Knowledge Graph and Knowledge Panels

Google’s Knowledge Graph is a structured database that stores information about entities and the relationships between them. An entity is any distinct, identifiable thing: a person, organisation, place, product, concept, or creative work. The Knowledge Graph stores facts about individual entities and the connections that link them: that a person founded an organisation, that a film was directed by a specific person, that a company is headquartered in a particular city.

This database is what powers knowledge panels, the information boxes that appear in search results when Google is confident it understands what the query is about.

How does the Knowledge Graph work?

The Knowledge Graph is built from structured and semi-structured data sources: Wikipedia, Wikidata, Freebase (now largely absorbed into Wikidata), government datasets, and Google’s own web crawl. Google extracts entity information from these sources, reconciles it across multiple references, and stores the consolidated record.

When a search query maps to a known entity with high confidence, Google surfaces the entity data directly in the results rather than (or alongside) the standard ranked list. This is the knowledge panel.

The Knowledge Graph operates at a fundamentally different level from keyword-based indexing. A page can rank well for a query without the entity being in the Knowledge Graph. But entities in the Graph have a distinct advantage: their information is displayed prominently, attributed to them directly, and used by AI systems when reasoning about that entity across queries.

Knowledge panels

A knowledge panel is the boxed summary that appears on the right-hand side of desktop results (or at the top on mobile) for entity queries. It typically includes:

• Entity name and description
• Key attributes (for businesses: address, hours, website; for people: occupation, birthdate; for organisations: founding date, headquarters)
• Images
• Social media profiles and official website links
• Related entities (associated people, parent company, similar topics)

The data in a knowledge panel comes from the Knowledge Graph, not from any single website. Google pulls from multiple sources and synthesises a consolidated representation. This means the panel may display information that appears nowhere on your own site, or may omit information that your site prominently features.

How do entities get into the Knowledge Graph?

There is no submission form for the Knowledge Graph. Entities enter through the accumulation of consistent, verifiable signals from credible sources. Google’s stated policy is that its automated systems, not manual requests, create and remove panels: “our current policy doesn’t manually create or delete Knowledge Panels.”¹

Wikidata is often the most practical route. Wikidata is an open, structured data repository widely used as a source of entity data for Google’s results, though Google does not publicly document the exact data flow. An entity with a well-maintained Wikidata entry, supported by citations from credible external sources, is more likely to be represented in Google’s Knowledge Graph. Wikidata requires notability: the entity must have verifiable coverage in independent reliable sources. Creating a Wikidata entry without supporting citations produces a stub that Google will not treat as credible.

Wikipedia carries significant weight. A Wikipedia article about an entity, maintained with cited sources, contributes substantially to Knowledge Graph representation. Wikipedia itself requires notability (coverage in multiple independent, reliable sources), so this is not a route available to all entities.

Structured data on your own site helps Google understand what your site and its content are about, but does not directly add entities to the Knowledge Graph. Organization, Person, LocalBusiness, and Product schema mark up entity attributes in machine-readable form. This makes extraction easier but does not guarantee Knowledge Graph inclusion. Google specifically documents the sameAs property, linking to your Wikipedia, Wikidata, and authoritative social profiles, as a recommended way to help it disambiguate your organisation.²

Consistent cross-web presence matters. The same entity referenced consistently across credible third-party sites (directory listings, industry databases, news coverage, social profiles, official website) creates the co-occurrence pattern Google uses to reconcile and verify entity identity. Google triangulates across corroborating sources rather than trusting any single page: consistent references raise confidence, contradictory ones lower it.³

What triggers a knowledge panel?

Knowledge panels appear when two conditions are met:

1. Google has an entity record with sufficient completeness and confidence
2. The query maps clearly to that entity rather than to a general topic

Organisations with a Google Business Profile, Wikipedia article, and structured data on their own site tend to get panels reliably. Public figures with Wikidata entries and consistent web references get panels. Generic topic searches and organisations with minimal third-party coverage typically do not.

The signals above are listed here as sources Google draws on. They also have a dependency order, because some are gated on others: Wikidata needs independent references to exist before an entry will survive, and your own structured data reinforces recognition rather than creating it. How to Get a Knowledge Panel sequences the work and covers the diagnostic traps, including why Knowledge Graph Search API records are not evidence of panel eligibility.

How do you claim and manage your knowledge panel?

If a knowledge panel already exists for your brand or entity, Google allows the owner to claim it. Claiming does not give you full editorial control over the displayed data. It allows you to suggest corrections and flag inaccurate information. Go to the knowledge panel and click “Claim this knowledge panel”. You will need to verify ownership of an official source (typically your official website or a social profile).

After claiming, you can:

• Suggest changes to factual information
• Suggest a replacement for an existing featured image (Google will not add one by request where none exists)
• Flag information that is outdated or incorrect

Google reviews suggestions but does not guarantee they will be accepted. The displayed data is drawn from what Google considers the most credible sources, which may not be your own website.

Creators, publishers, and brands have a second route. A Google Search profile is auto-generated for entities that already have a knowledge panel, and claiming it feeds the panel a controlled avatar, your latest content, and a link through to the profile, or can trigger a panel where none existed. Claiming is gated on follower count and was US-only at launch, so the route is closed to most site owners, but checking whether a profile already exists for you costs nothing.

Entity SEO and the Knowledge Graph

The Knowledge Graph is one reason entity SEO has grown in importance. A site that is clearly a resolvable, unambiguous entity with consistent signals across the web is:

• More likely to have a knowledge panel
• More likely to be cited accurately in AI-generated answers, where models reason about entities and their relationships
• Less likely to be confused with a different entity of the same or similar name (disambiguation)

Establishing entity clarity involves: consistent brand name usage across all online presence, a verified Google Business Profile (for local businesses), structured data on the homepage marking up the organisation, named and credible authorship on content, and citations from independent sources that reference the entity by name.

This connects directly to how AI search systems handle entities. When Google’s AI systems retrieve content to ground an answer, they can draw on the Knowledge Graph and entity co-occurrence patterns to determine which sources are authoritative on a given entity. Other AI platforms such as ChatGPT, Claude, and Perplexity have no access to Google’s Knowledge Graph and resolve entities by other means, but the same underlying property helps across all of them: a brand that exists as a clear, well-supported entity is cited more reliably and accurately than one that is poorly defined across the web.

The broader practice of entity-focused optimisation is covered in entity SEO.

Frequently asked questions

Can I add my business to the Knowledge Graph directly?
No. Google does not accept direct submissions. The routes in are through Wikidata, Wikipedia (subject to notability requirements), a well-maintained Google Business Profile, consistent structured data on your site, and sufficient third-party coverage that Google can verify the entity.

Does a knowledge panel help rankings?
Not directly. The panel is a display feature, not a ranking signal. Indirectly, the factors that produce a panel (strong off-page entity presence, credible citations, structured data) do contribute to E-E-A-T signals and AI citation likelihood.

What if the information in my knowledge panel is wrong?
Claim the panel, then use the feedback option to flag incorrect information. For serious inaccuracies, the most reliable fix is updating the underlying source data: correct the Wikidata entry (with citations), update relevant Wikipedia content where applicable, and ensure your official site’s structured data is accurate. Google draws on these sources continuously, so corrections there propagate into the panel over time.

Footnotes

1. About knowledge panels: manage your panel — Google Search Help ↩

2. Organization structured data — Google Search Central ↩

3. Google’s Knowledge Graph explained — Ahrefs ↩