Bent Flyvbjerg spent decades studying why ambitious projects disappoint, while journalist Dan Gardner turns that research into a readable operating manual. Their database covers more than 16,000 projects, and its headline result is brutal: only 8.5 percent were delivered on both budget and schedule, while roughly 0.5 percent also produced the promised benefits. The book argues that failure begins with the commitment fallacy. Leaders fall in love with an idea, announce it too early, and then use planning to justify a decision already made. Optimism bias understates difficulty, strategic misrepresentation keeps approval numbers attractive, and every month of slow execution creates more opportunities for inflation, political change, accidents, and small problems to combine. The alternative is to think slow and act fast. Start by defining the project's real purpose, study comparable outcomes through reference class forecasting, and reduce unknowns through experience and experiments. Pixar storyboards films repeatedly before expensive production, while Frank Gehry's teams use models and simulations to expose design problems early. Once execution starts, move quickly and build with modular, repeatable units whenever possible. A solar farm can add standardized panels one by one, while a first-of-a-kind nuclear plant concentrates too much novelty in one enormous bet.

What it gets right

• Replaces confident internal estimates with base rates from comparable completed projects
• Shows why early testing is not hesitation but the fastest route to reliable execution
• Connects modularity with learning curves, repetition, lower risk, and scalable delivery

What it overstates or misses

• The broad project database combines very different sectors, governance systems, and definitions of success
• Modularity is powerful but cannot fully solve projects dominated by site conditions, politics, or genuinely novel science
• The book gives less attention to power, procurement incentives, and organizations that benefit from optimistic forecasts

The evidence is unusually substantial for a management book, although the memorable case studies sometimes carry more explanatory weight than the statistics can support on their own. Reference class forecasting and modularity are the durable ideas; some of the surrounding project stories mainly make them easier to remember. The verdict: required reading before anyone approves a budget with more than one comma in it.