Applying HR Analytics to Employee Turnover: Evidence from the IT Sector in Guadalajara, Mexico

Abstract

This research examines the problem of employee turnover among the information technology (IT) sector of Guadalajara through IBM Human Resource Analytics Employee Attrition and Performance variables in a CRISP-DM model. The purpose is to determine the extent to which the 35 IBM variables can be applied to the Guadalajara city of Mexico and to come up with an empirical foundation for turnover prediction using AI in the future. The quantitative design was used on the sample of 1,154 employee records of two IT companies; these firms were selected because they operate in similar market segments which merged survey answers with the IBM-like fields of HR. The employees are fairly balanced with age (35.7% 25-35 years, 32.1% 35-45, 32.1% over 45) and well compensated, with a salary of 25 hours of overtime work monthly, 37.6 hours of annual training, and 3.2 years of service with the current manager. The 67.9% are not intending to change this year, 75% report job satisfaction, and 85.4% co-worker relations as very good or excellent. The correlation analysis reveals that there decrease in work-life balance with long working hours (r = -0.55), promotion delay is correlated to inferior performance (r = -0.45), and experience correlates with higher salary and tenure (r = 0.55 and r = 0.79). According to the findings, pay is important, with promotion opportunities, workload, and organisational climate being critical levers. The IBM variables and advises Guadalajara IT companies to include these measures in the data-driven retention and future AI-assisted attrition models.