Noindex vs canonical for thin pages: the choice changes your crawl economics

Noindex vs canonical for thin pages: the choice changes your crawl economics

Question: for near-duplicate or thin pages (filtered listings, light variants), is noindex or a canonical the better authority-preserving move?

Evidence: Log-file studies repeatedly show noindexed URLs still get crawled — Google must fetch them to see the tag — so noindex saves index space but not crawl budget. Canonicals consolidate signals to the target but are treated as a hint, and in one shared dataset ~15% of rel=canonical declarations were ignored when content differed too much.

Nuance: use canonical when pages are genuinely equivalent and you want equity to flow up; use noindex when pages are useful to users but should never rank. For pure crawl-budget pressure, neither beats not generating the URLs at all.

Method note: synthesis of two public log-analysis writeups plus one agency canonical-honor audit.

Caveat: Google's handling evolves; honor rates vary by site trust.

Confidence: medium