Structured Data
Schema Markup for SEO in 2026: The Complete JSON-LD & Rich Results Guide
Schema markup is structured data added to a page that tells search engines exactly what the content is, using the shared vocabulary at schema.org. Google's recommended format is JSON-LD placed in the head of the document.
Schema does not directly raise rankings. It raises comprehension — and comprehension is what earns rich results in Google and citations in AI answers.
Free tools used in this guide
Which Schema Types Actually Produce Rich Results?
- Article and NewsArticle — headline, date and author in news surfaces
- FAQPage — expandable question and answer blocks, and the most AI-extractable format
- Product with Offer and AggregateRating — price, availability and stars
- Recipe — cook time, ratings and image carousels
- Event — dates, venue and ticket links
- BreadcrumbList — path display instead of a raw URL
- Organization and WebSite — knowledge panel and sitelinks search box
- SoftwareApplication — category, price and rating for tools and apps
- LocalBusiness — hours, address and map presence
- HowTo, VideoObject, Course and JobPosting for their respective verticals
How Do You Implement JSON-LD Correctly?
Step 1 — Pick one primary type per page
Match the type to the page's real purpose. A blog post is an Article, not a Product, even if it mentions products.
Step 2 — Fill every required property
Missing required properties silently disqualify a page from rich results. Optional properties add detail and confidence.
Step 3 — Mirror visible content
Every value in the markup must appear on the page. Marking up content users cannot see is a policy violation.
Step 4 — Combine types with @graph
Use a single @graph block with linked @id values instead of scattering unconnected scripts across the page.
Step 5 — Validate before shipping
Run the markup through a validator to catch syntax errors and missing fields, then re-check after any template change.
Why Does Schema Matter for AI Citation?
AI retrieval systems must decide what a page is about before deciding whether to quote it. JSON-LD answers that question explicitly instead of forcing inference from HTML.
FAQPage markup is the highest-value format for GEO because its question-answer pairs match the shape of generated answers almost exactly.
Across 2.4 million audited pages, about 64% carry no structured data at all — which makes correct markup a straightforward competitive advantage.
What Are the Most Common Schema Mistakes?
- Invalid JSON — a trailing comma breaks the entire block silently.
- Marking up content that is not visible on the page.
- Using Review or AggregateRating markup for self-serving reviews, which Google penalises.
- Duplicate conflicting blocks of the same type on one page.
- Relative URLs in @id or url properties instead of absolute ones.
- Never re-validating after a template or CMS update.
Frequently Asked Questions
Does schema markup improve rankings?
Not directly. It improves how accurately search engines understand a page, which produces rich results and higher click-through rates — and those indirectly help performance.
JSON-LD, Microdata or RDFa?
Use JSON-LD. Google explicitly recommends it, and it keeps structured data separate from the HTML so template changes cannot break it.
Where should JSON-LD be placed?
In the head of the document is preferred, though Google can also read it from the body. Consistency across templates matters more than exact placement.
Can incorrect schema hurt my site?
Yes. Markup describing content that is not on the page is a structured data policy violation and can trigger a manual action removing all rich results.
How many schema types can one page have?
As many as genuinely apply. Link them in one @graph block — for example Article plus BreadcrumbList plus FAQPage on a guide.