High Bounce Rate – What to Do?

SEO Hangout Panel bounce rate issue

Recently the SEO hangout panel on Google+ has been getting a lot of great questions and even better answers, definitely a community to watch out for as you can learn a lot. A recent discussion peaked my interest and I decided to cover it a bit and share a few findings and advice from some great SEO experts here on our blog.

The discussion was started by Rick Eliason on topic of high bounce rate, the question was:

 “So, a high bounce rate is not necessarily a bad thing if you are attracting the right audience and giving them exactly what they are looking for. But how do the search engines know whether a high bounce rate is the good or bad kind? What supporting metrics are they using to make this decision?”

A great conversation started from this seemingly simple question, and some of the answers hit the spot. The first thing you need to understand before I give you some of the answers is that the bounce rate is a metric that we as users and webmasters see in Google Analytics, but it’s not something that Google uses for their records. They have stated the fact that they do not use GA data for their ranking methods, so what we see and what Google uses are two different things.

Back to the discussion, Rick first started with his understanding of how Google looks at certain cases and counts or discards the bounce rate issue:

“* When searchers close the window or search for something different entirely (problem solved)
* When users have a history or returning to your site over and over
* When users check out more posts/resources by the same author/brand (Google+)
* When users stay on the site for a sufficient amount of time (depending on length of content, query etc.)
* When branded search terms are included in the query
* When similar search queries/results produce similarly high-bounce outcomes” 

One of the first comments that drew my attention was from Lyndon NA:

“The “bounce Rate” that people reference is a metric from GA – so technically, G cannot be using it.

Yet, Google is looking at things like “dwell time” and “long clicks”.
If you use logic, good sites are those that people spend time on, return to etc.
So G msut be getting that from somewhere.
Be it that the simplistic BBtSERPs was to simple and thus denied, or if G are using things like using ISP/server traffic logs, or using bsic traffic models … G are likely doing something to monitor and consider traffic and it’s behaviour.”

More from Lyndon on bounce rate and Google’s possible use of a similar metric:

“Of course Google do not use Bounce Rate;
1) It’s a GA measurement, and G don’t use GA for direct ranking influence
2) Not enough people use GA
3) It’s a darn noisy signal
4) It’s not directly indicative of Quality or Value

No one will dispute that.
Wht you don’t seem to be recognising is;
A) G may have a different label for it
B) G are likely using a different Data Source
C) G are likely to be using Qualifiers/Disqualifiers (time/action/volume etc.)
D) G may be inverting the view (looking at Retention and Repeat hits, rather than abandonments or singular hits etc.)”

Gordon Campbell gave an interesting twist on the story by bringing a quote from this post:

“This is simply a gross misunderstanding of the actual metric that Google DOES look at, which is the ‘return-to-SERP’ metric. That metric, in a nutshell, is how often after clicking on a specific result users come back to Google’s search page and try a different result or a different query – i.e. the page they clicked on did not give them what they’re looking for. Pages with a high return-to-SERP metric are likely to be not relevant for the specific search query in question, and might as such lose ranking in Google.'”

So, is there such a thing as a “return to the SERP” metric”? Lyndon NA thinks it doesn’t exist:

“And yet a Googler has stated that they do Not use return to SERP(Bounce Back to SERPs) as a signal – as it would be to noisy!
(Again – I think this is because it is too broad – it needs additional Qualifiers/Disqualifiers to reduce the noise/amp the signal)”

We also covered the issue on our blog before where we busted the myth of return to serp metric, but some people still disagree.

The biggest question is whether bounce rate is a ranking signal or not? There is no official confirmation that it is, nobody at Google will give us the honors to confirm, on the other hand no one denies it, so it might be one of the many signals. Bottom line is that high bounce rate doesn’t necessarily mean a bad thing. One of the participants in this discussion gave an example where they changed a few things on their site which led to a great increase in calls from customers, increased customer usability, but at the same time bounce rate went sky high. It all depends on what you are aiming for, don’t be alarmed if your bounce rate is high, analyse the situation and try to find out why and how you can improve the user experience.

My advice, visit the panel, read the entire discussion, join other discussion and learn a bunch of new stuff, get your questions answered and stay on top of the SEO game.


1. https://plus.google.com/109130734805581958797/posts/3qXPjQAjQjf

2. http://www.custard.co.uk/is-google-using-bounce-rates-to-rank-websites/

3. http://dejanseo.com.au/seo-myth-busted-bounce-to-results-is-a-negative-signal/

Zac is a link building specialist and a seasoned SEO professional. He manages a team of link builders and actively promotes some of Australia's biggest brands. Zac is an active blogger and also maintains Dejan SEO's news and updates section.

More Posts - Website

12 thoughts on “High Bounce Rate – What to Do?

  1. Hey Zec…I also slightly looked at the discussion as it was happening but did not go into it, thanks for breaking it down.
    Question, could you please reference where to find this :”They have stated the fact that they do not use GA data for their ranking methods,”. I know about it but I don’t know where it comes from.
    “So, is there such a thing as a “return to the SERP” metric”? – Yes there is and I totally agree with Gordon…I think he sums up all pretty well.

  2. That’s too simplistic. What about a page of song lyrics that has a 90% bounce rate. Does that mean that you improperly optimized? No. Some content is deliverable and digested on a single page.

  3. Hey Gordon, thanks for this reference. Some time a go I was talking in person with someone that organizes a conference which includes Google people and re-known SEOs, etc…( and don’t want to tell his name, but he is well known on the international field and writes for SEL) whom have kind of “confirmed” very clearly that they do use it…I said “but Google already confirmed it publicly that they do NOT use it as ranking factor”…and then he answers back “then why would they have invested (and still do) so much money on analytics?”.

  4. agree Jeremy, but song lyrics maybe is not the best example as it usually contains long content which is consumed for some time, so not being considered bounce rate but exit rate in case they leave right after…I would say a “when did XYZ born” would be more of such a query in which there is a 90% bounce rate but the user was satisfied with the answer.

  5. Google can’t use GA data as a ranking factor for several reasons. For one thing, The vast majority of hobbist webmasters don’t have a clue what GA is.

    Many enterprise-level don’t use GA either. It’s pretty limited once you get into more advanced campaigns – the way it handles multi-touch attribution is pretty crappy by comparison, for example. Most large companies want their Analytics to encompass more than just website performance as well, so they can measure not just visits and revenue, but the cost of storage, staff time spent packing and sending, the effect of online marketing on offline, etc. You can’t track any of that in GA, while there are other (hugely expensive) solutions that can.

    If you started using GA data directly as a ranking factor, your throwing a hell of a lot of babies out with that bathwater. Not to mention that legal shitstorm Goog. would be wading into. Other platforms lawyers would have a field day.

    Where I suspect GA is used, is as a benchmark to test the result of algo changes. So the sort of insight they can get is “we implemented Panda 2.3, and exit rates decreased by 0.5% while time on site increased by 1.2 seconds on average across searches with a buying intent.”.

  6. I agree that Google does not look at the bounce rate. But, being a blogger it is really tough to see the bounce rate high on my blog, because i want the users to come at my blog and give some time to read the article.

    Therefore i mostly prefer to have a video along with the article. It helps me to reduce the bounce rate and people do spend some time on my blog page…..

  7. Adding a video does not lower the bounce rate. People could spend 20 minutes on your blog and still be classed as a bounce if they only view one page.
    Some simple things you could do to drop the bounce rate is to make sure you use read more breaks and to add useful links into your content, as well as give a user some interesting choices after reading with related posts or something like that.