The document discusses using logistic regression to identify customers of a major life insurance company in Asia who have a propensity to purchase additional insurance policies or riders. Logistic regression was used to analyze customer data and compute probabilities of customers purchasing different products. A cross-sell campaign was launched targeting customers identified as having a high probability of purchasing a specific additional product. The test group that was targeted based on the model outperformed a randomly selected control group in response rates and value of policies purchased.