Moderated User Testing

Testing with 5 to 8 participants per user segment is often enough to uncover most major issues.

• Include every segment. Invite participants from each user segment you identified, so you hear from the full range of people the project serves.
• Send invitations from the service owner. People are more likely to respond when the invitation comes from the department or service they already know.
• Working with the ITS UX team? The service owner or department builds the participant list and handles recruitment. The UX team provides invitation templates and scheduling links.

Send invitations with a scheduling link at least one week before sessions begin. Two weeks or more is better, especially during busy times of the semester.

• Plan time for follow-up. Send a reminder a few days after the first invitation to people who haven’t responded.
• Have a backup list ready. Not everyone will sign up. Identify extra participants from each user segment ahead of time so you can invite more people quickly if spots go unfilled.
• Confirm the day before. A short reminder before each session helps reduce no-shows.

We want participants to feel free to share what they really think, so we keep their identities private. Names and individual sessions aren’t shared with stakeholders.

That’s because user testing isn’t about what any one person said or did. It’s about finding the common themes, the needs and challenges that come up again and again across everyone’s experiences. Findings are shared as these overall patterns, so no single person’s feedback stands out.

A script keeps every session consistent, so results can be compared across participants. Use the same script for every session, and keep each session to about 30 minutes.

Suggested structure (30 minutes)

• Introduction (2–3 minutes): Welcome the participant, explain the purpose, and remind them you are testing the site, not them.
• Warm-up questions (about 5 minutes): Get to know the participant and note anything that could add context to their results, such as:
  
  • their role and how long they’ve been in it
  • how often they use the site or service, and how comfortable they are with it
  • any unique needs or ways they work
• Tasks (15–20 minutes): Three to six realistic tasks based on your experience priorities.
• Wrap-up (3–5 minutes): Ask open questions, such as what was easiest, what was hardest, and what they would change.

Write tasks that don’t lead

Describe a situation or goal, not the steps to get there. Avoid using the site’s labels, menu names, or button text in your tasks, because they point participants to the answer. Keep tasks informative but open on purpose, so you can see how people really approach them.

• Leading: “Click Request Hardware to order a new laptop.”
• Better: “Your laptop is getting old and you’d like a new one. Show me how you would go about getting one.”

Before Your First Session

• Get consent to record. Ask participants for consent before recording, and explain how recordings will be used and stored.
• Run a practice session. Test the script with a colleague to check timing, wording, and technology.
• Limit hosts and observers. Keep the number small, ideally one facilitator and one note-taker, so participants feel at ease. Observers should keep cameras and microphones off.

Stay focused, but stay curious

Keep the session on schedule and gently guide participants back if they drift too far from the tasks. At the same time, stay open to the unexpected. When a participant does something surprising, a short follow-up such as “What made you choose that?” can uncover issues you didn’t know to look for.

• Put participants at ease. Explain that you are testing the site, not them, and that there are no wrong answers. 
• Ask them to think aloud. Encourage participants to say what they are looking for, expecting, and thinking as they work. 
• Stay neutral and let users struggle. Take notes on behavior, not just feedback. Watch for body language and hesitation. 
• Close with open questions. Ask what was easiest, what was hardest, and what they would change.

• Write up your notes the same day. Details fade quickly. Fill in gaps while the session is fresh, and note anything surprising.
• Hold a quick debrief. Spend 5–10 minutes with the note-taker and any observers. What stood out? Where did the participant struggle? Did anyone notice something others missed?
• Send a thank-you. A short message within a day shows appreciation and leaves the door open for future research.
• Store recordings securely. Save recordings and notes in a protected location that only the research team can access.

• Organize notes by task. Use a shared spreadsheet with one row per task and one column per participant, so you can compare results side by side.
• Look for patterns, not one-off comments. Focus on issues that came up for several participants. One person’s struggle may be an exception. Three people struggling in the same place is a finding.
• Compare across user segments. Note where needs are shared and where different groups had different experiences.
• Connect findings to your framework. Group issues under your experience priorities and UX themes so findings tie back to what the research needed to learn.
• Check against other data. Compare what you observed with site analytics and support requests to confirm patterns and understand how widespread an issue may be.

• Rate each issue. Consider how severe it is, how often it happened, and how much it affects users’ ability to complete key tasks.
• Pull quotes and short clips. Real words and moments from participants make findings memorable. Only share clips with participant consent, and never include names.
• Recommend next steps. For each key issue, suggest an improvement and note its likely impact and effort.
• Share a clear summary. Lead with the top findings and recommendations, and tailor the format to your audience.

Report findings as overall patterns, not individual results. Remove names and identifying details from notes, quotes, and clips, and delete recordings once they are no longer needed.