AI VISIBILITY
Our methodology
Readiness is measured separately from actual AI mentions and citations. The crawler does not execute page JavaScript or bypass security challenges.
Scoring weights: technical 25, crawler policies 20, content 20, structured data and identity 15, discovery 10, visibility measurement 10.
Only passing and failing tested checks enter the score. Untested checks and errors reduce coverage. Not-applicable checks are excluded. Confidence reflects coverage, not a guarantee of outcomes.
Training-agent preferences are reported separately and are not rewarded for allowing training. Robots simulation is not verified firewall or server-log evidence.
JSON-LD checks validate syntax, not complete schema semantics. Content length is a heuristic and can differ between writing systems.
Official API samples are labelled with provider, model, date and search mode. Plain URLs in model prose are not treated as verified citations.