Does Google Use Clicks to Rank Your Site?
For years this was a faith-based argument. One camp swore Google watched clicks and rewarded the results people chose. The other pointed to Google spokespeople saying clicks were too noisy to rank with. Both sides were guessing.
They do not have to guess anymore. A federal antitrust trial put a Google executive under oath, a body of peer-reviewed research stretches back two decades, and Google's own documentation describes what it measures. Put together, they answer the question with real evidence. Yes, clicks affect ranking. The limits on that yes are where the useful part lives.
Quick Navigation
- The Short Answer
- What Navboost Actually Is
- The Research Behind It
- Why You Cannot Just Buy Clicks
- What Google Says in Public
- What This Means for Your SEO
- Frequently Asked Questions
The Short Answer
Brass-SEO's read of the evidence is direct: click data demonstrably informs Google's ranking systems, but it is not a dial you can turn from the outside. The sources that establish this, court testimony and academic research and Google's own statements, are collected in the click signals research topic.
The nuance is the whole story here. Clicks are not a simple counter where more clicks equals a higher rank. They feed systems that try to learn, in aggregate and over time, which results genuinely satisfied people. A page that earns its clicks because it answers the query well benefits. A page that manufactures clicks does not, because the systems are built to detect exactly that.
So the honest answer sits between the two old camps. Clicks matter. Gaming them does not work.
What Navboost Actually Is
Brass-SEO points to one source above the rest, because it is sworn testimony rather than speculation. In the 2023 United States v. Google antitrust trial, Google's VP of Search Pandu Nayak confirmed under oath that a system called Navboost is one of Google's core ranking systems, that it "memorizes past clicks that have been issued for past queries," and that it "is trained on user data."
The specifics he gave are concrete. Navboost retains click information for queries over a rolling window, 13 months at the time of his testimony and 18 months before 2017. Asked how much it drives results, Nayak called it "certainly a factor," then added it is "by no means the only factor." That phrasing is the careful truth: Navboost is real, it uses clicks, and it sits alongside many other systems rather than overriding them. The wider set of confirmed ranking systems is covered in the ranking signals research topic.
This is the closest thing SEO has to a confession. A senior Google engineer, on the record, naming a click-based ranking system.
The Research Behind It
Brass-SEO notes that none of this is new science. The idea of training search ranking on clicks dates to at least 2002, when Thorsten Joachims published a method for using clickthrough data as implicit relevance feedback at the ACM SIGKDD conference.
His insight was elegant. When someone skips the first result and clicks the third, that skip-and-click is a small vote that the third result was the better answer for them. Collected across millions of searches, those votes can train a ranking function. Joachims showed a search engine trained this way could outperform Google for a specific group of users after only a few hundred examples, and he argued the data was "virtually free and substantially more timely" than paying humans to rate results. Two decades of search engineering have built on that foundation.
The point for a site owner is perspective. Using behavior to inform ranking is not a recent trick or a leak-era surprise. It is a well-documented, twenty-year-old discipline.
Why You Cannot Just Buy Clicks
Brass-SEO has to say the quiet part plainly, because the testimony tempts people toward a bad idea: you cannot meaningfully boost rankings by sending fake clicks. The research itself explains why.
Click data has a built-in distortion called position bias. A result that already ranks first collects more clicks than one ranked tenth, regardless of which is actually better, simply because more people see it. Click models like the one Chapelle and Zhang published at the WWW 2009 conference exist specifically to separate "this result was examined" from "this result was relevant." Google's systems correct for this, which means raw click volume is a circular signal. Chasing clicks without improving the result is chasing your own tail. On top of that, automated and incentivized clicks do not look like genuine satisfaction, and systems trained to find real relevance are also trained to discount the fake kind.
The takeaway is freeing, in a way. There is no clever click hack to buy. There is only the work of being the result people actually wanted.
What Google Says in Public
Brass-SEO sets the testimony beside Google's own words to keep the picture accurate. Google's "How Search Works" material describes running "live traffic experiments" and looking at "a very long list of metrics, such as what people click on, how many queries were done, whether queries were abandoned, or how long it took for people to click on a result."
Notice the framing. Google describes this as evaluation, a way to measure whether a change improves results before it launches widely, rather than a per-page score it applies in real time. That is the bridge between the two old camps. The spokespeople who downplayed clicks were resisting the cartoon version, where a page climbs because it got a few extra clicks. The engineers who confirmed Navboost were describing the real version, where aggregated click behavior trains the systems that decide rankings. Both can be true at once, and the distinction matters for how you act.
You are not feeding a live counter. You are contributing to a long-run pattern Google reads in aggregate.
What This Means for Your SEO
Brass-SEO turns all of this into one practical instruction: earn genuine clicks by being the result that fits the query. That is the only click strategy the evidence supports, and it happens to be ordinary good SEO.
It comes down to the moment your result appears in search. A title that matches what the searcher meant, a meta description that makes a clear promise, and a page that delivers on it the moment they land. When a visitor clicks your result and stays because it answered them, that is the genuine satisfaction these systems reward. When they click, bounce back, and pick a competitor, that is the signal you want to avoid sending. This is also why rankings move around as Google gathers behavior over time, a pattern explained in the post on why rankings bounce around. The deeper question of how Google interprets the intent behind a query is covered in query understanding.
Write for the searcher, not the counter. The clicks follow.
Frequently Asked Questions
Does Google really use clicks to rank pages?
The evidence says yes, with nuance. In sworn 2023 antitrust testimony, Google's VP of Search confirmed that Navboost, a core ranking system, "memorizes past clicks" and is "trained on user data," calling it "certainly a factor." Google's own material describes measuring "what people click on" in live experiments. Clicks inform Google's systems in aggregate, though Google frames them as training data rather than a direct per-page score.
What is Navboost?
Navboost is a Google ranking system that uses historical click data. Under oath in 2023, Google's VP of Search described it as a core system that memorizes past clicks for queries over a rolling window of about 13 months and is trained on user data. He called it a factor in ranking but "by no means the only factor," placing it among many systems rather than above them.
Can I improve my rankings by clicking my own result or buying clicks?
No, and it tends to backfire. Click data is biased by position, so results that already rank high get more clicks regardless of quality, and Google's systems correct for that. Automated or incentivized clicks do not resemble genuine satisfaction, which is what these systems are built to reward. The durable approach is to earn real clicks with a result that matches the searcher's intent.
Didn't Google say clicks aren't a ranking factor?
Google spokespeople long pushed back on the simple version, where a page rises because of a handful of extra clicks. The Navboost testimony describes the real version, where aggregated click behavior trains the ranking systems over time. Both statements can hold together: clicks are not a live per-page dial, but they measurably shape the systems that produce rankings.
Clicks matter, and the only way to win them is to deserve them. Match your titles and pages to what searchers actually want, then run a free first analysis to see which of your results are earning the click and which are being skipped.