Predicting Term Deposit Subscribers for Banks

Optimizing Telemarketing Campaigns

Banks frequently use telemarketing campaigns to sell term deposits. However, calling every customer is inefficient. The goal is to predict which customers are most likely to subscribe based on historical data, demographics, and previous campaign interactions.

Classification Models

I benchmarked several classification algorithms including Logistic Regression, Gradient Boosting, QDA, LightGBM, and Random Forest. LightGBM emerged as the clear winner due to its ability to handle categorical features natively and its highly optimized gradient boosting framework.

Business Impact

The model achieved a top 15 rank out of 496 participants. Implementing such a model allows banks to drastically reduce marketing costs while improving conversion rates by targeting only high-propensity prospects.