Unlocking the power of multimodal online reviews: A multisensory perspective

Abstract

Multimodal online reviews have been an essential marketing tool for dominating consumer perception. However, the precise impact of visual, auditory, tactile, gustatory, and olfactory cues in multimodal content on consumers’ perceived review usefulness remains unclear. This study aims to fill this gap by exploring how text and image contents affect review usefulness from a multisensory perspective. To this end, we proposed a framework that combined machine learning and multimodal large language models to quantify the multisensory elements conveyed through review text and images. Based on the Poisson regression model, the estimation results show that while multisensory elements in text and images have a generally consistent impact on consumer perception, each sensory element carries distinct value. Furthermore, the sensory discrepancy in text and images has a negative effect on the perceived review usefulness. Our findings contribute theoretically to enhancing review persuasion effect and sensory marketing literature.