Key Takeaways
Treat yesterday's winning method as a hypothesis, not an identity.
Success can become a trap. A strategy that once worked can harden into a habit, even when the situation has changed. Unlearning means letting go of, reframing, or moving away from once-useful beliefs and behaviors that now limit results. It does not mean erasing knowledge or pretending experience has no value.
Serena changed her game. After injuries and a shocking first-round loss at the 2012 French Open, Serena Williams kept training as before, but her results kept sliding. New coach Patrick Mouratoglou identified problems with her balance and movement. Serena tried adjustments to her footwork and preparation, then won 19 consecutive matches and the Wimbledon and US Open titles. Her previous success had not vanished. It had become an unreliable guide to what she needed next.
The distinction between experience and attachment is crucial. Expertise is useful because it compresses prior learning into fast judgment, but that same efficiency can make old assumptions feel like facts. In cognitive science, this resembles the danger of overfitting: a model explains yesterday's data so well that it misses changed conditions. Serena's response was not to discard her craft, but to let fresh observation revise it. The approach has limits: change for its own sake can discard valuable expertise. The better test is whether a familiar method still produces the outcome the current situation demands.
Run a cycle: release stale habits, test alternatives, then scale evidence.
A repeatable process. The Cycle of Unlearning is a three-part method for adapting individual behavior and organizational practice. First, identify what is no longer working and become willing to let it go. Second, relearn by gathering new information and trying alternatives. Third, use the results to make a breakthrough, reflect, and begin another cycle.
Make learning practical. The cycle is not a motivational event or a one-time training session. A team can use it on a specific obstacle, such as a product customers do not want, before applying the lessons to larger challenges. Reflection asks what worked, what did not, and what should change next time. Each round should make the next experiment better informed, not merely repeat the same activity with a new label.
The cycle turns an abstract call to adapt into a feedback system: act, observe, revise. That logic connects with scientific inquiry and control theory, where systems correct course by comparing actual outcomes with intended ones. The emphasis on repeated loops is stronger than the familiar idea of learning as simply acquiring more knowledge. Yet a loop only improves the system if its measurements are meaningful and its feedback is honest. A team can cycle quickly in the wrong direction if it tracks activity rather than consequences, or filters out evidence that challenges its preferred explanation.
Make ambition measurable by naming the behaviors success would change.
Define the destination first. Before experimenting, state the challenge and describe what success would look like in observable behavior. A vague aim such as becoming more effective gives people little guidance. A clearer goal might be raising customer retention by a stated amount within a defined period, then identifying what customers, employees, or leaders would actually do differently.
Turn aspiration into evidence. The book recommends setting a constraint, such as time or investment, and tracking rates or ratios rather than relying only on totals. IAG leaders, for instance, imagined customers booking and boarding through digital services, and employees spending more time innovating instead of firefighting. If it is easier to describe failure, describe it first and reverse it into a target. The point is to replace opinion about progress with observable evidence.
Behavioral definitions make strategy discussable. They reduce the distance between an executive slogan and the daily choices that might realize it. This also echoes implementation intentions in behavioral science: people act more reliably when a goal is translated into a concrete response. Measurement, however, can distort the target. Once a metric becomes a reward, people may optimize the number rather than the underlying outcome. The safeguard is to combine indicators, review unintended effects, and keep the customer or system-level result in view instead of treating a dashboard as reality.
Start with experiments small enough to fail without wrecking the mission.
Make the first move easy. Relearning means testing new behavior, not simply collecting ideas. A safe-to-fail experiment is a limited trial whose downside is recoverable and whose results provide useful information. A large aspiration can be approached through small steps: a person training for a marathon might begin with a short walk, then increase the challenge as ability grows.
Match the behavior. The Fogg Behavior Model says behavior occurs when motivation, ability, and a prompt converge. If action is not happening, one or more may be missing. The book's related idea of behavior matching means finding an action suited to a person's skills, motivation, and situation instead of imposing one routine on everyone. Small tests let people discover which actions fit, while building confidence for harder steps.
Small trials serve two purposes: they limit risk and make learning visible before commitment becomes expensive. This resembles staged investment in engineering and venture design, where early evidence earns the right to spend more. It also counters the common tendency to demand motivation when the real obstacle is friction or poor fit. Still, small experiments are not automatically safe or informative. In tightly coupled settings, a local change can affect others, and a trivial test may not predict large-scale behavior. Good design specifies what the test can reveal and what it cannot.
Change what you do first; new experience can change what you believe.
Action updates perspective. Leaders often announce that a team needs a new mindset, but the book argues that speeches alone rarely change behavior. People gain new perspectives by acting differently, seeing the results, and reflecting on what those results imply. That revised understanding can make further behavioral change easier.
Measure outcomes, not busyness. The author describes executives who spend large portions of their workweek in meetings that they themselves consider ineffective. A small intervention is to reserve the final minutes for participants to say whether the meeting achieved its intended result. Likewise, a team should not stop at counting completed tasks. It should ask whether the work improved the outcome it was meant to affect, such as customer retention or service quality.
