Research on the Influence of User Characteristics on User-Generated Content

Abstract

The rapid development of the internet and social media has fundamentally transformed information ecosystems, giving rise to widespread user-generated content (UGC). With the maturation of e-commerce platforms, consumers actively share opinions, evaluations, and experiences regarding products and services through various formats, making UGC a central element of digital commerce. Users rely on it to obtain authentic, diverse insights for informed decision-making; businesses leverage high-quality UGC to enhance brand reputation and guide product development; and platforms treat valuable UGC as strategic digital assets that shape long-term competitiveness. Therefore, the volume, quality, and authenticity of UGC directly impact the interests of these key stakeholders, making its generation and management a core strategic concern.

Research has increasingly focused on strategies to improve the quantity, quality, and credibility of UGC. However, existing literature predominantly emphasizes external factors—such as platform design, merchant strategies, and social influence—while paying limited attention to the intrinsic characteristics of users themselves, including their identity attributes, psychological traits, and behavioral patterns. As the primary creators of content, users' individual characteristics fundamentally shape their likelihood of contributing and the nature of what they produce. Consequently, developing a systematic framework to examine how user characteristics influence UGC on e-commerce platforms holds significant theoretical and practical value.

This study focuses on two key dimensions of user characteristics: identity and behavioral traits, specifically investigating the effects of users' social class, stickiness, and batch content generation behaviors on UGC. Integrating theoretical perspectives from psychology, behavioral science, and marketing, this research constructs an analytical framework, employs text mining techniques to extract content features, and applies econometric and machine learning models for empirical validation. The specific investigations are outlined as follows:

First, this paper examines the impact of users' social class on UGC. Drawing on the social cognitive theory of social class, it theoretically analyzes and empirically tests the direct effects of social class on content characteristics, as well as its indirect influence on content helpfulness through these features. Empirical analysis based on UGC datasets reveals that lower-class users exhibit greater conformity in their evaluations and express less emotion in their textual content. The relationship between social class and content quality is moderated by platform social orientation: on platforms with low social affordances, lower-class users produce higher-quality content, whereas the reverse is observed on platforms with strong social features. Furthermore, social class indirectly affects content helpfulness—content exhibiting higher conformity, lower emotional expressiveness, or lower quality is perceived as less helpful.

Second, this paper investigates the impact of user stickiness—defined as sustained engagement with a product—on UGC. Grounded in attachment theory and self-determination theory, the study empirically examines how user stickiness influences content generation propensity and quality, while testing boundary conditions across user experience, product type, and merchant size. Results show that greater product stickiness significantly increases both the likelihood of generating content and its quality. This positive effect is stronger among experienced users and for privately consumed products. Additionally, the effect on generation propensity is further enhanced when the product is offered by small-scale merchants.

Third, this paper investigates the impact of batch content generation behaviors on UGC. By analyzing the temporal density of content creation, the study distinguishes between batch (multiple contributions created in close succession) and non-batch generation. Based on cognitive load theory and goal gradient theory, it examines how batch generation affects content quality, with moderation effects tested for time of day and device type. It also models the dynamic trajectory of quality within individual batch sessions. Findings indicate that batch content generation is negatively associated with content quality. This negative effect is amplified during nighttime and when content is created via mobile devices. Within batch sessions, content quality follows a nonlinear trajectory: in shorter sessions, quality declines monotonically with creation order; in longer sessions, quality initially declines but later recovers.

By elucidating the roles of users' social class, stickiness, and batch content generation behaviors, this research makes several contributions. Theoretically, it extends the framework of UGC antecedents, deepens understanding of the link between user characteristics and UGC, and pioneers the integration of social class and sequential creation behavior into UGC research. Practically, it offers actionable insights for e-commerce platforms to optimize UGC solicitation and presentation mechanisms, for merchants to develop targeted reputation management strategies, and for users to better evaluate the credibility and value of UGC.

Date of Award	11 Nov 2025
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Zili ZHANG (External Supervisor) & Chuangyin DANG (Supervisor)