What is an Interview Guide?

An interview guide (discussion guide, protocol) is the prepared document that structures a research interview: the objectives, the topic areas in a deliberate order, the key questions under each, the probes for going deeper, and the practical scaffolding (introduction, consent, timing, close). It is the instrument of the semi-structured interview, sitting between a fixed script (which produces comparable but shallow answers) and free conversation (which produces depth but no comparability), and its purpose is to let the interviewer be present with the participant instead of inventing the next question in real time. The guide serves the researcher; the participant should barely notice it exists.

The Anatomy That Works

1. Objectives first, written down.
Two or three questions the study must answer, at the top of the page. Every topic below earns its place by serving one; anything that doesn't gets cut, which is how guides stay short enough to actually run in forty-five minutes.

2. An opening that sets the terms.
Purpose in plain words, consent and recording notice, the reassurance that there are no wrong answers and that criticism is useful, and an easy warm-up question about the participant's context that gets them talking about themselves before anything evaluative.

3. Topics in a funnel.
Broad and past-tense before narrow and hypothetical: how they currently do the thing, a specific recent time they did it, where it went wrong, what they've tried; only then the reactions, concepts, and futures. The order matters because grounded stories are reliable and speculation isn't, the core lesson of the interviewing canon.

4. Key questions with probes underneath.
Each topic gets one or two anchor questions and a bank of neutral probes ("what happened next?", "can you show me?", "what made that hard?", "who else was involved?"). Probes are where interviews get their depth; a guide without them produces surface answers followed by awkward silence.

5. A close that catches what the guide missed.
"What haven't I asked about that matters here?" routinely produces the interview's best material; then thanks, next steps, and incentive logistics.

The Wording Rules

Ask about specific past behaviour, not general habits or future intent ("tell me about the last time you..." beats "how often do you..." beats "would you use..."). Keep questions open and single: one idea per question, no embedded assumptions, no leading adjectives. Never explain the hypothesis inside the question. Prefer silence to rescue: the pause after an answer is a probe. And avoid asking participants to design ("what would you want?") or to predict themselves; ask them to describe, and let the design inferences be yours. Every one of these rules exists because its violation manufactures agreeable, hypothetical, or performed answers that feel like insight and aren't.

Using the Guide Well

Pilot it once and rewrite; the first live run always exposes questions that don't land. Treat it as a map, not a route: follow the participant's thread when it leads somewhere valuable, and use the guide to come back. Revise between rounds in an exploratory study, logging the changes, and freeze it for comparability in a confirmatory one. Guides also transfer to unmoderated formats: the same objectives, funnel, and behaviour-first wording become the prompts for open-ended video answers in an asynchronous study (a Ballpark study asks the guide's questions on camera without a moderator present), which trades live probing for scale and the absence of interviewer influence.

The Takeaway

An interview guide is preparation in a form that frees the interviewer: objectives on top, an honest opening, a funnel from past specifics to future speculation, anchor questions with neutral probes beneath, and a close that catches the unasked. Word it for behaviour, pilot it, follow the participant, and revise with a log. The conversation should feel effortless precisely because the guide did the effort beforehand.