Main area
Agency
geo

Entity SEO and the Knowledge Graph

Redaktion ·

Type “Mercury” into Google. Do you mean the planet, the chemical element, the singer Freddie Mercury, or the Roman god? A pure keyword search can’t tell — it only sees the character string m-e-r-c-u-r-y. A search that understands entities knows: these are four different things in the world, each with its own facts and relationships. This very leap — from strings to things — is the core of Entity SEO.

An entity is a uniquely identifiable object: a person, a brand, a place, a concept, a product. The decisive part is the unique identifiability. An entity has an ID, not just a word — in the Google Knowledge Graph an MID (Machine ID), in Wikidata a QID. Through this ID, “Apple the company” is cleanly separated from “Apple the fruit,” no matter how the word is spelled or in which language. A keyword is a character string; an entity is a thing with an identity.

Things, not strings: the Knowledge Graph

Google made this paradigm shift public in 2012 — with the Knowledge Graph and the slogan “things, not strings.” The Knowledge Graph is an intelligent model that stores real-world entities and their relationships to one another. At launch it contained, per Google, more than 500 million objects and over 3.5 billion facts and relationships between them — many times that today.

What matters isn’t the volume but the structure: the graph stores not just “Marie Curie” but also that she was married to Pierre Curie, had children with him, and that both shared a Nobel Prize. The entities are nodes, their relationships the edges between them. This interlinking lets Google do three things pure keyword search can’t: resolve ambiguity (Taj Mahal the monument vs. the musician), display context-relevant facts as a knowledge box, and surface surprising connections between topics. The graph is fed from public sources like Wikipedia and Wikidata as well as continuous web analysis.

Why this matters for SEO — and especially GEO

For classic SEO, thinking in entities means: Google evaluates not just whether your text contains the keyword, but whether your page cleanly handles a recognizable entity and links it with the right other entities. That’s the foundation of Entity SEO — establishing topics and brands as clearly defined things, not as keyword clouds.

With GEO — optimizing for AI-generated answers — this becomes even more central. An AI system answering a question and meant to name or cite a brand needs a clear notion of what that brand is. If your brand exists as a clearly defined entity with authoritative facts and links, it’s recognized more reliably, categorized correctly, and cited more readily. If it exists only as scattered character strings without a clear identity, it blurs — and an AI system that, in doubt, prefers the unambiguous source will pass it over. Entities are the bridge between your content and the machine understanding behind AI answers.

The levers: how to build an entity

A strong entity doesn’t arise from a switch but from consistency and authoritative linking. The most important levers:

1. Consistent brand and NAP signals. Name, address, phone number (NAP), and the brand spelling must be identical across your entire web presence — website, imprint, social profiles, directories. Contradictory signals make it harder for Google to attribute unambiguously. An entity that appears the same everywhere is easier to recognize as one thing.

2. sameAs in the Organization/Person schema. The sameAs property in JSON-LD markup is the explicit linking lever. It links your own entity with authoritative profiles of the same entity — Wikipedia, Wikidata, LinkedIn, official social accounts. With it you tell machines unambiguously: “this organization here is the same as the Wikidata object Q… and that LinkedIn profile.” For people this is the foundation of the author schema with sameAs, which establishes authors as traceable entities.

3. Authoritative mentions and a Wikidata entry. Entities are also strengthened by what others write about them. Mentions in trustworthy sources, press, and specialist portals solidify the entity — related to the concept of brand mentions for GEO. A maintained Wikidata entry (where notability criteria are met) gives the entity a machine-readable ID that everything else can attach to.

4. Clear, consistent entity mentions in your own content. Address entities unambiguously in the text instead of hiding them behind vague paraphrases. Link them with related entities — that helps both readers and the machine understanding place the topic in the right semantic environment.

The long game

You don’t build entities overnight. A new Wikidata entry, consistent sameAs signals, and first authoritative mentions take time before Google condenses them into a stable entity. But it’s an investment with growing returns: the more AI systems dominate search and answers, the more the question “do you exist as a clearly defined entity?” decides whether you stay visible or vanish into the noise. Whoever builds a clean entity early gains a head start that can’t be caught up overnight.

FAQ

What’s the difference between a keyword and an entity? A keyword is a character string — apple. An entity is a uniquely identifiable thing with its own ID — “Apple the company” (an MID/QID), cleanly separated from “Apple the fruit.” Search engines that understand entities resolve ambiguity and evaluate content by the thing it handles, not by the bare word.

What is the Google Knowledge Graph? A knowledge model introduced in 2012 that stores real-world entities and their relationships — “things, not strings.” It enables knowledge boxes, ambiguity resolution, and cross-topic connections. It’s fed from sources like Wikipedia and Wikidata as well as continuous web analysis.

What is the sameAs property for? sameAs in the Organization or Person schema explicitly links your own entity with authoritative profiles of the same entity — Wikipedia, Wikidata, LinkedIn, official social accounts. With it you tell machines unambiguously that these profiles mean the same entity, solidifying its identity in the Knowledge Graph.

Why is Entity SEO so important for GEO? AI systems that generate answers and are meant to name brands need a clear notion of what a brand is. Clearly defined entities with authoritative facts are recognized more reliably and cited more readily. Blurry brands existing only as character strings are, in doubt, passed over.

How do I build my own entity? Through consistent brand/NAP signals across all profiles, sameAs linking in the JSON-LD schema to Wikipedia/Wikidata/social, authoritative mentions in trustworthy sources, and ideally a maintained Wikidata entry. This takes time but increasingly pays off with growing AI visibility.

Themenuebersicht

GEOGEO-GrundlagenAI-Suche