Algorithmic Dependency and Merchant Vulnerability in Social Commerce Ecosystems: Evidence from TikTok Shop Seller Reviews

Social commerce platforms increasingly serve as ecosystem orchestrators, coordinating technological infrastructures, financial systems, governance mechanisms, and algorithmic resource allocation. While prior research has predominantly focused on consumer behavior, limited attention has been given to merchant vulnerabilities arising from platform dependency. This study investigates operational vulnerabilities within the TikTok Shop ecosystem using 8993 negative merchant reviews collected from the Google Play Store. BERTopic, a transformer-based contextual topic modeling approach, was employed to identify latent vulnerability themes embedded in merchant complaint narratives. The results revealed 28 interpretable topics, which were aggregated into six higher-order vulnerability dimensions: Infrastructural Instability, Ecosystem Integration Vulnerability, Financial Vulnerability, Algorithmic Dependency, Coordination Breakdown, and Governance Asymmetry. Among these dimensions, Infrastructural Instability (13.59%) and Ecosystem Integration Vulnerability (13.44%) emerged as the most prominent sources of merchant dissatisfaction. The findings indicate that merchant vulnerability extends beyond isolated operational issues and is embedded within interconnected technological, financial, governance, algorithmic, and coordination structures. This study contributes to platform ecosystem research by providing empirical evidence and a structured interpretation of how platform-controlled resource orchestration may be associated with merchant vulnerability alongside value creation, while highlighting ecosystem integration as a significant source of risk during platform transformation.