Underwriting Beyond Data Enrichment and Machine Learning
Insurance companies are concerned about attracting new customers as well as boosting clients' retention rate. In this project, we applied machine learning methods to provide more accurate information to the high net worth policy underwriters. Our analysis focused on accurate prediction, feature engineering, customer segmentation, interpretation, and price elasticity.
Working along with a high-net-worth insurance company, we used advanced machine learning methods to identify the clients' characteristics that are more relevant for maximising customer retention and for increasing the chance of getting new clients. The analysis helped validate the methods that improved the company's income by up to 10 percent or more.
By conducting churn analysis on the company data set, we identified the group of the most profitable clients for the company. In addition, by applying a pool of different machine learning models, we identified the clients' main features that explained 80% of customers churn or leaving the company. The gain charts were used to identify the customers that are more susceptible to convert as well as the current clients that are less likely to leave.
