Discrete Event Simulation Model Performed with Data Analytics for a Call Center Optimization

Abstract

Optimization models enable organizations to find the best solution and respond to the demand from an uncertain environment and stochastic process promptly and with less engineering effort. This study aims to optimize the number of seasonal agents and customer prioritization needed for a call center system using big data analytics and discrete event simulations to improve customer satisfaction. The study was carried out based on data from a leading heating and ventilation company's call center. The K-means clustering technique was used to determine customer segmentation on 6-million-customer data. For prioritization, the making of a Recency-Frequency-Monetary (RFM) analysis was applied. The system was modeled using ARENA simulation software, and performance parameters were measured depending on the segments obtained. The results show that the simulation model performed with data analytics gives better results for a beneficial financial impact with numerical values in customer prioritization, reducing the average waiting time of the most prioritized customers by more than 90%, and for the least prioritized customers, it increased the average waiting time by approximately just 40%. However, with the company segments, the increase in the average waiting time of the least prioritized customers was approximately 300%.

Customer Segmentation For Digital Marketing Based on Shopping Patterns

“…The application of the machine learning method allows marketers to respond and adapt to the dynamics of changing shopping patterns and customer preferences so that marketers can deliver messages to the right customers and at the right time, thereby increasing the effectiveness of company promotions (Christy et al 2021). The combination of RFM and machine learning methods can generate various strategies for customer segmentation (Wong and Wei, 2018;Rahim et al 2021;Serper, Şen and Çalış Uslu, 2022;Salminen et al 2023).…”