Sociodemographic versus Geographic Proximity in the Diffusion of Online Conversations

Social media platforms are changing how people spread information via word of mouth by allowing individuals to rapidly disseminate information to virtually anyone. Yet little is known about whether this ability affects how contagion spreads within online social systems. While digital platforms may facilitate diffusion to those nearby, they also may facilitate dissemination to like-minded folk regardless of physical proximity. These different outcomes suggest competing drivers for the diffusion of online conversations: geographic proximity, sociodemographic proximity, or a blend of both. Identifying which of these drivers underlies virality is critical because they yield strategically different marketing implications. Based on a data set of 355,021 microblogs, we build a spatiotemporal model to capture the diffusion of conversations within the United States. When accounting for sociodemographic proximity, geographic proximity does not significantly govern the spread of online conversations. Rather, sociodemographic factors propagate online social contagion, suggesting that social media campaign success relies on sociodemographic segmentation-based targeting.