Expensive Mistakes: What Evidence Shows & What It Doesn’t

This article examines what the evidence actually shows about extremely expensive mistakes—from engineering failures to financial blunders—and, just as importantly, what it doesn’t show. You’ll learn the common patterns behind costly errors, the limits of post-incident analysis, and how to separate solid conclusions from speculation.

What the Evidence Shows About Costly Mistakes

Research across industries—from construction to software—reveals consistent patterns. Large projects are systematically over budget and behind schedule, with cost overruns of 50% or more being common. A well-known study of infrastructure projects found that 9 out of 10 had cost overruns. The primary causes are not random bad luck but predictable human and organizational factors.

Key evidence-backed findings include:

• Optimism bias: People tend to underestimate costs and timelines, a cognitive bias that persists even with historical data.
• Strategic misrepresentation: Sometimes project proponents deliberately understate costs to get approval, a form of deception that skews planning.
• Complexity and novelty: Projects that push technological or logistical boundaries have higher failure rates because there is less precedent to guide planning.
• Communication breakdowns: Many errors stem from miscommunication between teams, especially when handoffs are unclear.

For example, the construction of the Sydney Opera House was originally estimated at A$7 million and took 4 years; it ended up costing A$102 million and took 14 years. The evidence points to the novelty of the design and the lack of a detailed plan at the start. Similarly, in software, the Denver Airport baggage system fiasco—a highly automated system that failed upon launch—cost hundreds of millions in delays and ultimately was abandoned for a manual system. The evidence shows that the system was too complex and insufficiently tested.

What the Evidence Does Not Show

While we can identify patterns, evidence has limits. It does not show the exact probability of a specific mistake occurring in your project. Post-hoc analysis often suffers from hindsight bias—once we know the outcome, we see it as inevitable, but at the time many factors were uncertain.

Also, evidence from case studies is often anecdotal and not statistically robust. For every failed project, there may be a similar successful one that we don’t hear about. The evidence does not tell us the counterfactual: what would have happened if a different decision had been made? We can’t rerun history. So, while we know that overconfidence is a common factor, we cannot say that a particular decision caused a specific failure with certainty.

Moreover, evidence on “black swan” events—rare, high-impact mistakes—is inherently limited because they are rare. We cannot calculate their probability from past data alone. The 2010 Deepwater Horizon oil spill is a case in point: multiple safety systems failed in a chain, but the exact combination was unprecedented. Evidence from previous spills did not predict this event.

Common Patterns Behind Extremely Expensive Mistakes

Despite the limits, several patterns recur across industries:

• Inadequate risk assessment: Teams focus on the most likely risks but ignore low-probability, high-impact ones.
• Groupthink: In cohesive teams, dissenting opinions are suppressed, leading to poor decisions. The Challenger space shuttle disaster is a classic example where engineers’ concerns were overridden.
• Escalation of commitment: Once a course of action is chosen, people throw good money after bad rather than cut losses. The Concorde supersonic airliner is often cited: despite failing to become profitable, governments continued to fund it for prestige.
• Poor change management: Mid-project changes without proper impact analysis lead to rework and delays.

These patterns are not just theoretical; they are documented in project management literature and organizational psychology. For instance, a study of IT projects found that projects with frequent requirement changes were more likely to fail, but the study also noted that some changes are necessary in a dynamic environment.

How to Choose a Mitigation Strategy

Given the evidence, what can individuals and organizations do? The first step is to recognize that mistakes are inevitable, but their cost can be managed. Strategies include:

• Pre-mortem analysis: Imagine the project has failed and work backward to identify potential causes. This counteracts optimism bias.
• Redundancy and safety margins: Build in extra time and budget for the unknown. The evidence shows that projects that plan for uncertainty fare better.
• Incremental delivery: Break large projects into smaller chunks, each with its own testing and review. This limits the blast radius of a mistake.
• Independent oversight: Have a third party review plans and progress. This can catch strategic misrepresentation.

