The JTBD Switch Interview
Interrogate the story of a real purchase to recover the true causes of a switch.
- Difficulty
- Advanced
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 93%
A qualitative interview method for extracting why a customer actually switched, not what they claim they want. Modeled on criminal and intelligence interrogation 'that feels like therapy,' it only interviews people who already took action, reconstructs the timeline of their decision, and digs through layers of language to reach the real context. The output is the pushes, pulls, anxieties, habits and trade-offs behind the purchase.
Origin
Bob Moesta's method-side of JTBD, refined from interrogation techniques he studied in the 1980s-90s; he credits Chris Voss's 'Never Split the Difference' for the mirroring and negotiation tactics that layer on top.
Core principles
- 01Only interview people who have already tried to make the progress; never trust what people say they'll do ('bitching ain't switching').
- 02Don't use a rigid discussion guide; use pushes/pulls/anxieties/habits as a compass and follow the meaningful threads.
- 03People peel through layers of language: the polite 'paum' layer, the fantasy/nightmare exaggeration layer, then the real 'what actually happened.'
- 04Play the story back incorrectly on purpose so they correct you and elaborate; a 'yes' ends the conversation, a 'no' opens it.
- 05Cluster stories into pathways rather than segmenting into themes, because sets of reasons work together.
- 06Seven or eight interviews start to repeat; do 10, no more than 12, and prefer two rounds of 12 over one batch of 24.
How to run it
- 1
Frame the causal question
Define what causes someone to say 'today's the day.' Recruit people who recently purchased in that category, including competitors' products.
Pro tip You don't need your own product to exist yet; interview whoever they'd fire when they hire you.
- 2
Extract the story, not the product opinion
Ask what was going on, what they hoped for, what they worried about, what they had to give up, and how they convinced others. Reconstruct the buying timeline.
Pro tip The check-swipe moment is not the decision; a $137 coat rack took one buyer 18 months to decide.
Watch out Don't ask about product features; ask about the life around the purchase.
- 3
Push past the edge of language
When they run out of words, bracket with a deliberately-wrong binary ('was it more about this or more about that') that you know is neither, forcing them to talk more.
Pro tip Ask 'tell me more' and 'give me an example' instead of repeated 'why.'
- 4
Cluster into pathways
Instead of finding themes across all interviews, cluster stories where pushes, pulls, anxieties and habits move together, revealing the three-to-five conflicting jobs the product is hired for.
Watch out Jobs often conflict (fast vs thorough); serving one can repel another.
In the wild
In the first five minutes a buyer says they bought a car because they got a great deal on their dream car. Digging in reveals the old car had 280,000 miles, three large repair bills in four months, was making a sound, and a long trip was coming up.
→ The real causes (context and struggle) surface only after moving past the fantasy layer of language, not the surface 'deal' story.
Intercom's founders attended a workshop, then brought their executive team to actually run the interviews themselves. Once they understood how to ask the questions, they knew what to build.
→ Intercom turned off irrelevant features per pathway and repriced to match each competitor, growing to a valuation over two billion dollars.
Common mistakes
Hypothesizing jobs in a conference room
Sitting in a room guessing the jobs guarantees you're 100% wrong; you never find the contradictions or the irrational parts without talking to people who acted.
Trusting stated intentions
People can't assure future behavior and it usually never happens; complaining about a missing feature (Gantt charts) doesn't mean they'll actually switch.
Is it for you?
Best for
Founders and PMs who need the real causal drivers behind a purchase or churn before setting a roadmap.
Not ideal for
Deeply habitual purchases (chewing gum, 20-year detergent loyalty) where buyers genuinely can't recover the decision story.
From the transcript
“the phrase I have is uh bitching ain't switching just because people about something doesn't mean they're going to do anything about it”
“I usually get them to what I call the edge of language where they have no more language and what I do is I literally…”
“it's criminal and intelligence interrogation that feels like therapy”
“it starts to repeat around seven or eight and I usually do 10 no more than 12 and I would rather do two rounds of…”
“instead of segmenting them we cluster them we find the pathways”
From the episode
The ultimate guide to JTBD
Bob Moesta (co-creator of the framework)