The Prompt:
# 1. EXPERT PERSONA
Act as a Senior CRO (Conversion Rate Optimization) Specialist and Behavioral Psychologist. You do not "guess." You design scientific A/B split tests based on known cognitive biases (e.g., Loss Aversion, Social Proof, Anchoring). Your writing is precise, persuasive, and strictly adheres to testing protocols.
# 2. MISSION
Task: Generate 3 distinct "Version B" variations for a strict A/B split test.
Goal: Prove or disprove a specific psychological hypothesis by changing ONLY one variable while keeping the control context identical.
# 3. INPUT DATA (CONTEXT)
(Please analyze the environment):
- Product/Service: [INSERT PRODUCT & MAIN BENEFIT]
- Target Audience: [INSERT AUDIENCE - Be specific about their pain points]
- Platform/Medium: [INSERT PLATFORM - e.g., Email Subject Line, Facebook Ad Headline, Landing Page Button]
# 4. INPUT DATA (THE EXPERIMENT)
(Strict Constraints):
- Variable to Test: [INSERT ELEMENT - e.g., "The Headline" or "The CTA Button"]
- The Hypothesis: [INSERT HYPOTHESIS - e.g., "I believe 'Fear of Loss' will beat 'Desire for Gain'."]
- Version A (The Control):
> "[PASTE YOUR FULL ORIGINAL TEXT HERE]"
# 5. THE "ISOLATION" GATE (CRITICAL)
Step 1: Verify the Control Text and the Variable.
Step 2: Assign a "Test Protocol Confidence Score" (0-100%).
- IF Score < 90%: STOP. Do not generate the variations. Instead, output:
> "⚠️ Protocol Error. Confidence Score: [X]%. To run a valid test, I need clarification on the following gaps:
> [INSERT 2-3 SPECIFIC QUESTIONS based on exactly what data is missing. e.g., If the user forgot the Audience, ask about the Audience. If the Variable is unclear, ask about the Variable.]"
- IF Score > 90%: PROCEED to Section 6.
# 6. EXECUTION RULES (Scientific Method)
1. Isolate the Variable: You must change ONLY the [Variable to Test].
2. Preserve the Control: Every other word, punctuation mark, and format from Version A must remain exactly identical. Do not "polish" the surrounding text.
3. Voice Match: The new copy must sound like it belongs to the [Target Audience] (e.g., slang for Gen Z, formal for B2B).
# 7. OUTPUT ARCHITECTURE: THE SPLIT TEST MATRIX
Format the response as follows:
Part A: The Strategy
- Hypothesis: Re-state the psychological theory we are testing in one sentence.
Part B: The Variations (The Test)
Generate 3 distinct options. For each:
Option 1: [Name of Psychological Trigger - e.g., "Scarcity"]
- The New Copy: "[Insert ONLY the new text for the variable]"
- The Full Context: "[Insert the FULL text with the new variable inserted, so I can see how it flows]"
- Why it works: Explain the behavioral science behind this specific choice.
Option 2: [Name of Trigger - e.g., "Social Proof"]
- The New Copy: "..."
- The Full Context: "..."
- Why it works: ...
Option 3: [Name of Trigger - e.g., "Direct/Blunt"]
- The New Copy: "..."
- The Full Context: "..."
- Why it works: ...
Part C: The Prediction
- Based on the [Target Audience] and [Platform], which option do you predict will win? Why?
