Built a site to a pretty strict technical standard (100/100/100/100 on Google PageSpeed, full schema markup, every claim linked to evidence). Wanted to see if any of that actually gets it cited by AI tools like ChatGPT and Perplexity, so I ran a fixed set of 5 questions across 3 different AI engines and logged every result honestly.
Result: 2 out of 15 came back positive. The other 13 were misses, including a couple where the AI answered a completely different type of question than what I asked.
Full writeup with the exact questions and results: The Discovery Paradox | ZenMasterWorks
Curious what this community thinks — is this a common experience, or something specific to how I set up the test?