The essentials
- Definition: the Knowledge Graph is Google's database describing entities (people, companies, places, products, concepts) and the relationships between them, launched in 2012.
- The figures: several hundred billion facts about billions of entities; it feeds the knowledge panels, direct answers and, in part, the AI Overviews and Gemini.
- Sources: Wikipedia and Wikidata, Google Business Profile, the schema.org structured data on websites, and authoritative sources by sector.
- What's at stake: a company recognised as an entity gets a panel, enriched brand results, and better understanding by engines and AI.
The aim of Knowledge Graph is Google's knowledge base: a vast collection ofentities (people, organisations, places, products, works, concepts) connected by relationships ("is the founder of", "is located in", "belongs to the category"). Launched in 2012 with the phrase "things, not strings", it lets Google understand that a query names a specific reality and answer it directly, without being limited to the words. It feeds the knowledge panels to the right of the results, direct answers, carousels, and serves as the factual foundation for AI Overviews and Gemini.
What the Knowledge Graph displays
| Element | Where | The full destination URL |
|---|---|---|
| Knowledge panel | Right-hand column (desktop), top of page (mobile) | Name, logo, description, founder, headquarters, website, social networks, reviews |
| Direct answer | At the top of the results | "What is the capital of…", "Who founded…" |
| Entity carousel | At the top of the results | A list of films, brands or places |
| Local listing | Local pack, Google Maps | A local business entity, drawn from Google Business Profile |
| Context for AI Overviews | The AI block at the top | Facts about the entity picked up in the generated answer |
Where the data comes from
- Wikipedia and Wikidata: the main source for notable entities; a well-filled Wikidata entry is often what triggers a panel.
- Google Business Profile: the source of local entities (shops, practices, agencies); see Google Business Profile.
- Structured data: the schema.org markup on websites (Organization, Person, Product, LocalBusiness) describes the entity and its relationships in a language Google reads directly.
- Authoritative sources: company registries, directories, press, institutional sites, official social profiles — all confirming the entity's existence and attributes.
- The web as a whole: Google extracts facts from pages and cross-checks them; consistency of information (name, address, director, founding date) across every source is decisive; see NAP consistency.
Why it matters to a business
A company recognised as an entity in the Knowledge Graph gets a knowledge panel on brand queries, which occupies space, reassures, and raises the click-through rate to the official site. Google then understands the site's pages in the context of that entity: an article about SEO published by an agency identified as an SEO specialist benefits from that association, in line with the E-E-A-Tcriteria. And AI assistants draw on the same knowledge graphs and the same sources to describe a company: a well-established entity is described correctly by ChatGPT or Gemini, a blurry one is confused with another or ignored. How AI engines work is covered in AI engines explained.
Getting your company recognised as an entity
- A complete "About" page: exact name, history, founder, headquarters, activities, key figures, with Organization markup and links to your official profiles (sameAs).
- Google Business Profile: a claimed listing, a precise category, information identical to the site's.
- Wikidata: an entry can be created for an existing company with verifiable sources (a registry, the press), without Wikipedia's notability requirements.
- Consistency everywhere: the same name, the same address, the same description on your site, in directories, on LinkedIn, in registries and in the press.
- Mentions by third parties: press articles, professional directories, partners — all confirming the entity and its attributes.
- Claim the panel: once displayed, the panel can be claimed from Google search to submit corrections.
The limits and the mistakes
- The panel is not guaranteed: Google displays it when the entity is sufficiently established; a small local business gets a local listing, rarely an organisation panel.
- Contradictory data: two addresses, two trading names, a different director depending on the source; Google doesn't arbitrate, it doubts.
- Dishonest markup: declaring false attributes in schema.org risks having the markup ignored, or even a manual action.
- Confusion with another entity: a name shared with a better-known company; disambiguation comes through precise attributes and sources.
How GreenRed helps
Rather than juggling several tools, GreenRed's GEO and AI module brings these metrics together in a single dashboard, compares them over time and tells you which actions come first. You can try it free, with no card, from the Pricing.
Frequently asked questions
What is Google's Knowledge Graph?
Google's knowledge base, describing entities (people, companies, places, products, concepts) and the relationships between them. It feeds the knowledge panels, direct answers and carousels, and serves as the factual foundation for AI Overviews and Gemini.
How do I get a knowledge panel for my company?
By making the entity identifiable and consistent: a complete About page with Organization markup and sameAs links, a claimed Google Business Profile listing, a sourced Wikidata entry, identical information across directories, LinkedIn and registries, and press mentions. The panel is not guaranteed; Google displays it when the entity is sufficiently established.
What is the difference between the Knowledge Graph and structured data?
Structured data (schema.org) is the markup you add to your pages to describe an entity; the Knowledge Graph is Google's database, which aggregates that information with other sources. The markup is one possible way in, not a guarantee of inclusion.
Does the Knowledge Graph influence ChatGPT's answers?
Indirectly: AI assistants draw on similar sources (Wikipedia, Wikidata, official sites, the press) and on their own graphs. A consistent, well-documented entity is described correctly by AI; a blurry one is confused with another or ignored.