When we activate a monitoring system on our website, the main objective is to gather useful data to understand our site performance.
Among the classic KPIs which we will continuously record, for example, through a periodic report, there are sessions and users. However, sometimes, it is necessary to sacrifice one of these metrics for KPIs more relevant for our business.
We think that sessions can be a more reliable unit of measurement than the value of users. Below, we will explain why.
Sessions and Users: definitions
To motivate our preference for Sessions over Users, let’s start by stating the definitions (from the Google Analytics guide).
Sessions (or visits): session means the period between the first and last request made on a website. It lasts 30 minutes (after 29 minutes of inactivity it closes automatically) or expires at midnight.
Each session is characterized by the actions that the individual user performs in that time frame (precisely, 30 minutes). If after that time, the user selects a menu item or requests access to another page, then a new session will start.
Users (or unique visitors): consider the number of unique users who interacted with the site over a defined period. A single user may have generated multiple sessions during the same period (for example, one month), but will be counted just once in the reports. Similarly, it may have generated sessions from multiple channels, but it will always be counted once.
So the metric is affected by both the reference period and the dimension we are observing.
Practical examples of the difference between Users and Sessions
We invite you to carry out two experiments to personally test how these differences are presented in a daily analysis.
- Variation of the Users metric based on the reference period
Instructions:
- Open Google Analytics reports on Public > Overview
- Set the time range to the last 3 months and write down the number of users and sessions reported on the Analytics
- From the same report, set a month to month time range, writing down the number of users and sessions
- Add users and sessions extracted per each month
- Compare the arithmetic sum with the overall value reported by Google Analytics.
Test result: the arithmetic sum will be higher for users, but equal in sessions, compared to the overall data extracted from the Analytics. This is because Analytics recognizes that the same user has logged in multiple times within a specific period, and counts it only once while counting all the sessions.
- Variation of the Users metric based on the dimension applied
Instructions:
- Open Google Analytics reports on Acquisition> All Traffic> Channels
- Download the Excel of the tab showing the traffic divided by different channels
- Compare the arithmetic sum reported in Excel with the overall value reported by Google Analytics, both for users and sessions
Test result: As before, the arithmetic sum will be higher for users, but equal in sessions. This is because Analytics recognizes that the same user has logged in from different channels and counts it only once while counting all the sessions.
The calculation of the users described above is possible through the use of cookies. This measurement doesn’t usually take place correctly for various reasons, including:
- Different people using the same device;
- Different devices used by the same person;
- Different networks;
- Different browsers;
- Browsers that don’t allow the use of cookies.
Google Analytics offers some tools to record more accurate data on cross-device navigation of the same users, but it will be difficult to achieve perfection.
There are also the systems adopted by the various browsers and the operating systems to protect the user’s privacy, which requires cookies to be kept for a shorter period. One case is the ITP 2.1 of Safari.
