What is a knowledge graph?

A knowledge graph is a database that represents information as a network of entities (things) and the relationships between them. Each entity has attributes (properties with values) and is connected to other entities through typed relationships.

For example, in a knowledge graph:

  • “Hotel du Lac Lausanne” is an entity of type Hotel
  • It has attributes: address, phone number, star rating, amenities
  • It has relationships: located-in (Lausanne), near (Lake Geneva), category (boutique hotel)
  • “Lausanne” is a related entity of type City, with its own attributes and relationships

Google maintains the most widely known commercial knowledge graph, which powers Knowledge Panels, featured snippets, and increasingly, AI search responses.

How do hotels get into knowledge graphs?

Hotels get into knowledge graphs through three kinds of sources: structured data on their own website, authoritative platform data, and aggregated mentions across the web.

Structured data from websites. Schema markup on your hotel website tells search engines explicitly: “this entity is a Hotel, here are its attributes.” This is the most direct signal a hotel can send to a knowledge graph.

Authoritative platform data. Google Business Profile, Wikipedia, Wikidata, and official tourism board directories are high-confidence sources that contribute to knowledge graph entries.

Aggregated mentions. When your hotel name appears consistently across many sources (reviews, travel articles, directory listings) with coherent attributes, the knowledge graph’s confidence in your entity’s details increases.

AI systems, including ChatGPT and Google’s AI Overviews, draw on knowledge graphs when generating responses about real-world entities. A hotel that is in a knowledge graph with complete, accurate attributes is:

  • More likely to be recognised as a specific entity when mentioned
  • More likely to have its attributes cited accurately in AI responses
  • More likely to appear in response to attribute-specific queries (“hotels in Lausanne with a spa”)

A hotel not in a knowledge graph exists only as scattered text mentions. The AI can infer things about it, but less confidently.

How can a hotel get into a knowledge graph?

You cannot write directly into Google’s Knowledge Graph, but you can raise your odds of appearing in it through four actions: schema markup, a complete Google Business Profile, NAP consistency, and listings on authoritative sources.

  1. Implementing Hotel schema markup on your website
  2. Completing and verifying your Google Business Profile
  3. Ensuring NAP consistency across all platforms
  4. Getting listed on authoritative sources (tourism boards, Wikidata if your property merits it)

See schema markup and entity recognition.