The Rationality Illusion
CEOs are supposed to be supremely rational decision-makers. They're paid millions to exercise judgment that creates billions in value. They have access to the best information, the smartest advisors, and the most sophisticated analytical tools.
Yet CEOs make catastrophically wrong decisions with startling regularity.
The evidence is clear:
- 70-90% of acquisitions fail to create value, often destroying it
- Two-thirds of major strategic initiatives underperform expectations
- CEO overconfidence correlates with worse acquisition returns
- Experienced CEOs are often more susceptible to bias, not less
The problem isn't intelligence. The problem isn't information. The problem is cognitive bias—systematic patterns in how human brains process information that lead to predictably irrational decisions.
CEOs are not immune. In fact, the characteristics that make great CEOs—confidence, decisiveness, pattern recognition from experience—often amplify cognitive biases rather than reducing them.
The 12 Biases That Derail CEOs
Bias 1: Overconfidence
The most pervasive and dangerous bias affecting CEOs.
How it manifests:
- Overestimating the probability of success for initiatives you champion
- Underestimating time, cost, and complexity of projects
- Believing you can beat market returns through M&A
- Assuming you know more than you actually do
The research:
Studies consistently show CEOs are more overconfident than the general population—and more overconfident CEOs make worse decisions. Malmendier and Tate's landmark research found overconfident CEOs overpay for acquisitions and undertake value-destroying mergers.
Why CEOs are vulnerable:
Success breeds overconfidence. CEOs who've made good decisions believe they'll continue making good decisions. Past success is treated as evidence of skill rather than luck. The result: overconfidence compounds over a successful career.
The antidote:
Pre-mortem analysis (imagining failure before decisions), base rate consideration (looking at how similar decisions typically turn out), and structured devil's advocacy.
Bias 2: Confirmation Bias
Seeking information that confirms existing beliefs while ignoring contradictory evidence.
How it manifests:
- Soliciting opinions from people who agree with you
- Interpreting ambiguous data as supporting your position
- Dismissing contrary evidence as exceptions or methodological flaws
- Remembering information that confirms your views more easily
Why CEOs are vulnerable:
Power insulates from disagreement. People tell CEOs what they want to hear. Information flows are filtered to match CEO expectations. Contrary voices self-censor or are selected out.
The organizational effect:
Confirmation bias spreads through organizations. When CEOs signal their preferred conclusions, organizations produce supporting evidence. The CEO believes they're making data-driven decisions; they're actually making decisions that the organization learns to support with data.
The antidote:
Actively seeking disconfirming evidence, empowering devil's advocates, creating safe channels for contrary information, and regularly consulting skeptics rather than supporters.
Bias 3: Anchoring
Over-relying on the first piece of information encountered.
How it manifests:
- Negotiation positions anchored by the first number mentioned
- Strategic plans anchored by last year's plan
- Valuations anchored by initial estimates regardless of new information
- Performance expectations anchored by historical performance
Why CEOs are vulnerable:
CEOs receive information in particular sequences. First impressions carry disproportionate weight. Initial framings—by investment bankers, consultants, or internal advocates—anchor subsequent analysis.
The strategic effect:
Anchoring affects strategic planning profoundly. "Last year plus 10%" becomes the default rather than genuine reassessment. Acquisition targets are valued relative to asking prices rather than intrinsic worth. Budget negotiations are anchored by initial proposals.
The antidote:
Multiple independent estimates, explicit consideration of anchor effects, and deliberately generating alternative anchor points before negotiation or analysis.
Bias 4: Sunk Cost Fallacy
Continuing investments because of past commitments rather than future prospects.
How it manifests:
- Persisting with failing strategies because "we've invested so much"
- Continuing unsuccessful acquisitions to justify the purchase price
- Maintaining underperforming executives because of development investment
- Refusing to exit markets where sunk costs are significant
Why CEOs are vulnerable:
CEOs are personally identified with decisions. Admitting failure feels like personal indictment. Past decisions were often championed publicly—reversing them requires public acknowledgment of error.
The career effect:
CEOs often double down on failing strategies rather than admit mistakes. The psychological pain of loss acknowledgment exceeds the economic pain of continued investment. Result: throwing good money after bad.
The antidote:
Decision journals that separate past decisions from future analysis, "kill committees" specifically tasked with recommending discontinuation, and pre-commitment to decision review triggers.
Bias 5: Availability Bias
Overweighting information that comes to mind easily.
How it manifests:
- Recent events affecting judgment more than they should
- Vivid examples carrying more weight than statistical evidence
- Easy-to-recall successes informing strategy more than typical outcomes
- Dramatic failures creating excessive risk aversion
Why CEOs are vulnerable:
CEOs are bombarded with vivid, memorable information. The deal that went spectacularly well. The competitor that failed dramatically. The crisis that required personal intervention. These memorable experiences crowd out statistical reasoning about typical outcomes.
