Two ways to run a cohort analysis
Posted: Thu Feb 06, 2025 6:43 am
Then, if something was done in the early months to try to address churn i.e., better product features, easier onboarding, better training, etc., we’d want to know if those changes were successful. Cohort analysis allows us to do this by comparing the most recent cohort e. July in the chart above to January. The chart above shows that we made a big improvement in churn in the first month, from 15% to 4%.
There are two ways to run a cohort analysis: the first looks at the number of customers, the second looks at revenue. Each teaches us something different and valuable. The example graph below cameroon mobile database simply looks at the number of customers in each cohort over time:
cohort-retained
The example graph below shows how the MRR for each cohort evolves over time. This particular example illustrates what the graph would look like if there was very strong negative churn. As you can see, the revenue growth from customers still using the service easily exceeds the revenue lost from churned customers. It’s rare to see things look this good, but this is the ideal situation we’re looking for. For those wondering if this is achievable, one company in our portfolio, Zendesk, has even better numbers than those shown in the example below.
mrr-retained
Predicting Churn: Customer Engagement Scores
Since churn is so important, wouldn’t it be useful if we could predict ahead of time which customers are most likely to churn? That way we can have our best customer service reps working to save the situation.
There are two ways to run a cohort analysis: the first looks at the number of customers, the second looks at revenue. Each teaches us something different and valuable. The example graph below cameroon mobile database simply looks at the number of customers in each cohort over time:
cohort-retained
The example graph below shows how the MRR for each cohort evolves over time. This particular example illustrates what the graph would look like if there was very strong negative churn. As you can see, the revenue growth from customers still using the service easily exceeds the revenue lost from churned customers. It’s rare to see things look this good, but this is the ideal situation we’re looking for. For those wondering if this is achievable, one company in our portfolio, Zendesk, has even better numbers than those shown in the example below.
mrr-retained
Predicting Churn: Customer Engagement Scores
Since churn is so important, wouldn’t it be useful if we could predict ahead of time which customers are most likely to churn? That way we can have our best customer service reps working to save the situation.