Guide to the meaning of Bounce Rate in Analytics

How is a bounce rate recorded?

A bounce rate as it is initially set up in Google Analytics only looks at one factor: the number of events sent by the Analytics code to Google tracking. So that you understand how this works we’ll need to look at part of the code that you added to your website when you installed Google Analytics:

ga(‘send’, ‘pageview’);

Analytics works by sending a “pageview” or other event each and every time that a page in your website is loaded or an action occurs which is set to send data to Analytics. A bounce is recorded when a user’s session only has one event having been sent, the pageview above when the page was loaded. If any other event was sent in addition to this event, this isn’t recorded as a bounce.

How is the bounce rate misleading?

Not every website needs people to go beyond one page!

As an example, let’s say you own a restaurant website. A large percentage of people coming to your site are likely looking for a few pieces of content that are surely to be found on your homepage: your phone number, your address, a link to a map (external), your hours of operation… and once they’ve quickly found that information they don’t need to go elsewhere in your website, or they’re leaving your website.

You might have helped them in a micro-moment.

In this example the visitor’s session would register as a bounce (bad!) but the site has quickly and successfully provided the information sought by the visitor (good!). Which is it? Was the visitor session successful or a failure?

Another example could be a website with a blog. A visitor could come to the blog article, read the entire post, and then signup for the mailing list at the bottom of the page. Depending on how the mailing list functionality is set up, it could launch a new window to the subscription service, leaving the original page. In this scenario the blog post got a new subscriber (goal conversion, good!) but registered a bounce (bad?).

And one more example is a web page that has lots of interactive content within it such as embedded YouTube videos. A visitor comes to the webpage, views one or more videos within the page lasting over 30 minutes, and then continues further into the site and accomplishes actions. Since the page view lasted longer than 30 minutes without going to another page, that session had ended and a bounce was registered; a new session was started when the user clicked off of the page into a new page. A highly engaged visitor (good!) but a timeout bounce (bad?).

My bounce rate is REALLY LOW, does this mean my website is doing great?

Sometimes people come to me and tell me (in a really excited sort of way) “my website has a bounce rate of 8%! how great is my website!” Well, I’m sorry to tell you, but this is a mistake. Your website doesn’t really have a bounce rate that low. What’s happening is one of the following:

• You have Google Analytics installed multiple times (or perhaps GA and GTM both)
• You have something sending events (such as scroll depth or timing events) and those events are set as interaction events (as opposed to non-interaction events)

You’ll want to make sure you don’t have Analytics or Tag Manager both acting here, remove the duplicate tracking, and/or make sure that your extra events aren’t interaction events.

Setting up Events and Depth Tracking for greater detail

As the previous examples indicated, we need a way to go further and gather more information in order to differentiate a successful visit from a bounce. One of the ways that we can build additional knowledge into our pages is through Events in Google Analytics.

Events are snippets of information that you can trigger to be sent to Analytics using Javascript. You can trigger an event in lots of ways: upon clicking an object, hovering over an element, submitting a form, loading the page, based on a timer… lots of different ways. By attaching a Javascript function to any of these actions you’re able to send data to Analytics that can be analysed (giving you greater depth of understanding of user behavior), but also nullifies that bounce because the user actually did something that you wanted them to do.

For example,

• you can send an event when your phone number link is clicked on a mobile phone
• you can send an event when a person scrolls a significant way down your page (if it’s long, at intervals)
• you can send an event when an external link to a map is clicked

Using this you can begin differentiating true bounces (bounces where the person really bounced) as opposed to bounces where an event occurred within the page equaling a goal conversion, but would otherwise register as a bounce. For example, in the code block below, we use JQuery/Javascript to trigger the ga(‘send’) any time the telephone element within the webpage is clicked. It sends an event with the label “clickToCall” to Analytics.

$(‘#telephone’).on(‘click’, function() {
  ga(‘send’, ‘event’, ‘button’, ‘click’, ‘clickToCall’);
});

This event will nullify the bounce when clicked, but we can also set our Analytics to look for that event and use it to measure a “Phone call” goal conversion.

Comparing Segments of Bounces and Non-Bounces

Bounces can have a huge impact on your website’s metrics, especially when a big ZERO brings down your averages. If you have 50% of your visitors read 4 pages per session, and 50% bounce, rather than looking at 50% as being very engaged, your pages per session reads as an average of 2 pages, which is uninspiring.

Thankfully there’s an easy technical way to separate the Bounces from the Non-Bounces: you can create Segments.

Segments are a part of Google Analytics that enables us to only see (or not see) the data that meets a criteria or set of criteria. This could be first time visitors, repeat visitors, visitors in the United States, visitors originating in social media… lots of different factors, including Bounces or Non-Bounces.

What can we learn about bounces and non-bounces?

When you create a separation and the ability to view only bounced sessions or non-bounced sessions, you suddenly gain greater insight about the people that stayed on the site or had immediately left.

For example, you can investigate questions such as:

Are certain pages receiving a high percentage of bounces compared to the rest of the site?

A situation that recently arose with a site we know was that they received a ton of traffic from a referral link to a blog post. While lots of those visits were “flighty”, by factoring them (the bounces) out, we can evaluate those who stuck around beyond the immediate bounce. We can then have a better sense of the Behavior Flow.

You have bounces, but when you look at repeat visitors, they’re all repeating!
One of the things you might discover is nuances in visitor behavior. For example, you might discover if you have an active blog that you have a dedicated community of readers who come whenever you post new content, but they only come for that one new page of content. Dedicated readers, but bouncing each and every time.

The bounce rate is not a search engine ranking factor

This part is going to be a little confusing. Lots of advice sites will tell you that a bounce and your bounce rate will affect your website’s position in the search engine results page rankings, and that “lowering your bounce rate will improve your position in the Google SERP”. This is not entirely true.

As mentioned earlier, there is a something called the pogostick effect or Long Click that might be a ranking factor for Google.

The pogostick effect can be defined as

“going back and forth from a search engine results page (SERP) to individual search result destination sites. The behavior may indicate poor search results since the user hasn’t been satisfied by one or more of the SERP results” [source]

This behavior, which can often (but not always) be attributed to a bounce, can have a negative effect on your search engine rankings for specific pages and queries, indicating to the algorithm that your page was not a good result to show in that instance. Even when not a bounce, but a “Long Click” vs a “Short Click”, can often provide Google with similar information.

In short, what this means is that if a person runs a Google Search, clicks through one of the links to go to your website, then “bounces” back to the search engine result page, that Google is tracking that “pogostick” behavior.

Your duty in these cases is to understand why pages, and visitors arriving from a search engine query, have high or low bounce rates, and to focus on and optimize your page(s) for the inbound traffic that will capture the highest quality queries which we discuss in “Intent, Capture and Conversions“.

Are you interested in taking a deep-dive into your Bounce Rate? Then go ahead and click to install this Google Analytics dashboard that we’ve created for your viewing pleasure.

With this linked dashboard, we recommend installing it, but then utilizing different segments as a way of going deeper. Recommended segments might include New vs Returning, Local Traffic vs All, Mobile vs Desktop

The bounce rate is the beginning

If you’ve learned anything in this post it should be this: your bounce rate for your site or any piece of content should not be the important number; it should be the beginning of a quest to gain further insight.

Who bounced, who didn’t bounce, where did the bounces come from, what differentiates them from those who didn’t bounce…. this is what you should start looking into.

A percentage of website visitors will always bounce, and highly successful pages might have incredibly high bounce rates. The top level bounce rate is just the beginning of the questions you need to ask of your metrics, and of your insights to be gained.