We investigate the participation and effectiveness of paid endorsers in viral-for-hire social advertising. We conduct a field experiment with an invitation design in which we manipulate both incentives and a soft eligibility requirement to participate in campaigns. The latter provides a strong and valid instrument to separate participation from outcomes effects. Since likes, comments, and retweets are count variables, and since potential endorsers can self-select to participate in multiple campaigns, we propose a Poisson lognormal model with sample selection and correlated random effects to analyze variations in participation and effectiveness. There are three main findings. (1) Payments higher than the average reward a potential endorser received in the past (gains) do not increase participation, whereas lower payments (losses) decrease participation. Neither gains nor losses affect effectiveness. (2) Potential endorsers who are more likely to participate tend to be less effective. (3) Which endorser characteristics are associated with effectiveness depends on whether success is measured in likes, comments, or retweets. These findings provide new insights on how marketers can improve social advertising campaigns by better targeting and incenting potential endorsers.
Social media platforms allow users to connect and share content. The extent of information diffusion may depend on the characteristics of users’ connections, such as the overlap among users’ connections. We investigate the impact of network embeddedness (i.e., number of common followees, common followers, and common mutual followers between two users) on the information diffusion in directed networks. To accommodate the empirical observation that a user may receive the same information from several others, we propose a new hazard model that allows an event to have multiple causes. By analyzing the diffusion of sponsored ads on Digg and brand-authored tweets on Twitter, we find that the effect of embeddedness in directed networks varies across different types of “neighbors”. The number of common neighbors are not always conducive to information diffusion. Moreover, the effects of common followers and common mutual followers are negatively moderated by the novelty of information, which shows a boundary condition for previous finding on embeddedness in undirected networks. For marketing managers, these findings provide insights on how to target customers in a directed network at the micro level.