Customer research
Write usability test tasks without giving away the answer
Describe what the participant wants to accomplish in everyday terms, not the button or menu that does it. Build the scenario around a realistic goal, decide ahead of time what completion looks like, and leave space for observer notes.
A text to start with
I am testing [product or service] with people who [audience]. The workflow I want to observe is [workflow]. The interface uses these labels: [menu names, button text, page titles]. Write task cards, each with a short realistic scenario, a task stated as a goal, a completion signal I can observe, and a note field. Avoid revealing the navigation path or answer. Use ordinary domain words when needed to express the actual goal, even if the interface also uses them.
Example request. Change the details to fit your situation.Usability task card
Product or service tested: [ ] Participant description: [ ] Task number and order: [ ] Scenario, in everyday words: [ ] Task stated as a goal: [ ] Navigation hints removed: [ ] Domain terms needed to express the goal: [ ] Completion signal I can observe: [ ] Time allowed, if any: [ ] Observer notes (where they hesitated, what they tried first): [ ] Help given and when: [ ] Outcome (completed, partly, not completed): [ ] Follow-up question to ask afterward: [ ]
Hints hidden in wording
Compare task wording with the interface to check whether it gives away the route you want to observe. A natural word shared with a button is not automatically a hint; ask whether it tells the participant where to click. Use scenarios and information a real user could have. Define a completion signal tied to the goal rather than a convenient screen alone. If a task combines goals, decide whether to observe the whole workflow or separate the parts, and record that choice.
Ask another agent to check the result
Compare each task card with the list of interface labels and flag any wording that names or hints at the right button, menu or page. Point out scenarios that depend on information a real user would not have. Say which cards have no observable completion signal.
Rehearse before the session
Run through the cards with a colleague who did not write them, and note any moment they take a shortcut the wording handed them. Fix those cards. GOV.UK describes observing participants complete tasks and preparing realistic tasks that do not reveal the answer. Order the tasks, decide which are optional if time runs short, and agree who will take notes.