Conversational virtual shopping assistants (CVSAs) have emerged as innovative tools for product recommendation. However, there is still no broad consensus in the industry regarding their design, which has hindered their rapid diffusion. Moreover, theoretical understanding of how CVSA design affects downstream consumer behavior remains limited.
Drawing on the theoretical lens of design affordance, this research employs a mixed-methods approach to address these gaps. A grounded theory analysis of 40 semi-structured interviews identifies 2 key design affordances of CVSAs: modal affordance and aesthetic affordance. These affordances are further examined through 3 online experiments (N = 672) and an eye-tracking experiment to explore their influence on consumers' recommendation adoption intention.
CVSAs featuring multi-modal (vs uni-modal) responses and a warm (vs competent) aesthetic evoke significantly higher recommendation adoption intention. Perceived value – comprising both functional value and affective value – mediates these effects. Moreover, consumers' near field communication and their shopping goal moderate the impact.
This research innovatively applies a design science perspective to identify design affordances for CVSAs. It also highlights the mediating role of perceived value as a key explanatory mechanism.
