Voice analytics in business research: Conceptual foundations, acoustic feature extraction, and applications

Recent advances in artificial intelligence and natural language processing are gradually transforming how humans search, shop, and express their preferences. Leveraging the new affordances and modalities of human–machine interaction through voice-controlled interfaces will require a nuanced understanding of the physics and psychology of speech formation as well as the systematic extraction and analysis of vocal features from the human voice. In this paper, we first develop a conceptual framework linking vocal features in the human voice to experiential outcomes and emotional states. We then illustrate the effective processing, editing, analysis, and visualization of voice data based on an Amazon Alexa user interaction, utilizing state-of-the-art signal-processing packages in R. Finally, we offer novel insight into the ways in which business research might employ voice and sound analytics moving forward, including a discussion of the ethical implications of building multi-modal databases for business and society.