Mass entrepreneurship and innovation refer to encouraging the broad masses of the people, including industry, agriculture, commerce, education, and soldiers, to participate in entrepreneurship, encouraging all Chinese people to participate in innovation, which Premier Li put forward at the 2014 Summer Davos Forum in Tianjin. After seven years of development, the innovation orientation of mass entrepreneurship and innovation has become an important engine leading China’s economic growth in the future. This research aims to examine the effect of innovation orientation on enterprise innovation performance. Based on a survey of 378 high-tech SMEs in Sichuan Province, China, structural equation modeling (SEM) was employed to examine the research model. The result shows that innovation orientation and organizational flexibility significantly affect enterprise innovation performance. Among the effects of innovation orientation on enterprise innovation performance, capability and resource flexibility play mediators. This research disclosed that important factors affecting the innovation performance of high-tech SMEs enterprises include resource flexibility and capability flexibility. To give better play to the positive impact of innovation orientation, we should strengthen the flexible strategy of organizations. Thus, committed to more flexible market development, promoting the vigorous development of new technologies, new industries, and new formats, and realizing the growth of enterprise innovation performance.

Based on the theory of independent innovation and competitive advantage, this paper takes industry cluster as the independent variable, long-term high-technology small and middle size enterprises (high-tech SMEs) performance and short-term high-tech SMEs’ performance as the dependent variable, and introduces independent innovation as the mediator variable to explore the impact of industry cluster on high-tech SMEs’ performance. SPSS22.0 was used to test the reliability and validity of the questionnaire distributed to 310 high-tech SMEs in Sichuan, China. SPSS is used for statistical analysis, integrating data entry, organization, and analysis functions. Its basic functions include data management, statistical analysis, chart analysis, and output management. A confirmatory factor analysis was conducted. Amos 24.0 is the structural equation model analysis software. After using Amos 24.0 to construct the SEM (Structural Equation Modelling) to verify the hypothesis, it was found that industry cluster has a significant positive impact on long-term and short-term high-tech SMEs’ performance, independent innovation has a significant positive impact on long-term and short-term high-tech SMEs’ performance, and independent innovation plays a mediator role in the relationship between industry cluster and high-tech SMEs’ performance. Based on the research results, this paper puts forward the following suggestions: (1) attach importance to independent innovation, introduce relevant technical talents, and improve innovation; and (2) accelerate the formation of clusters to improve the high-tech SMEs’ performance of the whole industry.

Despite research showing that business model innovation (BMI) can enhance performance and create competitive advantages for small- and medium-sized businesses, these firms often do not achieve the expected outcomes. A business model can undergo irreversible fundamental changes due to Business Model Innovation (BMI), resulting in high levels of risk, uncertainty, and ambiguity. An analysis of 350 Chinese small- and medium-sized enterprises (SMEs) was conducted to determine the impact of innovating a business model on firm performance. Based on BMI calculations, this study investigates whether organizational capabilities and growth strategies impact business performance. Even though BMI is not directly correlated with firm performance, growth in efficiency and novelty, organizational capacities, and revenue are all fully mediated by growth in those factors. Revenue growth, organizational capacity, and efficiency affect a firm’s performance. The model is proven to be valid by this study. In addition to providing guidelines on implementing BMI based on a company’s strategy, this study contributes to the existing literature on BMI efforts for small- and medium-sized businesses.

In the context of the digital economy and cross-organizational collaboration, enterprises forming alliances to conduct data collaboration and joint modeling (e.g., cross-silo federated learning) have become a crucial means to enhance supply chain resilience and risk management capabilities. However, trustworthy data collaboration is not merely a simple data exchange: member firms have an inherent incentive to protect their data assets, and their contribution costs and compliance risks are highly heterogeneous and typically constitute private information. Simultaneously, constraints at the alliance level—such as incentive budgets, privacy leakage risks, and on-chain throughput/confirmation delays—exhibit characteristics of long-term accumulation and random fluctuations. Solely pursuing single-slot optimality or relying on centralized scheduling can easily trigger short-term runs, free-riding, and governance failures. To address these challenges, this paper proposes an online governance framework centered on long-term guardrails. We model the alliance collaboration process as a stochastic social welfare maximization problem with long-term average constraints, transforming budget, privacy, and congestion constraints into online regulatable price signals via virtual queues. Furthermore, we introduce Bayesian belief updating to learn the unknown member costs and contribution reliability, thereby adaptively adjusting incentive intensity under incomplete information. Finally, the per-slot drift-plus-penalty subproblem is formulated as an exact potential game with congestion pricing, enabling members to distributively solve it and achieve a consistent global improvement direction by exchanging only minimal price signals. Theoretically, we provide sufficient conditions for queue stability and long-term constraint satisfaction, characterizing the impact of potential game approximation errors and belief estimation errors on the long-term performance bounds. Reproducible experiments based on supply chain finance and data market trajectories demonstrate that the proposed framework significantly reduces the frequency of budget/privacy violations while enhancing long-term collaboration value and model quality, maintaining stable convergence and low communication overhead as the membership scales.