In this piece we’ll be debating bounce rate, bringing up some of our clients’ questions and investigating some of the misconceptions surrounding the topic. Here at the LeadByte office, we’ve recently had some clients debating the importance of Google Analytics’ ‘bounce rate’. One of our clients highlighted that during the testing process for a new feature on their landing page, the new feature took their bounce rate from averaging around 45-50% to a whopping 90+%. The feature didn’t change the page aesthetically and the site appeared to be running as expected, prompting the client to question the sudden increase in users bouncing from the page. To help our client understand and try to solve this problem, we wanted to put together a case study. In order to do this, we would go back to basics and investigate bounce rate from the ground up, meanwhile contributing to the bounce rate debate.
‘Bounce rate’ is defined as the percentage of users who navigate/bounce away from a page after viewing it with no interaction. Google Analytics states that “a bounce is calculated specifically as a session that triggers only a single request to the Analytics server, such as when a user opens a single page on your site and then exits without triggering any other requests to the Analytics server during that session” [1].
Our initial thoughts on the term ‘Bounce Rate’ were that if a user ‘bounces’ from your site you’ll be losing business. This could prove to be true, however, based on the study conducted by RocketFuel, only when dealing with high bounce rates. A high bounce rate could indicate that a page does not have the high-quality, relevant content that Google is looking for from its top ranked sites. Rocket Fuel’s study suggests that most websites will see bounce rates between 26% and 70%, further defined into their own bounce rate ranking system, as seen below [2]:
The information to inspect your own bounce rate is located in ‘Audience Overview’ tab in Google Analytics:
Depending on your page’s set goals, seeing a ‘high’ bounce rate may not be a worry. A ‘contact us’ page’s goal, for example, is to simply provide the user with contact information with no other call to action needed as they have got what they were looking for and ‘bounced’ from the page. If your page was more content driven and needed the user to complete an action to achieve a goal, then a high bounce rate would indicate an issue worth investigating.
If you are operating a website which should have a lower bounce rate, Search Engine Journal suggests that you should be making sure that each of your pages loads quickly, offers user-friendly navigation, avoids cluttered advertisements and features quality content [2].
Now that we have an understanding of bounce rate, we turn to a transcript from WebPromo’s Q&A with Google’s search quality senior strategist Andrey Lipattsev and Rand Fishkin CEO & Co-Founder of SEOmoz to investigate the correlation between bounce rates and search rankings. Rand stated that he has recently been running a few experimental tests with various participants ranging from 500 to a couple of thousand people. All participants were asked to take out their phones, laptops and other electrical devices and to carry out a specific search. Once the search listing had appeared, Rand asked participants to click a specific search engine listing at the bottom of a results page and then to ‘bounce’ (click away) from the resulting page [3].
Rand discovered various inconsistencies in his findings. For over half of the experiments, the ranking changed on the search engine results page, with the other half resulting in no change. These findings seem to create more questions than answers regarding bounce rate and the effect it has on search engine ranking. Time for a closer look.
How much does a search engine such as Google take into account bounce rate for its search engine ranking, if at all? A study by Search Engine Roundtable states [4]:
The goal to achieve good search ranking appears to be for your page to fulfill the user’s need, whether that includes interaction or not. The previous example of a ‘contact us’ page, wouldn’t require any action as the user is gathering their desired information and bouncing from the page. This is the expected behaviour for this type of page, therefore, is a good result and doesn’t appear to have a negative impact on your search ranking, according to Search Engine Roundtable.
So, if bounce rate doesn’t affect Google rankings as much as we first thought, how significant are other ranking factors?
Quality Score is one of the most significant factors when dealing with Google search ranking. Quality score as defined by Google Ads help is ‘intended to give you a general sense of the quality of your ads. The 1-10 Quality Score reported for each keyword in your account is an estimate of the quality of your ads and the pages triggered by them.’
Google also states the three factors that determine your quality score are:
Having a high ‘Quality Score’ means Google believes that your page is relevant and useful, therefore will be more inclined to show your ad and potentially lower the cost per click (CPC). This is a result of Google wanting to provide the most relevant and useful information to its users whilst using their search engine. So, does Google take bounce rate into account as an official metric?
Experts in the search engine optimisation (SEO) field, have been debating the relevance of bounce rate in Google Analytics to search engine ranking for years.
As a result, numerous theories and misconceptions have arisen surrounding bounce rate, such as page owners optimising for lower bounce rates in the hope that this would improve their search ranking. From our findings so far, the questions we found ourselves asking were, do bounce rates affect quality score? Does that then mean we can manipulate bounce rate to improve search ranking and effectively lower CPC?
