Chunking
Chunking is how a retrieval system splits your page into smaller passages before indexing it, so AI engines retrieve and cite chunks of a page rather than the whole document.
Chunking is the step where your page gets cut into bite-sized passages. Before a retrieval system can search your content, it divides each page into chunks — often a few sentences to a paragraph each — and embeds them individually. The engine then retrieves and quotes the chunk, not the page, which is why AEO optimizes passages rather than documents.
How your content gets chunked is partly out of your control, but your structure heavily influences it. Clear headings and self-contained paragraphs encourage clean chunk boundaries that align with complete thoughts; a wall of text gets split arbitrarily, sometimes mid-idea, producing a chunk that's incoherent on its own and unlikely to be retrieved. Writing each paragraph as a standalone answer — the extractability pillar — is effectively writing your own chunks.
Example. A 2,000-word guide might be chunked into 30 passages. If your key answer sits in one tidy, well-headed paragraph, it becomes one strong chunk an engine can lift whole; if it's spread across a rambling section, no single chunk captures it.
Relevant pillar
Related terms
- Passage RetrievalPassage retrieval is the practice of finding and returning specific relevant passages from within documents, rather than whole pages, which is why AI engines cite paragraphs instead of articles.
- EmbeddingsEmbeddings are numerical representations of text that capture its meaning, letting AI systems find passages that are semantically related to a query even when they share no exact keywords.
- RAG (Retrieval-Augmented Generation)RAG is the technique behind most AI answer engines, where the model first retrieves relevant documents from the live web or an index and then generates an answer grounded in what it found.