Key Takeaways
Design a better choice instead of optimizing between two bad options
A better kind of choice. Integrative thinking uses the tension between opposing ideas to create a new answer that preserves what matters in each, rather than settling for a compromise. LEGO CEO Jørgen Vig Knudstorp faced a choice between protecting the brand by controlling a film and giving filmmakers the freedom needed to make it good. Neither option was acceptable on its own.
Change the conditions. Knudstorp let the filmmakers shape the story, but brought them into contact with LEGO fans, designers, and customers so they could understand the brand from the inside. The resulting film earned more than $450 million worldwide and helped lift LEGO sales. The key move was not splitting control down the middle. It was designing a partnership that made creative freedom and brand care compatible.
The LEGO example makes a useful distinction between compromise and integration. Compromise divides a scarce resource, such as authority, while integration changes the structure of the work so that each side can contribute differently. This resembles good institutional design: rules and relationships can shape incentives more effectively than exhortations to cooperate. Still, the method depends on meaningful room to redesign those relationships. Some conflicts involve genuinely limited resources or incompatible rights, so a creative process cannot guarantee that every valued outcome can coexist.
Write down the model behind your certainty before trusting its conclusions
The mind edits reality. A mental model is a simplified picture of how part of the world works. It helps people act without processing everything, but it also leaves things out and often operates without conscious notice. In a study of a 1951 Princeton-Dartmouth football game, students at each university watched the same footage yet judged their own team’s fouls more leniently. Each group experienced its interpretation as the game itself.
Simple explanations can miss. Tata Motors assumed that families who could not afford a car would buy one if offered a very cheap model. Its Nano cost about $2,300, but first-year sales were roughly 60,000, far short of the 250,000 annual capacity planned. The model overlooked that a car could also signal aspiration. Naming assumptions makes it possible to test whether the explanation fits the world.
Mental models are not defects to eliminate. They are necessary compression tools, much like maps that preserve the features needed for a particular journey. The danger comes when a useful map is mistaken for the territory, or when a model built for one context is treated as universal. The Nano example highlights an especially important omission: consumer choices carry social meaning as well as functional value. A practical discipline is to ask what a model excludes, whose experience it represents, and what evidence would change it.
Treat opposing views as evidence about what your model leaves out
Dissent can reveal blind spots. People often interpret disagreement as proof that the other person is uninformed or acting in bad faith. That reaction shuts down inquiry and encourages groups to confirm what they already believe. In a health-care workshop, clinicians initially insisted that vaccines were not a matter for debate. A participant pointed out that parents were still refusing them, so the group needed to understand the reasons behind that resistance if it hoped to change behavior.
Understanding is not endorsement. The book describes the original claim linking the MMR vaccine to autism as discredited. The clinicians did not need to accept false claims to investigate why some parents valued choice, distrusted institutions, or feared harm. Taking those concerns seriously could improve communication and policy. The point is to examine an opposing model for what it reveals, not to pretend all claims have equal evidence.
This is a strong case for separating epistemic respect from factual equivalence. A person can be heard as a reasoning agent without granting equal weight to every assertion. That distinction matters in public health, where dismissive communication can harden mistrust, but uncritical accommodation can also expose communities to preventable disease. The useful question is not simply whether a dissenting position is correct. It is what unmet need, lived experience, or institutional failure helps explain its appeal, and how that insight can improve a response grounded in evidence.
Make thinking visible before asking a team to solve the problem
Three skills support better choices. Metacognition means noticing and examining one’s own thinking. Empathy means trying to understand another person’s experience, rather than assuming that person would feel as you do. Creativity means generating new answers instead of only evaluating existing ones. The book treats these as learnable capabilities that strengthen one another.
Trace the path to a conclusion. The ladder of inference is a tool for separating observed data from the interpretations and conclusions built on it. Teacher Beth Grosso used paper fish representing facts about a student. Facts such as slamming a door could suggest a bad temper, until a hidden fact revealed the student had been teased. The class could then reconsider its conclusion. Another interviewee, Kelli, presented a hunting rifle as her American artifact, but explained that it represented the freedom to speak amid disagreement. Listening uncovered a meaning that the interviewers had not expected.
