A usability test script example, from welcome to wrap-up
A usability test script is what the moderator reads and follows in every session, so each participant hears the same words and gets the same tasks. Below is a complete usability test script example for a moderated session on a fictional project-planning app, followed by how to adapt it to yours.
Usability · Published
In short
- A script keeps sessions comparable: the same welcome, the same tasks and the same prompts for every participant.
- Ask for consent to record before you start recording, and say the participant can stop at any time.
- Write each task in three parts: the goal in the participant’s words, where they start, and the sign that they succeeded.
- Read the task aloud and hand it over in writing. Keep the starting point and the sign of success in the moderator’s notes.
- Ask people to think aloud, then stay quiet. When you speak, use neutral prompts that don’t hint at the answer.
- Pilot the script once with a colleague and change it before the first session, not between sessions.
What a usability test script is
A usability test script is the moderator’s guide to one session. It holds what to say at each point (the welcome, the consent request, the warm-up questions, the tasks and the wrap-up) and what to watch for while the participant works. Most of it is read word for word. The prompts between tasks come from a fixed list.
The script matters most in the tasks. A task reworded between sessions measures something else, and a hint given in one session becomes part of its result. How to write usability test tasks and scenarios covers the tasks on their own.
The example below follows the order most published scripts use, such as Steve Krug’s and the US government’s Usability Starter Kit: welcome, consent, warm-up, think-aloud, tasks, wrap-up. Text in brackets is what you replace.
The example script
The product is Northwind, a fictional project-planning app. The participant is a team lead who plans work in spreadsheets today. The session is moderated and one to one, in person or on a video call.
Before the participant arrives
- Open the starting point for task 1: the Northwind pricing page, signed out, in a fresh browser profile.
- Have the staging account ready for tasks 2 and 3, with the “Spring launch” project already in it.
- Print or paste each task on its own card, numbered.
- Check the recording captures the screen and the voice.
- Brief observers: they stay silent and muted, and write down what happened, not what they think it means.
Welcome (read aloud)
Hi [name], thanks for joining. I’m [your name], and I’ll be with you for this session.
Today we’re looking at a website for planning projects with a team. I’ll ask you to do a few things on it, and I’d like to see how you go about them.
We’re reviewing the website, not you. You can’t do anything wrong here. If something is confusing or hard to find, that’s what we need to know, so please tell us.
I didn’t design what you’ll see, so you won’t hurt my feelings, and you won’t please me by being kind about it. Say what you think.
If you have questions as you go, ask them. I may not answer straight away, because I want to see what you’d do if I weren’t here. I’ll answer anything that’s left at the end.
Consent and recording (read aloud)
With your permission, I’d like to record this session. The recording shows your screen and your voice [and your face]. Only the team working on this product will watch it, to see where the website gets in people’s way, and we’ll delete it by [date].
[Number] colleagues are watching from another room [on this call, with their cameras off]. They’re here to learn, not to judge.
You can ask me to stop the recording, skip a task or end the session at any time, and you’ll still receive [the incentive]. Is it all right to start recording?
If they agree, start the recording, then ask again so the recording holds the answer:
I’ve started the recording. Can you confirm you’re happy to be recorded?
If they say no, carry on without recording and take notes.
Warm-up questions
Easy questions about the participant, not the product. They settle the person and tell you how to read what follows.
- What’s your role, and what does a typical week look like?
- How do you plan work with your team today?
- What tools do you use for it?
- Think of the last time a plan changed. What did you have to do to update everyone?
- Have you used Northwind, or seen it before today?
Thinking aloud (read aloud)
While you work, please think out loud: say what you’re looking at, what you’re trying to do and what you expect to happen. It feels odd at first, and that’s normal.
For example, if I were looking for a train time, I might say: “I’m looking for the timetable… I expected it under Travel, but it isn’t there, so I’ll look in the menu.”
If you go quiet, I’ll remind you. Any questions before we start?
Task 1: get started
Read aloud and hand over:
Your team of six wants to plan its work in one place instead of a spreadsheet. Set up Northwind for the team and create your first project, called “Spring launch”. Use the work email alex.rivera@example.com.
Moderator notes:
- Starting point: the pricing page, signed out, no account.
- Sign of success: a project named “Spring launch” is on the dashboard after a reload.
- Not done if: two accounts are created, or the project has another name.
- Watch for: which plan they choose and why, and whether they find where projects are created.
After the task:
- How did that go?
- Was anything different from what you expected?
Task 2: share the project
Read aloud and hand over:
Priya Shah from finance needs to see the Spring launch project, but she shouldn’t be able to change it. Give her access. Her email is priya.shah@example.com.
Moderator notes:
- Starting point: signed in as Alex on the staging account, on the Spring launch project.
- Sign of success: Priya is listed on the project once, with a view-only role.
- Not done if: she’s added with edit rights, or invited twice.
- Watch for: how they decide which role means “see but not change”.
After the task:
- How did that go?
- How sure are you that Priya can’t change anything? What makes you say that?
Task 3: move a date
Read aloud and hand over:
The launch has moved by a week. The press release was due on 3 March. Change it so it’s due on 10 March.
Moderator notes:
- Starting point: signed in as Alex, on the dashboard.
- Sign of success: the press release shows 10 March after a reload, and no other item’s date has changed.
- Not done if: the whole project moved, or the date reverts after a reload.
- Watch for: whether they find the item from the dashboard, and how they choose the date.
