Definition:
Cohort analysis is the behavioral analysis of a given segment of users who share a common characteristic over a period of time.
Cohort analysis works as a segmentation of users whose historical behavior is taken into account to detect patterns or changes in behaviors throughout the user’s life cycle. It is especially interesting for discovering trends and taking action.
For example, a cohort could be all users who have signed up on a given day. On this segment, an analysis would be carried out on how these users behave, what actions they perform and how their behavior differs from other groups of users.
Cohort analysis is a very important tool in digital marketing, it shows the effectiveness of the strategies worked and can help decision making.
Analysis of cohorts in ecommerce
Some of the cohort analyses can be applied to e-commerce to gain efficiency in digital strategy.
Transformation of Leads into Customers
Through Analytics you can generate a report that defines the leads that have finally become customers. To do this, you have to select a period of time of at least a few months. Select cohorts by date of first visit to the web to cross it with the metric number of customers. So you can check if you are managing to convert more customers on the first visit with the web
Revenue by geographic location
The cohort can be defined according to the location of the client. In this way you can check the difference in user habits in Andalusia or Galicia.
Define the best traffic channels
With cohort analysis you can check which channels work best. After grouping in cohorts by means of first visit, it is possible to study which channel has more user retention, which has a higher average cart value or from which medium the customers with the highest LTV are generated.
Behavior of registered users
With the cohort of registered customers you can check if they have a greater commitment than the rest of the clients with the web.
Cohort Analysis in Google Analytics
Google Analytics has a report called cohort analysis in which you can do web analytics studies such as user retention: which ones return two days after registering, a week later, or 30 days later.