Cohort Analysis - Google Analytics
How Do I Use The Cohort Analysis Report?
The purpose of Cohort analysis is to understand how a metric for a specific segment of users performs over time. A cohort is simply a segment of users based on the date of their first session, so all users with the same acquisition date belong to the same cohort.
<p">The Cohort Analysis report in Google Analytics - currently in beta - allows you to compare one cohort against another, and see how the performance of those cohorts compare to one another using metrics such as pageviews, revenue, and session duration.
There are four options at the top of the Cohort Analysis graph – Type, Size, Metric and Date Range. At the moment there is just one Cohort Type – Acquisition Date. Cohort Size means the time window you want to use for the Cohort Type, so captures all users who’s Acquisition day was within the date range selected.
Choosing a Metric allows you to select how to measure the Cohorts and is applied to all fields in the table, except Cohort Type which is a dimension.
You can apply up to four segments to your report and each segment creates a new table of data, allowing you to see how each segment performs against one another. This was possible before by using exporting segmented user data to a spreadsheet, but it’s really great to see it included in its own report within GA itself.
The tables automatically highlight spikes in activity according to date, so you can easily see where one segment of users might have had a higher session duration, for instance, and then begin to explore why.
Use Case Scenario
If we take an ecommerce site as an example, by using the Cohort Analysis report we can easily see transaction metrics by date and be able to see if there was a specific day where there were more visitors completing purchases.
By segmenting the report by mobile device we’ll be able to see if there was any difference between the number of transactions made by all users, and those made using a mobile device.
If the site also contained editorial content, by using a Daily Cohort analysis over a longer period we can examine how many days following a new article post resulted in a drop-off in visitor retention, and if this correlated with the resultant drop-off in transactions.
Cohort Analysis can be a fun and useful tool to help you compare your data to find those interesting correlations you had an inkling might have been there
