Turnover Management Method Based in Predictive Analysis

Context: IT companies struggle with the high statistics of job rotation every year. The high employee turnover in Brazilian software development companies hampers the execution of long-term strategic planning, as well as hinders new business opportunities and software projects due to the loss of intellectual capital. Problem: This article aimed to investigate the turnover data of an IT company in Brazil, and develop a data-driven method to predict voluntary turnover, to opportunize the company to act to minimize it. Solution: We propose a turnover management process, supported by a predictive model using artificial intelligence (AI). IS Theory: Contingency Theory, TAM(Technology Acceptance Model), Knowledge Management, and Organizational Justice are some of the most relevant theories in the relationship between Information Systems (IS) and turnover. Propper management policies can improve communication, engagement, and organizational justice — factors that have a direct impact on employee retention, as showed in this study. Method: The research is prescriptive in nature, and its evaluation was conducted through a case study. The results were analyzed using a qualitative approach. Summary of Results: the feasibility of applying the proposed process has been confirmed according to each stage of the turnover management process and the results obtained. Contributions and Impact on the IS Field: The primary contribution of this work is to assist companies in developing strategies for effective turnover management to minimize it, using a process based on a predictive model supported by AI.