Enhancing student retention through predictive analytics and outreach

A case study in early intervention in online postgraduate studies

Abstract

This paper presents a case study of a predictive analytics pilot aimed at improving student retention in postgraduate online learning. By leveraging Learning Management System (LMS) engagement metrics, students who had not accessed or submitted their first assessment at least five days before the due date, were flagged for early outreach. A tiered human-led interventions model via email, phone, and SMS were triggered to provide personalised support. The pilot was implemented across 12 postgraduate subjects in 2025 and benchmarked against 2024 cohorts. A 1.9 percentage point improvement in dropout rate was observed, equating to approximately 32 additional students retained. These findings highlight the practical value of low-complexity, data-informed interventions and the role of AI-human synergy in supporting online learners. This pilot demonstrates a scalable approach to integrating predictive analytics with personalised support in fostering adaptable learners and improving student outcomes, particularly in resource constrained settings.