Predicting SaaS customer churn

How to predict SaaS customer churn

In order to reduce your customer churn rate, we must figure out how to start predicting customer churn.

Predicting customer churn allows for SaaS leaders to strategically plan win-back campaigns that will re-engage at-risk customers.

Because customers churn for a variety of reasons, this is what makes predicting customer churn challenging.

Still, at Raaft, we believe we’ve found data-driven methods to help SaaS leaders make better predictions and even stronger customer win-back campaigns.

We’ve narrowed it down to the top two customer churn prediction methods to share with you: Customer Churn Time and Customer Churn Score.

Allow us to explain...

How to predict using customer churn time

Customer churn time measures the number of days from today that the customer will churn.

When using churn time to predict customer churn, you’ll need to consider the following factors:

Here are a few great online calculators that can help you get started:

• Date of first lead interaction
• Date of purchase or start of subscription
• Date(s) of check-in’s or post-sale engagements Date(s) of at-risk behaviour
• Average customer lifecycle lengths for similar buyers

When taking all these factors into account, churn time should help SaaS leaders predict the amount of time left until specific customers will likely churn.

Combined with your specific definition of customer churn, SaaS leaders can gain a better understanding of their customer lifecycle length. Understanding the customer lifecycle means strategies can be put in place to extend the customer lifecycle.

For instance, customers that have 500 employees average a 1 year lifecycle. In this case, you’ll want to consider when customers of this size start becoming a churn risk. Identifying churn behaviour correlated to customers of different sizes can help you reduce churn where it counts. Because chances are, 20% of your customers bringin 80% of your revenue. Does their user adoption tend to drop off after the first 6months?

Realizing this may help you learn, not only about the customer lifecycle but the pain points of your product or service.

PRO TIP: Don’t measure customer churn time using your entire customer base. It’s a simple way to look at churn time but can result in inaccurate predictions because your customers have varying demographics and needs.

Which brings us to our second customer churn measure: customer churn score.

How to predict using customer churn score

Customer churn score measures the likelihood of customer churn given their current health status.

When using churn score to predict customer churn, you’ll need to consider the following factors:

• Customer buy-in
• Customer adoption of your product
• Net Promoter Score (NPS)
• Date(s) of check-in’s or post-sale engagements Date(s) of at-risk behaviour
• Average customer lifecycle lengths for similar buyers

When taking all these factors into account, churn score gives SaaS leaders indicators of when a customer needs to be immediately connected with. Having these indication points will help you take action immediately, whereas previously, there were no pre- defined criteria to start reaching out to customers.

For instance, a 70% product adoption might sound great but the customer’s NPS might be low. This could indicate that your customer is enforcing the usage of your SaaS product but their employees are not happy to use it.

Since dissatisfied employees can result in customer churn in the future, the churn score will be reflective of both product adoption and NPS, indicating there is a need for concern. Having a churn score to assess the health status of all your customers, is a great way to measure the effectiveness of win-back campaigns.

And when done right, churn score can determine which campaigns have the highest ROI for retaining customers.

PRO TIP: When looking at investing in a customer churn solution, always make sure their features cover all the basics: customisable UI, metric tracking, automated win-back campaigns and easy implementation.