Are helpful reviews indeed helpful? Analyzing the information and economic value of contextual cues in user-generated images

Abstract

When shopping online, customers may find user-generated images (UGIs) where existing buyers share their product experiences in an actual setting. Drawing on the constructivist theory of visual perception, we propose a cognitive inference process in which shoppers utilize the background objects in UGIs that contextualize a product (e.g., a snow-covered mountain implying cold weather) to infer its features (e.g., the warmth of a jacket). As a result, the contextual cues in UGIs play a critical role in facilitating future buyers' purchase decision-making. Our empirical probes using data from an online outdoor gear and clothing retailer confirm this conjecture by demonstrating that product contextualization in UGIs increases a review's perceived helpfulness and improves sales. By contrast, the contextual cues in the review text only assist buyers' purchase decision process when they contextualize product functionality (e.g., the windproof of a coat). Yet, they do not work for aesthetic attributes (e.g., the color of a coat). We leverage an experiment to explore the relevant mechanism. In the sales analysis, we reveal that not all image content considered helpful would positively affect sales. For example, after we account for the contextual cues in the UGI, the mere presence of an image in product reviews does not affect sales. On the other hand, although illustrating product malfunction in a UGI does not increase its helpfulness, such content hurts sales. We offer managerial implications based on the empirical findings for review platforms that aim to assist online shoppers in making informed purchases.