This study aims to identify and evaluate the critical success factors (CSFs) shaping sustainable construction projects in Türkiye by distinguishing between perceived importance and actual causal influence within project systems.
An integrated AHP–DEMATEL framework was applied. Expert judgments were used to determine the relative importance of CSFs, while Decision-Making Trial and Evaluation Laboratory (DEMATEL) method was employed to analyze causal relationships and identify driving and dependent factors.
The results show that time–cost–quality performance, technical and contractor capability, and safety and risk management function as the main drivers of project performance. Sustainability and environmental sensitivity show a split causal profile: some environmental components act as causal drivers, while others, together with governance and learning-related dimensions, remain more dependent on upstream delivery, technical and managerial conditions.
The study is based on the judgments of eight experts and focuses on the Turkish construction context; therefore, the findings should be interpreted within these boundaries. Although consistency and threshold-sensitivity checks support the robustness of the results, future research could employ larger and more diverse expert panels, compare different national contexts, and apply fuzzy or interval-based AHP–DEMATEL approaches to examine uncertainty in expert judgments.
The findings suggest that improving sustainable construction performance requires focusing on driver dimensions while embedding sustainability, governance and learning into upstream project processes. AI-based decision-support systems, damage prediction models and structural health monitoring (SHM) tools may support more proactive project management.
By supporting more effective and resilient construction practices, the study contributes to safer built environments, particularly in disaster-prone regions.
The study advances construction management research by integrating prioritization and causal analysis and by showing that not all highly valued success factors function as system drivers.
