alexe-ev Develop Persona & Segment
Purpose
Help teams build research-grounded personas and segment definitions that are specific enough to guide product, design, and messaging decisions.
Skill type
Conceptual skill
Use this skill when
- The team doesn't have clear segments or has overly broad ones
- Personas exist but are based on assumptions, not research
- A new product area requires understanding a new user type
- Segment prioritization decisions need clearer user definitions
Do not use this skill when
- The segment is already well-defined and validated
- The goal is quantitative segmentation modeling (use data-analytics skills)
Required inputs
- Available research data or user observations
- Product area or context
Optional inputs
- Existing personas or segment definitions
- Behavioral data (usage patterns, feature adoption)
- Demographic or firmographic data (for B2B)
Upstream context
Works best when:
- Qualitative research has been synthesized
- At least some user interviews or feedback exist
If upstream context is missing
If no research data exists, produce a hypothesis persona clearly labeled as assumption-based. List what research is needed to validate it.
Downstream handoff
Output can feed:
- frame-insight-opportunity
- formulate-experiment-hypothesis (target segment)
- develop-positioning-messaging
Instructions
- Identify distinct user types emerging from research data.
- For each type, define: goals, jobs to be done, pain points, context of use.
- Add behavioral characteristics if data supports it.
- Avoid demographic stereotyping — anchor in behaviors and motivations.
- Assess which segments are most strategically important.
- Flag which elements are research-backed vs. assumed.
Output
For each persona / segment:
- Name and archetype
- Primary goals and jobs to be done
- Key pain points
- Context of use
- Behavioral patterns (if data supports)
- Strategic priority for the product
- Evidence quality: assumed / data-informed / validated
- Research gaps
Risks / caveats
- Do not build personas from demographic assumptions
- A persona based on one interview is a hypothesis, not a finding
- Avoid persona sprawl — 2–4 well-defined segments are better than 10 vague ones