Improving Customer Service Efficiency Using Generative Artificial Intelligence

Abstract

Social media has presented customer relationship management with a challenge while providing companies with a tool to maintain their relationships with customers. Effort and time in processing customers’ feedback may increase as the widespread use of smart mobile devices. The evolution of generative artificial intelligence (generative AI) is a possible solution to improve the efficiency of customer service. However, existing discussions mainly focus on how chat tools can benefit from generative AI. There is a lack of exploration of how generative AI can assist companies provide managerial responses for customers, which is an important part in customer relationship management and has different requirements from chat tools. This study examines the capability of generative AI in understanding customer reviews and generating managerial responses automatically by fine-tuning two language models. A dataset of customer reviews of restaurants is used. Results indicate that even with a relatively small model, generated responses are still satisfied. The current study also compares different methods to predict whether customers’ reviews would obtain managerial responses. It is observed that the patterns of making decisions about whether to reply to customer reviews differ across restaurants.