This is a useful correction to the assumption that persuasion must precede practice. In embodied learning, experience can revise belief because people interpret the world through what they do and notice. The meeting example also creates a low-cost feedback loop that can expose a leader's blind spots. The caution is that behavior alone does not guarantee insight: people can repeat a new ritual without examining its effects. Reflection needs candor, and the chosen outcome must be within the experiment's plausible reach, or participants may mistake noise for proof.
Give teams the destination, then move decisions toward the information.
Replace command with intent. Industrial-era management treated workers as executors of decisions made elsewhere. The book argues that knowledge workers often have better situational information than distant managers. Leaders should clarify what outcome matters and why, then let capable people determine how to reach it, with feedback loops to maintain accountability.
Build authority gradually. In the Ladder of Leadership, David Marquet's framework for shifting from orders toward employee initiative, people take increasing responsibility as their clarity and competence grow. As commander of the struggling submarine USS Santa Fe, Marquet began replacing commands with questions about the crew's intended action. The crew learned to make decisions instead of waiting for permission. The submarine later became one of the Navy's strongest performers. Delegation is not abandonment: leaders still set purpose, develop capability, and monitor results.
Decision authority works best when it follows relevant knowledge, a principle shared by distributed systems and modern incident response. Central leaders cannot maintain a complete picture of every local condition, so excessive approval layers slow adaptation and erase context. The book's emphasis on intent plus capability is important because decentralization without clarity can become fragmentation. Nor should every decision be pushed downward: irreversible, high-consequence choices may warrant tighter coordination. The leadership task is to calibrate authority to competence, uncertainty, and the cost of error.
Let customers expose the gap between your promise and their experience.
Hear reality directly. Internal reports can smooth over problems as they travel up an organization. The book calls a reassuring report that hides serious trouble a watermelon report: green outside, red inside. Leaders need direct, repeated contact with customers and employees to test whether their assumptions match actual experience.
The gap can be dramatic. In a Bain survey cited in the book, 80 percent of companies believed they delivered a superior customer experience, while only 8 percent of their customers agreed. T-Mobile CEO John Legere listened to customer-service calls and used social media to hear complaints firsthand. That input helped the company challenge industry practices customers disliked, including complicated contracts and fees. At IAG, testing a leader's favored booking-platform idea with customers repeatedly produced negative reactions, forcing him to reconsider the idea rather than blame the testers.
Direct contact is a form of epistemic discipline: it exposes decision-makers to evidence that organizational status and polished presentations can obscure. It also recognizes that users are not merely recipients of innovation but sources of knowledge about how a service functions in context. Still, one loud complaint or a small sample should not dictate policy by itself. Customer feedback needs interpretation, attention to who is absent, and comparison with broader evidence. Listening is valuable not because every request should be followed, but because assumptions should face contact with reality.
Make bad news safe to share before small deviations become disasters.
Fear corrupts information. Psychological safety means people can raise concerns or admit mistakes without expecting humiliation or punishment. The book connects this condition to stronger team learning: errors can be examined while they are still manageable, instead of hidden until they grow serious.
NASA learned through painful failures. Before the Challenger disaster, engineers' concerns about cold launch conditions were overridden. Before Columbia broke apart, recurring damage from foam strikes had come to seem routine. This pattern is called normalization of deviance: a departure from safe practice becomes accepted because earlier instances did not cause catastrophe. Later, NASA delayed Discovery's launch while engineers investigated a valve problem they could not explain. That response prioritized unresolved evidence over pressure to proceed.
The NASA cases show why psychological safety is not simply a matter of kindness. It is part of an organization's sensing capacity: people closest to a hazard must be able to transmit weak signals before they become emergencies. This connects with high-reliability engineering, where teams remain attentive to anomalies rather than treating past survival as proof of safety. Safety does not mean lowering standards or excusing negligence. It means separating useful reporting from blame so that accountability can address causes and controls, rather than teach people to conceal inconvenient facts.
Reward shared outcomes, not local numbers people can game.
Incentives shape attention. If employees are rewarded for a narrow number, they may optimize that number even when it harms the wider organization. The book describes Wells Fargo employees opening as many as 3.5 million unauthorized accounts under pressure to meet sales targets. The measurable activity displaced the intended outcome of serving customers.
Align effort to the whole. Capital One's cloud transformation combined visible measures of employee skills with measures of migration speed, quality, and cost. Certifications gave workers a concrete development path, while a Cloudometer dashboard connected that progress to broader outcomes. The author does not argue that every incentive should disappear. Recognition, career growth, and modest rewards can help, but measures should include hard-to-measure system results and be safe to discuss honestly.
This section applies a central insight from principal-agent theory: when one party can observe only a proxy for another's work, incentives based on that proxy invite gaming. The danger grows when complex work is compressed into one score. A balanced set of indicators can make trade-offs more visible, but no dashboard removes judgment or political pressure. Transparency is useful only when poor numbers trigger diagnosis and support, not retaliation. Otherwise, people learn to produce flattering metrics, and the organization loses the very information it needs to improve.
Build learning into everyday work, not annual transformation theater.
Make adaptation operational. Organizational learning requires routines for gathering information, sharing mistakes, testing changes, and improving the system. It cannot depend on a crisis or an occasional off-site. Leaders have to model these behaviors and create time, tools, and protection for teams to practice them.
