Jobs to Be Done Interviews: How to Uncover What Really Drives Customer Decisions
A practical guide to JTBD interviews, with example questions, timelines, and a framework for uncovering what drives adoption.
Why JTBD interviews work when feature questions fail
Most interviews stay at the surface. You ask what people want or which features matter, and you get tidy answers that rarely explain real behavior. People are bad at predicting behavior and at simplifying past decisions — "We need better reporting" doesn't explain why someone switched tools last month.
Jobs to Be Done interviews instead focus on a specific decision someone already made — adopting, switching, or deciding not to — and reconstruct it: what changed, what triggered action, what alternatives were considered, and why the timing finally felt right. That makes JTBD especially useful for understanding adoption, switching, stalled evaluations, and churn. (See also When to Use Surveys vs Interviews.)
The four forces
A JTBD interview explains movement: why someone left the status quo for a new solution. Four forces structure your analysis. Push is pain in the old situation; pull is the draw of the new solution; habit is what kept them in the old way; anxiety is what made switching feel risky.
Customers don't switch the moment they notice a problem. They switch when push and pull overcome habit and anxiety. A team may hate its workflow for months but stay because migration feels risky or the workaround is "good enough" for another quarter.
Who to recruit
Interview people who decided recently enough to remember clearly — usually those who switched or adopted in the last 30 to 90 days, before memory compresses into a clean story. Prioritize new customers, people who considered you but chose another option, recent churned customers, and the real decision-makers in complex B2B buys. Avoid long-time loyal customers for switch interviews.
For one segment, 8 to 12 strong interviews usually reveals repeated patterns. If buying process, alternatives, or stakes differ meaningfully — self-serve SMB vs. enterprise procurement — treat them as separate segments. See Recruiting Research Participants in 2026.
Structure the interview around the timeline
The biggest mistake is jumping straight to needs or feature requests. Start with the timeline. Open by lowering the pressure to be helpful — tell them you're not evaluating their satisfaction, just understanding how they decided to make a change, and that you'll ask specific questions about timing because the sequence is what teaches you.
Then work through the stages with specific, time-bound questions. Establish the switch (what they started using, when, and what came before), then find the struggling moment — when the old way stopped working, and why they didn't change earlier. Listen for concrete incidents ("We missed a client deadline because reporting broke"), not abstract dissatisfaction. Trace the timeline forward: what workaround did they try, and when did it stop being enough? This is where habit shows up. Probe pull and anxiety (What stood out? What felt risky? Who needed convincing?), then nail the decision point, often skipped: why move then instead of waiting — a deadline, a trigger? Close on outcomes, not satisfaction scores: what's improved, what's still harder than expected, and what would be hardest to replace.
Use follow-ups relentlessly — "Walk me through that," "What happened right before that?" See also How to Write a User Interview Guide.
From transcripts to decisions
Strong evidence is observable detail. Compare "We needed something more robust" with "We were onboarding three client accounts a week, and our spreadsheet broke every time two people edited at once." The second gives you context, trigger, scale, and cost.
After interviews, code verbatim quotes into the four forces, cluster repeated patterns, and mark differences by segment. Those patterns drive better outputs than a feature list: hiring criteria, sharper messaging that speaks to push and pull while reducing anxieties, and product priorities (if habit is strong, migration may matter more than new features).
The core principle
JTBD interviews aren't about collecting better opinions — they're about explaining behavior. Avoid the common mistakes: asking hypotheticals, leading with your framework, skipping the timeline, and ignoring non-buyers, whose stalled evaluations reveal anxieties your customers already overcame. Identify the struggling moment, map the timeline, and separate push, pull, habit, and anxiety — and you'll understand what actually gets customers to change.