Building a guide that avoids this takes more than writing good questions. It takes a specific structure, a clear sense of what each section is supposed to accomplish, and enough flexibility that the moderator can follow a thread without losing the session entirely.

Start With Objectives, Not Questions

The most common mistake is opening a blank document and starting to type questions. Questions come last. Objectives come first.

Write down, in plain language, what the client or research team needs to know by the end of the study. Not what they want to explore, not what would be interesting, but what they need to be able to decide or do differently once the research is complete. If you can't articulate a concrete decision this research will inform, the guide will show it.

A useful objective sounds like: "Understand which product features drive trial among first-time buyers and which create barriers to repurchase." A vague objective sounds like: "Learn more about how consumers feel about the product." The first one tells you what to ask. The second one doesn't.

Most studies have one primary objective and two or three secondary ones. The primary objective should take up the most time in the guide. Secondary objectives fill the remaining space. If you're finding that secondary objectives are competing for time with the primary one, drop them or run a separate study.

The Structure Every Discussion Guide Needs

A well-built guide has five distinct sections, each doing a specific job.

Introduction and ground rules (5-10 minutes). This isn't just housekeeping. How you open the session shapes what participants are willing to say for the next 90 minutes. Cover recording consent, the no-wrong-answers rule, the fact that the moderator didn't design whatever is being discussed, and that honest criticism is genuinely more useful than polite approval. The tone you set here determines whether participants perform or actually share.

Warm-up (10-15 minutes). Low-stakes questions that get participants talking without putting them on the spot. "Tell us your name and how long you've been using this type of product" is a warm-up question. "What's your biggest frustration with this category" is not, at least not yet. The warm-up builds comfort and gives the moderator a first read on who in the group talks easily and who needs drawing out.

General exploration (20-25 minutes). This is where you establish baseline attitudes and behavior before introducing anything specific to the study. If you're testing a new product concept, this is where you learn how participants currently solve the problem that product addresses. If you skip this and jump straight to the concept, you lose the context that makes reactions meaningful. The participant who hates the concept might already love a competitor's solution, or might never have had this problem in the first place. Without the general exploration section, you can't tell.

Core questions (30-40 minutes). The primary objective lives here. This is the longest section and should get the most deliberate attention when writing the guide. Questions should move from broad to specific, from individual experience to group comparison, and from unprompted reaction to prompted evaluation. This is also where stimulus material gets introduced if the study involves concepts, packaging, advertising, or prototypes.

Wrap-up (10 minutes). Two things happen in the wrap-up: you catch anything the guide missed, and you give participants a chance to say what they actually think matters. "Is there anything important we haven't covered?" consistently surfaces the most honest comments of the entire session, because participants have had 80 minutes to decide whether they trust the room.

Here's what a partial guide looks like in practice, using a consumer snack brand as the example:

INTRODUCTION (5 min) Thank you for joining us today. My name is [Moderator] and I'll be guiding our conversation. A few ground rules before we start: there are no right or wrong answers, we're here to hear honest opinions, and the session is being recorded for research purposes only. Nothing you say will be attributed to you by name.

WARM-UP (10 min) Let's start with some quick introductions. Tell us your name, where you're from, and what you typically snack on during the week. Probe if needed: Do you snack at specific times of day? At home, at work, on the go?

GENERAL EXPLORATION (20 min) When you think about buying a snack, what goes through your mind? Probe: What are you looking for? What do you try to avoid? Has that changed in the last year or two?

Walk me through the last time you tried a snack brand you'd never bought before. What made you pick it up? Probe: Was it something you saw in the store, or did you hear about it somewhere? What made you decide to actually buy it?

CORE QUESTIONS (35 min) I'm going to show you a product concept now. Take a minute to read it before we discuss it. [Distribute concept card. Allow 60 seconds of silent reading.]

What's your first reaction? Probe: What stood out to you? What do you notice first?

What does this product communicate to you about who it's for? Probe: Is that person like you? If not, how are they different?

What would make you want to try it? What would give you pause? Probe: If price wasn't a factor, what would change?

WRAP-UP (10 min) We've covered a lot today. Before we close, is there anything about this product or this category that feels important that we haven't talked about?

If you had to leave us with one thing, what would it be?

Notice the structure: each section has a stated time allocation, questions move from general to specific, and each core question has at least one probe written in before the session starts. The stimulus is introduced mid-session, after baseline attitudes are established, not at the start.

Writing Questions That Generate Discussion, Not Just Answers

Focus group questions work differently from survey questions. The goal isn't a clean, recordable answer. The goal is a discussion, a reaction, a chain of responses where participants build on or push back against each other.

Open-ended is the baseline requirement, but open-ended isn't enough on its own. "How do you feel about this product?" is open-ended and almost useless. "Walk me through the last time you bought something like this, what were you trying to solve?" is open-ended and generates an actual story.

