If you need to understand how customers perceive your product, message, or visual, a focus group remains one of the most accurate tools. It is not about statistics or hypotheses. It is about a real clash of opinions, formulations, resistances, and triggers in a live discussion. A well-conducted focus group gives what is impossible to get either through a survey or through analytics: immediate reactions, clarifications, group dynamics, and the client’s language in its purest form.
What Is a Focus Group in Marketing?
A focus group is a controlled discussion in which 6-10 people from your target audience speak out on a topic related to a product, brand, or marketing message. Unlike an interview, this is a group situation: participants respond not only to your questions but also to each other's opinions. This creates conditions for identifying contradictions, clarifications, and the social context of perception.
The goal of a focus group is not to record a percentage but to reveal arguments. Participants can discuss why they do not trust the brand, why the message sounds “salesy,” how they perceive the visual, and what exactly causes an emotional response. All this helps not to guess but to hear and observe.
Such groups are held in the early stages of launch, when changing positioning, before advertising campaigns, or during research into new segments. Real representatives of the target audience participate — those who already use the product or potentially can.
The session lasts from 60 to 90 minutes, is held offline or online, and is moderated by a specialist. The moderator's main task is not to influence but to provoke openness, clarify, and push for argumentation. After that, the team receives a video, transcription, and notes on key topics, quotes, and reactions.
A focus group is not a replacement for quantitative research, but a way to hear what numbers cannot show: language, associations, cultural barriers, and non-verbal reactions.
When and Why to Use Focus Groups
Focus groups are especially effective in situations where it is important to understand not what a person thinks but how they formulate it, with what words, and in what context. This is not a tool for measuring but for discovery - new insights, perceptions, and unexpected barriers. The ideal moment to use them is when you already have a hypothesis but are not sure how exactly it is perceived by the audience.
Here are the times when companies use focus groups:
- before launching a new feature or product
- when developing positioning or brand messaging
- when testing packaging, design, video, headlines
- to test reactions to controversial ideas or radical changes
- to identify barriers when entering a new market
- when reworking the customer journey
Example: if you want to change the tone of voice, a focus group can show which phrases are perceived as “artificial”, which sound “natural”, and which ones cause rejection. Or, for example, if you are testing a new value model, a group discussion will allow you to see which arguments sound convincing and which ones immediately repel.
A focus group also helps to identify social patterns. One person may not voice a doubt, but if he hears it from another, he will agree and develop the idea. Such dynamics form a more accurate picture of real objections and expectations.
It is important to understand that a focus group is not needed if you want to confirm a ready-made hypothesis. Its strength is in revealing the unknown. This is why it is often used before large-scale surveys or quantitative testing.
How to Structure and Conduct a Focus Group
The success of a focus group does not depend on the number of participants or the charisma of the moderator. It begins long before the session itself — with a clear understanding of why it is needed, who should speak, and what insights you want to get at the end. Without a structure, a focus group quickly turns into a superficial conversation that does not provide any benefit. Below is a structure that helps teams conduct focus groups consciously and effectively.
1. Define the goal and hypotheses
The wording “find out what users think” is not the goal. The goal sounds specific: “understand how the audience perceives the new message”, “identify associations with the brand after rebranding”, and “check whether the new price causes resistance”. Without a clear focus, it is difficult to build a scenario and useless to analyze the results.
2. Segment the participants
You don’t want “people in general”. You need representatives of a specific segment. For example, a focus group for early adopters and for customers who had a negative experience will give completely different signals. Participants can be selected based on demographics, behavior (for example, users who have made 3 or more orders), and status (active, left, interested but did not buy). A screener questionnaire is often used to exclude irrelevant ones.
3. Write a script - but not as a survey
A focus group script is not a list of questions. It is logical: from “warming up” to deep discussions. Start with simple topics (“Tell us what you use in this category”, “Which products do you think are successful?”), then move on to the main ones (“What do you feel when you see this message?”, “What associations does this design evoke?”), and at the end, leave room for open-ended reflection. A good practice is to use incentives: show a layout, a screenshot, text, a video.
4. Prepare a moderator
A good moderator is not someone who can speak but someone who can listen, clarify, and maintain focus. He should not “lead” participants to the desired opinion. His job is to spin, dig deeper, and involve everyone. It is important to control dominance: one participant should not crush the others. Rules are often used - do not interrupt, speak in turns, and join in the discussion.
