The article is devoted to the quantitative assessment of the impact of digital transformation and communication infrastructure on the productivity of small and medium-sized innovative enterprises in regions with varying levels of human capital saturation. The aim of the study is to identify the threshold conditions under which investments in digital infrastructure and ICT begin to generate a sustainable increase in gross value added and return on invested capital. The empirical base consists of panel microdata from the financial statements of 3,450 SMEs for 2018–2023 across 18 regions, supplemented by regional indicators of broadband access density and speed, penetration of digital services, and an index of human capital saturation constructed from employment structure, education level, and STEM graduate output. The methodology relies on a modified Cobb–Douglas production function expanded with variables of digital infrastructure, human capital, and their interaction, estimated on panel data using the GMM method, the construction of a digital maturity index for enterprises using the principal component method, DEA assessment of technical efficiency, and hierarchical cluster analysis that identified four types of regions (A–D) based on the combination of infrastructure quality and human capital. It is shown that enterprises in cluster A (high HC/developed IT infrastructure) demonstrate almost double the gross value added per employee and significantly higher ROIC and revenue elasticity with respect to IT expenditures compared with cluster D, whereas in cluster C (low HC/developed infrastructure) the accumulation of digital assets in the absence of sufficient competences leads to a decline in investment efficiency. The regression analysis of cost structure revealed a significant multiplicative effect of the interaction between broadband access speed and employee digital literacy on the reduction of administrative and commercial expenses, while the dynamics of working capital turnover indicators showed a reduction of the operating cycle by more than 20% and an increase in liquidity after the implementation of digital platforms. It is established that the impact of communication infrastructure on productivity exhibits a pronounced threshold nature and becomes statistically significant only upon reaching a certain level of the human capital index; the practical significance of the results lies in the possibility of using the obtained threshold values and cluster typology to adjust differentiated regional policy measures and revise corporate models for evaluating digital projects, with mandatory consideration of training and change management costs.