Making reasoning visible gives a team something more useful to discuss than clashing conclusions. It can expose where members selected different facts, applied different standards, or filled gaps with assumptions. This resembles clinical case formulation, where practitioners distinguish reported observations from interpretations before deciding on an intervention. The method is not a guarantee of accuracy: people can describe their reasoning selectively or rationalize a settled view. Its value lies in making those judgments discussable, especially when curiosity and genuine listening are part of the process.
Push choices to extremes, then find what each side is protecting
Start with a clear tension. The first stage is to frame a problem worth solving, identify two opposing answers, and describe what each would mean in practice. A financial-services team, for example, could place all employee training under a central office or give each region and business unit full control. Extreme versions make the underlying differences easier to see than a blurry middle-ground proposal.
Map the value before judging. A pro/pro chart records the benefits of each model for the people affected by the decision. Centralized training could lower costs and create consistency; local control could improve relevance and speed. The team identifies key stakeholders, then asks how each benefit works and why it matters. Temporarily postponing discussion of drawbacks helps members understand why someone might favor each model. It does not mean risks are ignored forever.
Extreme alternatives can work as a diagnostic device. They sharpen differences that ordinary proposals often blur, making it easier to notice which values are competing and which assumptions support them. The stakeholder emphasis also reduces the risk of treating an organization’s convenience as the only relevant measure of success. There is a limit: extreme models can become caricatures if participants do not recognize them as exploratory constructs. The discipline is to make them vivid enough to learn from, while staying willing to revise the framing when it misrepresents the actual problem.
Trace what produces the benefit, not just the benefit you want
Look beneath the outcome. In the second stage, teams compare the models, identify what they value most, surface assumptions, and examine the forces that produce valued results. A causal model is a diagram of hypothesized cause-and-effect relationships. It helps shift discussion from wanting an outcome to understanding how the current system generates it.
Turn inclusion into an asset. Toronto’s film festival was open to the public but struggled to become financially sustainable. The Cannes model attracted money and media attention through exclusivity and industry prestige. Festival director Piers Handling noticed that the Toronto audience itself could generate buzz: its size and diversity made it a useful test audience for films. The festival highlighted its existing audience-voted prize, the People’s Choice Award. By 2015, the festival drew about 400,000 attendees and had a budget above $40 million. Inclusion became a source of influence, not an obstacle to it.
The festival story shows why causal reasoning can open creative options. If prestige is assumed to be the sole cause of attention, an inclusive event seems destined to remain less influential. By asking what actually produces buzz, Handling found another mechanism: audience response that could signal a film’s wider appeal. This is systems thinking in practical form, though the causal links remain hypotheses rather than controlled proof. A crowd’s preferences may not predict every market, and success depends on context. Mapping causes is a way to generate and test possibilities, not to claim certainty.
Search for integrations by recombining, intensifying, or dividing the problem
Three routes open the search. The hidden gem approach combines one valuable element from each model and discards the rest. Tennis Canada joined the French system’s coordinated development with the American system’s flexibility, supporting young players through national training while leaving much of their development in local hands. The double down approach intensifies a favored model to create a missing benefit. Jack Bogle made Vanguard customer-owned and used low-cost index funds to pursue better net returns for investors.
Separate what need not stay joined. Decomposition breaks a problem into parts so each model can work where it fits best. Manitoba Hydro’s office design team separated heating and cooling from ventilation. The building used energy-saving systems for temperature control and natural airflow for fresh air, rather than treating both as one sealed climate-control problem. The building used about 90 kilowatt-hours per square meter each year, compared with 270 to 400 in the comparisons reported in the book.
