Google Analytics: Understanding and Lowering Bounce Rates

Google Analytics provides valuable intelligence into how visitors find, interact with and leave your website. This intelligence is central to improving both user experience and the profitability of your website. Google Analytics provides many useful metrics that help you do this and two of the most useful are bounce rate and exit rate.

The difference between a bounce and an exit can be confusing, especially if you are new to analytics. The goal of this article, then, is to demystify the two and explain why they are important. It also acts as a guide to interpreting bounce and exit data and how to lower them in order to improve the performance of your website.

Making An Entrance That Counts

Before you can understand and calculate bounce rate you need to know a little about entrance pages, also referred to as landing pages and entry pages. Google defines an entrance page as:

• Entrances: This metric identifies the number of entrances to your site. It will always be equal to the number of visits when applied over your entire website. Thus, this metric is most useful when combined with particular content pages, at which point, it will indicate the number of times a particular page served as an entrance to your site. Source

In short an entrance page is the first page a visitor lands on when visiting a website. Entrances are, as we will see, a key factor in calculating bounce rate.

What Is A Bounce?

A bounce is a single page visit. A bounce occurs when a visitor enters and exits a website viewing no other pages other than the entrance page.

What Is Bounce Rate?

If, for example, if 100 visitors enter your site via Page “A” and 20 of them leave without clicking through to any other page, page “A” would have a bounce rate of 20%.

Some of the reports Google Analytics generates will give site wide averages. The screen grab above has been taken from the Top Content report which can be found by clicking the Content tab in your Google Analytics dashboard.

The first thing you might notice is that when you add the average bounce rate and the average exit rate together the result is greater than 100%. If bounce rate and exit rate are measures of how many people leave your site, how can the total be greater than 100%. The answer is that it can’t.

You might be fooled into thinking that bounce rate is calculated as a percentage of Pageviews. A logical thought since it is figured in the report. However, when added together, bounces and exits would again be greater than the total Pageviews.

Bounce rate is not based on the number of visitors or the number of page views it’s based on entrances.

Will The Real Bounce Rate Please Step Forward

To get at the real numbers that contribute to bounce rate you need to dig a little deeper. The screen grab above has been taken from the Top Landing Pages report which can also be found by clicking the Content tab in your Google Analytics dashboard.

As you work your way down the report you can also view bounce rates for individual pages.

The Top Landing Pages report helps to identify pages with high bounce rates that might require further investigation.

You can clearly see from Figure 3 how bounce rate is calculated at for a single page: (283 bounces / 303 entrances) * 100 = 93.39939939934% which analytics has rounded up 93.40%. As interesting as this is, it tells us nothing about what is driving the bounce rate and what steps if any are required to lower it.

Bounce Rate Through The Looking Glass

Pages that fail to meet visitor expectations, don’t provide clear navigation, talk about features rather than benefits and content that’s not actionable all increase bounce rate. Not all visitors to your site are using desktop machines with ultra-fast connections and will abandon your site if page takes too long to download. If you have been over-zealously linking to your site, links from pages that are not closely related can also increase bounce rate. These are all things you can test for and fix to a degree.

Missing Timestamps And The Pages Time Forgot

Google Analytics reports the time visitors spend on pages by comparing timestamps. When a visitor lands on a page a timestamp is created which records the precise time they arrived.
If a visitor arrives at page “A” at 13.45 and clicks through and lands on page “B” at 13.47 two timestamps will be created. By subtracting the time the visitor lands on page “A” from the time they land on page “B” you arrive at time spent on page “A”:

13.47 – 13.45 = 2 minutes spent on page “A”.

If at 13.50 the visitor leaves your site completely no timestamp is created and there is no way to tell how long the visitor spent on page “B”.

Why was no timestamp created? If the page was outside the scope of your analytics account, on another domain for example, the timestamp can’t be accessed by your analytics account. Therefor and the time spent on that page can't be determined for that page view.

Similarly, the time spent on a page by visitor who enters a site and bounces without visiting any other page cannot be measured either.

Cookies, Sessions And Timeouts

Google Analytics uses cookies to track the activity of visitors to your pages and report those activities back to their server. Cookies enable Google to distinguish the activities of each visitor individually and track sequential page visits made by the same user during their time (session) on your website. This information is then reported back to you when you log into you Google Analytics account.

Every bounce or exit is the result of a session timeout. In Google analytics, a session will timeout after 30 minutes of browser inactivity. If a visitor navigates to another website, the session will still continue for a maximum of 30 minutes before registering a bounce or exit. As long as the visitor returns before the session times out and clicks through to another page of your website, it will not be considered as either a bounce or an exit.

• Each and every visit to your site culminates in a session timeout.
• A session that times out after a single page view is classed as a bounce.
• A session that times out after multiple page views is classed as an exit.

Have a look at the tabs open in your browser right now - how many have been open for more than 29 minutes without any activity? Even though the page is still open in your browser, some of the sessions associated with individual pages might have already timed out causing a bounce or exit. Similarly closing your browser, hitting the back button or disconnecting from the internet will all cause a session timeout that will likely be recorded as an exit or a bounce in someone’s Analytics.

In the next article in this series, Bounce Rate, Dwell Time And Panda, delves deeper into bounce rate and how dwell time influences post panda search rankings.