Customer discovery interviews are structured conversations that reveal how customers describe their problems in their own words, so you can validate or kill assumptions before you build. The single best next step is to schedule ten short, founder-led calls this week, each aimed at testing one specific assumption about the problem you think you are solving.
TL;DR:
- Use founder led calls before revenue or while seeking product market fit; once sales repeat, surveys or paid tests validate scale more efficiently.
- Set one objective, two or three testable hypotheses, and a narrow segment; identify 20 prospects and request 20 minutes to discuss workflows, not pitch.
- Ask about recent events, workarounds, costs, and decision makers; avoid hypothetical willingness to pay, leading prompts, and product pitches during interviews.
- Recruit from warm and cold contacts, explain data handling and withdrawal rights, and get separate explicit permission before recording any interview.
- After 8 to 12 calls, capture five phrases and three workarounds, group themes, then test the strongest hypothesis with a landing page or paid pilot.
Table of Contents
- Who should run these interviews and when to prioritise them
- What you actually learn from well-run interviews
- Plan the interview: objectives, assumptions and who to call
- Recruiting, consent and ethics: a practical checklist
- Sample questions and a reusable interview script
- Interview agenda, do's and don'ts for the live call
- Turn interview notes into validated insights
- Quick checklist after your first round of interviews
- Finding first conversations through curated founder networks
- How we help you run faster discovery conversations
- FAQ
- Sources
Who should run these interviews and when to prioritise them
Customer discovery interviews deliver the most value at the idea validation stage, before revenue, and in the early search for product-market fit. Once you have paying customers and a repeatable sales motion, the questions shift from "is this a problem" to "will you pay more for this", which surveys or paid pilot tests answer more efficiently.
Founders should lead these calls themselves whenever possible. You hear the hesitation in someone's voice, you notice which words they repeat, and you can follow an unexpected thread that a hired researcher might script around.
- Idea validation and pre-revenue founders get the clearest return from one-on-one calls.
- Early product-market fit teams use interviews to refine positioning and pricing language.
- Once you have a working funnel, surveys and paid A/B tests validate scale, not problem existence.
What you actually learn from well-run interviews
A good round of interviews gives you three concrete outputs: the exact phrasing customers use to describe their pain, the workarounds they already pay for or tolerate, and the specific triggers that push someone to look for a solution.
- Verbatim problem language you can reuse in landing pages and sales calls.
- Current alternatives, including manual workarounds, competitors and the dreaded "we just live with it."
- Named decision-makers and buying triggers you can test in your next experiment.
Founders who interview customers early and repeatedly tend to learn customer language faster and adapt their offers accordingly, according to teaching entrepreneurship research. That pattern holds whether the founder runs five calls or fifty: the language improves with repetition, not with a bigger sample alone.
Plan the interview: objectives, assumptions and who to call
Walking into a call without a clear objective wastes the conversation. Spend thirty minutes on setup before you book a single meeting.
- Write one primary objective for the round, such as "confirm that freelance designers struggle to track client feedback."
- List two or three testable hypotheses tied to that objective, phrased so a customer could prove you wrong.
- Define a narrow segment, not "small business owners" but "freelance graphic designers with 3 to 10 active clients."
- Name 20 specific people who fit that segment, using existing contacts, LinkedIn searches or founder communities.
- Draft a short outreach message that asks for 20 minutes to learn about their workflow, not to pitch anything.
A playbook worth following treats the first dozen conversations as a research sprint: name real people, talk to them directly, and capture their exact language before you touch paid acquisition.
Recruiting, consent and ethics: a practical checklist
Recruiting ethically protects both the participant and the quality of your data. Selection bias creeps in fast when you only call people who already like you, so mix warm contacts with cold outreach from your named list.
- State the research purpose, expected duration and format before the call starts.
- Make clear participation is voluntary and can be stopped at any point.
- Explain how notes or recordings will be stored and who can access them.
- Ask explicit, separate permission before recording audio or video.
A proper consent process should cover the research purpose, session format and duration, a withdrawal clause, data handling details and explicit recording permission, as outlined in this POPIA checklist. If you offer an incentive, research on recruiting bias notes it should be framed strictly as payment for time, since rewards can attract people who only show up for the reward. Keep personal data to the minimum needed and store it securely, consistent with how we handle data under our own privacy policy.
Pro Tip: If a participant declines recording, switch to handwritten notes immediately rather than pushing for a workaround.
Sample questions and a reusable interview script
Non-leading questions are the backbone of a useful interview. The Mom Test approach, widely taught in entrepreneurship programmes, asks about past behaviour and real stories instead of hypothetical opinions.
Opening lines:
- "Thanks for the time. This is for research, not a pitch, nothing to buy today."