These pathways are valuable because they turn creativity into a search process rather than a demand for sudden inspiration. Each asks a different structural question: which elements are worth combining, whether a favored approach can produce a missing result, or whether the problem itself can be divided more intelligently. The Manitoba Hydro example is especially powerful because it changes the unit of analysis. Yet a decomposition only works when the boundary between parts is real enough to guide action. Naming three routes can prompt inquiry, but it cannot substitute for understanding the technical and human conditions of a particular case.
Turn new ideas into rough prototypes, then test how they fail
Make the idea concrete. A prototype is a rough representation of a proposed solution, such as a story, drawing, role-play, or physical model. These forms expose gaps that remain hidden in abstract descriptions. They also help others understand what the proposal would actually change. During an innovation camp, a student built an elaborate recycling machine but struggled to explain how it would solve the school’s waste problem. Feedback revealed that the model needed a clearer link to user needs.
Test conditions, not confidence. For each possibility, identify what would have to be true for it to work, then design tests that could show where those conditions fail. Tennis Canada did not launch its new player-development system everywhere at once. It first tested whether players, parents, and coaches would support the model, then funded two players to try international training. The results helped the team refine its approach before expanding it.
Prototyping makes disagreement more productive because people can react to an observable proposal rather than a slogan. It also reduces the social cost of revision: a rough sketch is easier to change than a public commitment or a fully funded rollout. The emphasis on trying to falsify an idea is a useful counterweight to confirmation bias. In practice, however, tests often rely on proxies, and early pilots may not predict performance at scale. Good experimentation therefore asks both whether a design works in a small trial and what new risks emerge when conditions change.
Question the stance beneath your actions, not just the actions themselves
Inspect the source of behavior. Stance is a person’s broad understanding of how the world works and what role they have in it. The authors contrast single-loop learning, which adjusts an action after an outcome, with double-loop learning, which also examines the reasoning and assumptions behind that action. The second kind of reflection can reveal why the same response keeps producing the same result.
Act on a chosen view of responsibility. Unilever CEO Paul Polman believed the company should take a longer-term approach and accept responsibility across its supply chain. He stopped quarterly reporting and earnings guidance, while investing in the company’s longer-term plans. Its shares initially fell 8 percent. Polman was willing to lose short-term-oriented investors and attract shareholders more aligned with his strategy. The broader lesson is to examine the beliefs that direct decisions, then decide whether those beliefs are helping produce the outcomes sought.
The distinction between changing behavior and changing the assumptions behind behavior is consequential. A team can repeatedly refine its tactics while leaving intact the belief that speed, quarterly results, or consensus must dominate. Double-loop reflection asks whether that governing belief is appropriate, not merely whether execution was adequate. This connects personal learning to organizational incentives: leaders may sincerely favor long-term thinking while rewarding short-term metrics. Stance is not changed by insight alone, and the Unilever example involves unusual executive authority. Reflection becomes consequential when it is matched by changes to routines, measures, and commitments.
Analysis
The book’s central contribution is a shift in the unit of decision-making. Instead of treating a problem as a menu from which a leader selects the least costly option, it treats the menu itself as provisional. That shift matters because many management debates are not really disputes over facts alone. They are disputes over what counts as valuable, which stakeholders count, and what causal story explains how an outcome is produced. Integrative thinking gives teams a practical sequence for making those hidden commitments visible, then using disagreement as material for redesign.
Its strongest insight is that synthesis is not the same as splitting the difference. The examples work when they identify a mechanism that changes the relationship between competing goals: LEGO shaped the filmmakers’ understanding without controlling their script; TIFF used its public audience to generate attention; Manitoba Hydro separated systems that had been treated as one problem. These are design moves, not rhetorical compromises. The method also usefully treats prototypes as hypotheses, so an attractive idea must acquire evidence before it earns full commitment.
The important caution is that the method should not turn every conflict into a puzzle with a harmonious answer. Some trade-offs reflect actual scarcity, legal obligations, safety limits, or irreconcilable interests. A team can also use elaborate discussion to delay a necessary decision, or label an existing preference an integration after the fact. The process needs explicit decision rights, time limits, and honest accounting of who bears the costs.
