Is Your Voice Enough, Alexa? Assessing the Role of Digital Assistant Personality, Modality, and Product Involvement on Consumer Evaluations.

Abstract

This study dives into the relatively unexplored topic of digital assistants and the new phenomena of voice shopping. Specifically, we were interested in the different characteristics of said assistants on purchase intention. Given their relevance in this context, we also studied trust and perceived risk. We conducted a three-way between- participants experimental design with factors personality (social vs. intellectual), product involvement (low vs. high involvement), and modality (voice vs. voice and visual). With digital assistants and voice interactions being increasingly used in society, we also introduce a novel concept named “Need for Voice” (NFV) to determine whether there are idiosyncratic differences among individuals in their proclivity to enjoy and be affected by voice interactions. Through an online experiment (n = 641) we find, as expected, that low involvement products have a higher purchase intention than high involvement products. We did not find evidence in our data to suggest that personality or modality influence purchase intention. We do find initial evidence in support of a multidimensional NFV construct, however it requires additional validity checks and future testing. Our findings show that there is high risk associated with voice shopping, and that familiarity and previous experience is crucial to increase the rather low purchase intention. At this moment in time, we recommend manufacturers and third-party companies to be aware that selling low involvement products beyond the “reordering”-category might be challenging, and that normalization of the digital assistant as a sales channel will likely be of assistance. Regarding high involvement products, companies should be careful to sell these through voice shopping, as evidence in this study shows that these types of products have the lowest purchase intention. Finally, manufacturers should be cognizant of the possibility that a screen showcasing visual information in terms of a picture and text might not assist the user in voice shopping, and that future research is needed to evaluate its effect.