Persona Development
You are building user personas grounded in real user behavior and JTBD thinking — not demographic templates. A persona is useful only if it changes what you build or how you communicate. If a persona doesn't make a decision obvious, it's not sharp enough.
Frameworks: Bob Moesta / JTBD (demand-side thinking), Hilary Gridley (AI-embracer vs. skeptic segmentation), Lean Startup (build-measure-learn).
Step 1 — Load Context
Read memory/user-profile.md for product context and any existing persona notes. Read context/product/personas.md if it exists — understand what's already there and whether it needs updating or creating from scratch.
If research data exists (interview notes, support tickets, survey responses), use it. If not, build a research-grounded hypothesis persona that can be validated.
Step 2 — Persona Foundation
A good persona answers these questions from the user's perspective:
Who are they? (Role, situation, context — not demographics) What are they trying to do? (The JTBD — what progress are they making?) What's stopping them? (Pain, friction, workaround) What does success look like for them? (Desired outcome — functional, emotional, social) How do they currently handle this? (Current hire — the status quo) What would make them switch? (Switch trigger) How do they feel about AI assistance? (Embracer, neutral, or skeptic)
Step 3 — JTBD-Grounded Persona
Write each persona with a demand-side framing:
Triggering situation: "When [specific situation arises], this persona needs to [make progress]."
This is more actionable than "They are 32 years old and work at a startup." The triggering situation tells you when the persona needs the product, not just who they are.
Step 4 — AI Embracer vs. Skeptic Segmentation
Based on Hilary Gridley's segmentation: "The most meaningful segmentation for AI products is attitudinal: AI embracers vs. AI skeptics."
For each persona, identify their position:
AI Embracer:
- Excited about AI assistance; self-directed; tolerates setup friction
- Onboarding: jump straight to advanced features; show full capability surface
- Messaging: "What would you build if you had a senior PM teammate?"
AI Skeptic / Cautious:
- Trusts proven frameworks; fears losing craft or voice; needs to see before committing
- Onboarding: lead with framework credibility; show sample outputs first; emphasize they stay in control
- Messaging: "PM Copilot uses Teresa Torres and Marty Cagan's frameworks — not guesswork."
AI Neutral:
- Pragmatic; will use AI if it saves time; skeptical of hype
- Onboarding: show a concrete before/after; quantify time saved
- Messaging: "Run /write-prd and get a complete PRD in 10 minutes."
Step 5 — Persona Template
For each persona, fill in:
Name: [Descriptive label — e.g. "The Founding PM" not "Sarah"] Triggering situation: [When they need this product] JTBD: [What progress they're trying to make — functional, emotional, social] Current hire: [What they use today; what's wrong with it] Switch trigger: [What would make them look for something different] Pains: [3 specific frustrations with the current solution] Gains: [3 specific outcomes a new solution would enable] AI stance: [Embracer / Neutral / Skeptic] Onboarding path: [How to onboard this persona given their AI stance] Representative quote: [A real or composite quote that captures their core frustration] What makes them a bad fit: [Who this product isn't for — prevents over-targeting]
Step 6 — Anti-Persona
Write one anti-persona: the user who looks like a fit but isn't. This prevents wasted sales and support effort, and helps the team say no to feature requests from this segment.
Step 7 — Output
Produce:
- 2–3 personas (primary, secondary, and anti-persona)
- Each persona filled in using the template above
- A 1-page "user cheat sheet" that captures the essential differences between the personas for team use
Offer to save to context/product/personas.md and update memory/user-profile.md with any new persona insights.
