Marketing Hooks - Puzzle-Driven Content Model
Create attention-grabbing marketing content using proven psychological frameworks.
Core Principle: Questions Drive Engagement
Good content is driven by good questions. Good questions are puzzles.
When your opening creates a puzzle, the audience's brain automatically tries to solve it. This creates engagement and retention.
Four Question Types
1. How-Type (Method)
Pattern: "How to [achieve desired outcome]?"
Examples:
- "How to become wealthy quickly?"
- "From white-label to brand?"
- "How to start cross-border e-commerce?"
- "How to 10x your productivity with AI?"
Best for: Tutorials, guides, practical content
2. Why-Type (Reason)
Pattern: "Why [surprising fact/outcome]?"
Examples:
- "Why did they become CEO?"
- "Why 90% of AI tools fail?"
- "Why top creators don't post daily?"
Best for: Analysis, insights, thought leadership
3. What-Type (Definition)
Pattern: "What is [concept/term]?"
Examples:
- "What is supply-side reform?"
- "What is CPS?"
- "What is Product-Market Fit?"
Best for: Educational content, explainers
4. WoW-Type (Extreme Case) ⭐ HIGHEST ENGAGEMENT
Pattern: Create probability-defying extreme cases
Core insight: WoW = Manufacturing extremely low-probability extreme cases
Formula:
- Discover extremes → Get traffic
- Become extreme → Build brand
Examples:
- "Gen-Z makes $100K/month with AI"
- "10K followers in 3 days - the secret"
- "I replaced my entire team with one tool"
- "This 19-year-old built a $1M business with ChatGPT"
Why it works: Brain automatically fills in missing logic, creating engagement stickiness.
Visual/Info/Case combinations:
- Extreme visual + surprising data
- Impossible timeline + real results
- Tiny input + massive output
Content Tree Structure
Every piece of content should follow this hierarchy:
Question (Root)
↓
Viewpoint (Trunk)
↓
Reasoning (Branches)
↓
Evidence (Leaves)
Gan Method Content Model
Structure:
- Core viewpoint (Trunk)
- Sub-arguments (Branches)
- Evidence (Leaves) - Must be:
- Unexpected
- Memorable
- Vivid
Quality check: A good 15-min video or well-evidenced article = a tree with lush branches and leaves.
Platform-Specific Applications
Xiaohongshu (Little Red Book)
Best formats:
- WoW-type titles: "Gen-Z AI tools $100K/month"
- How-type for tutorials: "5 AI tools for 10x efficiency"
Structure:
- Hook (5-10s): WoW statement
- Pain point (10-15s): "Are you also facing..."
- Solution (120-180s): Step-by-step with demos
- Results (20-30s): Proof
- CTA (5-10s): "Follow for more"
X (Twitter)
Best formats:
- Start with puzzle: "Do you know why 90% of people fail with AI tools?"
- Then reveal answer in thread
Structure:
- Tweet 1: Puzzle/Hook
- Tweet 2-4: Reasoning + Evidence
- Final tweet: Conclusion + CTA
Newsletter
Best formats:
- Question (title) → Viewpoint (intro) → Reasoning (body) → Evidence (examples)
Structure:
- Subject line: WoW or Why-type
- Opening: Establish puzzle
- Body: Tree structure (trunk → branches → leaves)
- Closing: Actionable takeaway
Hook Generation Workflow
When asked to create hooks:
-
Identify the core message
- What's the main point?
- What outcome does the audience want?
-
Choose question type
- Practical guide? → How
- Surprising insight? → Why
- New concept? → What
- Extreme case? → WoW (prioritize this!)
-
Apply WoW amplification
- Can you make it more extreme?
- Can you add surprising numbers?
- Can you create contrast (small input → big output)?
-
Test the puzzle
- Does it make you think "why?" or "how?"
- Does it create information gap?
- Would you click it?
Examples Library
See EXAMPLES.md for detailed case studies and templates.
Quality Checklist
Before finalizing content:
- Does the opening create a puzzle?
- Is there a clear tree structure (root → trunk → branches → leaves)?
- Are the evidence points unexpected, memorable, and vivid?
- For WoW-type: Is the case extreme enough?
- Does it make the brain want to "fill in the gaps"?
References
- "Writing is a Craft" - "Questions are the engine of articles"
- Gan Method - Content tree structure
- WoW Psychology - Extreme case engagement mechanics