In the context of a new wave of technological revolution and industrial transformation, China has intensified its efforts in basic research and original innovation, with strategic emerging industries and future industry sectors experiencing rapid development. Innovation consortia—collaborative organizations led by enterprises and involving universities, research institutes, and other stakeholders—have emerged as critical vehicles for integrating the innovation and industrial chains and achieving breakthroughs in key core technologies. Despite significant progress in establishing such consortia—over 1,000 have been launched nationwide since 2019, including 24 at the national level coordinated by the State-owned Assets Supervision and Administration Commission—challenges remain. Many local policies lack detailed implementation guidance, and issues such as unclear mechanisms for risk sharing, benefit distribution, and intellectual property rights have led to phenomena of “alliance without integration” or “cooperation without sharing.” As a result, the actual effectiveness and development level of many innovation consortia remain suboptimal. To address this gap, this study systematically investigates the factors that influence the construction and performance of innovation consortia in China.

Building on the strategic triangle framework—which integrates resource, institutional, and industrial dimensions—this research adopts a mixed-methods approach combining grounded theory and machine learning to identify and prioritize key influencing factors. First, through open coding of policy documents, academic literature, and case studies, 36 initial factors were extracted and categorized into three dimensions: resource (e.g., scientific and technological human resources, innovation resources, market-oriented resources), institution (e.g., internal governance mechanisms, outcome sharing and protection systems, government support policies), and industry (e.g., industrial cluster advantages, industrial chain structure, regional upgrading demands). A theoretical model was then constructed to conceptualize how these factors collectively shape the development level of innovation consortia.

To validate and rank the importance of these factors, empirical data from 1,734 publicly announced innovation consortia were collected and analyzed. Among them, 356 consortia that include listed enterprises were selected as the analytical sample due to the availability of more reliable and quantifiable performance indicators. A random forest algorithm—a robust machine learning technique suitable for handling complex, non-linear relationships among variables—was applied to assess the relative importance of each factor. The results were further refined by incorporating subjective evaluations from domain experts to ensure contextual accuracy and practical relevance.

The findings reveal that scientific and technological human resources and industrial cluster advantages are the two most influential factors in enhancing the development level of innovation consortia. These are followed by market-oriented resources and innovation resources, which also play a substantial role. At the institutional level, the system for sharing and protecting innovation outcomes, internal governance mechanisms within the consortium, and government incentive and support policies are moderately important, indicating that well-designed institutional arrangements are essential for sustaining collaboration. In contrast, factors such as regional industrial upgrading needs and basic characteristics of the industrial chain exhibit relatively lower importance, suggesting that while macro-industrial contexts matter, they are less directly impactful than resource endowments and institutional design.

This study contributes to both theory and practice. Theoretically, it extends the application of the strategic triangle framework to the context of innovation consortia and bridges qualitative insight with quantitative validation through an integrated methodology. Practically, it provides evidence-based policy recommendations, including strengthening the foundation of high-quality talent and R&D resources, leveraging existing industrial clusters to foster collaborative ecosystems, and refining institutional mechanisms to enhance trust and coordination among consortium members. By identifying the hierarchical structure of influencing factors, this research offers a systematic reference for policymakers and practitioners aiming to improve the quality and effectiveness of innovation consortia in China.