Semantic SEO: Embeddings and Chunking for AI | Chris Green
With Chris Green, SEO consultant and UK Search Awards winner · Streamed live · 49 min
What this episode covers
Chris Green explains vector embeddings with a wedding seating-plan analogy, then covers cosine similarity, dot product and content chunking. He shares practical uses such as redirect mapping and checking how well a page matches a query, and concludes that chunking advice comes down to clear, concise writing.
Key takeaways
- ✦Structure pages into self-contained sections with clear headings, since AI retrieves chunks.
- ✦Repeat each AI query across several tools before judging whether your brand is visible.
- ✦Use embeddings for practical jobs like redirect mapping and page-to-query matching.
- ✦Build a test site to learn how traffic and rankings behave before advising clients.
Key moments
- 16:30What embeddings are
“Turning human readable words into kind of machine readable formations.”
- 22:17Cosine similarity explained
“If it's a cosine similarity of one, that means they're identical.”
- 28:45Chunking content
“Break your things up into like conceptually distinct chunks.”
- 34:48Embeddings for redirect mapping
“One of the ones I've used it the most for is redirect mapping.”
- 40:40One question, one answer
“One question, one answer. Small chunks, concise writing, do good for humans.”
Quotes are from the episode transcript, lightly cleaned of auto-caption errors.
Questions answered in this episode
- What are embeddings in simple terms?
- Chris describes them as turning human readable words into machine readable formations, numbers that machines can derive meaning from and use to understand similarities between words.
- How should I structure content so AI can retrieve it?
- Break content into conceptually distinct chunks with clear headings, answer one question per chunk concisely, and use standard semantic HTML such as headings, paragraphs and lists.
- Why can't I find my brand in ChatGPT results?
- Chris says LLM results are highly variable, so you should repeat the search several times across different tools and check which sources get quoted, then invest in PR on editorial sites.
More episodes on AI Search & LLMs

Keyword Research Beyond Tools for AI Search | Venkata Pagadala
Venkata Pagadala

AI SEO: How to Rank in AI Search | Ross Hudgens
Ross Hudgens

AI Search Optimization: How to Test ChatGPT and Gemini | Sanjay Singh
Sanjay Singh

SEO Reporting with AI and Search Console | Aaron Abbott
Aaron Abbott

SEO Reporting: Leading the Internal Search Narrative | Tyson Stockton
Tyson Stockton

LLM SEO: Track What AI Says About Your Brand | Bernard Huang
Bernard Huang