ESG and sustainability trends challenges and opportunities using PLS-SEM and unsupervised machine learning

This study investigates the links between Environmental, Social, and Governance (ESG) factors, sustainability practices, and stakeholder pressures in influencing corporate sustainability performance and long-term benefits. Using a quantitative approach, data were collected from 372 respondents in Indonesia. Partial Least Squares Structural Equation Modeling (PLS-SEM) was employed to test the hypothesized causal relationships and moderating effects. To extend the structural findings, an unsupervised machine learning approach was integrated through cluster analysis applied to the latent variable scores obtained from the PLS-SEM model. Theoretically, this study contributes by integrating variance-based structural modeling with configurational analysis, thereby moving beyond net-effect explanations toward the identification of heterogeneous organizational response patterns. Methodologically, the added value of cluster analysis lies in its ability to complement PLS-SEM results by uncovering distinct sustainability typologies that cannot be captured through linear structural paths alone. This hybrid analytical design enables the identification of hidden organizational patterns and natural groupings based on ESG awareness, commitment, and sustainability behavior. The results show that ESG factors significantly enhance both corporate sustainability performance and long-term benefits, while sustainability practices reinforce these effects. Stakeholder pressures moderate the relationship between ESG factors and corporate sustainability performance but no other links. The clustering analysis further reveals three ESG-sustainability profiles, highlighting differences in strategic responses and innovation levels among organizations. These typologies provide a more nuanced understanding of how firms translate ESG orientation into performance outcomes under varying levels of stakeholder pressure. These findings demonstrate how combining PLS-SEM with unsupervised learning can uncover structural relationships and latent behavioral typologies simultaneously. The study’s originality lies in applying this hybrid method within Indonesia’s emerging market context, providing both methodological advancement and practical insights for policymakers and business leaders seeking to strengthen ESG integration and achieve sustainable long-term outcomes. Practically, the identified clusters offer actionable guidance for managers and decision-makers to tailor ESG strategies, allocate resources more effectively, and design policy interventions that align with the specific sustainability maturity level of different organizational groups.