Which draft is ready to test?
Choose the smallest prototype that can answer the design question.
The question
ParcelNest needs to learn whether people can find the last collection time before entering an address. You have one afternoon and three AI-generated drafts.
- A · Map first
- Full-screen map, address search, collection cutoff behind a small information icon.
- B · Task first
- Collection cutoff beside the address field; explicit empty, lookup-error and ready states.
- C · Chat first
- A chat asks for an address, recommends a time and creates a booking with no review screen.
Invented drafts. Visual polish is equal; their task structure differs.
“Which draft is ready to test?”
Not saved by this site. Your browser may restore text when you navigate back.
A strong answer, step by step
Open each step in turn. This is one strong answer, not the only one.
Step 1: Name the test question
What would you do here? Open the move you’d make.
Which screen looks most modern?
Tempting, but it misses something. That does not resolve the booking uncertainty.
Can the person find the cutoff before committing?
The strong move. This connects the prototype to the observed barrier.
Which draft used the most advanced AI model?
Tempting, but it misses something. The production tool does not establish task quality.
See how a strong candidate says it
“Can a first-time user find the cutoff and decide whether collection today is possible?”
Why it works: A testable question gives the critique a criterion beyond taste.
Step 2: Choose the prototype
What would you do here? Open the move you’d make.
Choose B and test the task with plausible addresses.
The strong move. It exposes the needed information and meaningful states.
Choose A because users liked the map.
Tempting, but it misses something. The cutoff is still hidden behind the information icon.
Choose C because fewer visible controls mean less friction.
Tempting, but it misses something. It changes the interaction and commits without a review.
See how a strong candidate says it
“Draft B puts the critical information at the point of decision and includes the failure state.”
Why it works: Its structure makes the uncertainty observable with less work.
Step 3: Scope the afternoon
What would you do here? Open the move you’d make.
Polish animation before adding lookup failure.
Tempting, but it misses something. A missing failure state blocks a useful test.
Prototype all account and payment settings.
Tempting, but it misses something. That uses time without answering the current question.
Include the empty, failed and ready booking states.
The strong move. Participants can encounter and recover from the actual uncertainty.
See how a strong candidate says it
“Build the address, cutoff and lookup-error path. Defer animation and secondary account pages.”
Why it works: A small working slice can reveal task failures that a polished static frame hides.
Step 4: Respond to the test
What would you do here? Open the move you’d make.
Ship because B was the most defensible concept.
Tempting, but it misses something. Selection is not validation.
Observe the missed information and revise that part.
The strong move. This connects the next edit to task evidence.
Generate twenty more visual variants immediately.
Tempting, but it misses something. More variants do not explain the failure you just observed.
See how a strong candidate says it
“If participants still miss the cutoff, inspect its wording and placement before declaring the task solved.”
Why it works: A selected prototype is a hypothesis; its outcome still needs observation.
The answer, in short
- Decision
- Test Draft B as a small working slice.
- Because
- The draft exposes the critical decision and the states needed to study it.
- What would change it
- Observation shows a map or conversational route serves the target task better.
Common traps
- Equating fewer controls with a simpler task.
- Using visual preference as task-completion evidence.
Check your answer
- What uncertainty does this prototype address?
- Does it include empty and failed states?
- Can I explain why another draft is weaker for this task?
- What test result would make me revise it?
Last reviewed Nov 10, 2026