In order to optimize your company’s revenue, you need to understand what trends exist in your customer lifecycles. Cohort analyses are one way to identify those trends.
A cohort analysis breaks your customers into segments based on shared characteristics in order to bring to light behavioral patterns. You can use the data you’ve gained from them in different ways. But, in the B2B and SaaS industries, cohort analyses are used most commonly to help improve retention strategies.
There are two ways you can segment the cohorts for your analyses: acquisition time and based on behavior.
Acquisition Time Cohorts
When you conduct a cohort analysis based on acquisition time, you’re looking at what portion of the customers acquired you retain during a specific time month over month.
The acquisition time frame you use will depend on your company’s product, contract length and the spacing of important customer milestones. For example, if you have a retainer service that renews on a monthly basis, you might segment cohorts based on the month they started in, but if you have a freemium SaaS product where the renewal date is based on the customer’s start date, you might segment by day or week instead.
As you’re mapping out the month-over-month retention data for the cohort, you should specifically be paying attention to where major drop-off points are. Assuming nothing has changed with your product or service, these drop-off points can show where common friction points are or where there’s a lack of realization in the value promised during the marketing and sales process.
- What’s happening with the customer experience leading up to this drop-off?
- What’s happening with the customer experience at the time of this drop-off?
- Is this drop-off point consistent across multiple customer cohorts or unique to just one or two?
For example, you might notice that in general you have a significant drop-off point three months after a customer’s start date. You know that’s when your onboarding process officially ends, so you’ll want to dig deeper into what about that milestone is causing churn.
Or, you might notice customers acquired in June or later had a high drop-off point at the two-month mark that isn’t present in previous cohorts. June was when your company rolled out a new sales compensation plan, so it’s probable a shift in selling strategy is causing that churn and needs to be addressed.
To run a behavior-based cohort analysis, instead of segmenting based on when someone became a customer, you’re segmenting based on how they’re using your product or service.
These cohorts could be based on which product line they use, what product tier they purchased or which feature or set of features they leverage.
For example, HubSpot could run a behavioral cohort analysis based on their different Hubs and see how retention compares between the Marketing, Sales and Service Hubs. On top of that, they could also look at their Marketing Hub users to see if there a difference in retention trends between customers primarily utilizing their email marketing tools, their conversational marketing tools, their blogging tools and their social media tools.
When analyzing behavioral cohorts, ask yourself:
- Which product line or tier has the best and worst retention?
- Does the use of specific features correlate with better or worst retention?
- Which combination of features leads to the best retention?
Using Insights from Acquisition Time and Behavioral Analyses Together
Acquisition time cohort analysis primarily focuses on what’s causing churn. But if you perform behavioral analysis on the portion of the acquisition time cohorts that were retained, you might be able to identify some of the reasons why people continued to work with your company.
For example, maybe you find the customers you retain past the six-month mark all have high usage of a particular feature and most of the customers that churned didn’t leverage that feature. To increase retention, you can try to encourage the adoption of that feature earlier in the customer lifecycle.
When conducting cohort analyses, it won’t always be easy to find a clear cause of churn — there might not be one. It could be a combination of factors influencing the customer’s behavior. But, the more data you collect, the more likely you are to find some useful insights.
However, the insights you gain are just a starting point. After identifying correlations between behavior or time and retention, you’ll need to dig further into what the underlying causes are and experiment with different solutions to determine the best way to address them.
Guido is a Demand Generation Marketer for New Breed. He specializes in running in-depth demand generation programs internally while assisting account managers in running them for our clients.