Discussion guide: how to write one for interviews and focus groups, with a template

A discussion guide is the document a moderator uses to run a research interview or focus group. It holds the introduction you will read, the topics and questions in the order you plan to cover them, follow-up prompts, instructions for any activities and a checklist of what you need on the day (GOV.UK). It is also called an interview guide or topic guide, and focus group moderators sometimes call theirs a questioning route.

A guide differs from a script. Nielsen Norman Group explains that a script belongs to structured interviews, while a guide can be used flexibly: the interviewer can change the order, skip questions or ask ones that aren't in it (Nielsen Norman Group). That flexibility lets you follow an unexpected answer, which is why a good guide is mostly open questions with prepared follow-ups.

Why write one

NN/g names two risks of interviewing without a guide: making up questions on the spot, which tend to come out leading, and failing to cover the same topics in every interview. GOV.UK adds uses after the fact. A guide helps several interviewers cover the same ground, lets the team review sessions together and records what you did in each round.

What goes in a discussion guide

1. Research questions at the top, as a reminder of what the sessions are for.
2. Introduction: who you are, what the research is for, how long it takes, recording and confidentiality, and that they can skip any question.
3. Warm-up questions that are easy to answer.
4. Main topics, each with one or two starter questions and a few follow-ups.
5. Activities, if any, such as a task on a prototype or a document to review.
6. Closing: check your summary, ask what you missed, thank them and explain next steps.
7. Timings beside each section, so you know when to move on.

Pathfinder International's guide to in-depth interviews adds two rules worth keeping: no more than 15 main questions, and factual questions before opinion questions (Pathfinder International).

How to write one

NN/g's process, condensed:

1. Write your research questions. These are what you want to learn, not what you will ask. "How do supply chain teams decide which recalled items to pull first?" is a research question.
2. Brainstorm interview questions without judging them yet.
3. Broaden them. Replace clusters of closed questions with one open question that invites a story. Five yes-or-no questions about someone's workday can become "Walk me through your last workday," with the closed ones kept as follow-ups.
4. Check coverage. Every research question should have at least one interview question that serves it.
5. Put them in order. Chronological order usually suits an experience. Start with easy questions and leave reflective ones until people have recalled specific events.
6. Add follow-ups and probes, such as "When did that happen?" and "Tell me more about that."
7. Pilot the guide with someone like your participants, then reword confusing questions and cut what doesn't fit the time.

A worked template

This example is hypothetical. A company building recall-management software wants to learn how hospital supply chain teams handle a product recall, from the first notice to confirming every affected item is off the shelves. Each interview runs 45 minutes by video with a supply chain manager or director.

Research questions

• Who first learns about a recall, and how?
• How does the team find where affected items are?
• Who confirms the items have been removed, and how do they know?

Introduction (3 minutes)

Thanks for making time. I'm [name] from [company]. We're learning how hospital supply chain teams handle product recalls, and I'd like to talk through one recent recall you dealt with. There are no right answers. Please leave out patient information and anything confidential to your hospital, and feel free to skip any question. With your permission I'd like to record so I can focus on listening; only our research team will hear it. Any questions before we start?

Warm-up (5 minutes)

• What does your role cover, and how long have you been in it?
• How is your supply chain team organized?

Topic 1: The most recent recall (15 minutes)

• Think of the most recent recall your team handled. How did you first hear about it?
  • Follow-ups: Who told you? What did you do in the first hour? What information was missing?
• How did you work out where the affected items were?
  • Follow-ups: Which systems did you check? What did you have to check by hand?

Topic 2: Confirming removal (10 minutes)

• How did you know every affected item had been removed?
  • Follow-ups: Who signed off? Was anything found later?

Topic 3: Comparison (7 minutes)

• How did that recall compare with others you've handled? What made it easier or harder?
• If you could change one part of the process, what would it be, and why?

Closing (5 minutes)

• Summarize: "Let me check I have the sequence right…"
• Is there anything important about recalls that we didn't cover?
• Thank them and explain what happens next.

The order is deliberate: factual questions about one real event come first, and the opinion question about what to change comes last, after the person has recalled the details. The sections add up to 45 minutes, which leaves no slack, so mark one or two questions you can drop if a topic runs long. A 30-minute customer discovery interview script shows a more detailed version for discovery calls, and qualitative research interview questions covers writing individual questions.

Adapting it for a focus group

Richard Krueger's focus group guide for the University of Minnesota describes the same idea with parts specific to groups (Krueger, University of Minnesota):

• A standard introduction: welcome, an overview of the topic, ground rules, then the first question. His sample ground rules include that there are no right or wrong answers, one person speaks at a time and participants don't need to agree with each other.
• Five types of question, in order: an opening question everyone answers in turn, an introductory question, transition questions, the key questions that matter most and ending questions.
• Questions that look back. He recommends taking people back to an experience rather than forward to the future, and avoiding questions that can be answered yes or no.
• A three-step close: summarize and ask whether the summary is right, review the purpose and ask whether anything was missed, then thank everyone.

Focus group questions has examples of each type.

Adapting it for a usability test

For a usability session, the main section is a list of tasks instead of topics. GOV.UK's guidance on moderated usability testing describes a discussion guide with your introduction script, a description of each task with any instructions, and a planning checklist. Good tasks set a clear, believable goal without hinting at how to complete it (GOV.UK). Usability testing covers how to write the tasks.

Common mistakes

• Too many questions. Thirty questions in 45 minutes means rushing every answer. Keep only the ones that serve a research question.
• Leading questions. "How frustrating is the recall process?" assumes the answer. "How did that recall go?" doesn't.
• Opinions before facts. "What would you improve?" asked before the person has described a real event tends to get a generic answer.
• Reading it word for word. Read the introduction if you like; the rest should sound like a conversation.
• Skipping the pilot. One practice session usually reveals a confusing question or a section that runs long.
• Never revising it. NN/g says it's fine to update the guide as interviews go on. Note what changed so you can compare sessions fairly.

Your next step

Write your research questions at the top of a blank page. Under each, write one starter question about a specific recent event and two follow-ups. Add an introduction and a close, put a time beside each section, and run it once with a colleague before your first real session.

If you need people to interview, Instant Expert can find people who match a description you write, such as "supply chain directors at hospitals." You review who it finds, it sends your invitations, and you pay only for calls that get booked. The hospital supply chain experts directory page is one place to start, and in-depth interviews covers planning the whole study.