Short answer
Yes, in aggregate. In October 2023, Google’s head of search testified under oath that a system called NavBoost uses about 13 months of click data and is one of Google’s important signals. That is not the same as your bounce rate or time on site, which Google does not take from your analytics. Clicks shape rankings through patterns across many searches, not through any single metric you can tune.
For years Google gave carefully worded answers about whether clicks affect rankings. The answer most people heard was no.
Sworn testimony in a United States antitrust case changed that. It also matters for recovery, because it helps explain why a demotion rarely reverses overnight, even after the cause is fixed.
What was actually revealed
- On 18 October 2023, Pandu Nayak, then Google’s Vice President of Search, testified in United States v. Google.
- He described NavBoost, a system that uses aggregated click data for searches and results, as one of the important signals Google has.
- It works from a rolling window of about 13 months of data, reduced from 18 months around 2017, and has been in use since roughly 2005.
- Google’s own How Search Works pages say it uses aggregated and anonymised interaction data to judge whether results are relevant.
- Internal documentation that leaked in 2024 also referred to NavBoost and click-related measures.
None of this came with weights or formulas. What is confirmed is that click behaviour feeds ranking. How much, for which queries and in what way remains Google’s to know.
What it does not mean
- It does not mean your Google Analytics data is used. Google has said repeatedly that it does not take ranking signals from site owners’ analytics.
- It does not make your page’s click-through rate a single factor you can raise to climb. The data is aggregated across many searches and many results.
- It does not mean one bad week ruins a page. A 13 month window smooths out short swings in both directions.
Why bounce rate is the wrong number to watch
An analytics bounce is a visit that views one page and leaves. For many searches, that is exactly what success looks like.
Someone searching for a cooking time, a phone number or a definition finds the answer and goes. A high bounce rate on that page means it worked. Treating it as a ranking problem leads people to pad pages and bury answers, which makes them worse.
The behaviour that plausibly matters is different: whether people searching come back to the results to try something else, because the page did not give them what they needed.
What this means for recovery
Some of this is confirmed and some is reasonable inference. It is worth being clear which is which.
- Confirmed: click behaviour over roughly a year feeds ranking.
- Reasonable inference: after a demotion, a page earns fewer clicks, and that reduced history is remembered for months. That is one plausible reason recoveries tend to be gradual rather than sudden.
- Also inference: pages that genuinely satisfy searchers should build a better interaction history over time, which supports a recovery once it starts.
The measured pace of recovery is covered in how long Google updates actually take and my recovery stalled at partial.
Do not buy clicks
Every revelation about clicks produces a wave of services selling them. Bots or paid workers search for your keyword and click your result.
- It is a policy risk. Sending automated queries to Google falls under its machine-generated traffic policy, one of the sixteen in every Google spam policy explained.
- It works against the window. A short campaign is a small fraction of more than a year of real data.
- It fixes nothing. If real searchers are not satisfied, fake ones only delay the moment you find out.
What actually improves real clicks
- Write titles and descriptions that match the page. A misleading title can win the click and lose the searcher straight back to the results.
- Answer the question early. The first screen should tell the searcher they are in the right place.
- Match the intent behind the query. Someone ready to buy and someone researching need different pages.
- Make the page usable on a phone, fast and readable without fighting adverts.
- Offer the obvious next step, so a satisfied reader can keep going instead of searching again.
None of these are tricks. They are what a good page does anyway, which is precisely why they are safe to rely on.
This is not your cause if
- Your drop matched a confirmed update and spread across the whole site. Start with core update vs spam update.
- Search Console shows a manual action. That is a human decision with its own fix.
- Pages fell out of the index. That is technical, and clicks have nothing to do with it.
More claims like this one, each checked against what Google has actually said, are collected in Google SEO myths.
Pages ranking but not satisfying searchers?
A gap between what your titles promise and what your pages deliver costs clicks over time. Send your Search Console access and within 48 hours I will tell you where that gap is widest.
Get free recovery audit