- "Can I record this so I don't misquote you, or would you prefer I just take notes?"
Problem discovery questions:
- "Walk me through the last time this problem came up."
- "What did you try before, and why didn't it stick?"
- "How much time or money does this cost you in a typical month?"
Decision-process questions:
- "Who else is involved when you decide to change a tool like this?"
- "What would have to be true for you to switch today?"
A workable 25 to 30 minute script runs: two minutes of intro and consent, 10 to 15 minutes of story-based probing, five minutes on decision process, and three minutes to wrap up and ask for referrals, a structure also recommended in entrepreneurship interview guides.
Interview agenda, do's and don'ts for the live call
Keep the call time-boxed and resist the urge to fill silence. A pause often means the person is thinking, not that you need to rescue the moment.
- Paraphrase what you heard before moving to the next question.
- Ask for a specific story instead of a general opinion.
- Let silence sit for a few seconds after a question.
- Never pitch your product or describe features during the interview.
- Never ask "would you pay for X", since hypothetical answers rarely predict real behaviour.
- Avoid leading questions that hint at the answer you want.
When an answer feels vague, ask "can you give me an example from last week" rather than repeating the same question louder. When someone gets defensive, acknowledge the point and redirect to a concrete scenario: "that makes sense, what happened the last time you tried to solve it."
Pro Tip: Write down the exact words a participant uses for their problem; those phrases often become your best marketing copy.
Turn interview notes into validated insights
Raw notes are not insight until you organise them. Affinity mapping, grouping short verbatim quotes into five to seven recurring themes, turns scattered conversations into a visible pattern.
- Pull three to seven word quotes from each interview and cluster them by theme.
- Count how many distinct people raised each theme to gauge signal strength, since one mention is an anecdote and five is a pattern.
- Write three prioritized hypotheses based on the strongest themes, then pick one to test with a concrete experiment, such as a landing page or a paid pilot.
Triangulate wherever possible: a pattern that shows up in interviews, in support tickets and in a quick survey carries more weight than interview data alone.
Quick checklist after your first round of interviews
After eight to twelve conversations, consolidate what you learned before booking more calls.
- Capture five verbatim phrases customers used to describe the problem.
- Map three common workarounds people currently rely on.
- Write three testable hypotheses and schedule the next experiment to validate the strongest one.
Finding first conversations through curated founder networks
Cold outreach is often the slowest part of discovery, since most of your 20 named targets will not answer a stranger's message. Warm introductions inside a curated founder community shorten that gap because the connection already carries context about why you are reaching out.
We built Foundly around that idea: match founders with the people who can actually answer their questions, then suggest a coffee or short meetup instead of a cold email thread.
— Londolani
How we help you run faster discovery conversations
We match you with potential customers, co-founders and peers who fit the segment you are trying to learn from, with a clear reason for each introduction instead of a cold list. 
Rather than guessing who to message from your list of 20 targets, you can browse curated connections in founder communities such as Founders in Cape Town, Founders in Johannesburg or Founders in Pretoria, book a short coffee or co-working session, and start the conversation in person. Create a free account to see your first matches, and upgrade to Foundly PRO if you want deeper filtering as your discovery rounds scale up.
FAQ
What should you ask in a customer discovery interview?
Ask about past behaviour and specific stories rather than opinions or hypotheticals, such as "walk me through the last time this problem came up" or "what did you try before." These questions, drawn from Mom Test style interviewing taught in entrepreneurship guides, surface honest patterns instead of polite guesses.
What is customer discovery?
Customer discovery is the process of talking directly to potential customers to learn how they experience a problem, what they currently do about it, and what would make them switch to something new. It happens before you invest heavily in building, and it relies on real conversations rather than assumptions.
What are the five main questions asked in a discovery interview?
There is no single fixed set, but a strong round usually covers: how the problem shows up today, what workarounds exist, who else is involved in deciding to change, what triggers someone to look for a solution, and what a recent specific example looked like. The exact wording should stay open-ended and non-leading, as recommended in entrepreneurship interview resources.
How do you avoid bias when recruiting interview participants?
Recruit from a named list that mixes warm and cold contacts rather than only people who already like your idea, and if you offer an incentive, frame it strictly as payment for time. Research on recruiting bias notes that incentives can otherwise attract participants motivated mainly by the reward rather than genuine interest.
What consent details should a discovery interview cover?
A proper consent process explains the research purpose, session format and duration, the right to withdraw at any point, how data will be stored and accessed, and separate explicit permission to record. This structure follows the POPIA checklist for research-style interviews.
Sources
- POPIA checklist — Usercentrics
- Customer interviewing is valuable in any context — Teaching Entrepreneurship