Recently, I was faced with a situation where a client was confused by one phenomenon, which at first glance really looks very strange. Client employees set the visitor counter on just one page. That is, they set up a report on only one specific page that opens upon request, and saw that there are repeat visits. Then they casually glanced at the unique visits and were very surprised. The number of unique visitors was greater than the number of visits. Let's look at this fact and see what kind of explanation for such strangeness can be.
Note: if you see something incomprehensible, take a look at my
Web Analytics Primer page
.
How do visitors and visitors count?
')
Visits (Visits) - are counted by the first “hit” (request to the site) or by browsing during the session. During one visit the user can view a number of unique pages or pages on which he already was.
A visit is counted at the session level, and it may not be directly related to the total number of page views or the definition of unique visitors.
Unique Visitors (Unique Visitors) - this is a rather rough calculation of each visitor. In fact, there is no such term as “not a unique user”, but “a unique user” is just a GA term, and we will have to accept it))
The number of unique users is considered at the level of visiting the first page rather than the first session. This can be as many visits to this page at once (regardless of how or how many times this page was reloaded during the session of a unique visitor) or one visitor who has several sessions at once.
Puzzle solution
Analyzing what was written above, we can say that this situation arose because we decided to include in the statistics only visits and visitors who entered only one page, we filtered the rest. And when we remove this filter, the visitor counter and visitors will display clearer information. Given all the pages on the site together, the number of visits will always be greater than the number of unique visitors. Since this is normal, we expected the use of this dependency in all cases, even when the filter only considers one page. However, this was not the case, and the whole thing is in various ways of counting visits and visitors.
In this case, when we only have a counter on one page of the site, each visit means “a visit where a particular page was the first to visit the site.”
Unique visitors continue to be who they were - "any visitors who have ever visited our page or group of pages." That is why such a small difference between visits and entrances (in the aisles of 10%).
Note that this is different from entering the site in the general sense of the word, because the multiple entries to the site of one visitor can be considered as one visit if the session lasts no more than 30 minutes.
Invisible Visits - Sly Visitors
Let's look at the two visits shown above. Both visitors made two visits, where each of them looked at three pages. The first visitor went straight to page eight (green), while the second visitor went to page eight just like the second page (red) of the site.
The session of the first visitor began on page eight, and the visit was counted. We can see it as a visit - the page is green because it is counted.
The second user’s visit began on page five, and he looked at page eight later. When we click on page eight separately, this visit was not considered - the page is red because it is not counted.
However, both visitors, one way or another, looked at page eight and both were counted as visitors who were on this page. For these two visitors and the page, we counted two unique visits, one visit and two views.
The riddle is solved, we raise the level
All this shows us that statistics is often not what it seems when we start manipulating filters. The basic principle that must be adhered to is that indicators adjacent to one another: “visit / session, page” can influence each other. Regarding visitors and pages that are adjacent to each other, each level is the angle of the triangle, as you can see in the image above. So, you can search for another riddle.
Original article