Customer churn is a huge concern for every SaaS business. When your revenue is based on recurring monthly subscriptions, there's nothing more disappointing than a customer canceling their account.

GrasshopperGrasshopper, a SaaS company that provides a virtual phone system, noticed that many users were canceling their accounts within two months of sign up. Under the leadership of Mike Morris, who is responsible for Grasshopper's customer acquisition and retention, the company analyzed the behavior of these canceling customers, identified patterns in their activity, and implemented a solution that generated positive returns after just 3 months.

The Challenge

What was the challenge that you faced?

A significant percentage of customers canceled within two months of signing up.

How was this challenge impacting your company or team?

It is critical for us to retain customers as long as possible – this behavior impacted this goal – specifically, this behavior cost us marketing and customer support.

What, if anything, were you doing at the time to address this challenge? Why wasn’t it working?

We had done a number of things to reduce customer cancellations. However, without knowing which customers are actually going to cancel, our efforts were diffused over the entire customer base and were not as effective as targeted efforts.

The Solution

Can you describe the solution you implemented to address your challenge?

We implemented a neural network model to predict which customers would cancel within 60 days of signing up for our service. This solution looked for patterns in the customers who exhibited this behavior and used these patterns to predict which new customers might exhibit the same behavior.

In what ways did you use customer behavioral analyses to direct your business decisions?

The neural network model found customer behavior patterns based on a customer’s early interactions with us. For example, the analysis looked at phone usage within the first two weeks, the type of credit card used for signup, the plan the customer selected, and other factors, to find customers who were likely to cancel in the short term.

The Outcome

What results have you seen since beginning the project?

After running the model for several months we found that customers who the model predicted to be likely to cancel were in fact about twice as likely as the remainder of customers to cancel within the first two months. After reaching out to the customers, we found that, for the most part, they needed a little extra time to get our system set up. By giving them this time, we were able to reduce the cancel rate of this group to that of the customers who were not predicted as likely to cancel.

Bottom Line: What was the ROI impact of this project?

This project paid back its costs in about 3 months and generated a positive return thereafter.

How do you think organizations in general can benefit from understanding customer behavior?

By looking at customer behavior as opposed to customer feedback, surveys, focus groups, etc. , organizations can get a more accurate understanding of their customers' needs and better serve them because the behavior shows a customer’s true motivations and actions.

About The Author: Mike Morris manages Grasshopper's customer acquisition and retention activities including product management, marketing, and customer support. Mike's marketing and customer relationship experience spans more than 20 years including leadership positions in product management, consulting, and client management.

Learn more about Woopra's SaaS Analytics solution.

New Call-to-action

Subscribe to our Blog

New Call-to-action

Guest Post Blog Contributions

Are you a customer analytics expert? We're always looking for quality content to enhance our blog and inform our audience. Learn about becoming a Woopra Blog Contributor here


Let Us Know What You Thought about this Post.

Put your Comment Below.