Write the pilot promise in plain language
State what you will deliver, to whom, for how long and what is outside the test. Explain any manual steps, limitations or unfinished functionality that affect participants. Avoid advertising a feature as available if you are still exploring whether it can work.
Choose participants who match the situation you want to learn about. A public launch to everyone can create more support work than a small team can handle. Set a capacity you can realistically serve.
Check the whole path, including failure
- Can someone understand the offer without a founder explaining it?
- Can they start, complete and leave the pilot?
- Do they know how to report a problem?
- Is a person responsible for monitoring problems?
- Are limitations and any costs explained before participation?
- Have you checked how corrections, deletion requests and consent will be handled?
Use a test account and sample information to walk through the experience. Check a phone as well as a computer. This checklist does not certify security, accessibility or regulatory compliance; get the relevant expertise where your product needs it.
Collect evidence without turning the pilot into a survey marathon
Choose a small number of observations tied to the pilot’s purpose: where people stop, whether they complete the core task, which help they need, and what happens after completion. Ask about the specific experience rather than whether they “love the idea.”
Keep operational problems separate from questions about demand. A broken confirmation message can stop a user who otherwise wants the service. Conversely, a smooth signup alone does not show that the product solves a recurring problem.
Close the loop with participants and your team
Tell participants when the pilot ends and what happens to their access or information. Review promised follow-ups and unresolved problems. Do not quietly leave a test running without an owner.
In the team review, record what happened, what remains uncertain and whether to continue, narrow, change or stop. Avoid presenting a tiny pilot as proof of broad traction. Use the weekly review worksheet to make the next decision explicit.