Overview and details of the sessions of this conference. Please select a date or location to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).
A Text Analytic Approach for Intelligent Customer Routing in Online Service Centers
Authors: Noyan Ilk (Florida State University, USA), Guangzhi Shang (Florida State University, USA)
Customer service centers of medium-to-large sized firms commonly receive large volume of inquiries in a broad range of topics. At the same time, their service agents can resolve these inquiries at a higher speed and a better resolution rate when they are specialized in a narrow set of skills. These two characteristics of the service center business give rise to the need of a sorting mechanism that can help match customer inquiries with agent specialties and then route incoming customers to the correct agent departments. Unfortunately, these are not trivial tasks. Many real-world service centers are riddled with incorrectly routed customer inquiries due to a lack of understanding of the root causes of the problems and of agents' expertise. Customer-agent mismatches lead to inflated processing times and transferred rates that waste agent efforts, reduce service quality, and impair customer welfare. According to industry reports, these negative consequences could cost service centers tens of millions of dollars each year.
In this work, we develop a solution design for intelligent customer routing in online service centers. The proposed design emerges as a synthesis of computational linguistics and machine learning methods. At the core of this design is a text-analytic machine learning model that helps improve customer-agent matching and hence reduce transfer rates. The model utilizes customers’ open-ended responses to a problem description request (e.g., “please briefly describe your problem”) and combines five types of information – structural, stylistic, lexical, semantic, and sentiment, extracted from the raw input to automatically predict the correct agent department that can handle the customer’s problem.
To demonstrate the usefulness of the approach, we conduct a comprehensive case study on a dataset collected from an S&P 500 company. Our results indicate an over 15% accuracy improvement from the proposed solution over menu based routing. To assess the broader managerial implications of this improvement, we estimate potential reductions on agent service time and customer waiting time, as well as potential labor cost savings due to reduction in call transfer rates. We further benchmark the solution performance over a stylized human expert triage design using numerical experiments. Our work has implications for the design of routing policies in service centers, and more broadly for managing customer relations under emerging communication technologies such as live-chat, e-mail, and social media.