5. Organize logistics and recording
Focus groups can be held offline (in research rooms) or online (via Zoom, Teams, Lookback). The main thing is to ensure high-quality audio and video recording so that analysis can be carried out later. A stenographer or recording with subsequent transcription is essential. Participants need to be told in advance how everything will be arranged, how long it will take, and what they will receive in gratitude (payment, voucher, bonus).
6. Record reactions, not just words
It is important not only what is said, but also how. A smile, a pause, tension in the voice — all of this provides context. Take notes during the session. Separately record phrases that sound like ready-made quotes: “This sounds like an intrusive advertisement,” “I would trust you if…,” “This reminds me of something from the nineties.”
7. Analysis and synthesis
After the session is over, the main work begins. Review the recording, highlight themes, and make a map of arguments: what was supported, what was disputed, where there was agreement, and where there were conflicts. AI tools can help decipher speech, cluster phrases, and identify emotional peaks. But the point is teamwork.
A focus group gives the maximum if each stage is structured in advance. It becomes ineffective when it is held “just in case” or with an unclear focus. If there is a structure, the result will be accurate, lively, and useful.
AI in Focus Group Analysis
Conducting a focus group is half the battle. The real work begins later: when you need to transcribe the speech, highlight repetitions, record the wording, track emotional signals, find moments of agreement and conflict. Previously, this took days. Today, with the help of AI, the team can get a structured picture within an hour after the end of the session.
First of all, AI is used for transcription and the extraction of key phrases. Modern models recognize speech with high accuracy, including fillers, pauses, and intonations. But most importantly, they can automatically classify phrases as emotional, as objections, as rational arguments or as suggestions for improvement. This makes it possible to go through the meaning quickly without having to manually review each fragment of the video.
Next is topic clustering. AI finds patterns: which ideas were repeated, how often similar formulations were heard, which of the participants spoke on similar topics. This is especially useful when conducting a series of focus groups. The model can show that, for example, in three sessions in a row, participants with a “Pro” subscription mention price as a barrier, while “Starter” users do not raise this topic at all.
The third level is emotional analysis. Algorithms determine the emotional response to what is said: irritation, enthusiasm, mistrust. They can do this based on speech, facial reactions, response speed, and intonation. Such data helps to understand which messages “stick” and which ones cause rejection - even if a person does not voice it directly.
For example, Unilever used AI analysis of faces and speech to evaluate reactions to commercials. Models determined where participants lose interest, where they smile, and where they show irritation. This gave precise signals about what to change in the creative to increase engagement.
AI also helps to find quotes with high value. Not all statements are equally useful. The model can select phrases that are formulated vividly and logically and can be used in presentations and brand manifestos or to illustrate insights. This saves time when preparing reports and makes conclusions more convincing.
It is important that AI does not replace an analyst. It speeds up the initial processing, provides structure, and helps not to get lost in the flow of information. But the final interpretation, semantic categorization, and conclusions are still the responsibility of the research or marketing team. The model will not understand irony, will not recognize cultural references, and will not distinguish agreement from a diplomatic excuse.
Tools that companies use:
– Hume AI — analysis of intonation, micro-emotions, and timbre of speech
– Grain — automatic video tagging, citation search, topic building
– Transana, Dovetail — for transcription, visualization of topics, and report building
Connecting AI allows you to move from “feelings after the group” to specific conclusions: what was said, how it was said, how many times it was repeated, and how it relates to key segments. This saves time and increases accuracy, especially in situations where the findings influence decisions by the product, brand, and marketing teams at the same time.
Real-World Examples of Focus Groups in Action
Focus groups are not an academic methodology but a practical tool that is actively used by both large brands and B2B teams. Below are real-world examples of companies that used focus groups to test new solutions, find insights, and rethink strategy.
1. Google: Redesigning the Logo and Visual System
When Google decided to unify the visual images of its services, including the Gmail, Calendar, and Meet icons, the team conducted a series of focus groups in different countries. Participants were shown old and new versions, asked about perception, and tested for recognition and emotional reactions. The result: many users were confused by icons that were too similar to each other. Google adjusted the palette and enhanced graphic differences to improve readability in the mobile interface.
2. Lego: Reconnecting with Teens
Lego faced a problem: Teenagers were losing interest in their product. To find out the reasons, the brand held a series of focus groups with teenagers and parents. Research showed that the problem was not in the product itself, but in the positioning - it was perceived as “childish” and “outdated”. This became the impetus for the creation of the Lego Technic line with a different tone, packaging, and communication. Direct quotes from teenagers in focus groups were used in briefs for advertising agencies.
