On activation, and again whenever content, menus or the site title change (plus once a day), the plugin scans your site. The scan runs in the background through WP-Cron, never while a visitor loads a page.
| It reads | It uses it for |
|---|---|
| Site title and tagline | The greeting, such as “Hi there! Welcome to your site.” |
| Main navigation (classic menu, block-theme Navigation, or top-level pages) | Quick links in the mascot’s menu, unless you set your own |
| Published pages and posts (up to 500): title, summary, H2 and H3 headings, keywords, categories, reading time | Page lines, related suggestions, chat answers and 404 suggestions |
| Top categories | Lines such as “We have 12 articles about Research.” |
| Email and phone published on a Contact-like page or in the footer | Answers to “what’s your phone number?” (the admin email is never used) |
What it skips
Pages marked noindex in Yoast or Rank Math, and password-protected pages, are skipped. Pages built with Elementor are read from Elementor’s own data (headings, text and accordions).
Learn from more content
Developers can raise the 500-item limit or choose which post types are scanned. See Hooks and the JavaScript API.