Knowledge Graph

Google's Knowledge Graph is a vast database of entities, meaning people, companies, places, products and concepts, and the relations between them. It lets Google understand the world as things rather than text strings, and it fills the fact panels that appear in the results when someone searches for a known person or company.

Google launched the Knowledge Graph in 2012 under the motto "things, not strings". Where the search engine previously saw letters, it could now grasp that a footballer was born in one country, plays for a particular club and has won particular tournaments. The information comes from open sources such as Wikipedia and Wikidata, and from structured data on websites, which states unambiguously who is who.

For a website owner the Knowledge Graph has two faces. Fact panels answer more queries inside the results page, so the user never clicks through, the so-called zero-click searches. In return, a knowledge panel for your own brand is a strong trust signal when someone searches your name.

Entities became the currency of AI search

The entity understanding behind the Knowledge Graph is the same mechanic that carries AI search. Language models assemble answers from entities and relations: who the company is, what it offers, which topics it is connected to. A brand that exists as a well-defined entity, with consistent details across the web and mentions from credible sources, stands a better chance of being understood correctly and named in the answers.