Measure and decide

You now have a customer hypothesis, evidence to investigate, a route and a bounded test. Connect them to records you can trust.

Define each stage: relevant visit, evaluation started, useful result achieved, and the commercial next step if it applies. Decide whether you are counting people, accounts or events. Write where the records come from and how long you will observe them.

For the illustrative tutorial test, distinguish a repository visit, installation, first inspected delivery, repeat use and a team evaluation. A download or GitHub star does not establish that a developer ran the tool. Define the developer and account units separately, and record the buying step if team approval is required.

Complete the measurement plan. Schedule the review now. At that review, choose whether to repeat the test, change one part, stop, or collect more evidence.

If you have opted into the weekly email, we will keep sending practical studies, examples and tools. Otherwise, this completes the course.

Reply with the decision your test is designed to inform.

Jon