This thesis addresses the persistent challenge of cost estimation inaccuracies in the Dutch road construction industry, with particular emphasis on learning from previous projects to improve forecasting practices. Despite substantial progress in estimation techniques, cost overruns continue to undermine budget reliability and efficiency, frequently resulting in project delays and financial inefficiencies. To better understand and address this persistent issue, this research explores the following question: How can learning from previous projects be strengthened to improve the accuracy of cost estimation during the tendering phase? The study investigates the underlying causes of these inaccuracies and proposes targeted strategies to enhance organizational learning and subsequently improve the accuracy of cost estimation during the tendering phase.
The research identifies a substantial gap in systematically leveraging past project experiences. Even though practitioners recognize the value of reflecting on past projects, organizational conditions often prevent meaningful learning. Key barriers identified include time constraints, limited formal structures for knowledge capture, and cultural resistance characterized by blame avoidance and siloed information. These factors frequently result in repetitive estimation errors, regarding, for example, staffing costs, where analysis reveals an average deviation of 45% from initial budget estimates.
Methodologically, the study employs a mixed-method approach. A quantitative analysis was conducted on data from 18 recently completed road projects within a major Dutch contractor, highlighting significant deviations in staff cost estimations. Qualitative insights were gathered through 23 semi-structured interviews with professionals at the contractor involved in estimating, executing, and controlling road infrastructure projects. These interviews provided crucial context, uncovering practical and organizational reasons behind the observed deviations.
To explain where and how learning breaks down, the study builds on established theory in organizational learning. It introduces an integrated multi-level learning framework, drawing on Crossan et al.’s (1999) 4I model and its expansions by Jenkin (2013) and Wodnik et al. (2024). In this framework, projects are viewed as temporary organizations embedded within a broader coordinating structure. Learning is conceptualized as a progression through key processes: intuiting, interpreting, integrating, and institutionalizing. Two additional processes are included to reflect recent theoretical developments: interaction and incorporation.
The practical recommendations derived from this study are structured using the Plan–Do–Check–Act (PDCA) cycle, a widely used continuous improvement framework that supports iterative and systematic implementation of change. Key proposals include institutionalizing project evaluations by scheduling them at project initiation, creating structural ”learning slack” by allocating dedicated reflection time, simplifying data systems to enhance usability, and promoting stronger interaction between estimators and execution teams. These initiatives aim to foster an organizational environment where reflective practices become routine rather than exceptional. Emphasizing cross-regional knowledge exchange within the organization and establishing clearer accountability structures for capturing and applying project insights further complements the recommended strategies. In essence, this thesis finds that improving cost estimation accuracy hinges on institutionalizing a culture of structured reflection and learning within the organization.
Ultimately, the study underscores that improving cost estimation accuracy requires more than technical enhancements. Organizational commitment to structured reflection, effective feedback loops, accessible data systems, and cultural incentives for continuous improvement is essential. Prioritizing structured and systematic learning from previous projects offers a robust pathway toward achieving more reliable and accurate cost estimations, contributing significantly to the efficiency and sustainability of infrastructure development.