Detecting promising green ventures with foresight and data mining

There is growing consensus on the need to decouple economic growth from environmental degradation, with radical innovation playing a critical role in addressing environmental challenges. Green ventures are key drivers of such innovation, yet decision-makers lack systematic tools to identify ventures with transformative potential. This study advances computational foresight by integrating weak signal detection with venture-level performance analysis to identify promising green ventures and anticipate pathways of transformative green innovation. Using Crunchbase metadata, we identify strong and weak signals in green venture business descriptions (2013–2018) and link them to individual venture performance in 2019–2024. We further apply LDA topic modelling to examine how weak signals evolve into structured business opportunity domains. The results show that strong signals play a limited, context-specific role, primarily supporting early-stage funding, while weak signals exhibit heterogeneous and generally attenuated effects. Weak signals consolidate into distinct domains, with Sustainable Finance & ESG Platforms emerging as the most scalable pathway, whereas Green Mobility and Sustainable Consumption face greater commercialization challenges. Geographic variation further highlights differences in entrepreneurial ecosystem specialization and scaling dynamics.