Designing profitable seeding Programs: The effects of social network properties and consumer homophily

Abstract

This study explores the main and interaction effects of consumer social network properties and homophily on the profit impact of seeding programs. Social network properties capture the topology of the connections between consumers, while homophily captures the degree to which consumers share similar attributes with their social connections.We conduct extensive agent-based simulation experiments on eight empirical social networks mapped from Facebook, Twitter, and Emails, among others. The results reveal that seeding programs generate higher profits in clustered social networks, and are less profitable when the network diameter and average distance between consumers increase. The effects of social network diameter and average distance are only observed when homophily is high and are negligible when homophily is low. For random seeding and targeting social hubs, the effect of social network clustering on profits is stronger when homophily is high. For early adopters and experts, the effect depends on the attribute on which homophily exists in the market.We explain the findings by distinguishing between the effects of seeding programs on local diffusions—effects on direct acquaintances—and global diffusion processes—the impact across the social network. The findings offer critical implications for marketers aiming to maximize the return-on-investment in seeding programs.