- Crawlability first. A bot blocked in robots.txt cannot help you, no matter how good the page is. Rule access out before tuning anything downstream.
- All three bot types present — training, search, and user-query bots each map to a different visibility stage; a missing type names the gap. See AI bot types.
- Cross crawls against citations. The real insight comes from combining how often bots request a page with whether it shows up as a source.
- Technical: bots reach the page but heavy JavaScript blocks clean parsing.
- Content quality: accessible, but LLMs choose not to include it.
- Prompt-tracking gap: the page performs well on prompts you are not tracking — broaden coverage using the topics the crawled page covers.