There is no one-size-fits-all solution. The choice depends on the project’s risk profile. For a low-risk, routine project, heavy oversight may be overkill. For a high-risk, innovative project, more robust mitigation is warranted.

Costs of Mistakes: Direct vs. Indirect

The direct cost of a mistake is the rework: the labor, materials, and time to fix the error. But indirect costs often dwarf direct costs. These include:

• Opportunity cost: The revenue or benefits lost while the project is delayed.
• Reputational damage: A public failure can lead to lost customers and difficulty attracting talent.
• Legal and regulatory penalties: Fines and lawsuits can be substantial.
• Morale and productivity: After a major mistake, teams may lose confidence, leading to further errors.

For example, the Boeing 737 MAX grounding after two fatal crashes cost the company over $20 billion in direct costs and lost sales, but the long-term reputational damage may be even more significant. The evidence shows that indirect costs can be several times the direct cost, but they are harder to measure and often not captured in post-incident reports.

Safety and Accuracy in Post-Mortem Analysis

When analyzing a mistake, it is crucial to be accurate and avoid scapegoating. The goal is to learn, not to blame. A common error is to attribute a failure to a single “root cause” when in reality it was a chain of events. The Swiss cheese model, where multiple layers of defense have holes that align, is a useful metaphor. Each hole is a latent condition, and the accident occurs when they align.

To conduct a fair analysis, gather all the facts, interview people at all levels, and avoid leading questions. Distinguish between human error and systemic factors. For example, if a worker misreads a gauge, is it because of poor training, or because the gauge is poorly designed? The evidence often points to the latter.

Also, be aware of the limits of memory. People may recall events inaccurately, especially under stress. Use written records and data logs whenever possible.

Alternatives to Traditional Risk Management

Traditional risk management focuses on identifying and mitigating known risks. But for extremely expensive mistakes, which are often unprecedented, a different approach is needed. Options include:

• Resilience engineering: Instead of trying to prevent all errors, build systems that can absorb and recover from them. This involves training people to adapt to unexpected situations.
• Scenario planning: Develop multiple plausible futures and test your strategy against each. This helps prepare for the unknown.
• Real options thinking: Structure projects so that you can abandon or expand them at low cost if conditions change. This is like buying an option rather than a commitment.

These alternatives are not mutually exclusive; they can be combined. The key is to acknowledge that uncertainty is inherent and that no amount of analysis can eliminate it.

Frequently Asked Questions

What is the most common cause of expensive mistakes?

Evidence points to optimism bias and inadequate planning. People tend to underestimate complexity and overestimate their ability to control outcomes. This leads to unrealistic schedules and budgets, which then get blown when problems arise.

Can expensive mistakes be prevented?

Not entirely, but their frequency and impact can be reduced. Prevention involves better planning, independent review, and building in slack. However, some mistakes are inevitable, especially in innovative projects where there is no precedent.

How do you calculate the true cost of a mistake?

True cost includes direct rework costs plus indirect costs like lost revenue, reputational damage, and opportunity costs. To get a full picture, you must also estimate the value of time lost and future impacts.

What is the difference between a mistake and a risk?

A risk is a potential future event that could cause harm, while a mistake is an action or decision that leads to an undesired outcome. Risks can be managed; mistakes are realized risks. Not all mistakes are the result of poor risk management; some are due to factors outside anyone’s control.

Is it better to be fast or accurate?

It depends on the context. In high-stakes environments, accuracy is more important than speed. However, being too slow can also be costly. The evidence suggests that a balanced approach—moving quickly but with checkpoints—works best. The key is to catch mistakes early when they are cheap to fix.

Why do companies repeat the same mistakes?

Often because they fail to learn from past failures. Post-mortems may be superficial, or the lessons are not institutionalized. Also, new teams may not know the history. Creating a culture that encourages open discussion of errors is crucial to breaking this cycle.

Related

• Golf Ball Hit Events Schedule: Full Guide
• Virtual Desktop Services: What Users Access and Providers Host
• Conan Exiles Dedicated Server Launcher: How to Use