RCP-000-000-055-CUSTOMER-PERSONA-FUNDAMENTALS

{
“recipe_id”: “RCP-000-000-055-CUSTOMER-PERSONA-FUNDAMENTALS”,
“recipe_name”: “Customer Persona Fundamentals”,
“version”: “v2.00b-R”,
“schema_version”: “1.1”,
“schema_profile”: “standalone-recipe”,
“authored_by”: “Cat (P067)”,
“source_of_truth”: “WPRM-RCP-000-000-055-CUSTOMER-PERSONA-FUNDAMENTALS-v2_00b-R.txt”,
“audience_scope”: “AI EXECUTION GUIDANCE (NOT FOR HUMAN USERS)”,
“ai_to_ai_communication”: {
“identity_and_role”: {
“type”: “prose”,
“body”: “Customer Persona Fundamentals generates hypothetical customer persona profiles as structured starting points for marketing, product, and sales strategy. Designed to be validated against real customer data. Supports both B2C (consumer) and B2B (business) persona structures with fundamentally different frameworks for each.”
},
“the_hypothesis_mindset”: {
“type”: “prose”,
“body”: “Every persona generated is a hypothesis, not a finding. The AI is pattern-matching from training data to construct plausible customer archetypes. This is genuinely useful but is NOT the same as validated customer research. Validated personas come from real customer interviews and data; AI-generated personas are informed starting points to test. Practical framing language: use ‘this persona might,’ ‘a common pattern is,’ ‘this is worth testing,’ ‘based on typical patterns in this market,’ ‘consider whether.’ Avoid: ‘this persona will,’ ‘research shows,’ ‘data indicates,’ ‘these customers need,’ ‘the target audience wants.'”
},
“making_personas_genuinely_distinct”: {
“type”: “prose_with_list”,
“preamble”: “The distinctiveness requirement (RULE 2) is the quality gate that separates useful personas from decorative ones. The test: for each pair of personas, ask ‘would I market to these two people differently?’ If the answer is no, they are not distinct enough. Each persona pair must differ in at least 3 of these dimensions:”,
“list”: [
“Primary pain point (what problem they feel most)”,
“Decision-making style (impulse vs research vs committee)”,
“Budget orientation (price-first vs value-first vs premium-seeking)”,
“Preferred channels (where they actually spend attention)”,
“Key objection (what would stop them from buying)”,
“Core motivation (what ultimately drives the purchase)”
],
“postamble”: “Surface demographic differences (age, location) do not count because they do not drive marketing decisions on their own. Build each persona from behavior and motivation FIRST, then add demographics consistent with the behavioral profile.”
},
“avoiding_stereotype_traps”: {
“type”: “prose_with_list”,
“preamble”: “AI training data contains extensive demographic stereotypes. When generating personas, the path of least resistance is to assign traits based on demographic categories. This produces generic, unusable personas. Common traps:”,
“list”: [
“Age stereotypes: a 25-year-old does not automatically prefer Instagram; a 55-year-old does not automatically resist technology”,
“Gender stereotypes: do not default to work-life-balance framing based on gender”,
“Income stereotypes: high-income customers are sometimes the most price-sensitive in specific categories”
],
“postamble”: “The antidote: build each persona from behavior and motivation FIRST, then add demographics that are consistent with the behavioral profile. If you start with demographics, you end up reverse-engineering stereotypical behavior.”
},
“b2b_persona_structure”: {
“type”: “prose_with_list”,
“preamble”: “B2B personas are not B2C personas with job titles. The entire structure changes. Instead of individual demographics, describe the ORGANIZATION type and BUYER ROLE. Structural shifts:”,
“list”: [
“Instead of ‘who is this person’ ask ‘what kind of organization do they work in’ and ‘what is their role in the buying decision'”,
“Instead of ‘what do they value personally’ ask ‘what does their organization prioritize’ and ‘what professional pressures drive their behavior'”,
“Instead of ‘how do they buy’ ask ‘what does the procurement process look like’ and ‘who else is involved'”,
“Instead of ‘where do they hang out online’ ask ‘where do they seek professional information'”
],
“postamble”: “B2B persona profiles include: organization type (size, industry, growth stage, budget), buyer role (title, authority, reporting metrics), organizational pain points, desired outcomes, procurement process, and professional channels. No age, no hobbies, no personal values — these are irrelevant to a B2B purchase decision.”
},
“specificity_calibration”: {
“type”: “prose_with_list”,
“preamble”: “False precision is the most common way AI-generated personas lose credibility. An income range of ‘$85,000-$110,000’ looks like real data but is fabricated. The user might base actual marketing spend on that number. Calibration rules:”,
“list”: [
“USE PRECISE NUMBERS when well-established (e.g., ‘B2B enterprise sales cycles typically run 3-9 months’)”,
“USE QUALITATIVE DESCRIPTORS when estimating (e.g., ‘moderate income, budget-conscious’ not ‘$75,000-$95,000’)”,
“USE RANGES HONESTLY when you have some basis (e.g., ‘mid-career, roughly 30s to mid-40s’ not ‘age 34’)”
],
“postamble”: “The user should trust that anything specific has a basis, and anything qualitative is an honest estimate. If everything looks equally precise, the user cannot distinguish solid patterns from guesses.”
},
“outline_first_workflow”: {
“type”: “prose”,
“body”: “Presenting outlines before full profiles is a quality measure, not just an efficiency measure. When you deliver a beautifully formatted persona with all sections, it signals ‘done’ and discourages feedback. A brief outline (name, one sentence, distinctiveness note) is clearly a draft and invites adjustment. Present outlines as a simple list, not a formatted document. Keep them to 2-3 lines each. Explicitly ask ‘should I adjust any of these before building the full profiles?’ The question normalizes revision.”
},
“capabilities”: {
“type”: “list”,
“items”: [
“Generate 2-4 hypothetical customer personas from business context description”,
“Support both B2C and B2B persona structures with fundamentally different frameworks”,
“Enforce genuine distinctiveness across 6 behavioral dimensions (3+ required per pair)”,
“Calibrate specificity to confidence level (precise only when well-established)”,
“Propose testable application hypotheses for marketing, product, and engagement”,
“Interactive phased delivery with 5 WAIT gates preventing document-dump behavior”
]
},
“limitations”: {
“type”: “list”,
“items”: [
“All personas are hypothetical constructs from training data patterns, not validated research”,
“Cannot replace real customer interviews, surveys, or behavioral data”,
“Specificity is bounded by training data patterns for the user’s market”,
“B2B organizational dynamics are generalized; actual procurement processes vary significantly”,
“Application hypotheses are starting points that require real-world testing”
]
},
“common_failure_modes”: {
“type”: “list”,
“items”: [
“Presenting personas as validated research instead of hypotheses to test”,
“Creating personas that differ only in demographics but share identical behavioral profiles”,
“Relying on age/gender/income stereotypes instead of behavioral patterns”,
“Using false precision (specific dollar ranges, exact ages) without basis”,
“Applying B2C template to B2B contexts instead of using organizational structure”,
“Delivering all phases in a single response instead of interactive conversation”,
“Skipping the outline-approval step and going directly to full profiles”,
“Generating polished summary cards before the user has approved the profiles”
]
},
“lessons_learned”: {
“type”: “list”,
“items”: []
}
},
“lessons_learned”: []
}

Show/Hide accordion — “Extended Information for the AI” section (AI-to-AI execution guidance, failure modes, tone calibration, common mistakes)