Exit rate is the portion of users who left the website (i.e., didn’t do anything else on the website for more than 30 minutes). It can be measured for a single page or more than one page. Exit rate generally often appears next to bounce rate, which is the portion of users who exit the website without viewing a second page. The goodness or badness of a bounce rate depends on context, and with exit rate, this is even more true—a high exit rate may indicate that a page is turning off users, or it may indicate that users have gotten exactly what they wanted and are leaving the website happy.
The exit page is the last page a visitor opens before leaving the website. As with the land- ing page, you will mainly find exit page useful for segmenting users or to find pages that an unexpected large number of users leave the website after viewing.
Website Content Reports 119
“Site Speed” Report
Responsiveness is important for user interfaces, so it follows that the speed at which pages of a website load is important to the quality of the user’s experi- ence. Google made data about how quickly pages are loading available through Google Analytics in 2012 as a section in the Content area. It takes a sample of 1% of visitors to measure how long it takes for pages to load for them. This sampling is why you may not have data for every page of your website.
High load times make a website harder to use and can annoy users. The main value of the “Site Speed” report is to find out if any pages have a high page load time, or if page load times are high across the website. Within the “Site Speed” report, page timings is the metric that shows you information for specific pages. This report also shows you with the page load sample metric how many pageviews it used to calculate load time, which is useful to know because one unusually high data point will pull the average load time up.
The method for gathering page load time uses such a small amount of data that you would still want to confirm that a page loads slowly through other means.
In the end, the “Site Speed” report may be interesting to look at, but its scope is very narrow, so it may only warrant periodic viewing.
“In-Page Analytics” Report
The “In-Page Analytics” report, available as both as a standalone report in the navigation and as an option under the “All Pages” report, has historically been a seductive report. It displays actual pages of the website overlaid with markers denoting how many clicks each link received. In comparison to the
“Navigation Summary” report, the “In-Page Analytics” report can be a far eas- ier way to visualize the data.
Unfortunately, at the time of writing, the “In-Page Analytics” report can be misleading because, as discussed in Chapter 3, Google Analytics can’t tell where on a page users clicked. That means if multiple links go to the same place, it can’t tell which one users clicked on.
When you look at the “In-Page Analytics” report, it simply reports that the same number of users clicked on the same link. It also doesn’t report on more interactive features like search buttons or playing a video.
The “In-Page Analytics” report is of limited value, not just because of the data it lacks, but because most of the methods described in this book involve com- paring numbers. This visualization is less conducive to efficiently gathering metrics on where users click because it doesn’t present metrics in a tabular format. It can be useful for quickly looking up how many users click on a particular link on a page or for easily creating an interesting visualization to share.
KEY TAKEAWAYS
■ Content analysis will tell you where users go on your website and how long they stay there, what pages people enter your website on and from which pages they exit your website, and how they move from page to page.
■ The main purpose of content analysis is to look for potential problem areas that you can probe through other means, such as heuristic
evaluation and usability testing, as well as measuring the effectiveness of design changes.
■ When looking at page usage metrics, look for these patterns:
■ The highest values for a metric.
■ The lowest values for a metric.
■ Pages that have metrics that deviate from the average value.
■ A page may get many pageviews because many people are entering the website on that page, because links to it are easy to find and enticing (whether or not the page actually delivers on the promise of those links is a different matter), or it is really important to users.
■ On the other hand, a page may get few pageviews because it is difficult to find, links to it are poorly labeled, or because users don’t want to go there.
■ When pageviews are much higher than unique pageviews, it indicates that users are frequently revisiting that page.
■ Low time on page may indicate:
■ The content of the page doesn’t match what users thought they were going to get.
■ The content isn’t very interesting or is poorly written.
■ There isn’t a lot of content on a page.
■ The page is very well organized and users can quickly satisfy their goals.
■ The purpose of the page is to direct users to other pages, like a search results page.
■ High time on page only tells you that users are doing something on that page.
■ High bounce rate indicates that there are potentially problems on a page for people who enter your website through that page.
■ High % exit indicates a potential problem; there are some pages where a high % exit is appropriate.
■ Page value can indicate how often users view a page before going on to convert.
■ It is more meaningful to compare page metrics to other pages of the same type (or same template) rather than comparing two pages with completely different purposes.
121
Practical Web Analytics for User Experience. DOI:
© 2014 Andrew Michael Beasley. Published by Elsevier Inc. All rights reserved.2013
http://dx.doi.org/10.1016/B978-0-12-404619-1.00008-3
Click-Path Analysis
CHAPTER 8
INTRODUCTION
Getting insight into how users move from one page to another, also known as click-path analysis, is a powerful capability. However, there is one impor- tant thing to remember: there is no way that you can characterize a “ common path” unless you have a very small and/or very linear website. For many people, this is disappointing to learn because they want a simple story of how people move through their website. Consider Figure 8.1, a fictional example website.
A user comes to the homepage in Figure 8.1, clicks on a link to page 1, goes to page 4, goes back to page 1, goes to page 5, and then leaves. Another user comes to the homepage, goes to page 3, goes back to the homepage,
CONTENTS
Introduction ...121 Focus on
Relationships
between Pages ...122 Navigation
Summary ...123
“Visitors Flow”
Report ...126 Analyzing How Users Move from One Page Type to Another ...128 An Example:
AwesomePet
Toys.com ... 129 Key Takeaways ...134
Homepage
Page 1 Page 2 Page 3
Page 4 Page 5 Page 6 Page 7
FIGURE 8.1
The difficulty of click-path analysis is that users have a tendency to traverse your website in diverse ways.
This diagram is meant to convey a sense of how many websites have not just top-down hierarchical navigation, but also cross-linking between pages.
then to page 1, and then to page 4. Yet another user enters the website on page 2, goes to page 5, then to page 1, and then to page 6. What’s the most common path?
Obviously, this example is small and abstract, but the problem is the same even when the website is bigger and there are more users. Out of a thousand users, only a small percentage will actually follow the same sequence of pages through the entire website. To characterize how users move through a web- site, we must focus on navigation at a small scale—how users got to a page and where they go after leaving a page.