If you want to move your company toward one that is more effective, innovative and nimble, then you want to move your company culture toward one that truly values learning and growth.

As a leader, one way you can do that is to encourage people around you to treat more endeavors as efforts to learn, rather than mere efforts to execute on the current assumptions. Help people get curious. Help people to design experiments–with the rigor to ensure those experiments yield learnings and lead to improvements.

Add the Discipline of the Experiment Canvas

In my recent talk on Culture Change at the Agile Virtual Summit, I showed an Experiment Canvas (click to download). This version is a mashup of the various experiment canvasses I’ve encountered over time.

The left side is all about setting up the experiment.

  • Desired Learning/Background – How did we get here? What is the current state? What’s the bigger picture? What do we hope to learn? This first section is all about getting seriously into a learning mindset. We get to articulate curiosity.
  • Falsifiable Hypothesis – State our starting theory. What do we *think* might be true? How would we know it’s not?
  • Experiment Details – What actions, by whom, over what period of time? Get specific. The goal is to design something that will help us prove or disprove as fast as possible. What qualitative and quantitative data will we collect? How often will we retrospect? How will we know we are done? Done is important — companies grow weary of never-ended experiments.

The right side is all about what happens as we run and complete our experiment.

  • Safety & Recovery – What about your experiment method protects you from harmful side effects? This piece is important; by thinking about it in advance you (a) set yourself up to respond quickly, (b) improve your experiment design and (c) improve your ability to get support for the experiment.
  • Experiment Results – Capture the results. And include the analysis of how they compare to expected results.
  • Learnings & Next Steps – Too many experiments just continue or fade into the sunset. Here is where you practice the discipline to ensure you learn, communicate the learnings, and do something with the learning. For a proven hypothesis, do we now create a new standard, or pursue the new product direction? For a disproven hypothesis, do we stop, or develop a new experiment?

Try using this canvas to design your next experiment. Maybe you want to:

  • Try a process improvement
  • Test a design approach
  • Verify a feature direction

Then let me know how it goes!

In a future post I’ll provide some examples of experiments.