E-E-A-T, entities and knowledge graphs: how search engines and AI decide whom to trust

Where did link-based authority come from?

Brin and Page (1998) described PageRank as the stationary distribution of a random surfer who follows links with probability d and jumps to a random page otherwise: PR(A) = (1 − d) + d Σ PR(T_i)/C(T_i), summed over pages T_i linking to A with C(T_i) outgoing links (the paper's original, un-normalised form). Its lasting idea is that an endorsement counts in proportion to the endorser's own standing. Modern systems use many more signals, but mentions and links from respected, relevant sources remain a strong form of corroboration — for search engines and for AI assistants that summarise what the web says about you.

What is E-E-A-T, and how is it used?

E-E-A-T — experience, expertise, authoritativeness and trustworthiness — is the framework in Google's Search Quality Rater Guidelines. Google added the first E, experience, in December 2022 to recognise first-hand knowledge, and describes trust as the most important member of the family. Raters' assessments are used to evaluate how well ranking systems work; Google states they do not directly affect the ranking of individual pages.

For a publisher, the operational reading is: show who wrote or reviewed the content and why they are qualified, show evidence of first-hand experience, cite sources, keep content accurate and dated, and make the organisation behind the site easy to identify and contact.

What are entities, and why do they matter for AI?

An entity is a uniquely identifiable thing — an organisation, person, place, product or concept — as opposed to a string of words. Search engines maintain knowledge graphs of entities and their relations (Google introduced its Knowledge Graph in 2012), and language models learn associations between entity names and facts during training. When your organisation is described consistently and linked to authoritative identifiers, systems can resolve “SEOCells”, “seocells.ai” and “OmCells R&D Labs LLP” to the same entity and attach facts to it.

• Publish Organization markup (name, legalName, url, logo, address, contactPoint) and Person markup for named experts.
• Use sameAs to link to official profiles and authoritative records (company registry, LinkedIn, Wikidata if an entry exists).
• Use one short description of the organisation everywhere: site, markup, profiles and press boilerplate.
• Declare your preferred site name with WebSite markup so Google can show it correctly.
• Build an author page per expert, with credentials and links to their publications or academic profiles (ORCID, Google Scholar).

What is topical authority?

Topical authority is an informal term for a site's demonstrated depth on a subject: many interlinked pages that together cover a topic's sub-questions accurately, written by people with relevant expertise. It is not a single published Google metric, but it describes what helpful-content guidance rewards — comprehensive, original, first-hand coverage — and gives retrieval systems more relevant passages to choose from.

Why does brand memory in AI models lag behind the web?

What a model knows without searching reflects its training data up to a cut-off, so a new brand can be unknown to the model even when its site is indexed and retrievable. Retrieval-augmented answers can still cite the brand, but unaided recall (“What is SEOCells?” with search disabled) only improves after the brand is widely and consistently described on the web and later models are trained. This is why entity work is long-term, and why it should be measured separately: with and without search.

Questions answered

References

1. Brin, S. & Page, L. (1998). The Anatomy of a Large-Scale Hypertextual Web Search Engine. Computer Networks and ISDN Systems 30(1–7). doi:10.1016/S0169-7552(98)00110-X
2. Google Search Central Blog (December 2022) — E-A-T gets an extra E for Experience
3. Google — Search Quality Rater Guidelines (PDF)
4. Google Search Central — Creating helpful, reliable, people-first content
5. Google Search Central — Introduction to structured data
6. Google Search Central — Site names in Google Search

Search engines and AI assistants change often; check the linked documentation for the latest details. Nothing in this guide is a promise of rankings, citations or results.

Related guides

• How to get your business cited in ChatGPT, Gemini, Perplexity and Google AI Overviews
• How AI search engines retrieve, rank and cite sources: a technical primer
• International SEO and AEO: countries, languages, hreflang and local search engines