PM Customer Interviews
(2026 Edition)
5 prep moves, 6 questions that work, 5 to avoid, 6 listening techniques, and 6 biases to catch in yourself.
Build User Research Skills Daily — Free →5 Interview Prep Moves
Write 3 hypotheses you want to test — not features you want to sell
Write 10 open-ended questions — never yes/no
Prepare your probe questions: 'Can you say more?', 'What happened next?'
Test questions on a colleague first — refine before real users
Keep the interview to 30–45 min — fatigue beyond that
6 Questions That Work
Tell me about the last time you did [problem area] — walk me through it
What's the hardest part of [task] today?
Can you show me how you do [task] currently? (screen share)
What have you tried that didn't work?
If you could wave a magic wand, what would change?
Is there anything I haven't asked that would help me understand your world?
5 Questions to Avoid
Would you use a feature that did X? (Future intent is unreliable)
Do you like our product? (Leads to politeness)
How much would you pay for Y? (Stated WTP is wrong)
Is X important to you? (Everything sounds important when asked)
Don't you wish we had Z? (Leading, pushes them to your answer)
6 Listening Techniques
Silence is your friend — after they finish, wait 3 seconds before speaking
Note emotional words — frustration, delight, confusion — those are signal
Repeat back what you heard — 'So you're saying...' — confirms understanding
Resist the urge to explain your product — this isn't a demo
Record (with consent) — you'll miss quotes in the moment
Take sparse notes during, detailed write-up after — presence beats note-taking
6 Biases to Avoid
Confirmation bias — you'll hear what you expected
Leading questions — subtle cues shape their answers
Talking users — interviewing only people who say yes
Interviewing only power users — skewed picture of typical user
Small sample sizes — 2 interviews tell you nothing
Assuming stated = actual behaviour — what people say they do differs from what they do
FAQ
How many customer interviews should PMs run per week?
1 per week minimum, 3–5 per week during active discovery. Fewer and you lose pattern recognition; more and synthesis gets overwhelming. The goal is sustainable cadence. PMs who interview zero users per week lose intuition faster than they realise.
What's the biggest interview mistake PMs make?
Talking too much. Great interviews are 80% user talking, 20% PM. PMs who get excited and explain their product lose the user's real input. The discipline: treat it like therapy — ask, listen, let silence draw out the truth. Most PMs need to cut their talking by half.
Build User Research Skills Daily
Daily scenarios on interviewing, listening, and synthesising user insight.
Start Free Trial →