题　目：Amplifying Insights: Harnessing Voice Technology to Decode Customer Uncertainty in Marketing

嘉　宾：Tongyao Shen(申桐遥), Ph.D. Candidate, The Pennsylvania State University

主持人：张铄 副教授 BAT365唯一官网

时　间：2024年10月18日（周五）15:00-16:30

地　点：BAT365唯一官网 徐汇校区 安泰经济与管理学院B404

内容简介:

Verbal replies are frequently collected in various marketing contexts, such as voice surveys and focus groups. Thanks to advancements in voice technology, firms can now extract and utilize additional information from voice, including speaker identity, sentiment, and emotions, alongside linguistic content. However, the full potential of voice technology remains untapped, especially in the realm of response uncertainty—a significant marketing concept shown to help mitigate bias in customer responses. We trained and calibrated two groups of methods to infer uncertainty: one using raw data with deep learning models, and the other using key acoustic features (fundamental frequency) with machine learning models. The second group of methods preserve critical audio information while reducing privacy breach risks. After training and calibration, we applied three superior methods to two managerial applications: Net Promoter Score (NPS) and prediction polls. In a restaurant NPS survey, inferred uncertainty provided more accurate predictions at both the customer and restaurant levels. In football prediction polls, it improved prediction accuracy. These voice-based uncertainty inference methods should be valuable tools for practitioners, including CMOs, to obtain better customer insights.

演讲人简介：

Tongyao Shen is a doctoral candidate in Marketing at the Smeal College of Business, Pennsylvania State University. Her research interests focus on Voice Technology in Marketing, Marketing Leadership, Labor Markets, Virtual Reality, and Retailing.

欢迎广大师生参加!