Schema.org Structured Data

Schema.org is a shared vocabulary for embedding machine-readable metadata inside HTML, launched on June 2, 2011 as a collaboration between Google, Microsoft (Bing) and Yahoo!. Yandex joined in November of the same year, making the vocabulary the de-facto cross-engine standard. The founders' goal was to give webmasters a single set of type definitions rather than maintaining separate markup for each search engine. The initial release had about 297 classes and 187 relations; the vocabulary has since grown past 800 types covering common entities of the web. Three syntaxes are supported. <a href="/knowledge/Microdata" class="link" style="color: #D4A843;" title="Direct link to chunk">Microdata (HTML)</a> was the original encoding, designed by Ian Hickson for the WHATWG HTML5 effort, using inline attributes such as itemscope, itemtype and itemprop. <a href="/knowledge/rdfa" class="link" style="color: #D4A843;" title="Direct link to chunk">RDFa</a> (Resource Description Framework in Attributes), a W3C Recommendation since 2008, expresses the same triples through HTML attributes and was simplified to RDFa 1.1 Lite in 2012. <a href="/knowledge/json-ld" class="link" style="color: #D4A843;" title="Direct link to chunk">JSON-LD</a>, a linked data serialization of JSON that reached W3C Recommendation status in January 2014, has become the dominant choice. Google began explicitly recommending it in 2015 and made it the preferred format in 2017. JSON-LD won because it sits in a single <script> block in the document head, decoupling structured data from presentational markup, which is easier to template, easier to maintain, and survives front-end frameworks that mangle inline attributes. The vocabulary's most heavily used top-level types are Organization, Person, Article, Product, Event, Recipe, Review, LocalBusiness, Movie, Book and BreadcrumbList. Search engines parse these to render rich results: recipe cards with cook times and calories, product listings with price and stock, event panels with dates and venues, FAQ accordions, star-rated review snippets, and breadcrumb trails that replace raw URLs in the listing. Beyond per-page snippets, Schema.org markup is ingested into entity stores. The Google Knowledge Graph, launched in May 2012 with around 500 million entities and 3.5 billion facts, uses Schema.org as one of its primary structured inputs, alongside Wikidata and licensed data feeds; Microsoft's Bing Entity Graph plays the same role for Bing and Copilot. AI assistants and retrieval pipelines now consume the same markup to attribute answers to verified sources. Governance moved from the founding search engines to an open community process. The Schema.org Community Group develops the vocabulary through the public-schemaorg W3C mailing list and a GitHub repository, releasing the schema under a Creative Commons license.