Questions that work in focus groups tend to:

• Ask participants to recall a specific experience rather than describe a general attitude

• Put the locus of evaluation on the product or concept rather than on the participant's own preferences ("What does this communicate?" rather than "Do you like this?")

• Create space for disagreement by acknowledging that different people might see it differently ("Some people we've spoken to feel X, others feel Y, where do you fall?")

• Use projective techniques when direct questions trigger socially desirable answers ("If this brand were a person, how would you describe them?")

Plan your probes before the session. A probe is a follow-up question the moderator uses when an answer is interesting but shallow. "Tell me more about that" is a probe. "What makes you say that?" is a probe. Probes that are too vague produce more of the same shallow answer. Write two or three specific probes for each of your core questions, ones that push toward the specific information you need, not just more words on the same topic.

Handling Stimulus Material

Most market research focus groups involve showing participants something: a product concept, a packaging design, advertising executions, a prototype, competitive landscape imagery. How you structure stimulus exposure in the guide directly affects the quality of what you get back.

The standard structure is: establish baseline first, then expose. If you're testing three packaging concepts, spend time understanding how participants currently navigate the category before showing them anything. Once they've seen the stimulus, you can never get their unprompted response back.

When introducing stimulus, order matters. Sequential exposure (one at a time, fully discussed before the next is shown) produces more nuanced individual reactions. Simultaneous exposure (all options at once) produces clearer comparative rankings but can flatten individual response. Most market research guides use sequential first, comparative second, so you get both.

For concept testing specifically, the guide should include a read-through of the concept before discussion. Give participants time to read it individually before inviting reactions, because spontaneous verbal reaction to a concept that some people haven't finished reading yet produces a discussion that misrepresents what the concept actually communicates.

Timing and Question Count

A 90-minute session realistically covers 8 to 12 questions at full depth. A 120-minute session covers 12 to 16. These numbers assume real exploration of each topic, not a rapid-fire question-and-answer exchange.

Most first-draft guides are 30 to 40 percent too long. Write your first draft, then go through it and cut anything that serves curiosity rather than objectives. If a question doesn't connect to a decision the research is meant to inform, it's making the guide longer without making the research better.

Put your most important questions in the first half of the guide. The second half of a focus group always gets compressed, because interesting threads from earlier in the session take longer than planned. If your most important question is last, it will get 10 minutes instead of 30.

Pilot Before You Use It

Run a five-minute read-through of your guide with a colleague before the first session. Not to test the questions, but to time each section. Most guides take 20 to 30 percent longer to run than they look on paper, because real group discussion moves slower than a moderator's mental model of it.

If you're running multiple sessions, debrief after the first one and revise before the second. The first session almost always reveals one question that confuses participants and one topic that needed more time than you gave it. Fixing both in the guide before the second session means better data across the study.

From Guide to Transcript

The quality of your guide shows up directly in the quality of your transcript. A guide that moves logically from warm-up to core questions produces a transcript that codes cleanly because the data is organized by topic. A guide that jumps around produces a transcript where the same theme appears in six different places, requiring manual reorganization before any analysis can start.

If you're running sessions that will go to focus group transcription, let your transcription service know the guide structure upfront. At Qualtranscribe, knowing whether a session follows a structured guide or a more open ethnographic format affects how speaker labels and section markers are applied, which in turn affects how clean the output is for coding. Market research transcription that's formatted around your guide's structure reduces the time between transcript delivery and the start of real analysis.

Need the session captured accurately from the first word to the last? Get started here.

FAQ

How long should a focus group discussion guide be? For a 90-minute session, most experienced moderators use a guide that's 6 to 10 pages, including probes and timing notes. Longer than that and you're almost certain to run out of time on the most important sections.

Should the moderator follow the guide exactly? No. The guide is a map, not a script. A good moderator follows interesting threads when they emerge and returns to the guide when those threads run out. The guide ensures the session covers everything it needs to, not that it covers everything in a fixed order.

How many questions should a focus group discussion guide have? Plan for 8 to 12 substantive questions in a 90-minute session. Each question should have two or three prepared probes. The total question count matters less than whether each question actually serves one of your study objectives.

Should warm-up questions be included in the written guide? Yes. Warm-up questions are easy to improvise in the moment, but including them in the guide helps the moderator manage timing and gives a less experienced co-moderator or observer the full picture of what the session is supposed to accomplish.

Can the same discussion guide be used across multiple focus groups in the same study? Yes, and it should be, with minor revisions between sessions based on what worked and what didn't. Consistency across sessions is what makes cross-group comparison meaningful.

Related Reading

• How to Conduct a Focus Group

• Running a Successful Focus Group: What to Ask, Record, and Transcribe

• How to Ensure Diversity in Focus Group Participants for More Reliable Results

• Best Human Transcription Services for Focus Groups

• Understanding Timestamps, Speaker Labels, and Verbatim Formats