A regression model of the impact of digital transformation on labour productivity in IT companies

Abstract

The purpose of the study is to develop and empirically verify a model for quantitatively assessing the impact of digital transformation on the labour productivity of IT company personnel. Methods. The study was conducted based on data from Ukrainian IT companies for 2022-2024. An expert method was used to collect primary information. Econometric modelling was performed using the least-squares method, and correlation and regression analysis were performed using the MS Excel data analysis package. The quality of the constructed model was assessed using indicators including the multiple correlation coefficient, the coefficient of determination, Fisher's F-test, and Student's t-statistic. Results. The key factors of digital transformation that have the greatest impact on labour productivity in the IT sector were identified, namely the business process automation index, the level of use of cloud services, the intensity of implementation of artificial intelligence tools, and the index of digital personnel qualification. Correlation analysis confirmed an extremely strong relationship between labour productivity and each of the studied factors (correlation coefficients exceed 0.98). The model has high statistical characteristics (R²=1); however, multicollinearity was detected, which indicates a synergistic effect of digitalization. The use of cloud services has the greatest sustainable impact: a 0.1 increase in this indicator increases productivity by 850 thousand UAH per person. Conclusions. It is shown that digital transformation is a key driver of productivity in the IT sphere, requiring a comprehensive approach that takes into account infrastructure-technical, organizational-management, and competence factors. The proposed regression model can serve as a tool to substantiate management decisions on priority areas for investment in digital technologies. Further research should focus on expanding the empirical base, applying alternative modelling methods, and examining nonlinear effects and the specifics of the impact of digital transformation on labour productivity across companies of varying sizes and specializations.