The article featured on CXL ‘The SEO Impact of Bounce Rate (What You Need to Know)’ seems to provide the answers with the following points [5]:
Bounce rate is not a reliable measure of quality.
Apparently, Google Analytics’ bounce rate cannot accurately determine user engagement as it is not able to factor the users time spent on page. At the same time, bounce rate varies widely across industries as discussed previously regarding ‘contact us’ pages. It wouldn’t make sense for Google to negatively affect a page’s search ranking based on its bounce rate if the purpose of the page is not to lead the user through the rest of the website.
Google Analytics can be easily manipulated.
One of the clearest indicators of the ‘unreliable nature’ of Google Analytics is the number of posts across the web stating how to detect and filter out bot behaviour. If Google cannot filter out manipulated behaviour, then there is no reason for them to adopt this unreliable data into their ranking system. An example of unreliable data could be to have an interaction goal set for an enter/welcome button so as soon as a user visits a page, they immediately click the button and complete the set goal. This example lowers the possibility for Google to record a ‘bounce’, as the goal is more likely to be achieved due to less being asked of the user, resulting in a lower but more unreliable bounce rate.
Many websites do NOT use Google Analytics.
From a marketing perspective, it may be hard to comprehend not using Google Analytics for a page, with Orbit media studios providing us with the stat that 94.6% of marketing websites do use Google Analytics. However, by W3Techs’ estimates, only 54.3% of all websites use Google Analytics. Of the pages that do use analytical data, not all use Google Analytics with tools such as Clicky, New Relic, Amplitude, heap, Yandex Metrica, and WordPress jetpack. Therefore, Google Analytics’ bounce rate cannot be used as a ranking factor. So, why is it that so many people feel bounce rate does affect search ranking?
According to CXL’s article, Google doesn’t need to rely on Google Analytics to source data about traffic and the content of a page. Instead, Google employs its own tools to determine a more ‘accurate’ form of bounce rate referred to as the pogo-stick algorithm. This algorithm is seen as Google’s tool for tracking when a user clicks a link, realises it’s not what they’re looking for and returns to the search results to then click on another link. This is seen as pogo-sticking due to the user bouncing around from page to results to page to results etc.
Within Google Patents “search pogo-sticking benchmarks”, they state that Google tracks the number of times search results are selected before a user settles on a particular site as well as how many other links are clicked after a particular result. The key point is that, if a user clicks on numerous other links after visiting your page along with you having a high bounce rate, this algorithm concludes you’re causing users to pogo-stick. This negatively affects your search ranking as a result of leaving your users unsatisfied whether this be due to ranking for irrelevant keywords, using confusing page design or providing inadequate amounts of content.
Google does not measure the amount of pogo-sticking your page causes through bounce rate, but instead through looking at the number of long and short clicks your page generates (dwell time). A long click is defined by a user clicking on a result and remains on that page for a long time, without returning to the previous results page. A short click is defined by a user clicking a link then quickly returning to the results page (essentially a bounce). Google views long clicks as user satisfaction based on the content of the page, therefore is what they tend to optimise for [6].
The presumption can be made that to have efficient SEO, you shouldn’t dwell on bounce rate but instead aim to generate long clicks and keep your long to short click ratio high. Despite understanding why Google Analytics’ doesn’t use bounce rate as an official metric and the basics of the pogo-stick algorithm, we are still brought to a surprising conclusion, that bounce rate still matters.
Google's black-box (can’t see the inner workings) algorithm is difficult to dissect. However, we’re able to bypass this lack of knowledge by combining two factors that we’re now informed about: bounce rate and dwell time. If your page has a high bounce rate but the user on your page has a high dwell time, then your page is likely to provide a good long click percentage despite your bounce rate being damaged. So, by taking dwell time into account, you can avoid unnecessarily optimizing for bounce rate.
It appears that bounce rate doesn’t directly affect your page’s search ranking but still seems to be something you should take the time to understand and be able to improve upon if needed. High bounce rates (when calculated correctly) are often a sign of deeper problems such as user experience issues or poor targeting. Going back to our client’s issue regarding their new feature, a user may not have the patience to wait longer than a few seconds before bouncing due to potential poor loading time. These are the things you should be monitoring and prepared to investigate. If you work on deeper problems such as usability and customer targeting for your page, your SEO problems should see some improvement as well.
[1] Google. Analytics Help: Bounce Rate.
[2] Search Engine Journal. 10 Reasons For a High Bounce Rate.
[3] Search Engine Watch. Q&A with Google’s Andrey Lipattsev.
[4] Search Engine Roundtable. High Bounce Rates On Web Sites Don’t Necessarily Impact Google Rankings.
[5] CXL. The SEO Impact of Bounce Rate (What You Need to Know).