Google Analytics (GA) provides data that can help you understand your audience, their behaviors, and their interaction with your website. Even though GA provides quantitative data with clear values, the data can often seem confusing and inconsistent, leaving you to guess what the information is implying about the website’s performance. Many often think that a high bounce rate or exit rate on a page implies that it is not performing well, but this is not always the case. This is because we often look at the numbers first and try to make assumptions, which can lead us to incorrect conclusions.
By taking a different approach to evaluating analytics on an individual page basis, we can better understand how these metrics from GA can lead to insights on site success. We will establish a framework around how to view bounce rate and exit rate in the context of website performance, leading with an informed hypothesis rather than a guess.
Bounce Rate vs. Exit Rate
We must first define bounce and exit rate to understand the difference between them and how this data is helpful to understanding performance.
When a new user visits a site and leaves after viewing the page without any other interaction, this is considered a bounce. The bounce rate is the percentage of bounces in proportion to the total page visits. Note that the bounce rate is only applied if the user leaves the site after visiting their first page. Bounces can occur from a user typing in a new URL, closing their browser, or hitting the back button, among other scenarios.
The exit rate consists of the percentage of users who leave the website from a given page, after interacting with the site first. For example, if a user entered the site through the homepage, visits the About page and Blog, then the Contact page, then leaves, this would be considered an exit from the Contact page.
There is no ‘one size fits all’ for an ideal bounce rate or exit rates on webpages, it always depends on the context of the page. In this framework we will show how utilizing a structured approach, rather than looking at the numbers first, can revolutionize the insights we gather through Google Analytics.
1. Understand Your Audience and Their Journey
The first step in Google Analytics should be to identify your audience. Knowing who is visiting your site helps you understand how they may be interacting with it. You can look at demographics such as new vs. returning users, location, and age, among other factors. Once you know your audience, you can make assumptions about the journey they may take through the site and determine where you expect them to go. For example, if your audience is largely new users, you can expect an information-seeking path, compared to a returning user that would be able to access familiar pages more quickly. Identifying the key pages your audience visits, or that you want them to visit, will help in filtering the information as you begin to dive further into the analytics.
2. Establish the Purpose
Every page on your website has different goals in the user journey; the goals of the Contact page will differ from the News page or a Product page. In order to fully understand the data presented in GA, we need to determine the unique purpose behind each of these pages. We can begin by asking questions such as:
- What do we want users to accomplish on this page?
- Where do we want users to ultimately get to after viewing this page?
- Are there CTAs or links that should lead users to other places on the site?
- What information is the user seeing prior to this page that would lead them here?
Answering these questions can help us begin to form assumptions about users’ journey and their expected interaction on each page. We can establish if the page should be leading to further engagement or helping users along in their path.
3. Create a Hypothesis
Now that we understand the purpose of each page, we can begin to create a hypothesis about the expected performance and what determines success. I recommend writing down your hypotheses so you can easily compare assumptions to results. If you are having trouble determining what success will look like for a certain webpage, start with what you know to be true- continually referring back to your audience, the key pages in their journey, and the goal of those pages.
On each webpage, you will want to answer the question:
If this page is performing with the intended purpose, then what is success?
Here are some examples of this hypothesis formula in practice:
If the intent of this page is to get users to explore other areas of our site, then this should not be the last page in the user’s journey. Therefore, this page should aim for a low bounce and exit rate.
If the intent of this page is for users to find our company phone number quickly and call us, then users may end their journey on this page once they have found the information. Therefore, this page can afford to have a higher bounce and exit rate.
4. Review the Data
Once you have gone through all prior steps and fully established what success will look like, only then should you review the data presented in Google Analytics. Beginning with the key pages you established, review and document the analytics from that specific page before moving on to the next. Evaluating the bounce and exit rate on a page-by-page basis can prevent you comparing these values from incomparable pages and being distracted by subjective metrics.
5. Draw Conclusions
As you dive into GA, you can begin to decipher what the data is really telling you about page performance. Testing the analytics against your hypothesis on each page can help you determine which pages are accomplishing the goals you have for your users. If you find that the page is not meeting the expectations of your hypothesis, it may be necessary to conduct further user testing or re-evaluate the page design and benchmark the data from there.
Testing the Framework
Next time you are reviewing website performance in Google Analytics, try utilizing a new approach rooted in strategy, rather than a numbers-first approach. You may be able to gain new insights for future projects or re-evaluate data from the past. By working backwards and understanding goals prior to reviewing metrics, the bounce and exit rate will look less like analytics, and more like answers.