Loading...
Loading...
14Module
Optimizing WordPress sites for AI crawlers and recommendations. Plugins, settings, and best practices for the world's most popular CMS.
Available3 lessons18 min
WordPress renders pages on the server, so the substance normally arrives in the initial HTML and most of the extraction test is already passed; confirm it with curl on your five most important pages rather than assuming it. The failures come from three places. Content deferred by a page builder, a cache or CDN answering differently than your editor shows, and headings built from styling rather than markup. Two settings decide whether the site is present at all: Search Engine Visibility, which since WordPress 5.3 emits a noindex meta tag into the page head and no longer writes anything into robots.txt, so an audit reading robots.txt alone will miss it; and robots.txt itself, which WordPress only generates when no real file exists, meaning a physical file in the site root silently overrides every rule a plugin believes it is publishing. After that, sites lose answers to builder layouts that put the answer below three other blocks, thin tag archives competing with the page you meant, plugin schema that disagrees with the visible page, and bylines with no biography or traceable identity.
WordPress starts ahead: pages render on the server, so the substance is in the response and most of the technical pass is closed before anybody configures anything.
Which means the failures here are quieter than on a JavaScript site. A checkbox left ticked after a rebuild, a robots.txt file somebody uploaded years ago that silently overrides every plugin rule, and builder layouts that put three blocks between a heading and the sentence that answers it. This module is what to look for rather than where to click, because menus and plugin names change and a course that prints them goes wrong within a year.