The online community has emerged and developed into an influential part that can rebuild social relationships and redefine user performance. Various social communities have been formed to fulfill people’s different demands, including social commerce and prosocial activities (e.g., knowledge sharing and crowdfunding). Social relationship among members has a significant effect on the user behavior and performance in these communities.

The complexity of social relationships in online communities has been continuously resolved using in the impact process of user performance. Existing studies have analyzed the influence of social relationships on user performance from multiple perspectives. However, more effort is required to obtain clear mechanisms in the specific context of the social community.

In this thesis, I explored the impact of social factors on user performance by analyzing local social networks and social recognition. Specifically, I focused on two types of common social communities, including social commerce and prosocial aiding. I applied econometric models to show the influence of social factors on user performance in specific social communities.

In the first essay, labeled as “Social network in the social commerce community,” I studied the social network effect of sellers on their product promotion effectiveness in the social commerce context. Drawing from the social network and information diffusion theories, I studied the impact of topological characteristics of local social networks, also known widely as “friend circle,” on a seller’s promotion effectiveness. Using data collected from a large social commerce website in China, I built econometric models to investigate how local social networks can bring more likes to a product promoted by a social commerce seller. It’s concluded that local social networks with lower density, clustering (i.e., grouping), and centralization lead to more likes and favorites on sellers’ posted products. The friend circle with higher centralized sellers will damage the product promotion performance. Meanwhile, the density effect serves as a substitute for both the grouping and centralization effects.

In the second essay, labeled as “Social recognition in the prosocial community,” I concentrated on the social recognition impact among peers on users’ subsequent prosocial contribution to a mutual-aid community. I leveraged the data from a large online aiding community of truck drivers in China to explore the impact of the public social recognition from help seekers, as well as the moderated effect of appreciated group proportion on individuals’ subsequent prosocial contributions, which is rarely investigated but does make an impact. In this study, a dynamic matching method was developed to address the potential endogeneity issues and empirically reflect the direct impact of public appreciation on helpers’ prosocial actions. The obtained results indicate that social recognition from help seekers does activate helpers’ subsequent prosocial contributions to the community. However, a higher proportion of appreciated targets would hamper the positive effect of recognition on awarded helpers’ subsequent prosocial decisions. This study theoretically contributes to the research on prosocial communities and provides potential practical implications.

In summary, this study contributes to the research on online social communities. Specifically, the first study contributes to the literature on social commerce and e-commerce by deepening the understanding of determinants of social commerce success by exploring the local network. The results show that the seller-accessible local network plays a dominant role in driving promotional performance. The study provides insights of local network effect for future research on social commerce. The second study explores the effect of public social recognition on individuals’ prosocial actions. The results confirm the positive effect of social recognition and highlight the imperative role of social reputation. This study expands the research stream of the prosocial community by considering the unique one-to-many social recognition and combining specific offline activities. The findings also extend prosocial behavior research by uncovering the disadvantage of the appreciated group proportion on individuals’ motives. It was concluded that the appreciation effect would backfire if every helper is shown appreciation, which is rarely considered in the previous studies.