Entity Vector (Author Vector)
An Entity Vector (specifically Author Vector when applied to people) is the dense semantic embedding Google maintains for a real-world entity (person, organization, place) based on their accumulated web presence. Author Vectors are the 2026 backbone of E-E-A-T scoring — each author's vector encodes their expertise domains, citation density, and credibility, then influences every page they write.
Also called: Author Vector, Entity Embedding · Last updated: May 27, 2026 · By Joseph W. Anady
Why it matters.
Entity Vectors are how modern Google encodes the semantic identity of a real-world entity. They're calculated by embedding all content associated with the entity (pages they authored, mentions of them, their structured data) into a high-dimensional vector space. The vector clusters near other entities the system considers related (other authors in the same topic, organizations the entity works with, places they operate in).
How it works.
Author Vectors are built progressively. When Google identifies a new author (via Person schema with consistent sameAs chain across web pages), it starts accumulating an embedding for that author. Every new piece of content from the author updates the vector. Every external mention of the author updates the vector. Over months, the vector becomes a high-fidelity semantic representation of the author's expertise.
2026 reality check.
Author Vectors became the dominant entity-resolution backbone for E-E-A-T scoring in the March 2026 core update. Publishing 18 months of SEO content under 'Joseph W. Anady,' then suddenly a gardening post, would weaken the Author Vector for both topics. Topical discipline at the author level now matters as much as at the site level.
Data points
- Author Vectors became dominant entity-resolution backbone in March 2026 core update
- Established author vectors confer ranking lift over unknown bylines
- AI Overview citations favor pages with verified author entities (Leadgen Economy 2026)
- Wikipedia article is the strongest single signal for fast Author Vector establishment
- Topic drift weakens vectors — 18 months of SEO content + sudden gardening post dilutes both
First-hand insight from ThatDeveloperGuy.
ThatDeveloperGuy operates with a single consistent Author Vector for Joseph W. Anady across all 130+ client sites we operate, all 31 new TDG pages, all 13 city service-area pages, the /authors/joseph/ E-E-A-T hub, 175 dev.to articles, Zenodo papers, GitHub commits, and Wikidata Q139901957. The vector tightly clusters around: web development, SEO/AEO/GEO/AIO, Schema.org, federal contracting, SDVOSB. This tight clustering is by design — every published piece reinforces the centroid.
How TDG approaches it
TDG maintains Joseph W. Anady as the single author identity across the entire network. Person schema is identical across all sites. sameAs chain includes the same 20+ external references everywhere. The Author Vector clusters tightly around web development + SEO/AEO/GEO + federal contracting topics. We deliberately avoid topic drift to preserve vector tightness.
Common mistakes.
- Inconsistent author identity (different LinkedIn vs ORCID vs site bio = vector confusion)
- Mixing author topics (publishing AEO content then suddenly gardening dilutes the centroid)
- Missing sameAs chain in Person schema (Author Vector can't anchor without external references)
- Failing to identify the author at all (anonymous content has no Author Vector to work with)
- Using multiple author personas for the same person across sites
FAQ.
How do I know my Author Vector is being built?
Indirectly. Look for evidence that content under your byline ranks faster than content under unknown bylines on the same site. Author identity boosts ranking once the vector is established (typically 6-18 months of consistent publishing).
Can I have multiple Author Vectors for different topics?
Theoretically yes (different pen names per topic) but practically discouraged. The cost of maintaining multiple distinct identities (separate LinkedIn, ORCID, Wikidata for each) exceeds the benefit for most authors. Pick a topic focus and stick with it.
Does Author Vector affect AI Overview citation?
Yes substantially. AI Overview citations favor pages with verified author entities — the underlying mechanism is Author Vector recognition. Pages from established author vectors are cited more reliably than pages from unknown authors.
How long does it take to build an Author Vector?
6-18 months of consistent publishing under the same identity with a stable sameAs chain. Faster if you have a Wikipedia article (which immediately anchors the entity) or significant existing third-party press.
What kills an Author Vector?
Identity inconsistency (different bio data across sites), topic drift (suddenly publishing outside your domain), removed Wikipedia article, deleted Wikidata QID, removed sameAs targets, gaps in publishing (years of silence weakens the vector).
Maintained by Joseph W. Anady at ThatDeveloperGuy. Back to glossary · Suggest a term