Social Media / Viral Hooks
Generate 3-5 hook variations for the same content to test performance.
Why This Prompt Exists
Most creators test one hook — or none at all — and miss the best performer.
You get:
- one hook, one chance (no optimization)
- no data on what works for your audience
- same hook type every time (no variety)
- missed opportunities to improve CTR
- slow improvement from lack of testing
But A/B testing is not optional.
It is how you learn what your audience responds to.
- Variant A: curiosity hook
- Variant B: bold claim hook
- Variant C: relatability hook
- Variant D: question hook
- Variant E: pattern interrupt
Without A/B testing, you don’t know what works.
This framework forces AI to generate testable hook variations.
The Prompt
Assume the role of a hook testing specialist who creates A/B test variations. Your task is to generate hook variations for A/B testing. Generate: 1. VARIANT A — CURIOSITY HOOK - Opens a loop 2. VARIANT B — BOLD CLAIM HOOK - Surprising or impressive 3. VARIANT C — RELATABILITY HOOK - Names a specific pain point 4. VARIANT D — QUESTION HOOK - Engages the viewer 5. VARIANT E — PATTERN INTERRUPT (optional) - Unexpected statement PLUS: - TESTING RECOMMENDATIONS - Which 2-3 hooks to test first - Sample size needed - Test duration - WINNING CRITERIA - What metric determines the winner (CTR, retention, views) - HOW TO MEASURE - Platform-specific testing methods INPUTS: Content Topic: [WHAT IS THE VIDEO OR POST ABOUT?] Target Audience: [WHO ARE THEY?] Platform: [TIKTOK / INSTAGRAM / YOUTUBE / TWITTER/X] Goal: [CLICKS / VIEWS / RETENTION / COMMENTS] Brand Voice: [EDUCATIONAL / ENTERTAINING / BOLD / RELATABLE] RULES: - Each variant must be a different hook type - Keep all other variables the same (content, thumbnail, time) - Test 2-3 variants at a time (statistical significance) - Test duration: until you have enough data (not time-based) - Winning criteria: clear metric (not "feels better") - Document winners for future use
How To Use It
- Each variant must be a different hook type (curiosity, bold claim, relatability, question).
- Keep all other variables the same — content, thumbnail, posting time.
- Test 2-3 variants at a time — more than that and you won’t reach statistical significance.
- Test duration: until you have enough data, not a set number of days.
- Winning criteria: clear metric — CTR, retention, or views.
- Document winners in your hook bank for future use.
Example Input
Content Topic: How to raise freelance rates without losing clients
Target Audience: Freelancers earning $30-80/hour
Platform: TIKTOK
Goal: RETENTION (viewers watching past 3 seconds)
Brand Voice: EDUCATIONAL
Why It Works
Most creators don’t test hooks.
This framework improves outcomes by forcing:
- multiple hook types (variety)
- testing recommendations (methodology)
- winning criteria (clarity)
- measurement guidance (execution)
- documentation (learning)
Great hook testing doesn’t guess — it measures, learns, and improves.
Build Better AI Systems
Subscribe for advanced prompt engineering, AI social media tools, viral hook frameworks, and practical strategies for creators and marketers.