Google search representatives have consistently and clearly stated that they do not use data from Google Analytics to rank websites. But, there are still discrepancies between what Google says and what SEOs believe. Despite Google’s public statements, some search marketers continue to believe that bounce rate is in some way a ranking factor.
Why do they believe this? Is there any validity to the claims against Google’s public statements?
Does Google use your website bounce rate to rank webpages?
Does Bounce Rate Affect Search Rankings?
Google has stated that bounce rate is not a ranking factor for over a decade.
“Google Analytics is not used in search quality in any way for our rankings.” – Matt Cutts, Google Search Central, February 2, 2010.
“…we don’t use analytics/bounce rate in search ranking.” – Gary Illyes, Webmaster Trends Analyst at Google, Twitter, May 13, 2015.
“I think there’s a bit of misconception here that we’re looking at things like the analytics bounce rate when it comes to ranking websites, and that’s definitely not the case.” – John Mueller, Webmaster Trends Analyst at Google, Webmaster Central office-hours, Jun 12, 2022.
Why Google Doesn’t Use Bounce Rate As A Ranking Factor
There are technical, logical, and financial reasons why it is improbable that Google would use bounce rate as a ranking factor.
This can be summarized by looking at three primary facts:
- What bounce rate measures.
- Not all websites use Google Analytics.
- Bounce rate is easily manipulated.
What Does Bounce Rate Measure?
A lot of the confusion around bounce rate can be cleared up once people understand what bounce rate actually measures.
Bounce rate is a Google Analytics metric that measures the percentage of single-page sessions (no secondary hits) to your site divided by the total sessions.
Marketers often misinterpret this metric to mean that the webpage did not provide what the user was looking for.
But, all a bounce means is that a measurable event (secondary hit) did not occur. Technically speaking, Google can’t understand how long a user spends on a page unless a second hit occurs. If a user spends 2.5 minutes reading the webpage (as the Backlinko study found correlates with page rank) and then exits, it will count as a bounce because they did not send any subsequent hits to GA.
So, keep in mind that bounce rate does not necessarily indicate a bad user experience.
Users may click on a result, read it, and leave because their query was satisfied. That’s a successful search, and it doesn’t make sense for Google to penalize you for it.
This is why Backlinko’s study, looking at the time on the page, does not support the claim that bounce rate is a ranking factor.
Not All Websites Use Google Analytics
While Google Analytics is a widely-used analytics tool, not all websites use it.
If Google used bounce rate as a ranking factor, it would have to treat websites with the GA code differently than those without the GA code.
If websites without the GA code were not graded by bounce rate, they would theoretically have greater freedom to publish whatever content they wanted.
And if this were true, it would be illogical for any marketer to use the GA code. You see, Google Analytics is a “freemium” service. While most businesses use their service for free, large companies pay a monthly fee for more advanced features.
The paid version is called GA 360, and pricing starts at $150,000 annually.
There are 24,235 companies currently using GA 360.
That equates to $3,635,250,000 per year (on the low end.)
Using bounce rate as a ranking factor is not in Google’s financial interest.
The Bounce Rate Can Be Easily Manipulated
Some of you may still not be convinced.
You may have even noticed a correlation between average position improving and bounce rate decreasing in your daily practice.
While bounce rate and average ranking may correlate, they certainly are not dependent on each other.
What happens when you increase your bounce rate? Do the rankings fall back to where they were?
Bounce rate is easy to manipulate, and you can try this experiment yourself.
You will need to increase and decrease your bounce rate for this test while comparing the average position for a search query over time.
Remember that the bounce rate is sessions with zero secondary hits / all sessions. So, all you need to do to reduce your bounce rate is send a secondary hit.
You can add a second pageview event using Google Tag Manager.
Do not make any other changes on-page or off-page; chart your average rankings over three months.
Then remove this extra pageview tag.
Did your average rankings increase and decrease in unison with modifying the bounce rate?
