Use the fillable usability-test task worksheet to record the starting conditions, task and observations before the session. Download it without signing up, or use the plain-text version (Markdown) in your own document.
“Open Accessories and select the right adapter” already gives away part of the solution. You can see whether someone follows the instruction, but not whether they would have looked under Accessories themselves. To observe that decision, your task needs a goal and enough information to act without specifying the route.
Turn the research question into a situation
A useful task follows from the decision your study needs to inform. New category labels call for finding an offer, whereas evaluating size guidance requires measurements the participant can use. “Have a look at the page” leaves what you want to observe unclear in either case.
NN/g recommends realistic, actionable tasks that do not reveal the interface solution. Ordinary domain terms can appear in a task even if they are also menu labels. Do not supply the menu item or button the participant should choose.
If you rent cargo bikes and want to find out whether a new customer can choose one for a particular load, the task wording makes a clear difference:
| Too directive | More useful for the study |
|---|---|
| Open “Transport solutions” and filter for 80 kg. | You need to collect a 65 kg washing machine on Saturday. Find out whether you can rent a bike for it. |
| Click “Check availability”. | Check whether you can get your chosen bike on Saturday from 10 am to 2 pm. |
The second task requires test data that lets participants check those times. Otherwise, you are testing a broken study setup rather than the interface.
Keep success criteria in the facilitator’s guide
What counts as completion belongs in the facilitator’s guide before the session. Clicking any bike is not enough in this example: the load capacity must fit, and loading space or securing the item may need checking. If the page lacks that information, “I cannot decide” may be a correct conclusion.
Distinguish an unsuitable choice from missing information or an interaction failure. Those findings call for different changes. Record whether the person reached the result independently, after a neutral follow-up or after a specific hint. Assisted completion is not independent completion.
Do not let one task solve the next
The sequence also changes what you can learn: an early task can reveal the menu item whose findability you want to investigate later. Reviewing the tasks together lets you spot those clues and decide whether a selection should remain or be reset for the next activity.
For independent questions, varying the order may help reveal learning effects. For a connected workflow, the sequence is part of what you are investigating. Explain that choice in the test plan instead of improvising a different structure in every session.
Pilot the instructions
A pilot with someone uninvolved shows you whether the instructions and test access are understandable. Reading the tasks and trying the access can reveal missing information, inaccessible states or accidental hints before the study starts. This setup check does not replace sessions with suitable participants.
Keep the revised tasks together with success criteria and starting states. They underpin both usability testing and participant recruitment: better wording alone will not make someone a suitable participant if they would never perform the task in everyday life.
Prepare the complete test task
A well-written task is not enough if the participant starts with somebody else’s cart or an incorrectly prepared account. In the cargo-bike example, the test content needs to show load capacity and availability for the stated period. Record the opening page and whether a selection from an earlier task is allowed to remain.
The worksheet connects these preparations with the observations you will record:
| Field | Example for the cargo-bike test |
|---|---|
| Decision | Can new customers choose a suitable bike without calling the rental business first? |
| Starting conditions | Rental homepage, no bike selected, test inventory prepared. |
| Participant’s task | You want to transport a washing machine on Saturday. Find a suitable bike and check whether you can rent it for your intended period. |
| Additional information | Supply the washing machine’s weight and dimensions and the intended rental period with the task. |
| Moderator’s success criteria | Load capacity and cargo space are suitable; the participant can establish availability. |
| Observation | What information does the participant seek and find, and where do they need help? |
Use quantities and dates that match the prepared test content. If the booking system only supports whole days, the task should not casually assume an hourly reservation. Otherwise, you cannot tell whether you observed an interface problem, a limitation of the offer or a mistake in the test setup.
Analyse observations without deciding the finding in advance
Suppose the participant finds a bike with sufficient load capacity but cannot find the cargo platform’s width. They open the contact page. Recording only “failed the task” would lose the important connection: search worked, but information needed for the choice was missing. The first change to investigate would be a clear dimension in the relevant place, followed by another test.
If another participant repeatedly overlooks an existing dimension drawing, the question is different. The information exists but is not found or understood in the journey. Keep the observed sequence alongside your interpretation and record any help you provided. This lets the development team understand what supports the proposed change.
User interview or usability test: which worksheet fits?
An interview explores how people currently experience and organise a workflow. A usability test observes how they use a particular interface. If your team does not yet understand the task, start with the UX research interview guide. To investigate an existing website or prototype, prepare your tasks in the PDF worksheet and discuss a usability test with us.