The NHS case shows what small-scale change can deliver. After a costly national IT program faltered, Andrew Meyer led a 30-person team that rebuilt the NHS Spine service in-house, using smaller funding increments and frequent input from users. Spine 2 handled 45 million messages a day, had 99.999 percent availability, and saved the NHS an estimated £26 million annually. The team built a system that users needed, rather than delivering every feature originally requested.
The NHS example makes organizational learning concrete: not a cultural slogan, but a different procurement rhythm, team structure, and relationship with users. It also shows how constraints can enable experimentation when leaders protect a small team from reflexive bureaucracy. The reported results are striking, though a single case cannot establish that the same model will transfer unchanged to every public service. The transferable lesson is about mechanisms: small commitments, working increments, user feedback, and institutional permission to revise. Those mechanisms can be adapted; the exact operating model should not be copied blindly.
Analysis
Unlearn is a leadership and organizational-change book built around one paradox: past success can become a source of future failure. Its distinctive contribution is not simply urging openness to change, but linking that stance to an operating method. Define what is no longer working, set an observable outcome, test a different behavior in a bounded way, and feed the evidence into the next decision. Stories ranging from Serena Williams's coaching change to NASA's safety failures and the NHS Spine rebuild make the method legible across personal, managerial, and institutional scales.
The strongest argument is that adaptation depends less on telling people to think differently than on changing the conditions under which they act. Small experiments, direct customer contact, clear intent, and psychological safety make learning part of work rather than a separate training event. The book also correctly treats incentives and authority structures as behavioral design: people tend to follow the measures, permissions, and rewards that organizations create, not merely the values they announce.
A useful analytical distinction is between abandoning expertise and suspending its automatic authority. Unlearning should mean testing whether a known method still fits current conditions, not treating novelty as inherently superior. The same caution applies to metrics and experimentation. Measures are proxies, experiments have limits, and a safe-to-fail trial in one department may create risk elsewhere. A mature cycle therefore needs explicit boundaries, attention to who bears the downside, and multiple forms of evidence.
The framework is most persuasive when it treats learning as a property of a system rather than a heroic trait of an individual. Leaders shape whether bad news travels, whether decisions can be made near relevant information, and whether rewards favor enterprise outcomes or local wins. The practical test is not whether an organization uses the language of agility or innovation. It is whether people can notice a failing assumption, try a better approach, and revise course without having to wait for a crisis to make change politically acceptable.
Review Summary
Unlearn receives mostly positive reviews, with an average rating of 3.86/5. Readers praise its insights on breaking old habits and embracing change for success. Many find the concept of unlearning valuable for personal and organizational growth. The book's real-world examples and practical advice are appreciated. Some critics feel the content is repetitive or lacks depth. Overall, readers recommend it for leaders and innovators seeking fresh perspectives and improved performance.
People Also Read
Glossary
Cycle of Unlearning
Adaptation through repeated cyclesA three-step process for changing behavior when familiar methods stop producing desired results. A person or team first identifies and releases an outdated assumption or practice, then relearns through new information and experiments, and finally reflects on results to create a breakthrough. The process repeats as conditions change and new challenges emerge.
Fogg Behavior Model
Three conditions for actionA behavior-design model stating that a behavior occurs when motivation, ability, and a prompt converge at the same time. If the behavior does not happen, at least one condition may be missing. The model suggests making an action easier, finding an appropriate prompt, or addressing motivation rather than assuming that people simply need more willpower.
behavior matching
Fit actions to individualsThe practice of finding a behavior suited to a person's motivation, ability, skills, and preferences, rather than prescribing one approach for everyone. People can share an aspiration but need different actions to make progress. Small experiments help identify which behavior is workable for a particular person and situation.
Psychological safety
Safety to speak and experimentA team condition in which people can raise concerns, ask questions, admit mistakes, and try unfamiliar approaches without fearing damage to their standing or career. In the book, this safety supports candid information-sharing and experimentation. It does not remove accountability; it allows problems to be examined early enough to improve the work.
Ladder of Leadership
Growing employee decision authorityDavid Marquet's framework for moving employees from following detailed instructions toward taking greater initiative and ownership. Leaders increase responsibility as people demonstrate clarity about the mission and competence in making decisions. The progression gives managers a practical way to develop autonomy gradually instead of either retaining every decision or delegating without support.
normalization of deviance
Unsafe departures become routineA process in which deviations from expected or safe practice gradually become accepted because earlier instances did not produce immediate harm. The book uses NASA's shuttle program to show how recurring foam strikes came to seem tolerable before a later strike contributed to the Columbia disaster. The term warns that survival of previous incidents is not proof of safety.
mission command
Intent with decentralized actionA leadership approach in which leaders communicate the purpose of a mission and the outcome required, while giving people closest to the situation authority to decide how to act. It depends on clear intent, capable teams, and trust supported by feedback. It is not an absence of leadership, but a way to make decisions responsive to local information.
Download PDF
Download EPUB
.epub digital book format is ideal for reading ebooks on phones, tablets, and e-readers.