Small experiment planner

How to use this worksheet

1. Choose a question that could change your next decision. Avoid trying to prove that your preferred idea is right.
2. Design the smallest useful test. Set a time window, a resource limit and a clear owner. Explain the activity to participants and get their agreement.
3. Decide what evidence would support or challenge the idea before starting. Record behavior and feedback rather than relying only on impressions.
4. Review what happened. Separate observations from explanations, then choose whether to adapt, repeat, expand or stop.

A worked example

Illustrative scenario — not a client result.

Imagine a small team wondering whether a shared weekly planning session would improve coordination. Rather than changing every meeting at once, it tries one limited practice and reviews the experience.

What do you want to learn?

Would a short planning session help participants identify dependencies before they become urgent?

What is the smallest useful test?

Run one 20-minute session each week for two weeks with the team members who agree to take part. Keep other working practices stable during the test.

Who will take part?

One coordinator owns the test. Participants receive the purpose and format in advance and can raise concerns before agreeing.

What are the limits?

Limit preparation to ten minutes per person and use existing tools. Pause if the meeting adds work without helping participants make decisions.

What will you observe?

Record dependencies identified and actions agreed. Ask participants which decisions became clearer, which remained unclear and whether the time was useful. Do not assume positive comments alone demonstrate better outcomes.

When will you decide what comes next?

Review after the second session. Compare the notes, capture surprises and decide whether to change the format, repeat the test or stop.

What to watch for

A small test produces limited evidence. A useful result for one group does not establish that the same approach works everywhere, and a disappointing result can still reveal what to change.