Back to PM Skills
Research & Discovery

User Persona Generator

Create data-backed user personas that go beyond demographics to capture behaviors, motivations, and jobs-to-be-done.

# Drop into ~/.claude/skills/user-personas/
curl -L https://github.com/sunnyyang-hicks/pm-skills-for-claude/raw/main/user-personas/SKILL.md \
  -o ~/.claude/skills/user-personas/SKILL.md

Overview

A useful persona isn't a stock photo with a fake name and hobbies. It's a behavioral model that helps the team make better product decisions by asking "what would [persona] do?" Personas work when they're rooted in real data and capture what drives behavior.

Before You Start

Ask the user:

  1. What data do we have? — Interview transcripts, analytics segments, survey data, support patterns, sales notes.
  2. How many personas? — 3-5 is the sweet spot. More than 5 means the team will ignore most of them. Fewer than 3 means you're probably missing a segment.
  3. What decisions will these inform? — Feature prioritization, onboarding flows, marketing messaging, pricing tiers.
  4. Current assumptions — Who does the team think the users are? This helps identify blind spots.

Persona Structure

# [Persona Name] — [One-line descriptor]

> "[A quote that captures this persona's core motivation]"

## Overview
**Role/Title:** [professional context]
**Experience level:** [with the product/domain]
**Segment size:** [% of user base, if known]
**Revenue contribution:** [% of revenue, if known]

## Behavioral Profile

### Goals (what they're trying to achieve)
1. [Primary goal — the job they're hiring your product for]
2. [Secondary goal]
3. [Tertiary goal]

### Frustrations (what gets in their way)
1. [Primary pain point — with evidence]
2. [Secondary pain point]
3. [What they complain about but isn't actually the real problem]

### Behaviors (what they actually do)
- **Frequency:** [How often they use the product]
- **Key workflows:** [Top 3 things they do]
- **Workarounds:** [Hacks they've built around product limitations]
- **Tools in ecosystem:** [What else they use alongside your product]
- **Decision process:** [How they evaluate and adopt new features]

### Motivations (why they care)

**Functional:** [Practical outcome they need]
**Emotional:** [How they want to feel]
**Social:** [How they want to be perceived]

## Context & Triggers

**When do they reach for your product?**
- [Trigger moment 1] — [situation that creates the need]
- [Trigger moment 2]

**When do they NOT use your product (but could)?**
- [Missed opportunity 1] — [what they use instead]

## Success Metrics (from their perspective)
- [How THEY measure whether your product is working for them]
- [Not your metrics — their metrics]

## Relationship with Product
- **Acquisition channel:** [How they typically find you]
- **Aha moment:** [When they first see value]
- **Retention driver:** [Why they keep coming back]
- **Churn risk:** [What would make them leave]

## Design Implications
- [Specific product/design decisions this persona drives]
- [Features to prioritize FOR this persona]
- [Features to de-prioritize — they don't care about these]

## Anti-patterns
- [Common mistake teams make when designing for this persona]

Quality Checks

Before presenting personas, verify:

  • Behavioral, not demographic. Age, gender, and location should only appear if they genuinely drive different product behavior. "35-year-old in San Francisco" tells you nothing useful. "Power user who manages 50+ projects and needs keyboard shortcuts" tells you everything.
  • Evidence-backed. Every claim should trace to data. If you can't cite evidence, mark it as an assumption that needs validation.
  • Distinct. If two personas would make the same product decisions, merge them. The test: would the team build something different for Persona A vs. Persona B?
  • Actionable. Each persona should drive at least one product decision. If a persona doesn't change any decisions, it's not useful.

Output

Save as PERSONAS-[product-name].md. Create one document with all personas for easy comparison. Include a summary comparison table at the top.