The author details two examples of defining customer churn in contexts where it is not explicit: Retail and Banking.

Photo by Mantas Hesthaven on Unsplash

Retaining customers is easier and cheaper than acquiring new ones (here and here). That being said, retention strategies are not free, and the most effective ones can still cost a pretty penny. This underscores the importance for businesses to identify customers most susceptible to churning within their customer base and concentrate their efforts on this qualified audience. This is where the famed Attrition Model, or Churn Model comes into play. It is often one of the first models a data science team is tasked to produce in a marketing context.

Internet and phone providers, video and audio streaming sites, physical and online news publications, shave clubs… are all examples where customers explicitly tell the companies that they are leaving them — they do so by cancelling their subscriptions or memberships at a given time.

In many other cases, however, industries grapple with the challenge of defining churn when customers quietly disengage from services or products. It is harder to tell when a customer has churned from the service, product, or brand; how can we detect when a patron stops using a public library service? Or when a shopper abandons their favorite online clothing store? Even in banking, recognizing when a customer has left poses a significant challenge.

In this article, I will go through two real-life examples of defining customer attrition when it is not explicit, then I’ll provide some practical tips on applying attrition models effectively.

  1. Customer Attrition for Retail
  2. Customer Attrition for Banking
  3. Train and apply attrition models

By exploring these cases, my hope is to shed some light on potential strategies for defining attrition and assisting you in addressing your own churn challenges.