Federica Belfanti

Purpose: This study examines whether and how SMEs' post-adoption use of generative AI (GenAI) relates to firms' market-facing digital presence. Moving beyond binary adoption, it asks how three usage patterns, i.e. duration (continuity over time), volume (cumulative interaction) and frequency (regularity of use), are associated with externally observable signals of online visibility and credibility, including website content depth, social media audience/engagement, web traffic, and search visibility. Design/methodology/approach: We integrate three firm-level data sources for 175 Italian SMEs: (1) GenAI application usage logs (Dec 2023–Jul 2024) to construct duration, volume and frequency measures, (2) digital trace indicators from an SEO/competitive intelligence platform (Dec 2024) to capture five presence dimensions and build a composite index via PCA and (3) administrative controls from AIDA. We estimate OLS models with heteroskedasticity-robust standard errors and include industry and region fixed effects (final models: 138 firms with complete data). Findings: GenAI usage frequency is positively and significantly associated with the PCA-based digital presence index, while usage duration and cumulative volume are not. When duration, volume and frequency are entered jointly, none is statistically significant, consistent with overlap among post-adoption usage measures. The results indicate that routinised, regular GenAI use, rather than time since adoption or total interaction intensity, is the usage dimension most closely linked to stronger market-facing digital presence. Originality/value: The paper provides post-adoption evidence on GenAI in SMEs by linking behavioural usage logs to later, externally visible digital outcomes. It operationalises GenAI engagement through distinct continuity, intensity and regularity measures and connects them to a multidimensional digital presence construct derived from web and social media traces. By combining platform logs, SEO intelligence data and administrative firm records, the study offers a replicable approach for explaining heterogeneous outcomes after GenAI adoption and highlights the strategic relevance of routinisation for SME competitiveness.

Purpose: This paper aims to develop a scalable, patent-based framework to map clusters' repositioning towards environmental technologies and to distinguish current green technological specialisation from transition readiness. Design/methodology/approach: The framework is applied to Italy by combining PATSTAT patent data with the Italian Cluster Mapping Project. Green technologies are identified through CPC-Y tagging. The authors compute cluster-level Revealed Technological Advantage (RTA) in environmental technologies (2000-2019) and estimate Potential RTA (pRTA) as a forward-looking indicator based on technology relatedness networks at CPC subclass level. The authors compare alternative predictive strategies (zero-inflated beta regression, Artificial Neural Networks, Random Forests) and retain the best-performing model to generate pRTA for 2020-2024. Finally, the authors classify clusters into a four-quadrant typology combining RTA and pRTA. Findings: Green inventive activity is geographically concentrated in a small set of regions, while it is more dispersed across cluster categories. Current green specialisation (RTA) varies substantially across region-cluster combinations and does not fully overlap with transition readiness (pRTA). The combined mapping reveals four profiles: green pioneers, emerging green clusters, mature green and green laggards, enabling a trajectory-oriented interpretation beyond static rankings. Originality/value: The study offers one of the first cluster-level, forward-looking measures of environmental-technology transition readiness in Italy, combining cluster mapping with relatedness-based prediction to support more differentiated research and place-based policy design.

Entrepreneurial ecosystems are powerful engines of regional innovation, competitiveness, and transformation. While established hubs like Silicon Valley and Shenzhen are well studied, identifying emerging ecosystems with high growth potential remains a challenge. This study introduces the Surge Ecosystems Index, a forward-looking tool for detecting surging entrepreneurial activity at the sub-regional level. The index evaluates six dimensions - Innovation Capacity, Capital Fluidity, Skills Development, Market Reach, Entrepreneurial Culture, and Visibility - over a five-year period. Using Italian provinces (NUTS-3) as the empirical context, we integrate official statistics and web-scraped data to construct and validate the index. Results reveal geographical heterogeneity, with several provinces showing strong upward momentum despite not being traditional entrepreneurial hotspots. The index correlates positively with high-growth firm density, confirming its potential as an early-stage detection tool for policymakers and investors. This study advances entrepreneurial ecosystem theory by integrating temporal and micro-foundational perspectives, shifting from static assessments to dynamic, predictive analysis. The Surge Ecosystems Index offers a novel approach to identifying and supporting emerging entrepreneurial ecosystems, with implications for both research and policy. Future work should explore cross-contextual validation and refinement for broader applicability.