Visual presentation is critical in online marketplaces where customers rely on images to evaluate products without physical interaction. This study examines the primary and contingent effects of image aesthetics on product sales by analyzing 5,074 dish images and 506,553 orders from 71 merchants on a leading food delivery platform. We classify signals into promotional and reputational types, with product images serving as promotional signals that help reduce customers’ product uncertainty. Our hypothesis posits that additional information provided by the platform or merchants may either substitute or complement the value of these images, depending on the signal type (i.e., promotional or reputational). To address limitations associated with subjective ratings or low-level image features, we employ a deep learning algorithm to quantify image aesthetics more effectively. Our results show that image aesthetics positively influence sales. This effect is further strengthened by reputational signals (e.g., platform certification) but weakened by other promotional signals (e.g., advertisements). These findings highlight complementary and substitutive effects between different signal types, offering both theoretical insights and practical guidance for visual design strategies in e-commerce.