Exploring variance in SMEs performance: quantifying the business model effect

Understanding why SMEs differ in performance is crucial for entrepreneurship and strategy research, yet evidence on whether and to what extent business models explain performance heterogeneity beyond firm- and industry-level drivers remains limited. Drawing on the conceptualization of the business model as a system of interdependent activities, we classify a large sample of Italian manufacturing SMEs into distinct business models using financial statement data and a clustering approach. We subsequently estimate longitudinal multilevel Bayesian models to decompose the overall variation in firm performance into firm, industry, and business model effects. The results of Markov Chain Monte Carlo estimates show that business models explain 7.1% of the variance in firms’ ROA, which is smaller than the firm effect but outweighs the industry effect. The business model effect is stronger in high-tech industries, family firms, and high-growth entrepreneurial firms. Overall, the findings demonstrate that business models constitute a distinct level of analysis for explaining performance heterogeneity and a key source of competitive advantage for SMEs.