← All founder workflows

How to run customer interviews that produce information instead of encouragement

Run customer interviews by asking about the past rather than the future and never describing your idea until the end. Ask when the problem last happened, what they did, what it cost, and what they have already tried to fix it. Hypothetical questions — would you use this, would you pay — produce agreement, because people are kind and imagining is free. Ten interviews conducted this way will tell you whether a problem is urgent; ten pitches disguised as interviews will tell you nothing except that people are polite.

7 days $0 to run6 stepsUpdated 2026

The workflow

0/6 done

  1. Friends, colleagues and supportive founders will agree with anything. Find people who have publicly described the problem — in a forum, a review, a job post — and ask for fifteen minutes about how they handle it. Recruiting is most of the difficulty here, and the temptation to interview whoever is available is exactly what makes most founder interviews worthless.

  2. "When did this last happen to you? Walk me through what you did." Memory of a specific event is far more reliable than a general opinion, and it produces the details you need: who was involved, how long it took, what it cost, what they tried first. General questions produce general answers, and general answers cannot be built on.

  3. Ask what they currently spend on it — in tools, hours, or somebody's salary. A problem with existing spend has a budget line you can compete for; a problem solved by nothing has no budget and requires you to create one. This single question separates ideas that sell from ideas people merely agree with.

  4. "Would you use this?" is the most misleading question in startup research, and it is the one founders lead with. Hold your idea until the end, then describe it and watch for a specific reaction — a question about integration, a request for a timeline, an objection. Enthusiasm without a follow-up question is politeness, and politeness has ended more startups than competition has.

    Competitor research
    Input
    Interviewees keep mentioning they "just use a spreadsheet and a shared calendar" for client onboarding
    What comes back
    Named workaround confirmed across interviews. 6 paid tools exist at $25–$120/mo, all requiring migration of the whole client record. Complaint pattern in reviews: setup takes days. The spreadsheet is winning on switching cost — that, not features, is what a new entrant has to beat.

    What this step looks like when it is done properly — the free Idea Validator covers the same ground as part of its research.

  5. "If I build this, would you be a first customer at $X a month?" then stop talking. The answer sorts your interviews into two piles that look identical in your notes but behave completely differently later. Ask every time — an interview without a commitment question produces a transcript rather than a decision.

  6. After ten interviews, count how many people described your problem before you mentioned it. That number is your result. Five or more of ten is a strong signal; two means you have found something that interests you more than it bothers them, which is worth knowing before you spend a quarter on it.

    Idea ValidatorFull tool
    Input
    10 interviews on client onboarding. 6 described the problem unprompted, 8 use spreadsheets, 3 said yes to $49/mo pilot.
    What comes back
    GO, with a migration-first product. Six unprompted mentions and three pilot commitments is a strong ten-interview outcome. The binding constraint is switching cost from the spreadsheet — if the first version cannot import their existing sheet in one step, the three yeses will not convert.

    Run it yourself — free, no signup:

    Who it's for, what it does, what they pay. The more specific the sentence, the sharper the read.

Questions founders ask about this

What questions should I ask in a customer interview?
Ask about the past: when the problem last happened, what they did, what it cost, and what they have already tried. Those questions produce facts. Hypothetical questions about whether they would use your product produce agreement.
How many customer interviews do I need?
Ten is usually enough to see the pattern, and you can stop when new interviews stop teaching you anything. What matters is how many described the problem unprompted, not how many said the idea sounded good.
Should I tell people about my idea during an interview?
Only in the last five minutes. Describing it early turns the conversation into a pitch and the interviewee into a critic or a cheerleader, and neither role produces the information you came for.
How do I find people to interview with no network?
Find people who have already described the problem publicly — in forums, reviews, or job posts — and ask for fifteen minutes about their process. That framing gets far more replies than asking for feedback on an idea.
Is it a bad sign if everyone likes my idea?
Usually. Universal enthusiasm often means the questions were leading or the audience was friendly. Look instead for people who interrupt to describe the problem themselves, or who ask when they can have it.

Next, founders usually do this

The tools used above have their own pages — Idea Validator — and the SOP SOP: Validate a SaaS idea in 7 days runs the same ground in more depth. Also worth reading: the Nine Lives Doctrine, and real verdicts from ideas kitty has run this workflow on.

kitty.build runs this entire workflow for you

Every step above — the research, the competitor read, the numbers, the honest verdict — is what nine specialist AI boards do automatically when you feed her an idea. She will tell you to kill it if it deserves killing. First idea is free.

Feed her an idea — free

No card · Failed tasks are free · Your repo, your domain, your revenue