conduct-user-interview

Open questions with how, what, or tell me about, and avoid did/was/is phrasing except for factual confirmation. Use the participant's own words, probe with silence and echo techniques, and never supply your hypothesis or vocabulary mid-answer.

Six core probes: echo (repeat their key phrase), boomerang (return their question), Columbo (trail off naively), neutral expand (tell me more), significance (why does that matter), and silence (wait 5-10 seconds). Try silence first after every answer.

Interviews capture what people say — attitudes, memories, and self-reported behavior. Usability tests capture what people do while attempting tasks on a product. If the question is whether people can use something, moderate a usability session instead of interviewing.

Brief observers before the session: stay silent, mute cameras, and send questions to a back channel. Cap visible observers at one or two, filter and rewrite their questions neutrally, and ask them yourself at the end of the session.

People cannot predict their own behavior, so prediction questions produce unreliable data. Ask about past behavior instead — when they last did the related activity and what happened — and corroborate claims with behavioral data like logs or observation.

Terse answers mean you narrowed the funnel too fast. Climb back up to a broad grand-tour question, re-warm rapport, slow your own pace, and use comfortable silence to invite longer responses.