Return Frequency % of Total

Recently, a co-worker ran into an issue in Adobe Analytics related to Visit Frequency. The Visit Frequency report in Adobe Analytics is not one that I use all that often, but it looks like this:

This report simply shows a distribution of how long it takes people to come back to your website. In this case, my co-worker was looking to show these visit frequencies as a percentage of all visits. To do this, she created a calculated metric that divided visits by the total number of visits like this:

Then she added it to the report as shown here:

At this point, she realized that something wasn’t right. As you can see here, the total number of Visits is 5,531, but when she opened the Visits metric, she saw this:

Then she realized that the Return Frequency report doesn’t show 1st time visits and even though you might expect the % of Total Visits calculated metric to include ALL visits, it doesn’t. This was proven by applying a 1st Time Visits segment to the Visits report like this:

Now we can see that when subtracting the total visits (27,686) from the 1st time visits (22,155), we are left with 5,531, which is the amount shown in the return frequency report. Hence, it is not as easy as you’d think to see the % of total visits for each return frequency row.

Solution #1 – Adobe ReportBuilder

The easiest way to solve this problem is to use Adobe ReportBuilder. Using ReportBuilder, you can download two data blocks – one for Return Frequency and one for Visits:

Once you have downloaded these data blocks you can create new columns that divide each row by the correct total number of visits to see your % of total:

In this case, I re-created the original percentages shown in the Return Frequency report, but also added the desired % of Total visits in a column next to it so both could be seen.

Solution #2 – Analysis Workspace & Calculated Metrics

Since Analysis Workspace is what all the cool kids are using these days, I wanted to find a way to get this data there as well. To do this, I created a few new Calculated Metrics that used Visits and Return Frequency. Here is one example:

This Calculated Metric divides Visits where Return Frequency was less than 1 day by all Visits. Here is what it looks like when you view Total visits, the segmented version of Visits and the Calculated Metric in a table in Analysis Workspace:

Here you can see that the total visits for June is 27,686, that the less than 1 day visits were 2,276 and that the % of Total Visits is 8.2%. You will see that these figures match exactly what we saw in Adobe ReportBuilder as well (always a good sign!). Here is what it looks like if we add a few more Return Frequencies:

Again, our numbers match what we saw above. In this case, there is a finite number of Return Frequency options, so even though it is a bit of a pain to create a bunch of new Calculated Metrics, once they are created, you won’t have to do them again. I was able to create them quickly by using the SAVE AS feature in the Calculated Metrics builder.

As a bonus, you can also right-click and create an alert for one or more of these new calculated metrics:

Summary

So even though Adobe Analytics can have some quirks from time to time, as shown here, you can usually find multiple ways to get to the data you need if you understand all of the facets of the product. If you know of other or easier ways to do this, please leave a comment here. Thanks!