Guide
JSON-LD and metadata for AI discovery
Structured data will not “rank” you in ChatGPT by itself—but clear Organization, WebSite, Article, and FAQ schema help machines understand what a page is.
Page identity stack
- Title — human-readable, unique, includes the entity name
- Meta description — one or two sentences an answer engine could paraphrase
- Canonical URL — single preferred address
- Open Graph — title, description, image for link previews
- JSON-LD — typed facts (Organization, Article, FAQ, etc.)
Patterns that help citations
- Keep `@id` and canonical URLs stable over time
- Align schema names with on-page H1 and brand spelling
- Put FAQs in visible HTML and mirror them in FAQPage JSON-LD
- Avoid stuffing irrelevant schema types as a hack
What BatSignal evaluates
During the content-readiness portion of an audit, BatSignal looks for title, meta description, Open Graph, JSON-LD, headings, sitemaps, and llms.txt—alongside whether the HTML is a real document or an empty SPA shell.
FAQ
Which schema types matter most for a SaaS marketing site?
Start with Organization, WebSite, SoftwareApplication (or Product), WebPage/Article for guides, BreadcrumbList, and FAQPage where you truly have FAQs.
Does BatSignal require JSON-LD to score well?
JSON-LD is part of content readiness. Missing structured data lowers that pillar; it does not zero the entire composite score if other signals are strong.