Symptoms
Empty or repetitive text, missing headings, missing canonical titles, or key entities only present after client-side rendering.
Extractability.
Extractability is a practical measure of whether an AI crawler can reliably extract the primary content and meaning from your pages — without being derailed by rendering gaps, empty shells, or content hidden behind interactions.
Docs: Extractability metric · Example: /radar/domains/stripe.com
Why it matters
If your main content isn’t present in the raw HTML (or doesn’t survive rendering reliably), you’ll see regressions: missing facts, misclassification, or “thin content” signals.
Fixes
Server-render main content, avoid hiding primary text behind JS-only flows, and ensure stable titles/meta + headings exist in the initial response.
Related tools
Extractability improves when crawl access is clean and token bloat is controlled.
Crawl access
Make sure AI bots can fetch the content in the first place.
Token bloat
Reduce noise so the main content dominates.