Prompt · Engineering
6 frameworks. Performance-ready.
Why Standard Prompts Fail
"Write me 10 ad hooks for my product" delivers generic output. It is plausible, but not brand-specific and not performance-ready. Prompt engineering is not magic; it is structuring.
01
6
Frameworks
02
8-12
Usable Hooks per Session
03
30 min
per Session
The 6 Frameworks
- Pain-First: "Generate 12 hooks that open with a concrete buyer pain"
- Anchor-Twist: "Generate 12 hooks with an unexpected number plus subversion"
- Pattern-Interrupt: "Hooks that feel out-of-context in the feed"
- Insider-Reveal: "Hooks with industry insider knowledge"
- Comparison: "Hooks that contrast two options"
- Demo-First: "Hooks that show the product immediately, no pitch"
Example
Brand voice anchor plus framework choice plus format specification is the trinity. Remove any one element and output quality drops by 50%.
The Iteration Loop
Filter the first output wave and identify the top 3 hooks. Reverse-engineer what worked and incorporate it into the next prompt. 3 to 5 iteration loops per session produce a scalable hook bank.
„Prompt engineering is a discipline, not a tool.”