Taking a Faster Approach to Addressing Customer Churn

Share:
Customer Retention
Marketing
Customer churn is a big problem for companies, especially in industries such as telecommunications where annual churn rates average between 10 percent to as much as 67 percent, according to the Database Marketing Institute. And while technological advances have made it easier for marketers and other decision-makers to better identify and respond to the triggers for churn, most customer churn programs are reactive and don't leave marketers adequate time to deliver messaging or offers that might change a customer's actions.

Customer churn is a big problem for companies, especially in industries such as telecommunications where annual churn rates average between 10 percent to as much as 67 percent, according to the Database Marketing Institute. And while technological advances have made it easier for marketers and other decision-makers to better identify and respond to the triggers for churn, most customer churn programs are reactive and don't leave marketers adequate time to deliver messaging or offers that might change a customer's actions.Last week, I had a discussion with executives at Amplero which has developed a self-optimizing personalization system that continuously monitors customer behavior and generates up-to-date churn scores using machine learning to more quickly identify and respond to customers at risk of defecting.

Amplero, a division of Globys, has developed a sequential churn modeling technique that claims to provide a 300 percent improvement in churn prediction accuracy and up to 400 percent better retention marketing compared to traditional modeling techniques.

Using a test base of approximately 2.5 million customers for a telecom company in EMEA, the Amplero technology measured thousands of customer behaviors dynamically to automatically determine which customer actions are meaningful. Instead of taking months to aggregate customer attributes in a data warehouse and test their impact in a churn model, the Amplero system constantly updates churn scores and then uses machine learning to determine the optimal retention marketing actions to take with an individual customer.

"The ability to identify event-based drivers is part of the key to the success of this model," says Lara Albert, vice president of global marketing at Globys.

The sequential model can also be used to drill down on the first week to two weeks of a customer's relationship with a company and help determine actions that can be taken to strengthen engagement and retain customers early in the cycle.

Providing customers with consistent and personalized experiences has been shown to help brands to strengthen loyalty. Using customer data and technology to more accurately predict and respond to at-risk customers can also help marketers to become more effective at mitigating customer churn.

EXPERT OPINION
EXPERT OPINION