These prompts support the key analytical steps of mental model mapping — the stages where AI can save the most time without replacing the researcher’s interpretive judgment. Use them after you have conducted your listening sessions and have transcripts ready.
Extract behaviors from a listening session transcript
You are helping a UX researcher build a mental model diagram using Indi Young's method.
I will provide a transcript from a listening session about [domain, e.g., "how people manage their personal finances"]. Your job is to extract every distinct behavior the participant describes.
A "behavior" falls into one of three types:
- TASK: something the person does (action, activity, step)
- PHILOSOPHY: a belief, guiding principle, or value that shapes their decisions
- FEELING: an emotional reaction to a situation
For each behavior, provide:
1. A short present-tense phrase written from the participant's perspective (e.g., "compares prices at three stores before buying")
2. The type: TASK, PHILOSOPHY, or FEELING
3. The verbatim quote from the transcript that supports it
Format as a table with columns: Behavior | Type | Supporting Quote
Rules:
- Extract every behavior, even minor ones. Err on the side of including too many.
- Do not infer behaviors the participant did not describe. Only extract what is explicitly stated.
- Write behaviors in the participant's language, not in researcher jargon.
- If a statement is ambiguous between types, flag it as AMBIGUOUS and explain.
Transcript:
[paste transcript here]Suggest preliminary tower groupings
I have a list of [N] behaviors extracted from listening sessions about [domain]. I need to group them into towers for a mental model diagram.
A tower is a cluster of 5-15 related behaviors that describe the same narrow activity, concern, or area of thinking. Towers should be:
- Specific enough that each tower has a clear, descriptive label
- Broad enough to contain behaviors from multiple participants
- Organized around what people think and do, not around product features
Here are the behaviors:
[paste behavior list]
For each proposed tower:
1. Give it a descriptive name (2-5 words, in the participant's language)
2. List the behaviors that belong in it
3. Suggest which mental space (broader category) this tower might belong to
Also flag:
- Behaviors that could fit in multiple towers (note both options)
- Behaviors that do not seem to fit anywhere (these may indicate a missing tower)
- Towers that seem too broad and should be splitGenerate a gap analysis from diagram structure
I have completed the top half of a mental model diagram for [domain]. Below is the structure:
Mental Spaces and their Towers:
[list each mental space with its towers and approximate tower heights]
Product Features mapped to towers:
[list each feature and which tower(s) it supports]
Analyze the alignment between the mental model (top half) and the product (bottom half):
1. GAPS: Towers with no features beneath them. For each gap, explain what the unmet need might be and why it matters (consider tower height as an indicator of importance).
2. OVER-SERVED: Towers with multiple features beneath them. Note any potential redundancy.
3. MISALIGNED: Features that do not map to any tower — these may address needs users do not actually have.
4. PRIORITY OPPORTUNITIES: Rank the top 5 gaps by potential impact, considering tower height, which audience segments are affected, and strategic relevance to [company/product name].
Present findings as a structured report with evidence from the diagram.Synthesize cross-participant patterns
I have behavior inventories from [N] listening sessions about [domain]. Each inventory lists behaviors extracted from one participant.
Analyze the inventories for cross-participant patterns:
1. UNIVERSAL BEHAVIORS: Behaviors that appear (in similar form) across 70%+ of participants. These represent core aspects of how people think about this domain.
2. SEGMENT-SPECIFIC BEHAVIORS: Behaviors that appear frequently within one audience segment but rarely in others. Note which segment and hypothesize why.
3. SURPRISING OUTLIERS: Behaviors that only 1-2 participants described but that reveal an unexpected aspect of thinking about the domain.
4. EMOTIONAL HOTSPOTS: Areas where multiple participants described strong feelings (positive or negative). These often indicate high-stakes moments.
5. CONFLICTING PHILOSOPHIES: Cases where participants hold opposing beliefs about the same topic. These tensions are important for design — the product cannot assume one philosophy is universal.
For each pattern, cite the specific participants (by ID) and their verbatim quotes.
Participant inventories:
[paste inventories]