Cohort analysis for retention

Cohort analysis for retention

Though I already wrote a bit about cohort analysis beforehand, luckily the web is full of new material popping up all the time. This time I found some visualizations created in R through the AnalyzeCore blog of Sergey Bryl covering the retention aspect of the story. There is a different approach to this very same topic on toptal.com as well.

Just to recap what the basis of a cohort analysis is:

A cohort is a group of people who share a common characteristic over a certain period of time. cohortanalysis.com

When we look in specific of retention I would show you know these two approaches.

Retention rate and volume assessment

Cohort Analysis for retention cycle plot
Cohort Analysis for retention cycle plot – source: AnalyzeCore.com

The above chart and show the most usual approach to handle retention management through cohort analysis. These are as follows:

  • comparing retention rate and the number of customers for the newest and elder cohorts for the same period of their lifetime,
  • find trends both in general and within the exact year of the lifetime,
  • find outliers in both retention rate and the number of customers and look deeper for reasons.

On this example you can pretty well identify that there is a notable and continuous decline in retention rate, especially in the early and late lifetime brackets. In addition it also seems that there are some promotions or seasonal impacts which generate higher inflow of customers (in absolute terms), but their retention rate are not on par with business as usual.

Cohort analysis to identify revenue streams

Frankly I find this approach more appealing. Looking at retention rate and knowing that “the higher the better” is of course a good business metric. Though if you put it into perspective and show where your revenue and growth comes from you can impact your entire business modell.

 

Below are 3 examples on how retention rate vs. acquisition drive can impact your growth. All of them are showing a fundamentally different business modell.

Growth driven by improving acquistion90% retention rate builds a steady revenue growth10% retention rate stagnates business despite good acquisition
Cohort analysis: growth from acquisitionCohort Analysis: growth through strong retentionCohort analysis: stagnation through retention problem

These 3 situations highlighted require a totally different modell to bring in more growth:

  1. Once acquisition stagnates shift model into increasing retention to continue growing
  2. Explore if acquisition can be increased while keeping retention healthy
  3. Massive reserves available for growth in increasing retention

As you see not all versions consider improving retention as a must for growth. This is why I think this type of cohort analysis can inform business decisions much better than just looking at retention rate.

Summary

Yet again some examples on how useful cohort analysis can become. I do urge you to learn more about it and ask your analysts (or do it yourself) to produce some of these examples for you.

I would also consider it important to point out: it is better to formulate your analysis questions first to understand how a metric like retention rate impacts your entire business (see the 2nd example). While improving retention rates seems a “no-brainer” you might grow more if you put your resources elsewhere.


Leave a Reply