Success Stories/

Predictive Modeling

/Fitness Brand

Development of Churn Prediction Model for a Fitness Brand

How we helped a fitness brand to reduce their customer attrition rate

Overview

A fitness brand faced an increase in the customer attrition rate increased due to cancellations of subscriptions.

Merilytics partnered with the company to identify customers who are likely to cancel their subscription in near future so that they can proactively target them with appropriate preventive actions

Our Value Addition

  1. Built a predictive model that estimates a customer’s propensity to churn in a future period based on the historical characteristics and behaviour
  2. The model also helped in identifying factors that drive ‘churn’ (cancellation of subscription)
  3. Used advanced classification algorithms such as Random Forest, Adaboost and Logistic Regression to classify customer subscriptions
  4. Used SMOTE sampling technique as the data set had ‘churns’ as a very small minority class and the cost of misclassification of ‘churns’ was high

Impact

  1. The churn prediction model helped the client to identify top-100 ‘at risk’ customers with 66% precision
  2. The predictive model enabled the company to “treat” these customers through most suitable incentives and campaigns to prevent them from cancelling their subscription in the short-term

Results

66%
precision in risk identification
Increase in Life Time Value (LTV)
Reduction in Client Attrition
The analysis is fascinating. Everything is as expected with the models – we want to operationalize them and run them on a regular basis.
Co-founder

Intelligent Analytics

Distinctive Solutions

Contact Us
2nd & 7th floor, Gutenberg IT Park,
Kondapur, Hyderabad 500084,
Telangana, India
7924 Preston Rd Ste 350, Plano,
TX 75024

Information Security

ISO 27001:2013 Certified Processes for Information Security Management Systems

ISO 9001:2015 Certified Processes for Quality Management Systems

HIPAA Compliant Business Associate

GDPR compliant

© 2020 Merilytics Inc. All rights reserved.