Purpose – This study investigates the effectiveness of the Deep Digital Learning (DDL) model to address the failure of conventional digital learning to foster Higher-Order Thinking Skills (HOTS). Unlike prior fragmented approaches, the model proposes a novel conceptual synthesis of personalization, collaboration, authentic problem-based learning, and data-driven feedback to enhance critical thinking and problem-solving.Methodology – This quasi-experimental design employed 70 students from the Educational Technology study program. The experimental group (n = 35) used the DDL intervention via the SIDIA Learning Management System (LMS). In contrast, the control group (n = 35) used Conventional Digital Learning (CDL) as a non-equivalent control for seven weeks. Data were collected using validated rubrics for Critical Thinking (CT) and Problem-Solving (PS) skills tests, which were analyzed using Analysis of Variance (ANOVA).Findings – The ANOVA results statistically showed that the DDL group achieved significantly higher post-test scores for both Critical Thinking skills (F(1, 68) = 169.30, p < 0.001) and Problem-Solving skills (F(1, 68) = 140.65, p < 0.001). The mean difference confirmed the superiority of the experimental class in both skills (3.35 points for CT and 3.37 points for PS). This confirms DDL is more effective than CDL in enhancing students’ HOTS.Contribution – Beyond statistical significance, this study positions DDL as a strategic instructional blueprint in advancing HOTS. It provides Higher Education with a proven framework to strengthen digital transformation, ensuring the achievement of Outcome-Based Education (OBE) and 21st-century skills.