3. Airbnb: Localizing Marketing in Japan
When entering the Japanese market, the Airbnb team held a series of focus groups to test advertising messages. Participants responded to visuals, headlines, and communication style. The study showed that familiar Western communication (trust, freedom, personal experience) was perceived as intrusive. Participants preferred wording that emphasized safety, structure, and social approval. These signals formed the basis of the entire local marketing strategy.
4. Dove: Emotional Brand Perception
Dove actively uses focus groups to assess the emotional response to its campaigns. During the preparation of the “Real Beauty” series, researchers held discussions with women of different ages, body types, and backgrounds. The participants shared stories, associations, and feelings associated with the image of “real beauty.” These discussions helped the brand build a campaign that later became iconic and brought Dove 700% ROI compared to previous campaigns.
5. IBM: Testing New Digital Products for the Enterprise Segment
Unlike B2C, focus groups also work in B2B, especially at the early concept stage. IBM uses so-called solution rooms: formats similar to focus groups, but with the involvement of clients in discussing new interfaces, features, and use cases. This helps to identify mistrust, difficulties in perception, and points of misunderstanding between the client’s departments. Such groups are held before the launch of pilots to avoid large-scale rework.
These cases show that focus groups are effective in a wide range of contexts, from design and packaging to brand platforms and product interfaces. Their power lies in the user’s language, in nuances and reactions that cannot be quantified. And that’s why companies keep returning to this format — with every new product, market, or message.
Best Practices for Running Effective Focus Groups
Focus grouping is not just about collecting opinions. It is a controlled process, where it is easy to lose focus, spoil the dynamics, or get superficial answers. Below is a set of practices that help to run focus groups effectively, regardless of the format and topic.
1. A clear goal - before the start
The most common mistake is to go to a focus group “to listen to what they say.” Without a clear goal and hypotheses, the session quickly turns into a conversation about nothing. Determine in advance what exactly you want to find out: how the message is perceived, whether the interface inspires confidence, where barriers arise in the scenario. This will form the basis of the scenario and provide specific criteria for analysis.
2. Participants must be precise, not just a “target audience”
If you are making a product for marketers from SMB companies, you should not invite “just people with marketing experience.” Clearly define the criteria: role, level, geography, experience using specific solutions. It is better to hold 2-3 groups with a precise hit than 5 sessions with a vague sample.
3. The script is flexible, but structured
You don’t need to write the script like a questionnaire. Build the logic of the conversation: first warm questions, then the main topics, and at the end - a summary. Include places where you will show materials, ask clarifying questions, provoke a discussion. And be sure to leave room for surprises - the most valuable insights do not arise according to the script.
4. Moderator ≠ interviewer
The moderator should not drag the session on himself or suppress disputes. His job is to involve everyone, direct the discussion, clarify and intensify when necessary. Knowing how to work with a dominant participant, raise the silent one, return the focus - this is a skill that is critically important. Often the best moderator is someone who is not from the product team at all.
5. No hidden hints or marketing tone
Phrases like “we want to make the product more convenient for you” or “look how quickly it works” kill objectivity. Participants start playing along. Better: “How would you describe this interface?”, “What’s the first thing that comes to mind when you see this?” All demonstrations and wording should be neutral.
6. Record, record, note behavior
Video is a must. But it’s even more important to take notes: what caused laughter, a pause, a gesture, a doubt. Ask the second observer from the team to record not what is said, but how it is said. Often, this is where the strongest signal is.
7. Don’t put off analysis
You need to start analyzing immediately after the session, while the details are fresh. Conduct a short debrief — what sounded strong, where there were outbursts, what phrases should be included in the report. Then — work with transcription, categorization of topics, collection of patterns. AI tools speed up the process here, but the team must make the semantic interpretation itself.
8. One group is not the answer
A focus group is about depth, not representativeness. So don’t draw conclusions based on one group. It’s better to have 2–3 groups with different segments or scenarios to see where the signals are repeated and where they depend on the context.
A focus group is not “conduct” and “write a report”. It’s an iterative process: collect, listen, analyze, show to the team, change. And the more subtle the preparation, the more accurate the result.
Conclusion
Focus groups help to hear what is not visible in numbers: the client's language, hidden objections, emotional perception, and nuances that determine the choice. They are used at all stages - from testing ideas to the final evaluation of the message. With a competent structure and accurate segmentation, a focus group turns into a powerful tool for the product, marketing, and brand. And the inclusion of AI speeds up the analysis and makes it more transparent for the entire team.
.svg)
