Built a customer retention model for a large managed healthcare insurance provider
Case Studies:Mu Sigma
Published On: 29 April 2015
What We Did: A customer retention model was built to identify key drivers for targeted promotions to sustain market share.
The Impact We Made: Recommendations from the customer retention model helped the business revive 11% potential attrition cases as compared to almost no revival previously.
Summary – Risks of customer attrition identified
The healthcare insurance provider faced customer churn (switches to the competitor insurance) during the Annual Election Period (AEP) and Open Enrollment Period (OEP). Mu Sigma Team identified key drivers of retention to design targeted promotions for members at risk.
About The Client - A major healthcare insurance provider
The client is one of the largest providers of managed healthcare insurance services for individuals and families.
The Challenge - Understanding drivers of customer attrition
The client had several initiatives in place to understand and improve member retention and maintain market share. However, they were still facing significant customer attrition levels during certain periods over the course of the year. These initiatives were also proving to be heavy on the pockets. Many business questions like what type of customers are leaving, where are they going, etc., remained unanswered which impeded the efforts to understand customer retention. The challenge was to efficiently understand existing customers and identify factors contributing to customer attrition.
The Approach - Optimal prediction model algorithm
Mu Sigma worked with the client to implement a 3-phased approach:
Phase 1: Understand Existing Customers
Different sets of customers were profiled based on behavioral and demographic data to obtain a holistic view. They were then segmented based on product type, usage and payment patterns. In doing so we were able to create clear and distinguishable customer buckets. Promotional activities targeted at various customer buckets were analyzed for better understanding of behavior.
Phase 2: Determine the Optimal Predictive Model Algorithm
Three Predictive model algorithms were evaluated to identify the best fit: Survival Analysis Models, Logistic Model and Recurrent Approach Model. The model that allowed retention efforts to be focused for optimal use of available budgets was then chosen.
Phase 3: Customizing messaging to enable proactive targeting of Customers
We identified that targeting in a customized manner was required to lower customer attrition rates. Analyzing the data provided an understanding into the drivers affecting retention, customers who are likely to attrite, and means to maximize ROI of customers.
Additionally, we recommended steps to tailor the product mix and timing of promotions appropriately in order to achieve optimal ROI for an attriting customer.
The Outcome: Improved lift in customer retention strategies
Mu Sigma was able to devise a customer retention model for the client based on probability of attrition and key drivers for retention
The model based effort was able to achieve an 11% lift in targeting potential attrition cases
Key drivers of retention were identified to be a combination of customer profile and experience
Retention efforts were focused on high value targets to make the optimal use of available budgets and subsequently optimize ROI