Customer Interviews

Talking to potential users to learn about their problems, not to pitch.

Also known as: customer interview, user interview, discovery interview

A customer interview is a conversation to learn how someone works and what hurts — not a pitch, not a demo, not a validation of your idea. You ask about their past behavior, costs and workarounds; they talk, you listen and take notes. Fifteen such conversations teach more than a month of building.

bad:  "Would you pay for an AI dashboard for your inventory?"
good: "Walk me through the last time stock ran out. What did you do? What did it cost?"

Run them in batches with a fixed question list so answers compare. Record (with permission), write up the same day, and separate facts (“they export to Excel every Friday”) from interpretations (“they need automation”). Bring a co-founder: two people hear different things.

The classic mistakes:

Pitching disguised as research. The moment you describe your solution, answers become politeness. Mention the idea only at the end, if at all.
Leading questions. “Don’t you hate how slow invoicing is?” manufactures the answer. Ask neutrally and let silence do the work.
Talking to non-buyers. Users without budget authority describe a different problem than the person who signs. Interview the buyer and the user separately.
No notes, no synthesis. Ten conversations remembered as vibes. Write each up, tag the pains, count how many mentioned each — patterns beat impressions.
Stopping at five. Five interviews confirm your biases beautifully. Past ten, surprises start; past twenty, patterns stabilise.

Afterwards: extract the pains, rank by frequency × cost, and test the top one with a real commitment ask — a pre-sale, a pilot, a letter of intent. Attention is cheap; willingness to pay is the finding that matters.