Breaking the mold: how visual atypicality shapes brand-generated image virality
研究品牌在社交媒体上发布的图片的视觉非常规性（即与行业常规的偏离程度）如何影响其病毒式传播，发现对低传统品牌提升娱乐价值可增强传播，但对高传统品牌则因违背消费者预期而降低传播。
Brands frequently post images on social media, hoping consumers will share them to enhance product visibility and brand awareness. This research introduces visual atypicality (i.e., the extent to which an image’s visual presentation of a brand’s products deviates from prevailing industry norms) and examines its impact on the virality of brand-generated images. Employing machine learning techniques to analyze 60,669 images from 330 apparel brands and 4,087 images from 27 car brands, supplemented by an online experiment, this research reveals that the effectiveness of visual atypicality depends on brand heritage. For low-heritage brands, visual atypicality enhances virality by increasing entertainment value. Conversely, for high-heritage brands, visual atypicality reduces virality as it clashes with consumer expectations. These findings suggest that, in order to maximize social media engagement, brands should strategically consider the degree to which their content aligns with both industry conventions and their own brand identity.