SME performance is a multifaceted construct influenced by diverse factors, including financial management, digitalization, strategic orientation, and market share. These elements are critical for SMEs to navigate competitive landscapes and achieve sustainable growth (Na-Nan et al., 2017). In Indonesia, where SMEs play a substantial economic role, enhancing their performance is a priority.

Performance measurement for SMEs often extends beyond pure financial metrics. The Balanced Scorecard (BSC) framework is effective in aligning operational activities with strategic objectives by incorporating both financial and non-financial perspectives, specifically the financial, customer, internal business processes, and learning and growth dimensions (Kaplan and Norton, 2001; Muraba et al., 2024; Pineyrua et al., 2021).

SR involves measuring, disclosing, and accounting for an organization’s environmental, social, and governance (ESG) impacts. The main goal is to promote sustainable development and enhance long-term firm value by effectively communicating performance to stakeholders (Das et al., 2020a; Bunclark and Barcellos-Paula, 2021). For this study, SR is operationalized using formative constructs based on the GRI standards (2020), grouped into three primary indicators: economic value and financial literacy; social networks and sustainable innovation; and eco-efficiency and energy.

Financial literacy—encompassing both knowledge and application (Huston, 2010)—is crucial for sound financial decision-making and business growth. It enables SMEs to manage finances more effectively, secure better funding, and improve future welfare (Setyaningsih et al., 2024; Al-shami et al., 2024; Putri et al., 2025). Research shows a positive link between financial literacy and enhanced SME performance Furthermore, social networks and sustainable innovation are critical. Continuous innovation is fundamental for creating competitive advantage (Che and Chen, 2024; Dolšak et al., 2024), and social networks, ranging from internal to external ties, are vital catalysts for the diffusion of innovation and new product development (Leick and Gretzinger, 2020; Muafi, 2020). Finally, eco-efficiency involves minimizing environmental impact while significantly maintaining the economic value. Implementing green innovations and energy conservation not only yields environmental benefits but also reduces production costs and enhances brand image, thereby improving economic performance (Thompson et al., 2024; Dolšak et al., 2024). Based on this reasoning, the following hypotheses are proposed:

Supply chain management is a holistic approach to efficiently integrate suppliers, manufacturers, distributors, and customers to meet consumer demand. SCM is measured through five key indicators: product development, supplier relations, the production process, distribution, and customer returns management. Within SCM, product development serves as a strategic platform for innovation, enabling SMEs to ensure business continuity and competitiveness (Machado et al., 2020). Effective supplier relations and collaborative procurement are crucial for securing quality inputs and enhancing overall business performance (Chijioke and Vu Minh, 2021). Optimizing the production process enhances operational efficiency, reduces waste, and strengthens competitiveness (Simona et al., 2023). Furthermore, efficient distribution logistics directly improve market accessibility and operational outcomes (Nupus and Ichwanudin, 2020). Lastly, an effective system for managing customer returns is critical for maintaining satisfaction, loyalty, and operational efficiency (Rufina, 2022); (Lu et al., 2021). Consequently, the following hypotheses are advanced (Figure 1):

The study sample consists of SMEs enrolled in the Jakpreneur program, coordinated by the Provincial Government of Jakarta, Indonesia. This program was selected as the sampling frame because it provides participating SMEs with structured support—including mentoring, training, access to capital, marketing assistance, and business licensing—resulting in a relatively homogeneous population of over 300,000 active members. As an initial study, convenience sampling was employed for several reasons. First, this initial research examines how sustainability, SCM, and the performance of SMEs are interconnected. The goal is to better understand these relationships before carrying out more in-depth studies. In early-stage research like this, speed and flexibility are more important than making broad generalizations (Etikan, 2016; Saunders et al., 2023). Second, due to time constraints, data collection needed to be efficient to reduce the logistical burden associated with more complex sampling designs (Memon et al., 2025; Sunarsih et al., 2026). Third, a total of 203 SMEs voluntarily participated, which meets the minimum sample size recommended by the inverse square root method (Kock and Hadaya, 2016; Legate et al., 2023) based on predictive R² criteria. According to the minimum R-squared guideline (Kock and Hadaya, 2016), for a model where R² > 0.10 and the maximum number of arrows pointing at a construct exceeds 10, the required sample size is 189. Therefore, the sample of 203 satisfies the minimum threshold for robust partial least squares structural equation modeling (PLS-SEM).

