Continuous Improvement Experiment Template

A repeatable, interactive experiment worksheet (hypothesis, plan, measure, learn) frontline teams can use to run short, measurable tests, capture results, and save experiments for follow-up and sustainment.

Interactive Tool

Continuous Improvement Experiment

Run a short, measurable experiment that your team can own

Use this worksheet to turn an idea into a focused test: state the problem clearly, form a testable hypothesis, specify how you will measure impact, run the experiment for a defined period, and decide whether to adopt, adjust, or abandon the change. Save the completed form so results are visible to your team and to people who will help sustain successful changes.

Tips: Keep experiments short (1–4 weeks when possible), measure one main metric, include simple safeguards, and define success criteria up front. When in doubt, choose a smaller test you can run fast.

Describe the specific problem, where it occurs, who it affects, and why solving it matters (waste, downtime, quality, safety, cost, customer impact). Keep it 1–3 sentences.
Where the experiment will run (line, cell, shift, product family), who will participate, and any constraints. Helps limit variation.
A testable statement linking the change to the expected effect. Example: 'If we add a visual cue at the feed hopper, then changeover time will drop because operators can align parts faster.'
Describe the main benefit you expect (e.g., reduce cycle time, reduce defects, reduce setup time, increase throughput, improve safety). Be specific.
Choose the primary category for measurement.
Name the metric you'll compare (e.g., changeover minutes, rejects per 1000, uptime %, units per hour).
Current measure using the same units you'll use for results. If not available, describe how you'll estimate or collect it.
Period used to compute the baseline (e.g., last 2 weeks, last 30 shifts).
Define clear, measurable thresholds or rules that determine the decision after the experiment (e.g., >=10% reduction in setup time with no increase in defects).
How long the test will run. Keep it as short as practical while capturing meaningful data.
List the exact steps the team will take during the test. Include who does what and when so the test is repeatable.
List safety checks, quality checks, and what you will do if the experiment creates negative effects. Include a rollback trigger if applicable.
Explain how data will be collected and recorded (manual log, existing system, sample size, frequency). Assign who is responsible for each measurement.
Person accountable for running the test, collecting data, and reporting results.
People who will help run the experiment or be affected by it.
Link to photos, spreadsheets, video, or system records that support your findings. (Use a team-shared drive or attach later.)
Use this field to capture quick observations during the run (problems, unexpected effects, ideas). Update as you go.
Start date of the experiment (YYYY-MM-DD).
Completion date (YYYY-MM-DD). Fill when the experiment ends.
Summarize outcomes and compare measured results to the baseline and success criteria. Be explicit about numbers and units.
Primary measured result using the same units as the baseline.
Units for the measured outcome (e.g., minutes, % uptime, rejects per 1000).
Choose whether to adopt, adjust, or abandon based on the evidence.
If adopting, describe how the change will become day-to-day standard work (training, updates to work instructions, audits, owner, frequency of checks).
List follow-up actions (who does what, by when). Include owners for sustainment if adopted.
What did the team learn about the problem, the approach, and measurement? What would you do differently next time?
You can explore this tool now. Sign in or create an account to save your responses and return to them later.
Make this tool part of your work

Save a personal copy, bring it to your team, or tailor the questions and workflow to fit what you are hungry to improve.

Member customization and team collaboration are coming soon.

Discussion

Comments and conversation will live here.