Talk the 'Talk'?: Exploring the Congruence Between Text and Vocal Features in Enhancing Customer Satisfaction
(A2025-126017)

AUTHORS

Ye Zheng, The University of Sydney; Steven Lu, The University of Sydney; Jiang Qian, The University of Sydney; Jake An, UTS

ABSTRACT

This research proposal examines the relationship between textual and vocal features in customer service interactions and their impact on customer satisfaction. Utilizing advanced audio analysis, automated text analysis, and large language models, the study analyzes over 2,500 customer service calls from an Australian financial investment company. The research investigates vocal emotional features, textual characteristics, and their interactions, including control variables like task resolution and linguistic style matching. Preliminary analysis of 328 observations reveals that both congruence and incongruence between textual and vocal features can influence customer satisfaction, depending on interaction context. These findings challenge traditional communication theories, suggesting that strategic communication misalignment may sometimes produce superior outcomes compared to perfect consistency. The study contributes to marketing communication literature by exploring multimodal communication nuances and provides managerial insights for developing advanced customer service representative training programs. It demonstrates the potential of voice analysis in predicting customer satisfaction and potential churn rates.