# Chunking

> Chunking is the splitting of a document into smaller passages so a retrieval system can index and return the relevant part.

*Source: [https://contentcucumber.com/aeo-glossary/chunking/](https://contentcucumber.com/aeo-glossary/chunking/)*

*Topic: Getting found. Also called: Content chunking, Text chunking*

---

## Why it matters

Retrieval systems do the splitting themselves, and Google states that breaking content into tiny pieces is not required for its AI features.

## Example (illustrative, not a client result)

A retrieval system divides a long returns policy into sections and returns the section about refund timing for a question about refunds.

## What to check

Read each section of a key page apart from the others and confirm the section names its subject and keeps its conditions beside the claim.

## What this does not tell you

Each system sets its own chunk size, and Google lists chunking among the tactics a site can ignore for Google Search.

## Sources

- [Microsoft Learn: chunk large documents for vector search](https://learn.microsoft.com/en-us/azure/search/vector-search-how-to-chunk-documents) (Platform documentation)
- [Google Search Central: optimizing for generative AI](https://developers.google.com/search/docs/fundamentals/ai-optimization-guide) (Platform documentation)

*Evidence type: Documented concept. Sources checked on 2026-11-21*

## Related terms

- [RAG](https://llms.contentcucumber.com/aeo-glossary/rag.md)
- [Extractability](https://llms.contentcucumber.com/aeo-glossary/extractability.md)
- [Passage retrieval](https://llms.contentcucumber.com/aeo-glossary/passage-retrieval.md)
