Usability Study Planner

Run a study that gives you real answers

5 topics · 18 checks 0%

Goals

Write the research question State what you need to learn in one sentence. "Can new users complete checkout without help?" beats a vague "test the flow".
Decide what a result changes Know which decision the findings will inform. If nothing changes based on the outcome, the study isn't worth running yet.
Pick the study type Moderated for depth and follow-up questions, unmoderated for speed and scale. Match the method to the question, not to habit.

Define who you need Write a short screener that finds people who match your real users. The wrong participants produce confident, useless data.
Right number of people Around five participants per group surfaces most major usability problems in a qualitative study. Add a segment before you add more of the same.
Include disabled users Recruit people who use assistive technology. Accessibility problems stay invisible if everyone you test navigates the same way.
Sort incentives & consent Agree a fair thank-you and get informed consent to record. Handle this before the session, never during.

Realistic scenarios Give participants a goal and a context, not step-by-step instructions. "Book the cheapest flight next month" lets them find their own path.
Don't lead the witness Avoid using interface words in your tasks. If the task names the button, you're testing reading, not usability.
Define success up front Decide what completion looks like for each task before you watch anyone, so you're not grading on gut feeling afterwards.
Pilot the session Run one dry session end to end. It always exposes a broken link, a confusing task or a timing problem before it costs you a real participant.

Set people at ease Remind them you're testing the product, not them, and that thinking out loud helps. A relaxed participant shows you real behavior.
Stay quiet and observe Resist rescuing people the moment they struggle. The struggle is the finding. Ask "what are you thinking?" instead of pointing.
Capture the session Record with consent, or take timestamped notes with a second observer. You will forget the detail that matters most.

Cluster the observations Group what you saw into themes across participants. One person's stumble is a note; three people's stumble is a pattern.
Separate severity Rank issues by how much they block the task and how many people hit them, so the team fixes the painful problems first.
Share it so it lands Report findings with short clips or quotes tied to a recommendation. Evidence people can see moves decisions faster than a slide of numbers.
Close the loop Turn findings into tickets and decide what to retest. A study that doesn't change the product was just expensive theatre.

Progress auto-saves in this browser — no account, nothing leaves your device.