After the task:
- How did that go?
- If you had to do that every week, what would you want to be different?
Prompts while the participant works
Use only these, and note in the session notes each time you use one, so a prompt doesn’t pass for something the participant found alone.
When they go quiet:
- “What are you thinking?”
- “What are you looking for?”
- “Keep talking.”
When they ask a question (send it back):
- “What do you think?”
- “What would you do if I weren’t here?”
When they trail off (repeat their last words as a question):
- Participant: “This button is, hmm, not sure…” Moderator: “Not sure…?”
When they say they’re done:
- “Is that what you expected to see?”
When they’re stuck and the task can’t go on:
- “Where would you look next?” Then, if they still can’t go on: “Thanks, that’s helpful. Let’s move to the next one.” Mark the task not done, and note where they stopped.
Don’t say “good”, “right” or “almost”. Each one tells the participant something about the answer.
Neutral follow-up questions
Ask after a task, about what you saw. Each one should be answerable either way.
| Instead of | Ask |
|---|---|
| “Was that easy?” | “How did that go?” |
| “Did you see the Share button?” | “How did you decide where to go next?” |
| “Would the filter have helped?” | “What were you hoping to find on that screen?” |
| “Do you like the new layout?” | “What, if anything, would you change?” |
| “Why did you click the wrong tab?” | “Tell me about what happened on that screen.” |
Wrap-up (read aloud)
That’s all the tasks. Thank you.
A few last questions. Overall, how would you describe Northwind to a colleague? What stood out, good or bad? Is there anything you expected to be able to do that you couldn’t?
Do you have any questions for me? [Answer the questions you put off during the tasks.]
Your [incentive] will arrive by [date and method]. The recording stays with the team and is deleted by [date]. Thanks again for your time.
Stop the recording. Write down the three moments that mattered most before the next session starts.
How to adapt the script
- Your tasks. Replace the three tasks with yours, each with its goal, starting point and sign of success. Most moderated sessions hold four to eight tasks. Keep the order the same for every participant unless you mean to vary it, and avoid tasks that depend on each other. How to write usability test tasks.
- Remote, moderated. Add a line before the welcome asking the participant to share their screen, and check you can hear each other. Say who else is on the call.
- Unmoderated. Nobody is there to send a question back, so each task needs its full context in writing, and the think-aloud instructions need an example. Take out the prompts and follow-ups that depend on a moderator, and ask the after-task questions in writing.
- Participants who use assistive technology. Ask beforehand what they use and let them use their own device and settings. Allow more room in the schedule and never take over their device. In the warm-up, ask how they usually get around a website.
- Signed-in products. Prepare an account per participant with the data the tasks need, and reset it between sessions. Never use real customer data.
- Phones. Name the device and the orientation in the starting point. Record the screen from the device itself, or from a camera above it.
Pilot the script once with a colleague outside the team. Fix the wording, the timing and any task that gives away its answer, then freeze it for the round.
Where Variantly’s task reviews fit
Variantly’s task reviews start from the same three parts: you describe a task and how you’d know it worked. AI attempts then work through it from perspectives you describe, on your live product or a staging site. Each issue keeps the screens it happened on, and accessibility checks run on those same screens. A person decides each issue (Fix it, Not sure yet or Working as intended), and a recheck runs the same task after a change.
Attempts aren’t people. They can’t tell you what someone expected, how they felt or what they’d do next week, and they don’t replace sessions with participants. They can run your tasks between rounds, before you recruit, and recheck a fix against the task it came from. Task reviews.
Questions people ask
How long should a usability test script be?
Long enough to read every part the same way in every session. Most of its length is the tasks and their moderator notes. The welcome, consent and wrap-up stay short so the session’s time goes to the tasks.
Should the moderator read the script word for word?
Read the welcome, the consent request, the think-aloud instructions and the tasks word for word. Those must match across sessions. The warm-up and follow-up questions can be said naturally, as long as they stay neutral.
What’s the difference between a script and a study plan?
The plan says why the study runs: its questions, who takes part, how many sessions, and what counts as a result. The script is what the moderator says and does in one session. Write the plan first; the script’s tasks come from its questions.
What if the participant asks for help?
Send the question back with a neutral prompt such as “What would you do if I weren’t here?” If they still can’t go on, move to the next task and mark this one not done. Answer their questions at the end.
Do I need consent to record?
Ask before you record, say who will see the recording and when it’s deleted, and say they can stop at any time. Use your organization’s own consent form where it has one. Steve Krug publishes a sample recording consent form with his scripts.
How many participants do I need?
Five per group is a common planning convention for a qualitative round, not a rule. Report how many sessions hit a problem, not a percentage of a handful of people.
Sources
- Nielsen Norman Group, Usability (User) Testing 101 (Kate Moran)
- Nielsen Norman Group, Thinking Aloud: The #1 Usability Tool (Jakob Nielsen)
- Nielsen Norman Group, Talking with Participants During a Usability Test (Kara Pernice)
- Nielsen Norman Group, Turn User Goals into Task Scenarios for Usability Testing (Marieke McCloskey)
- Steve Krug, Downloads: usability test scripts, a recording consent form and “Things a therapist would say”
- Digital.gov, Usability Starter Kit: a usability testing script and recording consent form from the US government’s user experience program
See it on a product
- Read an example review, one task from start to finish.
- Ask for a review of one task on your own product.