Purpose This study developed a cluster-based framework for measuring MSME ecosystem performance that addresses the limitations of monolithic rankings by accounting for structural heterogeneity across Indian states. Design/methodology/approach A two-stage analytical framework is employed using panel data from 28 Indian states (2016–2021). First, K-means clustering grouped states into homogeneous ecosystem configurations based on 20 indicators across five facilitators: Regulatory Framework, Entrepreneurial Capabilities, Credit Support, Market Conditions and Cultural Factors. Second, TOPSIS with Shannon Entropy objective weighting generated performance rankings within each cluster. Temporal alignment analysis examined correspondence between ranking patterns and documented policy interventions. Findings The analysis identified two stable ecosystem configurations – less-industrialized and industrialized – validating regional heterogeneity hypotheses. Intra-cluster rankings demonstrated meaningful performance differentiation and temporal correspondence with state-level MSME policies. Bihar’s rise in the less-industrialized cluster aligned with agro-processing incentives; sustained leadership by Tamil Nadu, Andhra Pradesh and Karnataka in the industrialized cluster corresponded with fiscal and innovation policies. Research limitations/implications Secondary data limitations and contradictory cases prevent definitive causal attribution. Integrating firm-level longitudinal data through primary survey will strengthen construct validity and enable the development of a formal, operationalizable MSME Performance Index for entrepreneurial ecosystem heterogeneity and policy effectiveness. Practical implications The framework will enable state governments to benchmark performance against structurally similar peers so less-industrialized states should compare against Cluster 1 comparators rather than industrialized Cluster 2 states and to prioritize cluster-specific interventions focus should be on credit access for less-industrialized ecosystems and innovation infrastructure for industrialized ones. Shannon Entropy weights identify high-impact facilitators within each cluster, enabling targeted rather than uniform policy design. Cluster transition signals (such as West Bengal’s movements between configurations) provide early warning mechanisms for ecosystem instability, supporting proactive rather than reactive governance. Originality/value This study advances MSME ecosystem measurements by demonstrating that cluster-based performance assessment captures policy-responsive dynamics that uniform benchmarks obscure. This framework establishes a methodological foundation for context-sensitive policy evaluation in federal economies with substantial regional disparities, offering a replicable approach for developing countries pursuing MSME-led inclusive growth.
Mukherjee et al. (Wed,) studied this question.