Table 1 shows that the sample is composed of micro-enterprises, with 91.63% having fewer than five employees. This strongly reflects the typical structure of the SME sector, where very small, often owner-operated, businesses are the norm. In terms of business focus, the majority (74.38%) operate in the Food and Beverage sector, indicating that this is a primary industry for local SMEs. The retail sector denotes a significant minority (20.69%), while other services form a much smaller segment (4.93%). Regarding the respondents’ demographics, the largest group (45.81%) falls within the 25–45 years age bracket, representing the core working-age and entrepreneurial population. A substantial portion (32.51%) is also represented by younger entrepreneurs (18–25 years), suggesting active participation from a new generation of business owners. Respondents over 45 years contain 21.67% of the sample.

The measurement instruments for this study were developed by adapting and integrating well-established constructs from the literature to fit the specific context of SMEs in Indonesia. All items were measured using a five-point Likert scale (1 = Strongly disagree to 5 = Strongly agree). A detailed overview of the constructs, dimensions, and sources is provided in Table 2. The initial scale development involved content validity assessment by five experts in management and finance to ensure the relevance and clarity of the proposed items.

R was modeled as a formative higher-order construct comprising three key dimensions, with items adapted from academic frameworks and the global reporting initiative (GRI) standards. Economic value and financial literacy were measured using three items that assess the understanding and application of accounting principles, bookkeeping, and financial planning, drawing on (Bunclark and Barcellos-Paula, 2021; Al-shami et al., 2024), and aligned with relevant GRI standards (e.g., GRI 102, 201). Social Network and Sustainable Innovation was measured with three items evaluating community involvement and stakeholder relations, informed by the same sources and GRI standards on local communities (e.g., GRI 413). Eco-Efficiency and Energy were measured using three items covering environmental management practices related to resource use and materials, based on Bunclark and Barcellos-Paula (2021) and Al-shami et al. (2024), and corresponding GRI categories (e.g., GRI 301, 302).

Supply chain management was also modeled as a higher-order formative construct, comprising five dimensions derived from an integrated SCM framework. Product/business development (three items), procurement process (six items), production process (six items), shipping and distribution (three items), and customer returns management (four items) were established by adapting scales from Machado et al. (2020) and Oliveira et al. (2022). These items measure core SCM capabilities, including innovation, supplier relations, operational efficiency, logistics, and post-sales service.

SME Performance was measured formatively according to the four perspectives of the Balanced Scorecard (BSC), as defined by Kaplan and Norton (2001). The financial perspective used three items focusing on revenue and profit stability. The customer perspective included four items related to customer growth and satisfaction. The internal business process perspective comprised four items assessing product diversity and quality control. The learning and growth perspective was measured with seven items covering employee skills, training, and organizational effectiveness.

The pilot test results confirm the measurement model’s internal consistency (see Table 2). All constructs demonstrated strong internal consistency, with Cronbach’s Alpha values exceeding the accepted threshold of 0.70. These results, detailed in Table 2, indicate that the adapted scales are reliable and valid for measuring SR, Supply Chain Management, and SME Performance in the context of this study.

This study employs a two-stage PLS-SEM analytical procedure. The analysis was conducted using SmartPLS software, well-suited for this research due to its predictive orientation, the use of formative constructs, and the relatively small sample size. In the first stage, a higher-order construct evaluation was executed for SME performance. Since performance was conceptualized as a formative second-order construct comprising the four BSC perspectives (financial, customer, internal business process, and learning and growth), latent variable scores were generated. These scores were produced by creating composite indices from the first-order dimensions, resulting in a single, robust composite score for overall SME Performance to be used in the subsequent path analysis. The second stage involved the primary PLS-SEM analysis. The structural (inner) model was evaluated to test the hypothesized relationships among all eight specific dimensions of SR, SCM, and SME performance. This analysis assessed the individual influence of each dimension—economic value, social network and innovation, eco-efficiency, product development, supplier relations, production process, distribution, and customer returns management—on the performance outcome, in line with hypotheses H1–H8.

The measurement model demonstrated strong psychometric properties, confirming the reliability and validity of the scales employed in this study. First, indicator reliability was established, as all factor loadings exceeded the recommended threshold of 0.70, confirming that each item is a reliable measure of its corresponding latent construct. Second, internal consistency and convergent validity were confirmed. All constructs exhibited excellent internal consistency, with Cronbach’s alpha (α) and composite reliability (CR) values surpassing the accepted level of 0.70. Convergent validity was established, as the average variance extracted (AVE) for every construct exceeded the benchmark of 0.50. Collectively, these results confirm the robustness of the measurement model, providing a solid foundation for proceeding with the structural model analysis. Thirdly, the discriminant validity was verified using the heterotrait–monotrait ratio (HTMT) analysis, where all values stayed below the conservative cutoff of 0.90 (highest HTMT = 0.72) (Franke and Sarstedt, 2019; Legate et al., 2023). These findings validate the reliability and justify the measurement model for the following structural analysis (Table 3).

