Encode and interpret senses: how sensory nature of recommendation tasks influences consumer preference for algorithms

Abstract With the rapid development of algorithmic recommendations, unearthing contexts in which consumers accept algorithms more demands greater attention. This study delves into how the sensory nature of recommendation tasks influences consumer acceptance of algorithms. We distinguish tasks based on their primary reliance on distal senses versus proximal senses. Distal senses enable people to perceive from afar, like seeing or hearing, while proximal senses require close physical proximity to smell, taste, or touch. Through four well-controlled experiments (including two pre-registered studies; N = 1,050) and a pretest, we illustrate that consumers prefer the algorithm for tasks tied to distal senses and are more skeptical about its effectiveness for tasks linked to proximal senses. The underlying mechanism is that consumers perceive algorithms to be more capable of processing distal sensory information, but lack the capacity to process proximal sensory detail. This research provides both theoretical contributions and practical implications for algorithmic recommendations.