An AI-driven approach to improve talent retention and mobility

Abstract

This research is an effort to develop a holistic AI-driven approach in talent management, which uses AI to discover hidden information in the data leading to improve talent mobility and retention within organizations. In addition to extract insights and values from data, the approach is designed to mitigate biases involved in decision-making process in talent mobility and retention. Therefore, required procedures and mathematical considerations have been taken into account to reduce risk in all steps from collecting data to implement ML methods. The results show a robust performance for ML methods which receive data including 31 features as input and predict overall employees satisfactions, their intention to be mobilized internally, and retention risk as outputs.

The presented theoretical framework in the thesis is the backbone of the implemented AI flow and it makes sure that the approach is holistic, meaning that it takes into account all the key factors which form the employee’s experience and does not narrow down to a specific factor. It takes different societal and psychological theories to extract key factors forming employees' experience in the workplace; trust, relevance and empowerment, work culture, compensation and rewards, performance, competencies and growth. Then the data collected through 31 statements designed based on key factors are fed into an ML pipeline testing various supervised ML methods. Finally, the best model is selected based on the performance. For all cases the best performance was for Random Forest which has shown a superior performance for predicting the overall satisfaction, intention to mobility and retention risk. The data is collected through a completely anonymous survey on LinkedIn platform to mitigate geographical and sector biases. The collected data have been stored and processed responsibly to be in line with EU ethical practices like GDPR and responsible AI development.

The best model out of the AI flow has been used to find out the importance of the inputs. Although trust is the most influential on satisfaction, intention to mobility and retention risk, other key factors such as empowerment, work culture and fair compensation have remarkable effects on forming the overall talent experience and affecting the intention to mobility and retention. Moreover, the analysis uncovers that beyond simply identifying who is satisfied or dissatisfied, AI can detect more nuanced subgroups, such as those who are satisfied yet open to new roles. Leveraging minimal personal data allows organizations to tailor engagement efforts accordingly, though safeguards must be in place to prevent misuse of these insights.

Finally, by adopting a comprehensive, AI-driven perspective, organizations can enhance talent management decisions and reduce biases. This holistic approach not only helps lower turnover but also promotes a work environment where employees feel motivated to stay, develop, and contribute.