NavBoost & Clickstream Signals

**Sources:** Content Warehouse leak attributes · US v. Google antitrust trial testimony (Pandu Nayak, 2023) · SparkToro / iPullRank analysis.

System overview

**NavBoost** is an internal Google system that adjusts rankings using aggregated user navigation behavior from search results. It was referenced in leaked API docs and confirmed in federal antitrust proceedings.

Key leaked / trial-referenced attributes

Attribute	Interpretation (analyst consensus)
`goodClicks`	Positive engagement clicks from SERP
`badClicks`	Quick return / dissatisfaction clicks
`lastLongestClicks`	Dwell / longest click patterns on result set
`unsquashedClicks`	Raw click counts before normalization
`navBoost`	NavBoost scoring container / re-rank feature

Google public position vs. leak

Topic	Public messaging	Leak / trial signal
Click data	Often minimized as ranking input	Long-running click-based re-ranking
User behavior	"Helpful content" abstractly	Quantified good/bad click classes

Operational recommendations (crypto / YMYL)

1. **SERP snippet honesty** — Title/H1/meta must match page content; bait titles create bad clicks on regulated topics.

2. **Intent satisfaction** — Exchange pair pages: rates, limits, steps above the fold; reduce pogo-sticking to competitors.

3. **Brand queries** — NavBoost favors sites users recognize; invest in brand search volume and consistent NAP/entity.

4. **A/B titles carefully** — Winning CTR with misleading claims increases bad-click risk on YMYL pages.

Audit mapping

Audit item	Why
EE18, EE19	Trust / transparency reduces bad clicks
YY02, YY03	Harm prevention — misleading claims → dissatisfaction
Checklist title/meta items	titlematchScore alignment
Pro bot / render gap	If Googlebot sees different content → engagement mismatch

Сливы Google + рекомендации по аудиту — Quickex