Knowledge Graph

Overview and context

A knowledge graph is a knowledge base that models entities — people, places, things, and concepts — as nodes and the relationships among them as edges, weaving knowledge into an interconnected web. This structure is what lets a search engine stop matching the literal characters in a query and instead understand the actual "thing" those words refer to, along with its context. Ask "How tall is the Eiffel Tower?" and getting the fact back directly is the knowledge graph at work.

Google introduced it on May 16, 2012, through a blog post by then–head of search Amit Singhal titled "Introducing the Knowledge Graph: things, not strings." The central message was that search was shifting from matching strings to understanding things. Google describes the Knowledge Graph as a system that understands facts and information about entities from material shared across the web — in its own words, "a giant virtual encyclopedia of facts."

How it works and where it's used — entities and semantic search

The unit of a knowledge graph is the entity, not the keyword. Each entity is uniquely identified and connected to others through relationships (edges) such as "is the director of," "is located in," or "is the author of." Data structured this way becomes the foundation of semantic search, allowing the engine to disambiguate terms that share a spelling and to interpret a query's intent at the level of meaning.

Common applications include:

• Factual question answering: returning a direct answer to fact-seeking queries like "Where were the 2016 Summer Olympics held?"
• Knowledge panel generation: the information box that appears alongside search results (the knowledge panel) is generated automatically by pulling from the Knowledge Graph.
• Voice and assistant responses: voice assistants also consult the knowledge graph when answering factual questions.
• Entity understanding in AI search: it serves as foundational data that generative search and AI answers use to identify and connect topics and entities.

Knowledge graph vs. knowledge panel

The two are frequently confused, but they operate at different layers. The knowledge graph is the underlying structure and knowledge base that holds the data, while the knowledge panel is the SERP display surface that shows that data on the results page.

Evidence and examples

The scale figures Google has disclosed show how the Knowledge Graph has grown into core search infrastructure (sources: Wikipedia, "Knowledge Graph (Google)"; Google's official blog and knowledge panel help).

Its data sources are equally varied. Google explains that the Knowledge Graph draws on "hundreds of sources across the web" — public material such as Wikipedia, the CIA World Factbook (discontinued in February 2026), and Freebase, which formed the basis of the early graph; licensed data such as sports scores, stock prices, and weather; and structured data submitted by content owners. At the same time, the fact that knowledge-graph answers are often presented without source attribution or citation has drawn criticism on the grounds of verifiability.

SEO and GEO implications — entity optimization

As search moves from keywords to entities, the focus of optimization shifts from "targeting a specific keyword" to "making sure search engines clearly understand an entity — a brand, product, or person — and connect it within the knowledge graph." This is known as entity SEO. Structured data is not a direct ranking factor, but it is a key lever for rich results, citations in AI answers, and entity recognition in the knowledge graph.

A practical checklist:

• Mark up your core content (organizations, people, products, and so on) with structured data using the schema.org vocabulary, in Google's recommended JSON-LD format.
• Reference internal entities consistently with @id, and use sameAs to link to authoritative external sources such as Wikipedia and Wikidata, so each entity is clearly identified and disambiguated.
• Keep core facts such as NAP (name, address, phone) consistent across channels so an entity's identity and relationships stay coherent.
• Build an entity's authority with trustworthy information that aligns with E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness).

The following JSON-LD example defines an organization entity clearly and links it to external sources.

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "@id": "https://example.com/#organization",
  "name": "Example",
  "url": "https://example.com",
  "sameAs": [
    "https://www.wikidata.org/wiki/Q12345",
    "https://en.wikipedia.org/wiki/Example"
  ]
}

References

• Google — Google's Knowledge Graph and knowledge panels (blog.google)
• Google — How Google's Knowledge Graph works (Knowledge Panel Help)
• Wikipedia — Knowledge Graph (Google)
• Schema.org — Official structured data vocabulary documentation
• Schema App — What is Entity SEO and How Do I Implement It?
• Ahrefs — Google's Knowledge Graph Explained: How It Influences SEO & AI Search