Skip to content
Prompt

AI prompts for focus groups: discussion guides, screening, and analysis

Ready-to-use AI prompts for planning focus groups, writing screening questionnaires, analyzing transcripts, and creating stimulus materials.

How to use

Copy and paste into your AI assistant chat

These prompts help you prepare for and analyze focus group research using an LLM. Each prompt includes placeholders in [brackets] — replace them with your project specifics before running. The prompts are designed for ChatGPT, Claude, or any general-purpose LLM.

Discussion guide creation

You are a senior qualitative researcher preparing a focus group study.

Context:
- Product/service: [describe your product or concept]
- Research objective: [what you want to learn from participants]
- Target participants: [who they are, how many per group]
- Session length: [90-120 minutes]
- Format: [in-person / remote via Zoom / hybrid]

Write a complete focus group discussion guide with:
1. Opening (5-10 min): ground rules script, consent reminder, icebreaker question related to the topic
2. Broad exploration (20-30 min): 3-4 open-ended questions about participants' current experiences and attitudes in the problem space, each with 2-3 follow-up probes
3. Focused discussion (30-40 min): 2-3 specific concepts or stimuli to react to, with instructions to write individual reactions before group discussion
4. Interactive exercise (15-20 min): a ranking, dot-voting, or card-sorting activity with clear instructions
5. Wrap-up (10 min): closing question, thank-you script, next-steps explanation

Rules:
- All questions must be open-ended
- Include moderator notes on when to ask for written responses before verbal discussion
- Mark approximate timing for each section
- Include specific probes for managing dominant speakers and engaging quiet participants

Screening questionnaire for recruitment

I need to recruit focus group participants.

Context:
- Product/concept being discussed: [describe]
- Research goal: [what you want to learn]
- Target users: [behavioral profile, not demographics]
- Number of groups: [how many], [how many per group]

Write a screening questionnaire (10-12 questions) that:
1. Filters for behavioral criteria relevant to the research goal
2. Includes at least two disqualifying questions (e.g., works in the industry, participated in a focus group in the last 6 months)
3. Does not reveal which answers qualify — avoid transparent criteria
4. Groups participants by attitude or experience level to compose balanced groups
5. Ends with availability, consent to be recorded, and confirmation of incentive

Cross-session thematic analysis

I conducted [N] focus groups about [topic]. Below are the transcripts or detailed notes from each session.

[Paste transcripts or session notes, clearly labeled Group 1, Group 2, etc.]

Analyze this data:
1. Identify 5-8 themes that appeared across multiple groups
2. For each theme: one-sentence insight, supporting quotes (labeled by group), and how many groups raised it
3. Note themes that appeared in only one group — are they outliers or early signals?
4. Highlight points where groups disagreed with each other and possible reasons
5. Flag instances of potential groupthink: rapid consensus with little debate, or agreement that contradicts individual written responses
6. Suggest 3-5 actionable recommendations based on the findings
7. Identify gaps where additional research (interviews, surveys, usability tests) is needed

Format each insight as: "Observation: [what emerged]. Implication: [what it means for the product]."

Stimulus material creation

I am running a focus group about [topic] and need stimulus materials to prompt discussion.

Context:
- Product/concept: [describe]
- What we want reactions to: [specific aspect — positioning, feature set, naming, pricing model]
- Participants: [who they are]

Create [3-4] distinct concept descriptions (each 80-120 words) that:
1. Present the same core idea with different framings, value propositions, or emphasis
2. Are written in plain language participants would use, not marketing copy
3. Are neutral in tone — none should be obviously "better" than the others
4. Include a brief instruction for the moderator on how to present them (all at once vs. one at a time, individual written reaction first vs. immediate group discussion)