AIWorkDex
Customer Research Practice

Write a neutral, event-based interview guide

Produce a short guide that asks about real events and keeps the preferred solution out of the answer.

Preparation
A guide-writing and peer-review session
Difficulty
Moderate
Task
Remove cues from interview questions
Back to the Task: Remove cues from interview questions

Before you start

Get what you need before you start

  • Write the research question and the assumption most likely to bias the interviewer.
  • Choose a shared review document without participant data.

Steps

Work through the method

Use the listed inputs and tools. Check the evidence when you need to verify a step.

  1. Anchor the interview in a recent event

    Open with a request to describe the last time the participant faced the target situation.

    Why it matters: A concrete event gives the conversation an inspectable context.

    Input
    The research question.
    Output
    An opening event question.
    Tools in this stepGoogle Sheets

    Evidence for this step

  2. Cover context and behaviour

    Ask what triggered the event, what the participant did, which tools were involved, and what happened next.

    Why it matters: Behaviour and context show the current workflow without requiring a proposed solution.

    Input
    The opening event question.
    Output
    A chronological core question set.
    Tools in this stepGoogle Sheets

    Evidence for this step

  3. Remove leading words

    Replace yes-or-no wording, praise, criticism, and solution names with open factual questions.

    Why it matters: Leading wording can cue the answer the interviewer expects.

    Input
    The first guide draft.
    Output
    A neutral guide draft.
    Tools in this stepGoogle Sheets

    Evidence for this step

  4. Use one short question at a time

    Split compound questions and leave space for the participant to answer before a probe.

    Why it matters: Short questions make the answer easier to attribute to one prompt.

    Input
    The neutral guide draft.
    Output
    A guide with one idea in each core question.
    Tools in this stepGoogle Sheets

    Evidence for this step

  5. Prepare neutral probes

    Add probes for sequence, consequence, workaround, and missing detail without suggesting an answer.

    Why it matters: Prepared probes support comparable interviews while preserving unexpected evidence.

    Input
    The core guide.
    Output
    A separate probe list.
    Tools in this stepGoogle Sheets

    Evidence for this step

  6. Run a bias review

    Ask a reviewer to mark every question that reveals the desired product, judgement, or answer.

    Why it matters: The founder’s preferred solution needs an explicit check before the first session.

    Input
    The complete guide and stated assumption.
    Output
    A reviewed guide and a record of removed cues.
    Tools in this stepGoogle Sheets

    Evidence for this step

Success checks

Check the result before you finish

  • The guide starts from a recent event.
  • No core question names or praises the preferred solution.
  • The same core guide can be used across the round.

Failure modes

Watch for these problems

  • Rewrite questions that ask whether the participant would use an imagined feature.
  • Record a new spontaneous probe before reusing it in later sessions.
  • Treat the guide as a bias check, without claiming that it removes all bias.

Tools

Choose the tools you need

Sources

Read the sources behind this practice

Check what each source supports and where the advice has limits.