Rapidly prototyping behaviour


One of the most useful ideas from the agile computing world is the idea of rapid prototyping. It's well described here: rapid prototyping in an agile age.

The essential idea is illustrated by Keith Sawyer who tells the story of two teams. The first team is asked to work on producing one really excellent vase. The other team is just asked to make as many vases as possible. The second team, by producing lots of vases, gradually gets more skilled, so by the end of their time they are able to produce quite good work. The team aiming for excellence, however, ends up producing one not very impressive vase.

In our work, we've taken this idea, and mixed with activities from improv theatre to help people rapidly prototype behaviour.  We call it action storming.

The central idea is to explore relationship dilemmas - tricky conversations, challenging customers etc - through improvisation. 

We ask for real world examples of situations that participants have struggled with, and choose one. The person who suggests it, we’ll call them the client, now gets to re-enact it, playing themselves. Fellow participants play the other parts in the scene.

We normally create a scene with a very few lines of dialogue. We aim to find the parts that really fire the hot buttons of the client, who starts by playing him/herself in the scene.

After this, we replay the same short scene again, and ask the client to try something different in their response. We usually encourage them to try something crazy that would never work, as a way of loosening up the creative synapses.

We’ll play the same scene again and again, each time asking the client to try something else. There may be a little debrief between scenes but not too much. We are trying to downplay the role of analysis in favour of plenty of experimentation.

Over time, we experience moments that seem to change the dynamics in some way. Many of these are not really solutions - but they could be part of a solution. There are no guarantees but eventually someone will assemble some of the fragments into what feels like a breakthrough response.


This approach seeks to avoid the temptation of teaching five rules for dealing with X. It values discovery and surprise over the comfort of predicting what people will learn. 

As we are fond of saying, the clue about those difficult relationships is in the name. They are difficult and require us to try new things.

Action storming allows people to try ideas out in an environment where it is safe to fail and learn. 

Here's Viv's slideshare on action storming

(Photo by Sidney Pearce on Unsplash)