TL;DR
Key Definition
Schema markup turns the visible content of a page into facts a machine can read. This page is an article, this person wrote it, this product costs that much. Google recommends it in JSON-LD and uses it to understand pages and display rich results. Schema markup still guarantees neither a rich result nor an AI citation, and Google says no special markup is required for its AI features.
Importance
Why Does It Matter?
- Schema markup removes ambiguity. It says that a number is a price, or that a name is the author, instead of leaving the engine to guess.
- Structured data makes a page eligible for some Google rich results, such as breadcrumbs, products, or events.
- With the Organization type, Google can tell your company apart from a namesake more easily, and the logo shown in results and the knowledge panel can depend on it.
- When structured data matches the visible text, the page reads more clearly, and AI engines that read the indexed version benefit too.
- Google, Microsoft, Yahoo, and Yandex founded schema.org together. One markup therefore speaks to several engines.
How Does It Work?
Schema markup puts a layer of metadata on top of the HTML. During crawling, the engine reads that metadata and checks it against the visible text. It then decides whether to use it to understand the page or for an enhanced display.
- 1
The schema.org vocabulary
Schema.org defines types (Article, Product, Organization, Person, Event, BreadcrumbList) and their properties (headline, author, price, sameAs). Each type inherits from a more general one, so an entity can be described with precision. The rule is simple. Pick the type that fits the real content of the page and add nothing that doesn't apply.
- 2
The JSON-LD format
Google accepts JSON-LD, Microdata, and RDFa. It recommends JSON-LD, which lives in a script tag apart from the HTML and is easy to maintain. A minimal example for an article looks like this. <script type="application/ld+json">{"@context": "https://schema.org", "@type": "Article", "headline": "What is schema markup?", "author": {"@type": "Person", "name": "Timothe"}}</script>. The block can go in the head or the body, and injecting it with JavaScript is fine as long as the rendered page contains it.
- 3
What Google does with it
Google reads structured data to understand content and for some rich results, but never promises they will appear. Since August 2023, FAQ rich results are limited to well-known, authoritative government and health websites, and HowTo rich results are gone. FAQPage is still valid schema. It just no longer triggers a special display for most sites.
- 4
Schema markup and AI engines
For AI Overviews and AI Mode, Google says there is no special structured data to add. Standard SEO best practices apply, provided the markup matches the visible text. Other AI engines haven't published an equivalent rule. Treat schema markup as a clarity gain, then, and never as a shortcut to being cited.
Metrics
How Do You Measure It?
- Run each page template through Google's Rich Results Test to confirm the markup is read and eligible.
- The schema.org validator catches malformed types and properties, including the ones Google ignores.
- In Search Console, the enhancement reports list valid items, warnings, and errors for each structured data type.
- Compare in Search Console the CTR of pages that earn a rich result with that of similar pages shown as plain results.
- On every release, have a crawler confirm the JSON-LD is in the rendered HTML, and not only in the source code.
Pitfalls
What Are the Common Mistakes?
- Marking up content that isn't on the visible page breaks Google's guidelines and can earn a structured data manual action.
- Adding FAQPage everywhere to get a FAQ display is wasted time. Those rich results are now limited to a few authoritative sites.
- Counting on schema markup to get cited by an AI pulls effort away from the content, which is still what decides.
- An outdated price, rating, or date in the JSON-LD contradicts the page, and the engine trusts your data less.
- Copying a generic template without matching the type to the content leaves you with articles marked up as products.
What Should You Do First?
Follow these steps to get started.
- List your page templates and give each one the most precise schema.org type.
- Put Organization on the home page with name, logo, url, and sameAs links to your official profiles.
- Mark up editorial content as Article, with author, publication date, and modification date.
- On deep pages, BreadcrumbList makes the site hierarchy explicit.
- Generate the JSON-LD from the same data as the visible content, so the two never drift apart.
- Validate each template in the Rich Results Test and the schema.org validator.
- After every release, check the Search Console reports.
Examples
E-commerce product page
A store adds Product and Offer to its product pages, with price, currency, and availability pulled from its catalog. When a price moves, the JSON-LD moves with the page. No engine ends up showing an outdated price.
SaaS vendor and disambiguation
A software vendor's name is also a common word. It adds Organization with sameAs pointing to its Wikidata item, LinkedIn page, and GitHub organization, and the engine has an easier time linking the site to the right company rather than a namesake.
Glossary and FAQPage
A hypothetical site marks up its glossary pages with Article and FAQPage. Google gives it no FAQ display, which is normal since 2023. Its markup stays valid and mirrors the questions visible on the page.
Frequently Asked Questions
Is schema markup required to appear in Google?
No. Google indexes and ranks pages without structured data just fine. Markup helps it understand content and makes some pages eligible for rich results, with no guarantee. It's worth the effort on page types that still get an enhanced display, such as products, events, or breadcrumbs.
Does schema markup help you get cited in AI Overviews?
Not directly, according to Google. AI Overviews and AI Mode rely on the regular search index and need no special structured data. Schema markup still helps general page understanding, as long as it matches the visible text. Content quality and clarity carry far more weight.
Should you still use FAQPage markup?
You can, but don't expect a FAQ display. Since August 2023, Google limits those rich results to authoritative government and health websites. FAQPage is still a valid schema.org type for describing a question page. If your questions are visible and useful, the markup does no harm. It just won't produce a special display.
What is the difference between schema markup, structured data, and JSON-LD?
Structured data is any information formatted for machines. Schema.org is the most widely used vocabulary for writing it, with its types and properties. JSON-LD is a syntax, standardized by the W3C, in which that vocabulary is expressed. Schema markup, then, is structured data written with schema.org, most often in JSON-LD.
How do I check that my schema markup is correct?
Google's Rich Results Test checks eligibility, and the schema.org validator checks full syntax. After that, watch the Search Console reports over time. For a wider look at how readable your pages are to AI engines, Hikoo's Site agent and the free audit at /tools/free-audit also review your structured data.
