Cohort analysis refers to the practice of analyzing the performance of a group of visitors who share common characteristics over a specific time-period, selected from within an organization’s defined customer life-cycle.
An effective cohort analysis will often zoom-out and look into several similar cohorts side by side, examining each group’s performance during the same part of the customer life-cycle (e.g. onboarding) across different time periods. As such, a well-executed cohort analysis is a great way to understand customer behavior over time (usually focusing on quarterly or YoY behavior) and identify patterns and trends.
Cohort analysis is very similar to the practice of Behavioral Segmentation in the sense that visitors who share common characteristics are grouped and analyzed together, but in cohort analysis the focus on specific parts of the customer lifecycle across different time periods is integral.
As an example, a marketer conducting a cohort analysis can look at the purchase rate of visitors who recently registered for membership during any month of 2016-2017, segmented by signup location:
- Website footer
- Promotional exit-intent pop up message offering 25% discount
- Purchase checkout form
Visitor segmentation happens on two levels: membership origin and month of sign up. The purchase rate of every cohort is then tracked and examined for 24 months. Once enough time has elapsed, the marketer can glean from the analysis the segment yielding the highest purchase rate over time, and the month most ideal to sign up said visitors for membership. For example, visitors who register for membership via promotional messages have the highest purchase rate during the next six months, especially if signup takes place in November.
Cohort analysis is a great way to understand the customer lifecycle, anticipate certain behaviors, calculate lifetime value, and optimize each cohort’s experience in order to reduce churn and improve retention.
