A prototype before you build
Hand one flow to five strangers and see if it finishes without a guided tour.
UX RESEARCH LAB
Guerrilla testing planner
A guerrilla usability test is a short public session: you hand one task to a few strangers and watch whether they can finish it without a tour of the product. This page keeps the plan, each person’s result, and how severe the breaks were.
Write the task, the success line, and the place. Then log whether each person finished alone, needed a hint, or quit. The severity chart and the stop rule tell you to end the round or fix one problem first.
Plan the round
People
Read the round
Three rules before you stop strangers
Guerrilla testing is not a deep interview. It only tells you whether this task breaks in a stranger’s hands.
- One task, one success line, and ten minutes at most. Two tasks means two rounds.
- Do not explain the product. Say the task and stay quiet unless they are fully stuck.
- When one serious problem shows up three times, do not recruit a sixth person. Fix that problem first.
Notes stay on this device. Nothing is sent to Nova Lab servers.
Go deeper
Why you’ll love this
This guerrilla testing tool is for the round you run in a cafe or on campus instead of a lab. You write one task, a success line, a place, and a time box, then log whether each person finished alone, needed a hint, or quit. A quote and a severity stay with that person. The chart, the repeated problems, and the stop rule say whether to recruit again or fix one blocker first. It fits a UX class and a rough prototype. It does not replace a deep interview or a recruited panel.
How to jump in
-
Write one task
Use the sentence you will actually say, plus one line that defines finished.
-
Lock the place and the time box
Cafe, campus, station, lobby, or sidewalk — usually five to ten minutes and a small thanks, such as a coffee.
-
Log each person before the next one
Finished alone, needed a hint, or quit, plus seconds, one quote, and a severity for each problem.
-
Read the stop rule
Close the round when one serious problem shows up three times. Move to the next task when most people finish alone.
What you walk away with
- One task and a success line, not a long script
- Severity chart and repeated problems
- A stop rule around five people
- Stored on your device, with CSV, JSON, PNG, and PDF
Where it shines
A usability class
Open the campus sample and discuss why the reserve control became a major problem three times.
After a screen review
Take the main action from a heatmap into a public place so guessed attention is not treated as behavior.
A stop-or-continue decision
Print completion and severity instead of the team’s hunch, then close the task or fix the blocker.
Quick answers
How is guerrilla testing different from a lab usability test?
Guerrilla testing is short, public, and cheap: one task, about five people, and a few minutes each. A lab session buys more time, tighter recruiting, and a deeper interview. Guerrilla testing finds obvious breaks. It does not prove a statistic.
How many people do I need?
Five people in the same situation is the usual first round for one task, enough for obvious problems to repeat. This tool stores the round on your device. It does not recruit anyone for you.
Should I help when someone gets stuck?
Stay quiet first. If they fully stop, give one short hint and log the result as needed a hint, not finished alone. Log a quit separately.
Where is the round stored?
In this browser only. Export JSON if you need a backup for class.
Which Nova Lab tool should I open next?
Open the heatmap if the break is one zone on a screen. Open the journey map if the path is longer. Open the empathy map if the person is still unclear.