Key performance indicators for optimizing academic performance and course design in online educational platforms

This study investigates how Key Performance Indicators (hereafter KPI) can enhance academic performance and inform course design on online educational platforms. As part of the AI4Ed project (Erasmus+), it bridges educational theory with computational analytics to develop data-driven, adaptive e-learning systems. This study employed a quantitative, correlational design to examine the relationship between students’ interaction patterns on a digital learning platform and their academic performance, and data from 63 postgraduate students were analyzed through Moodle interaction logs. High-frequency tasks, forum contributions, and resource downloads were evaluated as predictors of performance. The results highlight the predictive strength of active engagement and suggest practical strategies for individualized support, with activity peaks near task deadlines and gradual declines indicative of academic fatigue. The findings emphasize the importance of adaptive engagement strategies and demonstrate how KPI can support personalized learning and real-time course adjustments, enhancing online learning environments and fostering self-regulated learning. These findings have practical implications for educators and institutions aiming to enhance adaptive learning systems, optimize student engagement, and design data-informed interventions in online education. The study contributes to the growing field of learning analytics by proposing a KPI-driven framework grounded in cognitive and didactic theory.