QUALIFICATION LEVEL, WORK EXPERIENCE, AND COMPENSATION IN IT: AN EVIDENCE-BASED ANALYSIS USING ML MODELS (2024)

Abstract

This paper presents a practical aid for managerial decisions in information-technology compensation and staffing: a salary-modelling calculator together with a quantitative ranking of ten determinants of pay derived from an empirical dataset and a supervised predictive approach whose coefficient of determination equals 0.65. The hierarchy of effects is transparent. Seniority contributes the most to salary variation (28.4 percent), followed by total years of experience (21.5 percent). Domain of work (12.8 percent) and primary programming language (9.8 percent) introduce consistent market premia that should be reflected in budgeting and portfolio planning. Specialisation across back-end, front-end, and quality assurance (6.9 percent) and the designated functional role (5.7 percent) further refine ranges within job families. Region (4.8 percent) and company size (4.7 percent) add moderate adjustments linked to local competition for talent and benefit structures. English language proficiency (4.4 percent) provides an incremental uplift through access to higher-value engagements, while formal higher education (1.1 percent) shows the weakest direct effect compared with practical skills and accumulated experience. Managerial implications are threefold. First, the evidence supports standardised grading with clear, competency-based progression rules rather than reliance on tenure alone. Second, realistic budgeting requires anticipating premia associated with domains and programming languages when planning statements of work and resource mixes. Third, leaders can use targeted levers-focused upskilling, development of English proficiency, balanced role composition, and flexible regional policies - to optimise cost and risk without eroding delivery quality. The calculator operationalises these insights: managers can estimate individual compensation corridors, compute the total cost of an illustrative team under alternative collaboration formats such as time-and-materials, fixed-price, and outstaffing, and explore the consequences of changing domain, technology stack, seniority level, or geography before commitments are made. For clients, the same tool offers a fast and reproducible way to evaluate the cost of the requested team against a specific technical brief and to compare cooperation formats with transparent assumptions.