Schema Markup for AI Search: What Actually Helps
Structured data won’t get you cited on its own, but it removes guesswork about who you are and what a page contains. Here is what to add, what to skip and how to check it.
What you’ll learn
- What AI systems can and can’t get from structured data
- The schema types worth adding for most sites
- Common mistakes that make markup useless or harmful
- How to test your JSON-LD
Structured data is machine-readable information about a page, usually written as JSON-LD with the Schema.org vocabulary. Search engines have used it for rich results for years. In AI search its role is narrower but still useful.
What it does and doesn’t do
- It clarifies facts. Who publishes the site, which product a page is about, its price, author and dates. These are the details AI answers get wrong most often.
- It feeds knowledge graphs. Google’s Knowledge Graph and other entity databases combine schema with sources such as Wikidata and Wikipedia.
- It powers shopping and local answers. Product, offer and business data flow into Google Shopping and AI shopping agents.
- It doesn’t replace content. Language models read the visible text of a page. Markup that describes things not on the page won’t make them appear in answers, and Google treats it as a policy violation.
Types worth adding
| Type | Where | Why |
|---|---|---|
Organization with sameAs |
Home page or About page | Connects your site to your Wikidata, LinkedIn and social profiles, so AI systems recognise one entity |
WebSite |
Home page | Site name used in results |
Article or BlogPosting |
Articles and news | Headline, author, publish and update dates |
Product with Offer |
Product pages | Price, currency, availability, reviews |
LocalBusiness |
Local businesses | Address, hours, phone |
BreadcrumbList |
Most pages | Site structure |
FAQPage |
Real FAQ sections | Question-and-answer pairs that match the visible text |
Common mistakes
- Markup that doesn’t match the page, such as a price in schema that differs from the price shown.
- Several conflicting Organization blocks from a theme, an SEO plugin and a custom script.
- Missing dates. Without
datePublishedanddateModified, AI systems have no reliable signal for how fresh a page is. - JSON-LD added only by JavaScript after the page loads. Many AI crawlers don’t run JavaScript and never see it. Put structured data in the HTML the server sends.
- Invalid JSON. One missing comma makes the whole block unreadable.
Keep entity data consistent
AI systems trust facts more when several sources agree. Use the same name, logo, founding date and description in your Organization schema, your Wikidata entry and your About page. Link them together with sameAs.
Check your markup
The Schema Markup Checker reads the JSON-LD a crawler receives, without running JavaScript, and lists what’s missing or wrong for each type. The AI Entity Checker compares your Organization data with Wikidata and Wikipedia, and the AI Crawlability Checker shows the text an AI crawler actually reads from the page.