Google: Ranking documents based on user behavior signals
How dwell time, click patterns, and refinement behavior feed into modern ranking decisions.
This patent family describes ranking adjustments driven by aggregated user interaction — not the click itself, but the pattern of clicks over time.
What the patent claims
The system observes sequences of clicks and refinements across many users, aggregates them into behavioral signals, and adjusts ranking based on which results resolved sessions cleanly.
The unit of measurement is the session, not the click. A single click is noisy; a satisfied session is not.
Why it matters for content
Content that resolves the job end-to-end quietly outperforms content that only wins the click. The implication is architectural: build pages that answer the whole question, not just the query.
Retention and internal navigation are secondary — the primary signal is 'did this session end here'.