The Google Analytics users metric is used by many to track the number of visitors on their websites. I have even seen marketing departments using this to verify their own work.
Is this a viable solution? My answer is a definite NO. Not before knowing the caveats. In the following article I will try to explain why, and come up with approaches to measuring your website traffic.
The technical stuff
I can almost sense you browsing past this section, but bear with me while I try to explain this as best I can.
When talking about how the Google Analytics users metric works, we also need to talk about cookies, and not the ones your mom makes. The way Google Analytics keeps track of you is by setting a cookie once you enter a site that has Google Analytics enabled. They also assign a unique identifier to you, in order to keep track of new and returning users. So everytime you revisit a site, Google Analytics will look for a cookie stored in your browser, and marks you as a returning user.
Google Analytics also uses this cookie to track what you do while on the site. Which sub pages you visit, how much time you spent on each page and so on.
So what are the problems with the approach Google uses, as explained in the “Technical stuff” section? There are many, and I will try to go through a few of them here.
One of the problems is that cookies are not shared between browsers. So if you start using a new browser Google Analytics cannot detect that you are a returning user, and will treat you as a new user. This issue also holds true when you access from different devices. Users can also disable, or delete cookies in their browser, leaving you with little chance to track anything.
Another problem that could have a great impact on the accuracy of your data is the new GDPR directives regarding cookies. For more information regarding cookies and GDPR compliance read this.
One major issue has already had an impact on the Google Analytics users metric, maybe unbeknownst to you. In the early 2017 Google changed their algorithm that calculates the users metric. This affects data prior to 23rd of August 2016, so if you try to get data before, chances are that you will see something similar to this: