Focus Group Discussion: What It Tells You and What It Doesn’t

A focus group discussion is a moderated conversation with a small group of participants, typically six to ten people, designed to surface attitudes, perceptions, and motivations that surveys and analytics cannot reach. Used well, it gives you texture and language that quantitative data cannot provide. Used poorly, it gives you false confidence in the wrong direction.

The method has been around for decades, and it remains one of the most misused tools in marketing research. Not because the format is flawed, but because most teams misunderstand what it is actually measuring.

Why Focus Groups Get a Bad Reputation They Half Deserve

I have sat through a lot of focus groups over the years, on both sides of the glass. Early in my agency career, I watched a client spend a significant chunk of their research budget running six groups across three cities to evaluate a new campaign concept. The groups were enthusiastic. The campaign launched. It performed poorly. The client was confused.

What happened was not unusual. Participants said they liked the concept because saying you dislike something in a group setting is socially uncomfortable. The moderator, under pressure to keep the session moving, did not probe hard enough on the hesitations that surfaced briefly and then got buried. The debrief was a summary of what people said, not an analysis of what it meant. The client heard what they wanted to hear, and the research process helped them do it.

This is not an argument against focus groups. It is an argument against running them badly. The method itself is sound when applied to the right questions. The problem is that it gets applied to questions it cannot answer, by teams who are not equipped to run it properly, and then treated as validation rather than exploration.

If you want to understand the full landscape of qualitative and quantitative research methods available to marketing teams, the Market Research and Competitive Intelligence hub covers the range of tools and approaches worth knowing.

What a Focus Group Discussion Is Actually Measuring

The core output of a focus group is language and attitude. You are listening to how people talk about a category, a brand, a product, or a problem. You are watching which words they reach for, which emotions surface, and where their reasoning becomes inconsistent or uncertain.

This is genuinely valuable. When I was working with a financial services client on a repositioning project, we ran a series of focus groups not to test the new positioning but to understand the vocabulary customers used to describe their relationship with money. The language in those transcripts shaped the copy, the tone, and the channel strategy in ways that no survey data could have. People do not describe anxiety about debt in the same clinical language that appears in a multiple-choice questionnaire. They use metaphors. They hedge. They contradict themselves. That texture matters.

What focus groups are not measuring is what people will actually do. Stated preference and revealed preference are different things, and the gap between them is often wide. When someone says they would pay more for a sustainable product, they may mean it in the room. At the point of purchase, with a price differential in front of them, the behaviour frequently tells a different story. This is not dishonesty. It is the gap between how people think about themselves and how they act under real conditions.

The Forrester piece on best practices makes a related point worth reading: what works in controlled conditions does not always transfer to real-world behaviour. That principle applies directly to focus group findings.

The Group Dynamic Problem

There is a structural flaw in the focus group format that no amount of good moderation fully eliminates: people change what they say when other people are listening.

Social desirability bias is well documented in research methodology. Participants tend to express views they believe will be well received by the group, and they tend to converge toward the most confidently expressed opinion in the room. In a group of eight people, one assertive participant with a strong view can pull the discussion in a direction that does not reflect the majority’s actual position. By the end of the session, three or four people have softened their dissenting views, and the moderator’s notes suggest a consensus that never really existed.

I have watched this happen in real time. A participant who described a brand as “a bit corporate and cold” in the pre-session screener spent the group agreeing with a more vocal participant who loved the brand’s reliability and professionalism. The screener response was more honest. The group response was more socially comfortable.

There are techniques that reduce this effect. Asking participants to write down their views before the group discussion begins captures individual responses before the social dynamic takes hold. Paired depth interviews remove the group pressure entirely. Online asynchronous focus group formats, where participants respond independently before seeing others’ contributions, reduce conformity significantly. None of these are perfect, but they are worth considering when the research question depends on genuine individual attitudes rather than group consensus.

How to Write a Discussion Guide That Actually Works

The discussion guide is the document most teams spend the least time on and the one that most determines the quality of the output. A weak discussion guide produces a polite conversation. A strong one produces insight.

The most common mistake is writing a guide that mirrors a survey. Questions like “How satisfied are you with the product?” or “Would you recommend this brand to a friend?” are closed, evaluative, and produce answers that tell you nothing you could not have got from a Net Promoter Score. The value of a focus group is in the open, exploratory territory. The guide should push into areas where you do not already know the answer.

A good discussion guide structure for a sixty to ninety minute session typically follows this shape. The first fifteen minutes are warm-up: broad, low-stakes questions about the category that get participants comfortable talking. The next thirty to forty minutes are the core exploration: open questions about behaviour, experience, and attitude, with probing questions built in to follow unexpected threads. The final fifteen to twenty minutes are stimulus response: showing a concept, a piece of creative, or a proposition and exploring reactions in depth.

Probing questions are where most moderators underperform. “Can you tell me more about that?” is a probe. “Why do you say that?” is a probe. “You mentioned X earlier, and now you’re saying Y. How do those two things sit together for you?” is a probe that a skilled moderator uses to surface genuine tension in someone’s thinking. Build these into the guide explicitly, not as an afterthought.

The guide should also specify what the moderator should not do. Do not lead with the brand name if you want unprimed reactions. Do not introduce the concept until the warm-up has established baseline attitudes. Do not let one participant dominate without a structured re-engagement of the rest of the group.

