SEO Perception
Help search and AI understand your site
AI discoverability is not one switch. It is a set of clear signals: who may crawl, what your brand is, what each page is about, and what answers visitors can read.
- Crawl rules that match your intent
- Structured data that matches visible content
- Product checks after connect and crawl
Four signals that matter
Work these in order. Schema without crawl access, or markup that does not match the page, does not help.
-
Who may crawl
robots.txt should say which AI-related user-agents you allow or block. Blanket blocks hide pages you may still want discovered.
-
What the brand is
Organization JSON-LD and clear identity on the homepage help systems name who publishes the site.
-
What each page is about
Article and FAQ schema, honest titles, and visible answers beat empty markup that does not match the page.
-
What AI can read
llms.txt and crawlable HTML give models a calm entry point. Possibilities flag missing or thin signals after a crawl.
Start here
Free tools and docs for the signals above. Paid product checks flag gaps after you connect a site and crawl.
- JSON-LD generator Build Article, Organization, or FAQ schema and copy a script tag.
- SEO Perception MCP Bring Search Console keywords and Possibilities into Cursor and Claude (paid plans).
- JSON-LD basics What structured data is and where it lives in HTML.
- llms.txt possibility When we flag a missing or thin /llms.txt on the homepage.
- AI crawlers in robots.txt Blanket blocks for common AI-related user-agents.
Inside the product
Connect a site, run a crawl, and open SEO Possibilities. We flag missing structured data, organization schema, question queries without FAQ markup, and related signals. Use Generate schema where offered.
- Organization identity from website settings into Organization JSON-LD
- Editorial pages: Article JSON-LD with honest author signals
- Possibilities that stay tied to crawl and Search Console evidence
Editorial pages
For articles, combine Article JSON-LD with honest author signals. Schema should match what a human can read on the page.
Keep exploring
Make the signals, then verify
Generate schema for free, connect a site when you are ready, and let Possibilities flag what is still missing after the crawl.