Mapping the spatial evolution of entrepreneurial ecosystems: a clustering perspective on global investment flows
使用K-Means聚类分析2004至2024年全球超过11,670条投资记录，揭示创业生态系统投资空间格局从简单集中到复杂网络的演化，对学者、投资者和政策制定者有参考价值。
Purpose The spatial distribution of investments across entrepreneurial ecosystems reveals patterns of entrepreneurial growth, regional disparities and emerging financial centers, offering a nuanced perspective on the spatiality of entrepreneurial ecosystems. In this article, we investigate how these spatial dynamics have evolved globally from 2004 to 2024. Design/methodology/approach Using an unsupervised machine learning approach – K-Means clustering, chosen for its efficacy in detecting latent regional patterns – we analyze trends in funding volumes, investment valuations, growth metrics and investor density. Drawing from a global dataset of over 11,670 investment records from Crunchbase, we identify shifts in the spatiality of investments across entrepreneurial ecosystems, demonstrating some evolutionary nuances that outline an increasingly complex relational geography in investment flows. Findings During the first decade (2004–2014), spatial investment networks observed across entrepreneurial ecosystems presented much simpler topologies, mainly concentrated in developed economies. In contrast, the period from 2015 to 2024 saw a proliferation of clusters with increasingly complex network topologies and differential elements when comparing different groups. These features shed light on the phenomenon of entrepreneurial ecosystems’ spatial fluidity, a concept that is related to the growing embeddedness of regions in geographical connections that amplify their territorial scope and interconnectedness. Originality/value This study makes a novel theoretical contribution by integrating geospatial methods into entrepreneurial ecosystems literature and offers insights for scholars, investors and policymakers aiming to foster sustained development in ecosystems at different stages of development.