The book’s most durable practice may be modest: articulate each model fairly, ask what it protects, and test the assumptions connecting actions to outcomes. That routine can improve a decision even when no superior synthesis appears. It replaces premature certainty with disciplined curiosity, without requiring permanent indecision. The approach is especially apt when familiar choices repeatedly produce dissatisfaction and when stakeholders possess different but relevant knowledge. It is less a universal answer than a method for discovering whether the apparent dilemma is fixed, or merely badly designed.
Review Summary
Creating Great Choices receives mostly positive reviews for its practical approach to integrative thinking. Readers appreciate the step-by-step guidance, templates, and exercises for problem-solving and decision-making. The book is praised for its clear explanations of complex concepts and its emphasis on considering multiple perspectives. Some find it particularly useful for management roles and challenging organizational problems. A few reviewers note that the book can be slow or academic at times, but overall, it's seen as a valuable resource for developing creative solutions and breaking out of limited thinking patterns.
People Also Read
Glossary
Integrative thinking
Creating a superior third answerA problem-solving approach that uses the tension between opposing models to develop a new model that preserves valuable elements of each and aims to outperform both. It differs from compromise, which usually accepts a weaker mixture of existing options. The process is treated as a learnable practice, often carried out by a team.
Mental models
Simplified maps of realityRepresentations people use to understand how the world works and to anticipate outcomes. Mental models filter complexity, making action possible, but they are incomplete and may operate outside conscious awareness. Because people select and interpret information through these models, different individuals can experience the same event in sharply different ways.
Metacognition
Examining one’s own thinkingThe ability to understand and regulate one’s own thinking. In the book’s decision-making method, it includes making the data, interpretations, assumptions, and reasoning behind a conclusion explicit. This enables a person or team to examine not only what it believes, but how it arrived at that belief and where its reasoning may be incomplete.
Ladder of inference
Tracing data to conclusionsA framework associated with Chris Argyris that depicts how people move from the many facts available to selected observations, interpretations, conclusions, and beliefs. Because these steps often happen automatically, people can mistake an inference for an observed fact. Tracing the ladder helps make reasoning visible and reveals where additional information might change a conclusion.
Pro/pro chart
Mapping benefits of both modelsA tool for describing the benefits produced by two opposing models, without listing their drawbacks at the initial consideration stage. Teams identify the important stakeholders and ask what each model provides for them, how it produces that benefit, and why the benefit matters. The purpose is to understand each model before evaluating or combining them.
Hidden gem
Combine one valuable element eachOne pathway for creating an integrative solution. It selects a deeply valued building block from each of two opposing models, discards the remaining structure, and designs a new model around the selected elements. The chosen benefits must be compatible enough to work together, while new components replace the parts that were left behind.
Double down
Extend one model for a missing benefitOne pathway for integration that intensifies a favored model so it can produce an important benefit associated with the opposing model. The team first identifies what the preferred model lacks, then investigates how that benefit is generated elsewhere and whether a different mechanism can create it within the extended model.
Decomposition
Divide the problem into distinct partsOne pathway for integration that reframes and divides a problem so opposing models can be applied to different components, rather than forced into a single blended approach. It requires a meaningful boundary between the parts. Without that boundary, asking people to do both can leave them balancing incompatible demands without clear guidance.
Stance
Beliefs about world and roleA broad set of assumptions about how the world works and how a person fits within it. Stance influences what someone notices, how they interpret events, and which actions seem appropriate. The book argues that examining stance can help explain recurring patterns of behavior and outcomes, and that practice can gradually alter it.
Double-loop learning
Questioning the reasoning behind actionA learning process that examines not only whether an action produced a desired result, but also the reasoning and assumptions that led to the action. It contrasts with single-loop learning, which adjusts behavior while leaving underlying beliefs intact. Double-loop learning can reveal when repeated failures arise from the way a problem is understood, not just from execution.
Download PDF
Download EPUB
.epub digital book format is ideal for reading ebooks on phones, tablets, and e-readers.