Optimisation of R&R contests for a life insurance company using predictive and prescriptive analytics¶
Project done at Indian Institute of Management, Bangalore for an Indian large life insurance company.
Technologies used: R, Excel, Google Colab Timeline : Dec 2019 - April 2020
Awards: Awarded Highly commended project for the BAI batch of 2019-20 at IIM Bangalore.
Project report: Download from DCAL IIMB website
Problem Statement: Building optimal Reward and Recognition contests for agents of a large life insurance company.
Forecasting sales of an agent
- Built a regression model which can explain 97% variation in sales
- Quantified the lift generated due to different contest parameters
Clustering of agents based on their capacity
- Identifying the factors which affect the maximum capacity of sales of an agent and clustering the agents based on them
Simulation and optimization of contest parameters
- Simulated the cumulative sales for different contest parameters in each cluster of agents
- Identified the most optimal parameters based on budget and other constraints