Blog
Semantic SEO
Writing for meaning: entities, definitions, attribute-value facts, semantic distance and pages that actually answer.
Semantic SEO is writing for meaning rather than for keywords. Engines model things, their attributes and the relations between them. A page that names the right entities, defines them and states their attributes precisely is cheaper for an engine to understand and easier to trust.
This topic covers the building blocks: entities, definitions, attribute-value facts, semantic distance, and the difference between a relevant page and a page that actually answers.
Start here: the reading path
- 1Entities and the Knowledge Graph: How Search Engines Understand Things, Not Strings
Entities: how engines understand things, not strings.
- 2Writing Definitions Machines Can Parse
Writing definitions a machine can parse and quote.
- 3Entity-Attribute-Value: The Data Model That Organizes Knowledge
Organizing facts as entity, attribute and value.
- 4Relevance Is Not Responsiveness: The Distinction That Reshapes Content
Why a relevant page can still fail the searcher.
- 5SERP clustering: let the search results decide which keywords share a page
Let the search results decide which keywords share a page.
- 6Keyword cannibalization: when your own pages compete for the same search
Find and fix pages that compete with each other.
- 7Cost of Retrieval: Why Easy-to-Process Content Wins
Why content that is easy to process wins ties.
- 8How Google AI Overviews choose the pages they cite
How the same principles decide which pages AI answers cite.
All articles in this topic
12 articles in this category
SERP clustering: let the search results decide which keywords share a page
SERP clustering groups keywords whose top results overlap, so each group gets one page. How it works, how to pick a threshold, and when to overrule it.
Keyword cannibalization: when your own pages compete for the same search
Cannibalization is two of your pages chasing one search. It isn't a penalty, but it splits signals. How to spot it in Search Console and fix it safely.
Mixed-intent results: when Google shows guides and shops for the same search
Some searches return guides, shops and maps on one page. How to read a mixed result, decide which intent to serve, and when to build two pages instead.
How Google AI Overviews choose the pages they cite
What Google says about the pages AI Overviews and AI Mode link to, what that means for your site, and the controls you have. With Gulf and Egypt examples.
Getting cited by ChatGPT, Gemini and Perplexity
How AI assistants find web pages, which crawlers to allow in robots.txt, and what makes a page worth citing. Facts from OpenAI, Google and Perplexity docs.
Query fan-out: writing for the questions behind the question
AI search splits one question into many smaller searches. How query fan-out works, how to find the sub-questions, and how to cover them without thin pages.
Writing Definitions Machines Can Parse
Anatomy of a machine-parseable definition: declarative genus-differentia sentences, microsemantics, and markup that let engines lift facts without inference.
Entities and the Knowledge Graph: How Search Engines Understand Things, Not Strings
Entities are singular, unique, well-defined, distinguishable concepts. How definitions turn strings into knowledge-graph nodes engines can match, lift, cite.
Cost of Retrieval: Why Easy-to-Process Content Wins
The cost of retrieval is the effort a search engine spends to process a page. How technical and semantic cost shape rankings, crawling, and eligibility.
Semantic Distance: How Far Is Your Content From the Query?
Semantic distance measures the conceptual gap between a query and a document. What widens it, how search engines estimate it, and how publishers narrow it.
Relevance Is Not Responsiveness: The Distinction That Reshapes Content
Relevance and responsiveness are scored as two different tests: one connects a document to a query, the other decides whether the query ends satisfied.
Entity-Attribute-Value: The Data Model That Organizes Knowledge
Content hallucinates because it has nothing real to say. Entity-Attribute-Value is the structural fix: verifiable facts stored as records claims trace back to.