Short, focused customer interviews are one of the fastest ways to learn whether a problem, workflow or product idea has real traction. The risk is bias: leading questions, loaded prompts, or selective note-taking can create the illusion of validation. This guide shows how to run a tiny (10–20 minute) interview that minimizes leading, collects useful evidence, and keeps you honest about what you actually learned.

When to choose a tiny interview

Use this format when you need quick directional evidence rather than a formal discovery study. Good contexts: early-stage idea checks, recruiting users for prototypes, triage on feature priority, or validating observed behavior before a deeper study. The aim is not to prove product-market fit but to collect observations you can act on.

Before the interview: consent, scope and single-goal framing

• Consent wording (read verbatim): “Thanks for joining. This is a short, informal conversation to understand your experience with X. There’s no sales pitch; your answers are confidential and we’ll use them to improve our thinking. You can stop at any time. Is that okay?”
• Define a single learning goal for this interview (example: “Understand how you currently manage receipts for business expenses”). Keep every question tied to that goal.
• Limit duration: tell participants 10–20 minutes. Tiny interviews are shorter for a reason — they force focus.

Neutral question set: the script to avoid leading

Use open, behavior-focused prompts and avoid “would you” or “do you like” phrasing. Follow up with requests for concrete examples.

1. Warm-up (30–60 seconds): “Can you tell me briefly what you did last time you handled [topic]?”
2. Process probe: “Walk me through what happened, step by step.”
3. Frequency and context: “How often does that happen, and where are you when it happens?”
4. Pain and workaround: “What, if anything, is hard or annoying about that?” Follow: “Can you show or describe the last time it was hard?”
5. Cost and consequence: “What does it cost you — time, money, stress — when that happens?” Request a concrete number or estimate if possible (illustrative numbers are fine).”
6. Alternatives: “What have you tried instead? Why did you pick that?”
7. Closing check: “Is there anything I didn’t ask that you think matters?”

Avoid: “Would you pay for X?” or “Do you like feature Y?” — both lead participants to imagine solutions rather than describe reality.

How to take unbiased notes

Assign two roles if possible: interviewer and note-taker. If you’re alone, adopt a two-column note format:

• Left column: Verbatim quotes or actions (use exact wording when possible).
• Right column: Observations and hypotheses (clearly labeled as such).

Time-stamp key moments (e.g., “2:15 — participant described manual workaround”). If you paraphrase, mark it with brackets [paraphrase]. This helps separate what was actually said from your interpretation later.

Worked example (illustrative numbers)

Learning goal: discover how consultants track project time.

Consent: read verbatim and confirm. Duration: 15 min.

Key exchange (verbatim left / observation right):

Left: “Last week I used an Excel sheet I copied from a friend. I open it weekly and paste times from my calendar.”

Right: [Uses Excel template, not a dedicated time-tracker; manual copy step exists; likely low frequency].

Follow-up: “How long does it take per week?”

Participant: “Maybe 30 minutes total.”

Observation: [30 min/week ≈ 2.5 hrs/month. If consultants bill $75/hr, that’s $187.50 in time cost — useful for rough value calculation, illustrative only].

Resulting hypothesis: There’s friction in manual transfer from calendar to Excel; a simpler capture tool could save ~30 minutes/week per consultant. Next step: test this with three more consultants or run a micro-prototype.

Quick reusable template (copy-paste for use)

Consent: “Thanks — 10–20 minutes to talk about [topic]. No sales, confidential, stop any time. Okay?”

Learning goal: [one sentence]

1. “Tell me about the last time you [activity].”
2. “Walk me through what happened, step by step.”
3. “How often does that happen? Where are you?”
4. “What’s the hardest part? Can you describe the last time?”
5. “What do you do instead? Why?”
6. “Anything I missed?”

Note-taking: Left = verbatim / Right = observations & hypothesis; timestamp key moments.

Two common challenges and how to mitigate them

Challenge 1 — Participant tries to help by offering suggestions or asking if you’ll build their idea: Politely redirect. Say, “Thanks — I’m trying to understand how you handle this today. Can you tell me about the last time it happened?” This brings the conversation back to behavior.

Challenge 2 — Interviewer drifts into solution talk: Use a visible checklist with your single learning goal. If you notice solution talk creeping in, pause and explicitly say, “Can we go back to how you currently do this?”

Use interviews as learning rather than proof

Ask participants what happened recently and how they handled it; do not treat agreement with your product idea as a purchase commitment. Remove identifying information from notes when possible, explain whether recording is optional and honor requests not to record. A few interviews are not a representative survey and should not be presented as proof of demand. Next, use the Decision Compass or a small reversible test to decide what deserves further research.

Sources and scope

This post gives practical editorial guidance for short customer-discovery interviews. It is not a substitute for formal market research or legal advice. The methods are drawn from common qualitative research practice: behavior-based questioning, separate note-taking, and small-N hypothesis testing. If you want a beginner’s path into broader user-research techniques, start here: Start Here.

FAQ — short answers to likely questions

Q: How many interviews are enough? A: For directional insight, run 5–10 tiny interviews. You’ll see patterns quickly; use more interviews if responses are mixed.

Q: Can I record the interview? A: Only if you have explicit consent. Offer to stop recording if they prefer. Recording helps avoid note bias but must be handled securely.

Q: Should I offer payment? A: Small compensation respects time and can improve participation. It must not influence answers — treat it as a thank-you, given after the interview.

Checklist (copy for the session):

• Single learning goal written on a card
• Consent script ready
• 10–20 minute timer
• Neutral question script
• Two-column note sheet (verbatim / observations)
• Post-interview summary slot: capture 3 quick bullet insights

Run these tiny interviews regularly and use the resulting micro-evidence to avoid false positives. When interview data consistently points the same way, you have stronger grounds to design experiments or prototypes — but never treat a single interview as proof of product-market fit.