Negative examples ("don't do this") are surprisingly powerful for context loading. They create boundaries that focus AI's responses.
When to Use Negative Examples
1. When avoiding common mistakes "Don't use corporate jargon like 'synergy' or 'leverage'"
2. When style is crucial "Avoid the overly enthusiastic tone common in sales emails"
3. When precision matters "Don't include general productivity tips—focus only on remote work"
4. When differentiating from competitors "Unlike HubSpot's everything-included approach, we're intentionally minimal"
The Do/Don't Template
Structure negative examples clearly:
```text DO: Write like you're explaining to a smart friend DON'T: Write like a textbook or manual
DO: Use specific numbers and data DON'T: Use vague phrases like "significantly improved"
DO: Address reader objections directly DON'T: Pretend obstacles don't exist ```text