Skip to main content
Search & Retrieval

Schema markup

Machine-readable annotations that describe the structure and meaning of content.

Also known as: Structured data markup, Schema.org markup

Definition

Schema markup is a standardised vocabulary of machine-readable annotations added to web content or documents that describes their structure, type, and meaning. Using formats like JSON-LD, microdata, or RDFa, schema markup tells search engines and other systems what a page contains — not just the text, but the semantic meaning: this is an article about a legal concept, that is a definition, this is a publication date, those are related terms. Schema markup bridges the gap between human-readable content and machine-interpretable structured data.

Why it matters

  • Search engine understanding — schema markup helps Google and other search engines understand page content, enabling rich results (featured snippets, knowledge panels, FAQ cards) that increase visibility and click-through rates
  • SEO value — pages with schema markup are more likely to appear in enhanced search results, which consistently outperform standard blue links in click-through rates
  • Internal search improvement — beyond external search engines, schema markup helps internal retrieval systems understand content structure, enabling more precise filtering and categorisation
  • Interoperability — schema markup uses shared vocabularies (primarily Schema.org) that enable different systems to understand and exchange content descriptions consistently

How it works

Schema markup is implemented by adding structured data annotations to web pages or documents:

JSON-LD (JSON for Linking Data) is the recommended format. A JSON-LD block is embedded in the page’s HTML <head> or <body>, describing the page’s content in a structured format. For a glossary page, this might specify: the term being defined (name), the definition (description), the category (about), related terms (relatedLink), and the page’s language (inLanguage).

Schema.org vocabulary provides the standardised types and properties. Common types for a legal AI website include:

  • DefinedTerm — for glossary entries (term name, description, category)
  • Article — for blog posts (headline, author, datePublished, articleBody)
  • FAQPage — for pages with question-and-answer content
  • WebSite — for the overall site with search functionality
  • Organization — for the company with contact information

Implementation involves identifying what content types each page contains, mapping them to appropriate Schema.org types, and generating the JSON-LD markup. For a multi-language site like a Belgian legal platform, the inLanguage property distinguishes between Dutch, French, and German content.

Validation uses Google’s Rich Results Test or Schema.org’s validator to verify that markup is syntactically correct, uses valid types and properties, and is likely to generate enhanced search results.

Schema markup does not change the visible page content — it is metadata consumed by machines, not by human readers. However, the information it conveys should accurately reflect the visible content; misleading markup violates search engine guidelines and may result in penalties.

Common questions

Q: Does schema markup directly improve search rankings?

A: Schema markup is not a direct ranking factor, but it enables rich results that increase click-through rates, which indirectly supports SEO. More importantly, it helps search engines understand content semantics, which can improve how and when pages appear in search results.

Q: How much schema markup should a page have?

A: Enough to describe the primary content of the page. A glossary page should have DefinedTerm markup. A blog post should have Article markup. Over-marking incidental content (navigation, footers, advertisements) is unnecessary and can confuse validators.