Digitalisation is often perceived as a driver of operational performance in manufacturing, but the mechanisms by which advanced digital skills influence productivity remain poorly understood. Digitalisation processes are heterogeneous in nature and are shaped by regional factors. The study aims to explore how workers’ digital human capital affects the performance of production systems in the metallurgy sector considering differences in regional digitalisation contexts. The research methods are based on multigroup analysis of partial least squares structural equation models (MGA PLS-SEM), in which the dependent variable is the performance of production systems. The research measured accumulated human capital as a stock of relevant digital basic and specific skills using a survey of 2 570 employees conducted in 2022 in Sverdlovsk, Chelyabinsk, Rostov, and Volgograd oblasts, which differ in their levels of digitalisation, innovation, industrial specialisation, and gross income. The findings indicate that advanced digital skills not only complement basic ones but also significantly enhance production performance, as the standardised path coefficients are ranging between 0.4 and 0.7. Specifically, the industrially advanced Chelyabinsk oblast shows a more significant impact of basic digital competencies on Industry 4.0 skills, though path coefficients are still less than 0.2, suggesting a moderate overall effect of Industry 4.0 skills on performance across all regions. This study contributes to the contextual economics perspective by demonstrating the heterogeneous nature of digital human capital accumulation within a single industry.