Google Analytics engagement reports can provide great insight into user behaviour on your site. However, it’s not obvious how to read them - and when you figure out how to read them, it’s not always obvious what’s good!
The visitor frequency report is found under Audience > Behaviour > Frequency and recency in the Google Analytics menu. The report shows the sessions, by the number of visits for each visitor.
Google’s explanation of how to read the frequency graph is nice and clear. However, their simplified example leaves out something important: returning users. In the graph, the visit count shown is for the whole user history, not just the period of the report. So if a visitor has come three times before the period their session will show in the three band.
Similarly, if a user has visited the site five times in the past and then visits the site twice during the period, they will in count of sessions as 6 and 7.
The example below is from a newly released site, where nearly all the visits are from the developers, who have been many times before - so there are no visitors in the 1,2,3,4 or 5 visit bands for this site in this period.
The fact that the banding is based on the whole user history, not just behaviour during the period, can make the report much harder to interpret - you can't easily see the drop off in repeat visitation if an unknown number of return visitors at some point in their many visits history are also coming in.
Fortunately, Google Analytic's segmentation capability comes to the rescue!
For example, you can find out about the returning behaviour of visitors who came the first time during the period in question, with the segment - the settings are in the screenshot below. Note that you need to change the segment to filter users, not sessions, otherwise you will just create a more complex version of the built-in new users segment!
Here’s an example of the sort of thing you might see with the segment, showing 3 segments:
- All Session (built in)
- 'User sessions = 1' - custom segment for users having their first session in the period
- New Users (built in)
- For the visit count = 1 band, the session count is the same across all three segments,
- For the visit count = 2 band, (and above) the number for all sessions is higher, because it includes all the users who came in on their second visit during the period. The number for the user sessions = 1 segment, is lower because it includes only the users who had their first visit during the period. The number for new users is zero because users are not new on their 2nd visit
As you can see, the custom segment makes it possible to see the real return rate of new visitors in the period, narrowing to visitors who came the first time in the period in question.
From this example, you can also see how misleading it would be to naively interpret the default 'all sessions' segment for, say, four sessions as the number who returned four times during the period - clearly there is a large number who have previous visits outside the period of the report. Note that none of the segments in the example actually gives the number of users who returned four times during the period - this is actually really difficult to obtain.
Leaving that question aside for now, to extract some real insight from the approach of segmenting in the frequency report, combine that segment condition with a goal, say ‘transactions per user > 0’ - then you can see how many new users went on to a transaction, and how many visits they made during the period.
Need help to set this up or have any questions? Get in touch with our team of experts and we'd be happy to answer any questions!
This is a valuable segment to monitor and analyse - how many users have gone from first visit to a transaction this week, and how many sessions did they make along the way?