NavBoost

NavBoost is a Google ranking system that uses data about how people have clicked on search results in the past to adjust future rankings. Its existence became public through testimony in the US government’s antitrust case against Google in 2023, and it also appears in the Google documents that leaked in 2024.

Written by Paul Hoda, SEO consultant since 2005. Last updated 8 November 2026.

Where the information comes from

For years Google gave mixed messages about whether clicks influence rankings. That changed in late 2023, when Google search executive Pandu Nayak testified in the US Department of Justice antitrust trial. He described NavBoost as one of Google’s important signals and explained that it learns from user interactions with search results.

NavBoost was later mentioned many times in the Google search documents leak of 2024. Those documents list data fields but do not explain how they are weighted, so they confirm that NavBoost exists without revealing exactly how it works.

What is known about how it works

Based on the trial testimony and evidence:

  • NavBoost uses historical data on which results searchers clicked for a given query.
  • It considers about 13 months of click data. Testimony said this window had earlier been 18 months.
  • The data can be broken down by factors such as location and device type, so results can differ for mobile and desktop or for different areas.
  • It helps refine a set of candidate results rather than finding pages from scratch.
  • A related system called Glue was described as doing similar work for other features on the results page.

NavBoost is not new. Testimony suggested it has existed in some form since the mid-2000s.

What it does not mean

The evidence supports the idea that user signals feed into rankings, but it does not mean every click on every page moves rankings in real time. NavBoost works on aggregated, historical patterns across large numbers of searches. Pages with little search traffic will have little click data for it to use.

Common community theories, such as an exact value for dwell time or a specific “good click” threshold, are not confirmed. Treat precise claims about NavBoost’s internals with caution.

What it means for SEO

If Google learns from how searchers behave, the best long-term approach is to satisfy them. In practice that means:

  1. Write title tags and descriptions that accurately describe the page, so the right people click.
  2. Match the search intent behind the query so visitors find what they expected.
  3. Answer the question quickly, without forcing people to scroll past filler.
  4. Reduce reasons for searchers to return to the results unsatisfied, known as pogo-sticking.
  5. Build a recognisable brand, since people are more likely to choose a name they know.

Common mistakes

  • Trying to manipulate clicks with bots or click farms. Google has long worked to filter fake interactions, and this breaks its spam policies.
  • Writing clickbait titles that win the click but disappoint the visitor.
  • Assuming NavBoost replaces content quality and links, rather than working alongside them.

Key Takeaways

  • NavBoost uses past search click data to refine Google’s rankings.
  • It was described under oath in the 2023 US antitrust trial.
  • Testimony said it uses about 13 months of click data.
  • It works on aggregated patterns, not individual clicks in real time.
  • The best response is to match intent and genuinely satisfy searchers.

Related Terms

Not sure how your own site measures up? Request a free SEO review, or keep learning in the SEO Encyclopedia.