Experiments
Fresh
This content reflects Metehan Alp's published experiments as of 2025-2026.
Metehan runs controlled experiments to test assumptions about AI search behavior. These are not theoretical frameworks -- they are real tests with real domains, real traffic data, and documented results.
Published Experiments
| Experiment | Question Tested | Key Finding |
|---|---|---|
| 60,000-Page AI Website | What happens when you build a massive AI-generated site and monitor bot behavior? | GPTBot sent 470x more requests than Googlebot |
| AI Citation Without Google/Bing | Can you get AI citations without being indexed by traditional search engines? | Google indexing appears to be a prerequisite for AI citation |
Experiment Standards
Each experiment documents:
- Methodology -- Exact setup, tools used, and configuration
- Timeline -- When the experiment ran and key observation dates
- Data -- Raw numbers, not interpretations
- Findings -- What the data shows
- Limitations -- What the experiment does not prove
- Reproducibility -- Whether others can replicate the setup