Claude Customer Persona Builder Prompt
Build detailed, research-based customer personas that actually guide marketing and product decisions.
Category
📣 Marketing
Difficulty
Intermediate
Models
3
Last Updated
2026-06-28
Works with
📄 Example output
⚠️ Common Mistakes
❓ FAQ
⚙️ Fill in your variables
📋 Prompt
You are a user research specialist who has built customer personas used to guide $10M+ marketing budgets.
Business type: [business type — SaaS/e-commerce/service/marketplace]
Product/Service: [product or service]
Existing customer insight: [existing insight — describe your best customers or paste anonymised customer feedback]
Number of personas: [number of personas — 1, 2, or 3]
Task: Build [number of personas] detailed customer persona(s):
For EACH persona:
IDENTITY:
- Persona name and archetype (e.g., 'The Pragmatic Manager')
- Demographics: age range, role, seniority, company size
- Quote that captures their worldview in one sentence
SITUATION:
- Day-to-day role: what they actually do each day
- What they're responsible for and judged on
- Tools and workflows they currently use
PAINS & FRUSTRATIONS:
- Top 3 professional pains (specific, not generic)
- What they've tried before that didn't work
- The language they use to describe their problem
GOALS & MOTIVATIONS:
- Professional goal they're trying to achieve
- Personal motivation underneath the professional goal
- What success looks like to them in 12 months
BUYING BEHAVIOUR:
- How they discover solutions (channels)
- What triggers a purchase decision
- Who else is involved in the decision
- Deal-breakers that kill a purchase
MESSAGING:
- Headline that speaks directly to this persona
- Value proposition in their language
- Channels and content types that reach them
Business type: [business type — SaaS/e-commerce/service/marketplace]
Product/Service: [product or service]
Existing customer insight: [existing insight — describe your best customers or paste anonymised customer feedback]
Number of personas: [number of personas — 1, 2, or 3]
Task: Build [number of personas] detailed customer persona(s):
For EACH persona:
IDENTITY:
- Persona name and archetype (e.g., 'The Pragmatic Manager')
- Demographics: age range, role, seniority, company size
- Quote that captures their worldview in one sentence
SITUATION:
- Day-to-day role: what they actually do each day
- What they're responsible for and judged on
- Tools and workflows they currently use
PAINS & FRUSTRATIONS:
- Top 3 professional pains (specific, not generic)
- What they've tried before that didn't work
- The language they use to describe their problem
GOALS & MOTIVATIONS:
- Professional goal they're trying to achieve
- Personal motivation underneath the professional goal
- What success looks like to them in 12 months
BUYING BEHAVIOUR:
- How they discover solutions (channels)
- What triggers a purchase decision
- Who else is involved in the decision
- Deal-breakers that kill a purchase
MESSAGING:
- Headline that speaks directly to this persona
- Value proposition in their language
- Channels and content types that reach them
PERSONA 1: 'The Pragmatic Marketer'
IDENTITY:
Name: The Pragmatic Marketer | Age: 28–42 | Role: Content Manager / Head of Marketing | Company: B2B SaaS, 10–200 employees
Quote: 'I don't need AI to replace my team — I need it to make my team 3× faster.'
SITUATION:
Day-to-day: Oversees content calendar, briefs writers, reviews copy, manages social and email. Spends 30–40% of time in editing mode.
Judged on: organic traffic, lead quality from content, email open rates, and 'did the content feel on-brand'
Tools: HubSpot, Google Docs, Notion, ChatGPT (used inconsistently)
PAINS:
1. AI output is 'almost good but not quite' — requires more editing than starting from scratch sometimes
2. No reliable prompt system — team uses AI differently every time, produces wildly inconsistent quality
3. Can't trust AI with SEO content — has been burned by inaccurate facts and keyword-stuffed output
BUYING BEHAVIOUR:
Discovers: Twitter/X, newsletter recommendations, specific searches ('ChatGPT prompts for SEO', 'how to write SEO content with AI')
Trigger: When a piece of AI content embarrasses her or when she sees a competitor's content and wonders 'how did they do that so well?'
Deal-breaker: Any paid tool that requires another monthly subscription
MESSAGING:
Headline: 'AI prompts that your whole team can use — and trust'
Value prop: '200+ tested prompts for every content task. Each one shows you the output before you commit to it.'
IDENTITY:
Name: The Pragmatic Marketer | Age: 28–42 | Role: Content Manager / Head of Marketing | Company: B2B SaaS, 10–200 employees
Quote: 'I don't need AI to replace my team — I need it to make my team 3× faster.'
SITUATION:
Day-to-day: Oversees content calendar, briefs writers, reviews copy, manages social and email. Spends 30–40% of time in editing mode.
Judged on: organic traffic, lead quality from content, email open rates, and 'did the content feel on-brand'
Tools: HubSpot, Google Docs, Notion, ChatGPT (used inconsistently)
PAINS:
1. AI output is 'almost good but not quite' — requires more editing than starting from scratch sometimes
2. No reliable prompt system — team uses AI differently every time, produces wildly inconsistent quality
3. Can't trust AI with SEO content — has been burned by inaccurate facts and keyword-stuffed output
BUYING BEHAVIOUR:
Discovers: Twitter/X, newsletter recommendations, specific searches ('ChatGPT prompts for SEO', 'how to write SEO content with AI')
Trigger: When a piece of AI content embarrasses her or when she sees a competitor's content and wonders 'how did they do that so well?'
Deal-breaker: Any paid tool that requires another monthly subscription
MESSAGING:
Headline: 'AI prompts that your whole team can use — and trust'
Value prop: '200+ tested prompts for every content task. Each one shows you the output before you commit to it.'
🏆
💡 Pro Tips
Best model for this prompt
Claude
Claude (Opus 4 / Sonnet 4)
Base personas on real customer data whenever possible — internal assumptions about customers are often wrong in surprising ways
The 'quote that captures their worldview' is the single most useful line in a persona — if team members can cite it from memory, the persona is working
Pain points should use the customer's exact language, not your product's language — 'I can't get consistent AI output' not 'they need better prompt engineering'
Revisit personas every 6–12 months — your customer base changes as your product and market evolve
Too many personas — 2–3 is the maximum before personas stop driving decisions and start creating arguments
Generic demographics ('35-55 years old, college educated') that don't differentiate between your best and worst customers
No buying behaviour section — knowing who the customer is without knowing how they buy leads to wasted marketing spend
Building personas in isolation without validating them against real customer conversations
- How many customer personas should I have?2–3 is the practical maximum. With more than 3, teams argue about which persona a decision serves rather than making the decision. If you genuinely have 5 different customer types, serve 2 well before expanding.
- How is a persona different from an ICP?ICP (Ideal Customer Profile) describes the type of company or account. Persona describes the specific human within that company who makes or influences the buying decision. For B2B, you need both.
- What data should I use to build personas?In priority order: (1) 5-10 interviews with actual customers, (2) customer support ticket analysis, (3) sales call recordings, (4) NPS survey open-ended responses, (5) analytics and behavioural data. Avoid building personas from assumptions alone.
- Can AI generate accurate customer personas?AI generates the structure and prompts the right questions, but the specific details must come from real customer data. An AI-generated persona based on 'SaaS product for marketers' is too generic to drive decisions. Feed it real quotes, support tickets, and feedback for meaningful output.