We Launched non-Cohorted Retention....Use at your own Risk

As part of our most recent product launch we revamped our “Retention” chart to show non-cohorted retention rates.

To review, cohorts are groups of customers with the same sign-up period (eg October 2020). To date SaaSGrid has tracked retention, whether it be Net Dollar, Gross Dollar or Logo, by cohort. For example: what is retention for customers that signed up in October 2020 after 3 months, 6 months, 12 months, etc. By taking the weighted average of all cohorts at the same age (eg month 12) you can get a good overall idea of your customers’ retention behavior.

Non-cohorted retention works differently. Instead of grouping customers by cohort, we group all active customers at a certain point (eg October 2020) and then look at retention for those customers 1, 3, 6, or 12 months later. For example, you can look at trailing 12 month retention in October 2021 so see how active customers in October 2020 behaved over the past year. Many people find non-cohorted retention more intuitive because the x-axis is dates instead of the nebulous “Month 12”, and because there’s just one number for the current month, versus having a different retention number for each cohort. Trailing 12 month retention also makes easy to double-click on what's happened just in the last year, versus across the entire history of the business.

However, there is a danger to non-cohorted retention. Because it combines data from companies at different stages of their customer journey it can hide what’s going on within your customer base. The most acute example of this is if you look at retention for the trailing 1, 3, or 6 months (adjustable using the “Trailing Period” chart setting). Most SaaS companies have annual contracts, so customers are really only at risk of churning every 12 months. This means that if you look at customer retention compared to 3 months ago only a small percentage of your customers have had the opportunity to churn, while all have had the opportunity to expand, allowing you to simultaneously have high churn and high Net Dollar Retention. Especially if you’re growing fast and your newer cohorts are larger than older ones, non-cohorted retention may obscure a real churn issue in the business. Only when growth slows with the full extent of this “leaky bucket” issue become apparent.

To minimize the risk of missing a churn issue, SaaSGrid’s retention chart looks at the trailing 12 months by default, meaning all customers included have been through a renewal cycle. However, even looking at the trailing 12 months you can miss insights. It’s easier to spot customers acting differently at their second renewal than their first, or if recent cohorts are over or under-performing historical averages, by looking at cohorted retention.

Of course, non-cohorted retention is still valuable (if it wasn’t, we wouldn’t have built it). If you make a change in your business, for example hiring a Customer Success Manager, it can be helpful to see what the impact is 3 or 6 months later. Additionally, a weakness of cohorted retention is that it can be difficult to determine expansion relative to churn in a specific month or quarter (ie whether your business organically expanding). Non-cohorted retention makes this kind of analysis easy. Just make sure you fully understand the two methods of tracking retention so churn never takes you by surprise.