This study investigates the direct relationships among the eight key dimensions of SR, SCM, and SME performance. The structural model was assessed by examining the significance of the standardized path coefficients (using a bootstrapping procedure with 5,000 subsamples), the model’s explanatory power (R²), the effect sizes (f²) of the predictor variables, and its predictive relevance through the Stone-Geisser Q² test.

The analysis results are depicted in Table 4. Diagnostic checks confirmed the absence of multicollinearity, with all variance inflation factor (VIF) values below the threshold of 5. The model demonstrates exceptionally strong explanatory power, explaining 82.4% of the variance in SME performance (R² = 0.824). Predictive relevance was confirmed, as the blindfolding procedure (omission distance D = 7) yielded a Q² value of 0.492 for SME performance [calculated as Q² = 1-(1-R₁²)], which is significantly above zero.

Six of the eight proposed hypotheses were statistically significant. The dimensions of production process (β = 0.268, p < 0.01), distribution (β = 0.195, p < 0.05), and product return management (β = 0.177, p < 0.05) demonstrated the most potent positive effects on SME performance, with small to medium effect sizes. Networking innovation (β = 0.128, p < 0.05), product development (β = 0.164, p < 0.05), and economic value (β = 0.083, p < 0.05) were also significant, supporting hypotheses H2, H4, and H1, respectively. In contrast, the paths from eco-energy (β = −0.083, p > 0.05) and supplier relation (β = 0.119, p > 0.05) to SME performance were not statistically significant; therefore, hypotheses H3 and H5 were not supported.

This study evaluated the out-of-sample predictive validity of the proposed model using the PLSpredict procedure (Hair et al., 2020; Legate et al., 2023). The results, presented in Table 5, demonstrate strong overall predictive performance. All indicators of the SME performance construct yielded positive Q²predict values, confirming the model’s predictive relevance. Furthermore, the PLS-SEM model consistently yielded lower root mean square errors (RMSEs) than the linear model (LM) benchmark for each indicator, as indicated by the negative ΔRMSE values. This pattern confirms that the specified structural relationships enhance prediction accuracy. A nuanced observation is made for the Financial Perspective indicator, which shows the smallest Q²predict value (0.359) and the most modest improvement in RMSE (ΔRMSE = −0.024). This proposes that while the model retains predictive power, its capacity to forecast this specific financial dimension is relatively constrained compared with other performance aspects, such as learning and growth, which showed the most significant gain in accuracy (ΔRMSE = −0.070).

These findings are further substantiated by the cross-validated augmentation test (CVAT). As shown in Table 5, the prediction error for the SME performance construct is substantially lower in the PLS-SEM model, with a PLS loss value (0.404) significantly less than the IA loss benchmark (1.013). This difference is statistically significant (t = 4.384, p < 0.001), confirming that the specified structural relationships meaningfully enhance the model’s predictive ability (Sharma et al., 2023). In sum, these results affirm that the integrated framework of SR and SCM dimensions offers a robust and valid model not only for explaining but also for reliably predicting SME performance.

This study examined the influence of SR and GSCM practices on the performance of SMEs. Based on data from 203 SMEs within Jakarta’s Jakpreneur program, the PLS-SEM analysis revealed that six of the eight hypothesized dimensions significantly enhanced SME performance. These positive drivers are Production process, distribution, product return management, networking innovation, product development, and economic value (financial literacy). In contrast, eco-energy and supplier relations did not show a statistically significant impact. The integrated model demonstrated strong explanatory power, accounting for approximately 82.4% of the variance in SME performance.

These findings are contextualized and enriched by qualitative insights, revealing a gap between formal concepts and on-the-ground practice. Notably, while economic value was a significant driver, interviews showed SME owners’ unfamiliarity with formal SR terminology, aligning with observations by Chen et al. (2024) regarding limited reporting capacity of SMEs. This suggests that the measured relationship stems from a foundational understanding of finance rather than advanced reporting. Similarly, the non-significant result for econergy is clarified by qualitative data showing that SMEs prioritize basic environmental actions over complex energy efficiency, reflecting the “green cost dilemma” for resource-constrained firms discussed by Das and Rangarajan (2020). The lack of support for supplier relations aligns with descriptions of transactional, cost-focused supply chains, a gap between partnership theory and SME reality noted by Chijioke and Vu Minh (2021).