Service research is characterized by complex theoretical frameworks involving latent variables (constructs), complex structural relationships, including higher-order models, and a growing emphasis on prediction. Under these conditions, partial least squares structural equation modeling (PLS-SEM) has gained increasing prominence as a methodological approach well suited to provide answers to research questions in the service domain. As a result, PLS-SEM has become widely adopted across a broad spectrum of applications in the service domain.
These applications span technology-enabled services—such as consumers' responses to service robots, smart voice assistants, AI-based services, and chatbots—as well as market-related phenomena, including ownership perceptions in the sharing economy. Moreover, PLS-SEM has been extensively used to examine employee- and organization-related issues, such as frontline employee characteristics and servitization, service failure and recovery, employee–AI collaboration, customer experience management, value co-creation, and leadership styles.
At the same time, methodological developments in PLS-SEM have progressed rapidly. Recent advances include its combination into a necessary condition analysis framework, advanced types of mediation analysis, model comparison techniques, endogeneity assessment, and predictive model evaluation. Frameworks for assessing the robustness of results have further expanded the methodological toolkit available to service researchers.
Together, these developments offer significant opportunities—but also challenges—for service research. They call for more transparent, theoretically grounded, and methodologically purposeful applications of PLS-SEM that clearly articulate whether models are intended to explain, predict, or both. This special section aims to leverage these opportunities by advancing the methodological sophistication and substantive contribution of PLS-SEM-based research in the service domain.