The strategy effect:
Strategic decisions get distorted by memorable examples. A competitor's successful acquisition leads to imitation regardless of whether circumstances transfer. A memorable failure in a market leads to permanent avoidance. Vivid beats probabilistic.
The antidote:
Systematic data collection that supplements memorable experiences, base rate consideration for all major decisions, and deliberate consultation of less vivid but more representative evidence.
Bias 6: Status Quo Bias
Preferring current state over change, even when change is beneficial.
How it manifests:
- Requiring stronger evidence for action than for inaction
- Framing decisions as "whether to change" rather than "what's optimal"
- Underweighting opportunity costs of inaction
- Treating existing strategies as defaults that require justification to change
Why CEOs are vulnerable:
Current arrangements have constituencies that defend them. Change creates identifiable losers. Inaction losses are diffuse and invisible while action losses are concentrated and obvious.
The strategic effect:
Status quo bias leads to strategic stagnation. Necessary changes are delayed because the bar for action exceeds the bar for inaction. Gradually declining businesses continue because change feels riskier than staying the course.
The antidote:
Framing decisions as choices between alternatives rather than change-versus-status-quo, explicitly calculating opportunity costs of inaction, and regular zero-based strategic reviews.
Bias 7: Halo Effect
Allowing one positive attribute to influence judgment of unrelated attributes.
How it manifests:
- Assuming successful executives in one role will succeed in others
- Believing companies with strong share prices are well-managed overall
- Judging candidates' competence based on confidence or appearance
- Attributing multiple positive qualities to people we like
Why CEOs are vulnerable:
CEOs evaluate people constantly—executives, board candidates, potential hires. Strong first impressions create halos that color all subsequent evaluation. A confident presentation creates assumption of underlying competence.
The hiring effect:
The halo effect distorts executive hiring and promotion. Candidates who interview well are assumed to be competent. Executives with one visible success are assumed to have broad capabilities. Result: promotion and hiring decisions based on surface signals rather than capability.
The antidote:
Structured evaluation separating different competencies, multiple independent assessors, and evidence requirements for each evaluation dimension rather than global impressions.
Bias 8: Attribution Error
Attributing others' outcomes to character while attributing own outcomes to circumstances.
How it manifests:
- Assuming competitor failures reflect incompetence rather than bad luck
- Believing personal successes reflect skill while failures reflect circumstances
- Judging executives' past performance without adequate situational context
- Overconfidence about ability to succeed where others failed
Why CEOs are vulnerable:
CEOs have detailed knowledge of their own circumstances but only surface knowledge of competitors' situations. Their own struggles are understood as externally caused; others' struggles are assumed to reflect capability.
The competitive effect:
Attribution error leads to competitive arrogance. Competitors who fail are assumed to be incompetent, leading to overconfidence about ability to succeed in same markets. Executives who struggled in previous roles are assumed to lack capability, regardless of circumstances they faced.
The antidote:
Systematic analysis of situational factors for all outcome evaluation, explicit consideration of alternative explanations, and humility about how circumstances contributed to personal success.
Bias 9: Groupthink
Collective rationalization that suppresses dissent and critical evaluation.
How it manifests:
- Executive teams reaching consensus too quickly
- Absence of meaningful debate on important decisions
- Self-censorship by team members who disagree
- Collective overconfidence in team decisions
Why CEOs are vulnerable:
CEOs set the tone. When CEOs signal preferred directions, teams often converge on those directions rather than genuinely debating. The stronger the CEO's reputation, the more team members defer rather than challenge.
The decision effect:
Groupthink produces decisions that feel unanimous but reflect suppressed disagreement. Critical perspectives don't surface. Risk assessments are optimistic because no one wants to seem negative. The team feels confident in decisions that no individual would make alone.
The antidote:
Structured dissent processes, CEO speaking last in discussions, anonymous input mechanisms, and explicit assignment of devil's advocate roles.
Bias 10: Escalation of Commitment
Increasing investment in failing courses of action.
How it manifests:
- Continued investment in struggling acquisitions
- Throwing resources at failing strategies rather than pivoting
- Refusing to exit positions despite mounting evidence of failure
- Interpreting negative signals as reasons to try harder rather than stop
Why CEOs are vulnerable:
CEOs are publicly committed to decisions. Reversal requires admitting error. The psychological and reputational costs of abandoning commitments often exceed the economic costs of continuing.
The organizational effect:
Escalation compounds through organizations. Once the CEO is committed, the organization mobilizes to execute. Negative information is filtered. Success is redefined. The failing strategy becomes the strategy the organization is organized to support.
