How it works
Paste your rendered HTML (view source of a live page, or the output of your SSR build) and we run 11 structural checks against it. Each check reports pass, warn or fail with a specific fix, and we roll the weights into an overall citation-readiness grade A to D.
Why rendered structure is load-bearing
- Assistants skim, they don't read. The ranker looks for structural anchors (H2, list, table, schema) and passes only those passages forward to the generator.
- Hydration gaps kill scores. If your content only appears after JavaScript, most AI crawlers never see it. Always validate the rendered HTML, not the source template.
- Structure compounds. A page with H1 + 5 question-shaped H2s + a list + JSON-LD beats a page with better writing but flat prose. Every time.
Pair with
Fix weak headings with the headings optimizer, chunk long prose with the content chunking tool, and validate your JSON-LD with the structured data validator. Full playbook: write content AI assistants cite.