Below is a graph of a quick version of this study on my own site; one that shows no correlation between bounce rate and average position.
No, bounce rate is not a Google ranking factor. Bounce rate is not a reliable measurement of the relevance of web pages – and Google has repeatedly said it does not use it for rankings.
With big industry names like Rand and Backlinko putting their weight behind bounce rate as a ranking factor, confusion is understandable.
Experts have tested this user signal with varying results.
Does The Myth Have A Basis In Fact?
As recent as late 2021, recognized resources have perpetuated the myth that a website’s bounce rate is a ranking factor.
Rand Fishkin, Founder of MOZ, tweeted in May 2020 that “…Google uses (relative) bounce rate (or something that’s pretty darn close) to rank websites.”
Backlinko published an article (June 2020) about bounce rate saying that “bounce rate may be used as a Google Ranking factor.”
They cite an industry study they ran and claim it found a correlation between first-page Google rankings and bounce rate.
Later the same year, Semrush reinforced this claim in December 2020, saying, “Bounce rate is an important ranking factor.”
They did not provide evidence to back up the claim.
HubSpot included bounce rate in a rundown of “all 200 ranking factors” ( in a cheat sheet to Google’s known ranking factors in July 2021.
Bounce rate is included as a factor twice under “site-level factors” and under “user interaction,” with no supporting evidence for their claim.
So, let’s take a look at the evidence, shall we?
Could Bounce Rate Be A Google Ranking Factor?
In “How Search Works,” Google says, “…we use aggregated and anonymized interaction data to assess whether search results are relevant to queries.”
The vague wording here has led to many assumptions about what “interaction data” Google uses to inform its machine learning systems.
Some marketers believe the “interaction data” includes the bounce rate of a website.
They use a handful of studies to support this hypothesis.
The Backlinko study mentioned above ran a subset of domains from their own data set through Alexa to determine a site-wide time on site.
They discovered that the average time on site for a Google first-page result is 2.5 minutes.
The study goes on to clarify:
“Please keep in mind that we aren’t suggesting that time on site has a direct relationship with higher rankings. Of course, Google may use something like time on site or bounce rate as a ranking signal (although they have previously denied it). Or it may be the fact that high-quality content keeps people more engaged. Therefore a high time on site is a byproduct of high-quality content, which Google does measure. As this is a correlation study, it’s impossible to determine from our data alone.”
Brian Dean confirmed in reply to a comment that the study did not actually look at bounce rate (or pageviews).
The Backlinko study, which supposedly found a correlation between first-page Google rankings and bounce rate, did not look at bounce rate.
Rand Fishkin stated that Google uses relative bounce rate to rank websites, and discussed this topic with Andrey Lipattsev, Search Quality Senior Strategist at Google Ireland, in 2016.
Rand described tests he had been running where he would ask people to do a search, click on the seventh result, and then observe over the next 24 hours what happened to that page’s ranking for that query.
The results were inconclusive.
In seven to eight tests, rankings improved for a day or two. Rand said the rankings did not change in four to five tests.
Andrey responded that he believes it’s more likely that the social mentions, links, and tweets (which are basically links) throw Google off temporarily until they can establish that the “noise” is irrelevant to the user intent.
Both the Backlinko study and Rand’s experiments helped shape the bounce rate myth. But the study didn’t look at bounce rate, and Rand’s experiments did not prove a causational relationship between user behaviour and ranking.
Summary
Some experiments may have demonstrated a correlation between bounce rate and SERP rankings in certain situations.
Other experiments haven’t done that, but people reference them as if they’re proof.
“Confirmed ranking factor” requires a high degree of evidence. No one has proven a causal relationship.
You need to watch out for this in SEO, even when reading trusted sources.
SEO is complicated. Google representatives and industry pros love to joke that the answer to every SEO question is: “It depends.”
We’re all looking for ways to explain success in SERPs.