Skip to content
KatafactsBeta

Voice of the customer

Customer interview guide

Customer interview guide is licensed CC BY 4.0. Attribution: Katafacts (katafacts.com).

Customise with AISkill file ↓

1 · What it is

What it is

An interview guide is what turns 'go talk to some customers' into a real, repeatable instrument — a topic-by-topic question bank built around a stated research goal, with genuine follow-up probes woven in, not just a list of first questions. State who you're talking to and what you're trying to learn; a full guide gets drafted, and it's yours to edit before you take it into the room.

2 · When to use it

When to use it — and when not to

Use it when

  • You're about to run a general discovery interview and want a real instrument, not an improvised conversation.
  • You have a stated research goal — even a rough one — that a structured question bank would serve better than winging it.
  • You want interview-technique reminders specific to this conversation, not generic interviewing advice.

Not when

  • The interview is happening in context, watching real work as it's done — that's the gemba interview protocol, built around what's observed in the moment rather than recalled afterward.
  • The interview is specifically about a closed deal (won or lost) — that's the win-loss interview guide, built around a decision that's already been made.
  • You don't have a real research goal yet, just a vague sense you should 'talk to customers more' — sharpen the goal first, or the guide will be as vague as the goal.

3 · How to fill it in

How to fill it in

Research goal
What are you actually trying to learn? Specific beats broad — 'understand why onboarding takes longer than expected' drafts a sharper guide than 'understand the customer experience.'
Topics and questions
Drafted from your research goal and any seed topics you already had in mind — add, remove, or reword anything before the interview.
Interview technique reminders
Drafted for this specific guide's context — real interview-technique discipline, not generic platitudes.
How to run this interview
Drafted from the research goal — how to actually run this conversation and what to listen for.

4 · What good looks like

What good looks like

The example below is a general onboarding-discovery guide: four topics building from background through to open frustrations, real follow-up probes within each topic, and interview-technique reminders grounded in the coaching stance this whole product tries to model — ask before answering, let silence sit.

Same example, as a downloadable xlsx workbook.

Download .xlsx

Interview guide

New-customer discovery interview guide

Dana Whitfield, Product Lead · 2026-02-01

Team: Marcus Lee, Support · Sponsor: Alex Romero, VP Product

Research goal

Understand the real, lived onboarding experience well enough to feed the VOC evidence log with genuine quotes, not assumptions about what's hard.

Topics and questions

4 topics · 9 questions

Background and context

  • Tell me about your role and how your team uses the product day to day.
  • How long have you been a customer, and what changed for your team when you first started using it?

First-week experience

  • Walk me through your first week — what happened, in order?
  • What almost made you give up, if anything?
  • Tell me more about that moment — what were you thinking at the time?

Ongoing use

  • What does a typical week look like when you're using this?
  • What have you built a workaround for, and why?

Wishes and frustrations

  • If you could change one thing, what would it be?
  • What's something you've mentioned to a colleague but never told us?

Interview technique reminders

  • Ask before answering — if they ask what you think, turn it back: "What's your take on it?" first.
  • Let silence sit for a few seconds after a question before rephrasing it.
  • Ask "tell me more about that" instead of moving on to the next question.
  • Avoid leading language — "was it frustrating that..." plants the answer; "how did that feel" doesn't.

How to run this interview

This guide is built for genuine discovery, not confirmation — the goal is real quotes for the evidence log, not validation of an assumption already held about what's hard in onboarding. Spend the most time in 'First-week experience': it's where the most specific, evidence-rich stories tend to surface, and the follow-up probe there is doing real work, not decoration.

5 · Common mistakes

Common mistakes

  • Writing questions that are really statements looking for agreement.

    "Wasn't onboarding frustrating?" plants the answer before the customer has said anything — open-ended questions surface what's actually there, leading questions confirm what you already believe.

  • Treating the question list as a script to read verbatim rather than a guide to work from.

    The real value of an interview comes from the follow-up that isn't written down — a rigid script misses exactly the specific detail a genuine 'tell me more' would surface.

  • Covering too many topics to go deep on any of them.

    Same vital-few discipline as everywhere else in this catalogue — a focused guide that goes deep on three topics beats a shallow pass across eight.

6 · What it connects to

What it connects to

upstream

    downstream

    • VOC evidence log

      The quotes and observations this interview produces are exactly what feeds the evidence log — log them as they come in, not from memory afterward.

    • Affinity diagram

      Once you've run several interviews, an affinity-diagramming session is where the individual quotes get synthesized into real themes.

    7 · Where AI helps

    Where AI helps

    Judgement — stays yours

    • Deciding whether a drafted question actually fits the real conversation, or reworking it before the interview
    • Knowing when to deviate from the guide because something more interesting just came up

    Analysis — AI helps

    • Drafting open-ended questions and real follow-up probes from a stated research goal
    • Drafting interview-technique reminders specific to this guide's context

    Drudgery — automated

    • Counting topics and questions
    • Exporting to xlsx in the house format

    9 · Rate this kata

    Rate this kata