One of the most common questions I get asked is about using the “sessions” metric in a Google Analytics custom report. In standard Google Analytics reporting, “sessions” are only available in a limited range of reports. Sessions are available in audience reports (for example, geography and mobile device usage)and acquisition reports (for example, traffic sources and campaigns), but not within content reports that directly look at the individual pages of your site. So, if you want to see sessions as a metric in a report that has page as a dimension, you’ll need to use a custom report.
“Sessions” is such an elemental part of Google Analytics data that it can seem a little strange when another key report completely ignores it. However, when trying to calculate how many sessions land on a certain page, Google Analytics has to make a compromise about what data it does and does not show.
For example, let’s say you have three pages on your site which we’ll call “Home”, “Video” and “Content”. Two users come onto your site and visit all three pages, but in different orders:
- User A: home > video > content
- User B: content > home > video
In GA, each of these user actions would be counted as a single session, with 3 pageviews each. However, let’s say we want to see how many sessions landed on each page. Logically, we’d expect each of these pages to have 2 sessions each, since all 3 pages appeared in both sessions. However, this is how the numbers would look in a report.
- Home: 1 session
- Content: 1 session
- Video: 0 sessions
Even though we know 2 sessions visited each of those pages, we’re only seeing at most 1 session for each of those pages, and for the slideshow page we’re not seeing any sessions at all. In this instance GA is only attributing a session to a page if that page was the first page accessed within a session. So, since since User A landed on “home” and User B landed on “content”, each page has one session attributed to it. In contrast, the “video” page is not attributed to any sessions since it was not the first page in either session.
There’s a reason the data is displayed this way. Imagine, for example, that GA did count 2 sessions visited for each of the aforementioned pages. This would reflect what you’d expect individually for each page, but it would also mean that the sum of all sessions from all pages would equal 6; triple the amount of sessions that actually occurred. Thus, to make sure that the sum of all sessions per page is equal to the number of sessions sitewide, GA tries to assign it so that each session is attributed to only a single page.
Hopefully this helps clear up how Google Analytics uses the “sessions” metric in a custom report. While the number you get might not be based on what you expect, there is a logic at work that can be followed in any report you pull.