Recruitment: The Variable That Kills More Focus Groups Than Bad Moderation

I have seen more research projects undermined by poor recruitment than by any other single factor. If the people in the room do not represent the audience you are trying to understand, nothing else matters. The insights will be real, but they will be insights about the wrong people.

The typical failure mode is recruiting people who are easy to find rather than people who are right for the brief. Panel providers offer fast turnaround and low cost, but the people who join research panels tend to be more engaged with research, more articulate about their opinions, and more comfortable in group settings than the average consumer. They are also, in many categories, not typical buyers. A panel-recruited group for a low-involvement FMCG product will frequently over-represent people who think carefully about their purchases in that category, which is not how most people in that category actually behave.

For B2B research, the recruitment problem is more acute. The people who agree to participate in focus groups are rarely the senior decision-makers you actually need. They are often more junior, more available, and less representative of the buying process. I have run B2B research where the recruited participants were nominally in the right job function but had no actual authority over the purchasing decision we were researching. The entire exercise was mapping the attitudes of people who would influence but not decide.

Proper screening is non-negotiable. Write a screener that tests actual behaviour, not just job title or demographic category. If you are researching attitudes toward a financial product, ask about recent purchase behaviour in the category. If you are researching B2B software decisions, ask specifically about their role in the last purchase of that type. Exclude anyone who has worked in marketing, research, or the relevant industry in the past three years. The cost of proper recruitment is real. The cost of recruiting the wrong people is higher.

Online Versus In-Person: What the Format Change Actually Affects

The shift toward online focus groups accelerated considerably over the past few years, and it has produced a genuine debate about whether the format change compromises the method. My view, having run both, is that it depends entirely on what you are trying to learn.

Online synchronous groups, where participants join a video call together, replicate the social dynamic of an in-person group fairly well. The moderator can read body language to some degree, participants can hear each other’s responses, and the conversation flows in a recognisable way. The practical advantages are significant: you can recruit nationally without travel costs, participants are in their own environments which can reduce social pressure, and the logistics are simpler. The disadvantages are that technical issues disrupt flow, participants are more easily distracted, and the physical cues that a skilled moderator reads in person are reduced.

Online asynchronous formats, where participants respond to questions over a period of hours or days without seeing each other’s responses until later, are genuinely different in character. They reduce social conformity significantly. They allow participants time to reflect before responding, which can produce more considered answers. They work particularly well for sensitive topics where participants would self-censor in a live group. The trade-off is that you lose the spontaneous conversational dynamic that sometimes produces the most revealing moments in a live session.

For concept testing and proposition development, I tend to prefer online asynchronous for the first round and a live group for follow-up once you have a clearer picture of the territory. For language and emotional texture research, live groups, whether in-person or synchronous online, produce richer material.

Synthesis: Where Most of the Value Is Lost

The debrief and synthesis stage is where focus group research most consistently fails to deliver its potential. Teams sit through the sessions, gather pages of notes and hours of recordings, and then produce a report that summarises what people said. That is not analysis. That is transcription with formatting.

Genuine synthesis means identifying patterns across groups, noting where the stated attitude diverges from the emotional response, flagging where different segments of your audience are in fundamentally different places, and translating all of that into implications for the brief, the strategy, or the creative direction. It means being willing to write down findings that contradict the client’s assumptions, including findings that suggest the research question was the wrong one to begin with.

One of the most useful things I learned from judging the Effie Awards is how rarely the research behind effective campaigns is presented as a neat confirmation of the original hypothesis. The cases that win tend to show a moment where the team found something unexpected and had the discipline to follow it rather than explain it away. That kind of finding almost always comes from qualitative research, and it almost always gets buried in the debrief unless someone is specifically looking for it.

Build the synthesis process into the project plan before the groups run. Decide in advance who is responsible for analysis, what the output format will be, and how findings will connect to the next decision. If the focus groups are feeding into a campaign brief, the synthesis document should speak directly to the brief’s questions. If they are feeding into a positioning exercise, the output should map directly to the positioning framework. Research that does not connect to a decision is expensive opinion collection.

The keyword gap analysis approach outlined by Moz is a useful analogy here: the value is not in the raw data, it is in identifying the specific gaps that have strategic implications. The same principle applies to focus group synthesis. You are looking for the gaps between what people say, what they feel, and what they do, and then working out which of those gaps matter for the decision in front of you.

When to Use Focus Groups and When Not To

Focus groups are the right tool when you need to understand language, explore unfamiliar territory, or develop hypotheses that quantitative research can then test. They are not the right tool when you need to measure incidence, predict behaviour, or validate a decision that has already been made.

Use them early in a research programme, before you have written your survey, to understand the vocabulary and the territory. Use them when you are entering a new category or audience segment and do not yet know what questions to ask. Use them when your quantitative data has thrown up a result you do not understand and you need to explore what might be driving it. Use them when you are developing creative and you want to understand which emotional territory resonates before you invest in production.

Do not use them to decide whether to launch a product. Do not use them to set a price point. Do not use them to validate a strategy that has already been approved internally. In each of these cases, you are asking the method to answer a question it is not built to answer, and you will either get misleading data or spend money confirming what you already believed.

The broader point about research methodology, and about using the right tool for the right question, runs through most of what we cover in the Market Research and Competitive Intelligence section. The focus group sits alongside a range of other methods, and its value depends almost entirely on how it fits into the wider research design.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions