Customer interview questions: the Mom Test, for founders

Customer interview questions and the Mom Test

Customer interview questions that follow the Mom Test share one rule: talk about the other person’s life, ask for concrete past episodes, and keep your idea out of the room until you have heard how they already cope. You are not collecting compliments. You are looking for repeated pain, money already spent, and workarounds they built without you.

This page goes deeper than a general validation checklist. For where interviews sit next to landing pages, concierge tests and pre-sales, see how to validate a startup idea. Here the focus is the question list, the failure modes, and what a good conversation still cannot prove.

What the Mom Test changes in your script

The usual mistake is to describe the product and ask for a verdict. “Would you use this?” “Is this a problem for you?” “What would you pay?” Those prompts invite politeness. Friends, former colleagues and friendly buyers in Switzerland often answer in your favour because the meeting is social, not because they will change suppliers next quarter.

The Mom Test pushes you toward behaviour you could verify if you had receipts: last time the problem happened, what they did, who else was involved, what it cost in time or money, and whether they searched for a fix. Hypothetical futures (“If we built X…”) are weak evidence. Past episodes are stronger, even when the sample is small.

A short run of questions you can reuse

Pick one thread and stay on it for twenty minutes. Jumping across problems feels productive but blurs the notes.

Problem episode

One real week or one real project, not “generally”.

Workaround

Spreadsheet, agency, intern, manual export: proof the pain was worth a fix.

Commitment

A next step with cost: intro, data sample, paid pilot, not “sounds interesting”.

1. Open on their job, not your slide

   ”How does your team handle [outcome] today?” No product name yet.

2. Anchor in the last occurrence

   ”When did this last slow you down?” Follow the story chronologically.

3. Quantify in their units

   Hours, fines, rework, headcount, software seats, not your guessed price.

4. Ask who pays and who blocks

   In Swiss B2B, procurement, IT and the line manager often disagree.

5. Close with a concrete next step

   Second interview with a colleague, a redacted export, or a paid trial week. If they decline, note that too.

Who should skip long interview rounds

Interviews are the wrong first move when you already have repeat usage and invoices from strangers. They are also a poor substitute when the risk is technical feasibility, not demand: you may need a spike in code, not another coffee in Zurich.

Skip or shorten the programme if your testers are almost all people who know you personally, if you cannot name a segment beyond “SMEs”, or if you are using interviews to postpone a clear landing-page test that would answer the same question faster. Interviews shine when the problem is messy, the buyer is not obvious, or the status quo is a bundle of tools nobody advertises as a single product.

What interviews do not prove

A stack of polite meetings does not show that strangers will pay your price, that legal and security review will pass, or that retention will hold after onboarding. It does not replace watching people use a rough product, a concierge delivery, or a page with a costly call to action.

A made-up example: scheduling software for trades firms

A made-up example: a founder targets small plumbing and electrical firms in German-speaking Switzerland. She books five conversations through a trade association mailing list, not friends. She does not show a mock-up.

In the first call she asks how jobs get from the customer call to the calendar today. The owner describes WhatsApp voice notes, a paper diary and a part-time admin who retypes addresses into a generic calendar app on Friday afternoons. Last month a missed slot cost a half-day of travel because two fitters drove to the same village on different days. They never bought dedicated scheduling software; they once paid a web agency CHF 2’400 for a contact form that nobody maintains.

In the third call a larger firm says IT would need to approve anything that stores customer addresses in the cloud. That is not a “no” to the idea; it is a buying path she had not mapped. After five calls she has three similar stories, two dead ends where the pain is real but spend is zero, and one introduction to an office manager who agrees to a paid two-week manual scheduling trial. None of the five said “great idea”. The useful output is the repeated workaround and one paid next step, not enthusiasm.

When a validation package fits

The product validation package fits when interviews have pointed at a segment and a painful workflow, and the open question is whether people will use a real product and whether acquisition can be repeated. Weeks one to four put software in their hands, not a slide deck. Weeks five to twelve add positioning, a landing page, paid traffic and measured pivots, ending in a Go, Pivot or Stop read with what the data still cannot prove.

If you are still choosing which segment to interview, stay with discovery calls and a written problem statement. Book the package when you are ready to trade build time for behaviour data you cannot get from questions alone.