This paper integrates a customer segmentation method with a discrete event simulation model to bridge the gap between identifying customer behaviors and using this knowledge to respond to customers and make the best use of resources. Three strategies are proposed and examined to improve the operation efficiency of a ticket-booking system. Their objective is to assist high-value customers in obtaining the tickets they want and/or reduce cancellations and failure-to-pays from low-value customers. Our simulation results demonstrate that the high-value, customer-friendly strategy beats all in assisting high-value customers and simultaneously improves railway operation performance. Additionally, the indirect, low-value customer abandonment strategy also has improved slightly in all aspects. Applying these strategies is expected to result in a decrease in complaints regarding booking system rejections and an increase in high-value customer satisfaction. On the other hand, the direct abandonment strategy to reject all low-value customers does not make any improvement.