold fast to being totally rigorous with the guidelines, so I was OK with some groups of five, and there didn’t seem to be any problem with that.

I’m looking forward to running this exercise again in the future.

posted under a Creative Commons Attribution 4.0 International License