Published March 19, 2026
There are two ways to implement schema across a page. Standalone schema uses multiple separate JSON-LD script tags — one for the Organization, one for a Service, one for a WebPage, each independent. Connected schema uses a single @graph block that links all entities together, giving the AI crawler one structured object to read instead of several disconnected ones.
The AIFDS blueprint library uses the connected @graph approach across every blueprint. The reasoning is practical: make the AI crawler's job as easy as possible.
When a page has separate JSON-LD script tags, the crawler reads each independently and infers how they relate. With a single @graph block, those relationships are explicit. The Organization has an @id, the Service references it, the WebPage references it too. The crawler reads it once and understands the full picture.
This is the same principle behind page weight as a structural advantage. Connected data in a single block is easier to process than the same information scattered across multiple blocks requiring assembly.
The less work AI has to do to understand your page, the more reliably it processes what you are giving it.
Every AIFDS blueprint uses a single script tag containing an @graph array. Each entity has an @id — a unique identifier other entities reference.
The homepage Organization gets an @id like https://yourdomain.com/#organization. A service page's provider field references that @id. An about page's Person entity includes worksFor pointing back. Every page is structurally connected to the business entity on the homepage.
This is why homepage schema is the single most important implementation. The homepage Organization is the anchor every other page references. Without it, there is nothing for the rest of the site's schema to connect to.
The sameAs field links a business entity to profiles on other platforms — Google Business Profile, Yelp, LinkedIn. Whether those connections change AI behavior depends on competition.
In a crowded market, AI needs to verify which entity is which. sameAs links confirm that this Organization is the same entity as this Google Business Profile, this Yelp page, and this LinkedIn company page. The connections reduce ambiguity and build confidence.
In a market with less competition, sameAs matters less. The WhatsMyArtWorth experiment demonstrated this — a site with no social media profiles and no sameAs links earned AI citations within 33 days because it was the only site filling that specific completion gap.
sameAs linking is not required for AI citation. But in competitive markets, the more competitors in your space, the more verification signals matter.
Multi-page sites with multiple entity types. Each page's schema references the Organization, creating a coherent web of structured data AI reads as one entity.
Businesses in competitive markets. Connected schema with sameAs verification is easier for AI to identify and trust than disconnected blocks requiring inference.
Healthcare and professional services. Industries where individual providers matter — healthcare, legal, financial — benefit from Person entities connected to the Organization via @graph.
New sites filling a clear gap. AI will cite based on the core fields regardless of whether external profiles are linked.
Single-page sites. Nothing to connect. The @graph structure still works, but the connection benefit is realized when additional pages are added.
One script tag per page. All entities connected. The Organization anchor on the homepage, every other page referencing back. A page with one connected @graph block is easier to process than four disconnected script tags — and making AI's job easier is a structural advantage that compounds across every page on the site.
David Valencia writes about how AI systems find, parse, and cite websites.