The antidote:
Pre-commitment to review triggers, independent evaluation of ongoing initiatives, and separation of decision-making from decision-justifying roles.
Bias 11: Recency Bias
Overweighting recent events in forming judgments and predictions.
How it manifests:
- Recent quarters affecting strategy more than they should
- Current market conditions assumed to persist indefinitely
- Recent employee performance weighted too heavily in evaluations
- Last interaction affecting relationship assessment disproportionately
Why CEOs are vulnerable:
CEOs are evaluated on recent performance. Recent information is most vivid. Current conditions are most salient. The result: judgment distorted toward whatever happened recently.
The cyclical effect:
Recency bias amplifies business cycles. Good recent performance leads to aggressive expansion. Poor recent performance leads to excessive contraction. Decisions appropriate for current conditions prove wrong as conditions change.
The antidote:
Systematic historical analysis, explicit consideration of how current conditions compare to historical ranges, and deliberate weighting of longer time horizons.
Bias 12: Self-Serving Bias
Attributing successes to personal capability while attributing failures to external factors.
How it manifests:
- Taking credit for favorable outcomes regardless of luck's role
- Blaming external factors for unfavorable outcomes
- Believing personal involvement improves project success rates
- Overestimating personal contribution to team successes
Why CEOs are vulnerable:
CEO incentives reinforce self-serving attribution. Compensation depends on taking credit. Reputation depends on being associated with success. The self-serving narrative is professionally advantageous.
The learning effect:
Self-serving bias prevents learning from failure. If failures are always externally caused, there's nothing to learn. If successes are always personally caused, overconfidence compounds. The result: CEOs who don't improve from experience because they attribute all outcomes to factors that require no personal change.
The antidote:
Post-decision reviews that explicitly analyze luck versus skill, external perspectives on personal contribution, and deliberate effort to learn from failures attributed to external causes.
Why Experience Doesn't Solve the Problem
Conventional wisdom suggests experienced CEOs should be less biased. They've seen more, learned more, developed judgment through practice.
The research suggests otherwise:
Pattern matching: Experienced CEOs develop pattern recognition that can become pattern imposition—seeing patterns where they don't exist.
Success attribution: Experienced, successful CEOs attribute past success to skill, reinforcing overconfidence.
Confirmation environments: Powerful, experienced CEOs create environments where their views are rarely challenged.
Cognitive entrenchment: Long experience can create rigid mental models that resist updating.
The paradox:
The experience that creates CEO capability also amplifies the biases that distort CEO judgment. The confidence that enables bold action also enables overconfident action. The pattern recognition that accelerates decision-making also creates pattern imposition.
What Can Be Done
Individual Strategies
CEOs can develop personal debiasing practices:
Decision journals: Documenting reasoning before decisions enables accurate review of judgment quality.
Pre-mortems: Imagining future failure before decisions helps identify overlooked risks.
Devil's advocate assignment: Deliberately seeking contrary perspectives counters confirmation bias.
Base rate research: Investigating how similar decisions typically turn out combats overconfidence.
Decision pause protocols: Building in reflection time counters hasty, intuitive decisions.
Organizational Strategies
Organizations can create structures that counter CEO bias:
Independent review: Strategy and M&A decisions reviewed by parties without stakes in outcomes.
Anonymous input channels: Mechanisms for information to reach the CEO without filtering.
Structured debate: Formal processes that require genuine consideration of alternatives.
Kill committees: Groups specifically tasked with recommending discontinuation of failing initiatives.
External perspectives: Regular input from outsiders not subject to organizational conformity pressure.
Board Strategies
Boards play critical roles in countering CEO bias—a topic explored in depth in a companion article.
The Bottom Line
CEO cognitive bias is not a character flaw—it's a feature of human cognition. The same mental shortcuts that enable quick, effective decision-making also create systematic errors.
The cognitive reality:
- Biases affect all CEOs, regardless of intelligence or experience
- Success often amplifies biases rather than reducing them
- Organizational dynamics typically reinforce rather than counter biases
- Debiasing requires deliberate structures and practices
What CEOs should do:
Accept vulnerability: Acknowledge that you're subject to the same cognitive biases as everyone else. Intelligence doesn't immunize.
Build countermeasures: Implement personal practices (decision journals, pre-mortems, devil's advocates) that counter predictable biases.
Create organizational structures: Design information flows and decision processes that surface contrary perspectives.
Welcome challenge: Encourage and reward people who disagree with you. They're more valuable than those who confirm.
The best CEOs aren't those who somehow transcend cognitive limitations. They're those who recognize their limitations and build systems to compensate.
That's the only path to consistently good judgment.
In a role where single decisions can create or destroy billions in value.