Examples are context in action. They show rather than tell, calibrating AI's understanding instantly.
The Three Types of Context Examples
1. Output Examples (What you want) "I need email subject lines like these high-performers: - 'Your competitor just raised $50M (here's what it means for you)' - 'Quick question about your Q3 roadmap' - 'Spotted this problem on your website (+ easy fix)'"
2. Situation Examples (Similar scenarios) "This is similar to when Slack pivoted from gaming to communication. They had existing technology but needed to reframe for a new market. We're in a similar position with our AI tool."
3. Anti-Examples (What to avoid) "Unlike typical corporate communications that say 'We're excited to announce our strategic initiative to leverage synergies,' we want direct, plain English that actually means something."
The Example Sandwich Technique
Structure your context with examples as "bread":
```text Top Bread (Example of desired outcome): "Like how Netflix's recommendation engine explains: 'Because you watched Stranger Things' → simple, clear, personal"
Filling (Your specific needs): "We need to explain our AI's decision-making to non-technical users in our healthcare app"
Bottom Bread (Example of current problem): "Currently we show: 'Based on algorithmic analysis of multivariate factors' → confusing, intimidating" ```text