The Experiment Canvas, created by Chris Stone, is a practice and artifact that helps teams design focused experiments with clear guardrails around scope, involvement, timeframe, and success criteria. It's a great tool to use when a team has:

a) a problem they want to solve

b) an idea they want to implement

If you get the sense the team doesn't have quite enough of a firm grasp of the problem, you might try running something like an Abstraction Ladder practice to help the team explore and map out the problem space.

When a team fills out an experiment canvas together, it forces conversations that require the team to make decisions about specific areas that make up a well-designed experiment.

By the end, all teams members have a clear sense of what's being done, thereby improving alignment, clarity, and speed while instilling a culture of ideation, collaboration, and experimentation as the vehicle to continuous improvement 💡 👩‍🔬 📈

Prepare a blank Experiment Canvas on a wall or in your digital whiteboarding tool of choice (Miro, Mural, Jamboard, etc).

With the team, go through each section of the canvas 1 by 1 and fill in the required information, one idea per stickie.

You can use the following recommended order to fill out the canvas, but it's not required:

1. The Problem

What problem are we trying to solve?

Why do we need to experiment with something new?

Tips:

• If the team already has a problem statement, insert it here. If they have multiple, have the team list all problem statements, then do a quick voting session to prioritize them and focus on the top-voted problem.
• If the team is struggling to articulate the problem, it might be helpful to run an Abstraction Ladder activity to help them map out the full context of the problem space first. If you choose to do this, take the top problem the team votes on at the end of the Abstraction Ladder and insert it into this portion of the Experiment Canvas.

2. The People

Who will be affected by our experiment? (Who experiences our problem statement?)

Who do we need to engage and involve?

Tips:

• Ask the question to the team and just have a group conversation. Capture ideas mentioned, 1 per stickie, then when ideas have stopped, playback what's captured with the team.
• Don't focus on the core team members yet - that comes in "The Scientists" section. This section should focus more on beneficiaries and people you may have to make aware that this experiment is happening or even occasionally collaborate with while the experiment runs....but the people in this section shouldn't be assigned to this experiment full-time. That's who The Scientists are.....we'll get to that section later!

3. The Outcome

What are our desired outcomes?

What are we hoping to achieve through the experiment?

Tips:

• Another group conversation.
• Aim for 3-5 outcomes. If there are more, have the team vote on the top 3-5 to bring focus.

4. The Hypothesis

What's our hypothesis?

Tips:

• This section will likely take the longest, but it's very important!

• The facilitator will lead the discussion, creating stickies, then opening up to the team for discussion. With smaller groups, you can have the team start ideating as you capture ideas. For larger groups, or groups getting a bit stuck, you can facilitate this by nominating options and using your idea(s) as a starting point for the group to discuss.

• IF:

  • reword the problem statement as if it were accomplished

• BY:

  • explain the solution (if the team doesn't have a clear solution in mind, you might consider running a "How Might We..." practice to generate and vote on ideas)

• WILL:

  • the main outcomes the solution will achieve if successful (look to "The Outcome" section of the canvas for this)

• BECAUSE:

  • a belief statement about why you think the solution will achieve the outcome. It helps to start this stickie with "We believe..."

5. The Results

How will we know our hypothesis is true?

How will we know our hypothesis is false?

• Another group conversation. Try to get the team to identify specific metric goals to hit that would indicate pass or fail (specific numbers are great to use, here...like "a 25% increase in conversion")

6. The Litmus Test

What metrics, data, or KPIs do we need to measure?

• Another group conversation.
• It's helpful to ask "what metrics do we need to capture in order to determine what we agreed on in "The Results" section?

7. The Trials

When will we measure results?

What are the next steps to bring this experiment to life?

• Another group conversation.
• Be sure the group identifies a specific timeframe to run the experiment.
• Assign names to action items to ensure someone is owning each one.

8. The Scientists

Who is carrying out the experiment?

What are our roles & responsibilities?

• Another group conversation.
• Ask, "Now that we know what we want to do and the time we have, who should be part of this core team driving this experiment?"
• This is different than "The People" section because "The Scientists" are going to have the most time dedicated to running the experiment

Next Steps

• Have the owners of each "next step" carry out their step to get the experiment ready.

• Once prepared, begin the experiment and schedule a follow-up meeting as soon after the end date as possible.

• Facilitate a team retrospective during the follow-up meeting to view the data collected and compare it to what was identified in "The Results"

• Ask the team what worked and what didn't work well

• Review all information and make a team determination to...

  1. Success = the experiment achieved the target outcomes. You can either move onto another experiment, or scale the solution to other teams, if applicable.
  2. Partial success = the experiment didn't quite achieve the results the team wanted, but we all feel it's on the right track. In this case, ask how the experiment might be improved in order to achieve the results, update the canvas with these changes, and determine a new timeframe in "The Trials" to run a second iteration of the experiment.
  3. Failure = the experiment was way off on achieving the target outcomes, and the team feels it's best to try a different solution instead. In this case, start a new canvas with the new solution and run another experiment.