Checklist: How to spot (and stop) pretend improvement

Use this quick checklist to see if your improvement work is real—or mostly cosmetic. For each item, tick Yes if the statement is true for your team.

1. Experiments are time boxed and owned. Yes / No
2. There is a single, simple measure for each experiment. Yes / No
3. Sponsors clear blockers quickly rather than only approving budgets. Yes / No
4. Active experiments are visible to the whole team (board or log). Yes / No
5. Regular rituals are short, predictable, and focused on decisions or learning. Yes / No
6. The team shares failures as learning, not blame. Yes / No
7. There are no more than 2–4 active experiments per team at once. Yes / No
8. Data collection is reliable enough to support the experiment’s conclusion. Yes / No
9. Improvements that work are adopted and briefly documented for others. Yes / No
10. Improvement work is resourced as part of regular schedules, not only as optional overtime. Yes / No

If several answers are No

Prioritize the top three statements that are No, and choose one to address this week—name an owner, pick one metric, and run one short test. Incremental improvements to the process matter as much as individual experiments.