How Big Things Get Done by Bent Flyvbjerg and Dan Gardner: Study & Analysis Guide

Megaprojects—from high-speed rail lines to Olympic Games—fail at a staggering and predictable rate, routinely delivering less value for far more money and time than promised. In How Big Things Get Done, Bent Flyvbjerg and Dan Gardner dissect this global phenomenon using the world’s largest dataset of big projects, revealing that failure is not bad luck but the result of systemic planning fallacies. Their analysis provides a powerful, evidence-based framework for any leader, manager, or individual undertaking a significant endeavor, offering tools to transform ambitious visions into reliable successes.

The Iron Law of Megaprojects: Why Big Things Go Wrong

Flyvbjerg’s core thesis, drawn from analyzing thousands of projects, is that megaprojects—defined as large-scale, complex ventures costing billions and impacting many people—consistently fall victim to what he terms the "Iron Law." This law states that over 90% of megaprojects exceed their budget, schedule, or both, with cost overruns averaging 45% and schedule delays averaging 40%. The primary culprit is not technical complexity but human psychology and political pressures. Planners and promoters suffer from optimism bias, believing their project will defy the odds, and strategic misrepresentation, where they deliberately underestimate costs and overstate benefits to gain approval—a behavior Flyvbjerg calls "lying." This creates a "survival of the unfittest" scenario where the projects that look best on paper are often the least likely to succeed in reality.

Think Slow, Act Fast: The Core Mindset for Success

To counter these destructive tendencies, Flyvbjerg advocates for a fundamental shift in planning philosophy: "Think slow, act fast." This paradoxical mantra is the book’s central strategic framework. "Thinking slow" means investing disproportionate time, energy, and resources in the front-end planning and design phase. This involves thorough exploration, prototyping, scenario planning, and de-risking before a single dollar is committed to execution. The goal is to create a detailed, modular, and foolproof plan. "Acting fast" then refers to the execution phase, which should be as swift and decisive as possible once planning is complete. Prolonged execution exposes projects to inflation, shifting political winds, and unforeseen disruptions. By decoupling these two phases and prioritizing deep, deliberate planning, you compress the vulnerable period of implementation and dramatically increase the odds of on-time, on-budget delivery.

The Antidote to Guesswork: Reference Class Forecasting

The most powerful practical tool Flyvbjerg introduces to enable "slow thinking" is reference class forecasting (RCF). This is a decision-making technique that combats optimism bias by using historical data from a reference class—a group of similar past projects—to predict future outcomes. Instead of basing estimates on the unique details of your specific project (which invites optimistic assumptions), you ask: "How did projects like this one typically turn out?"

The process involves three steps:

1. Identify a Reference Class: Find a statistically significant group of comparable past projects (e.g., urban light-rail systems, major software platform rollouts, corporate mergers).
2. Establish a Probability Distribution: Determine the range of outcomes for those projects, especially regarding cost overruns and schedule delays.
3. Predict Your Project's Position: Statistically place your specific project within that distribution, adjusting for any verifiable unique aspects.

For example, if the reference class shows that 80% of similar software projects overran budgets by 50-100%, you would base your forecast and contingency plans on that distribution, not on an idealistic bottom-up estimate. RCF forces realism into planning by grounding predictions in empirical reality rather than wishful thinking.

Applying the Framework Beyond Infrastructure

While the book’s data is rooted in physical megaprojects, its insights are universally applicable. The principles of "think slow, act fast" and reference class forecasting are crucial for success in other domains.

For Software Projects: The tech industry is notorious for agile development that sometimes privileges "acting fast" over "thinking slow." Flyvbjerg’s framework suggests a hybrid approach: spend extensive time on system architecture, selecting the right technology stack, and building detailed prototypes or MVPs (Minimum Viable Products) to test core assumptions. Execution (coding, integration) can then proceed rapidly in sprints against a stable, well-understood plan. RCF can be used by looking at the historical performance of similar in-house development projects or vendor implementations to set realistic timelines and budgets.

For Organizational Change: Major initiatives like digital transformations, mergers, or cultural shifts are classic "big things" that often fail. "Thinking slow" means conducting exhaustive readiness assessments, piloting changes in one department, and securing genuine buy-in before a full-scale rollout. "Acting fast" means implementing the change decisively across the organization once the plan is validated to avoid change fatigue. Leaders can use RCF by studying the success rates and pitfalls of similar change programs within their industry.

For Personal Endeavors: Whether writing a book, launching a startup, or planning a major life event, the same rules apply. "Thinking slow" involves meticulous research, skill-building, and creating a detailed, phased plan. "Acting fast" is about focused, relentless execution against that plan. You can practice a form of RCF by seeking out case studies and biographies of people who have achieved similar goals to understand common hurdles and realistic timeframes.

Critical Perspectives

While Flyvbjerg’s framework is robust, applying it requires critical adaptation. A key limitation is the availability of data for reference class forecasting. For truly novel projects (e.g., the first commercial quantum computer or a unique organizational structure), a clear reference class may not exist, requiring more analogical reasoning. Furthermore, the "act fast" imperative must be balanced with the need for adaptability. A rigid plan executed quickly can be disastrous if based on flawed assumptions. The solution lies in the "slow thinking" phase: building modularity and feedback loops into the plan itself so that execution can be swift and responsive to valid new information.

Some critics also argue that the political dynamics Flyvbjerg describes are intractable; the incentives for strategic misrepresentation are so powerful that better forecasting tools alone cannot overcome them. This points to the necessity of complementary reforms in governance, accountability, and funding structures to align political and project success criteria.

Summary

• Megaprojects fail predictably due to optimism bias and strategic misrepresentation, not technical misfortune. Recognizing these systemic forces is the first step toward better outcomes.
• Adopt the "think slow, act fast" mindset. Invest heavily in exhaustive planning, prototyping, and de-risking upfront to enable rapid, decisive, and less vulnerable execution.
• Use reference class forecasting to combat optimism. Base your forecasts on the statistical outcomes of similar past projects, not on the unique details and hopeful assumptions of your current plan.
• The framework extends beyond construction to software, organizational change, and personal goals. The core discipline of empirical, deliberate planning is universally valuable for any significant undertaking.
• Successful application requires critical adaptation, especially regarding data availability for novel projects and maintaining necessary flexibility during the execution phase.