Key Takeaways
Change behavior by redesigning the workplace, not by blaming the people in it.
Behavior is the hinge. Organizational change succeeds only when people regularly do something differently. Vague ambitions such as improving collaboration do not tell employees what to do, or make the new behavior practical in their daily work. The authors argue that leaders should specify the actions that would show the change is real, then shape the conditions around those actions. Their Work Systems Model is a way to diagnose and redesign the environment that influences work behavior.
Halloran’s warning. A manufacturer outsourced supply-chain logistics, but plant managers bore program costs while corporate received the savings. Local purchasing agents also lost authority and supplier relationships. Leaders had promoted a sound strategy without making its local mechanics work. After 36 months, the program was canceled, its executive sponsor had departed, and competitors had moved ahead. The lesson is not that employees always resist change. It is that a system can make the old behavior more sensible than the new one.
The shift from individual blame to situational diagnosis echoes a well-established finding in social psychology: observers often explain behavior by personality while overlooking the circumstances shaping it. That correction can prevent leaders from cycling through employees when the actual obstacle is a broken workflow or conflicting incentive. The approach also has a useful accountability test: if a desired action is routinely absent, inspect the conditions before judging motivation. Yet environment is not destiny. People interpret rules, exercise judgment, and sometimes oppose changes for principled reasons, so diagnosis should include their accounts, not just management’s design.
Write the future as a scene people can recognize, not a slogan.
Make tomorrow observable. A scene is a concrete account of how work should unfold after a change. It follows particular people through specific tasks, decisions, conversations, and tools. Instead of promising better communication, a leader can describe who contacts whom, what information prompts the exchange, and what they decide together. The authors recommend imagining a sufficiently distant future, then telling its story as if the desired change has already taken hold.
IKEA made purpose tangible. Göran Carstedt, then president of IKEA North America, described a family getting ready for the day and explained how the company’s products should make that routine easier. The story helped executives see the customer and IKEA’s role in her life, rather than merely review charts. Scenes can also reveal conflicts: one imagined team’s success may depend on a decision or resource another scene assigns elsewhere. Revise the picture before implementation exposes the contradiction at greater cost.
Scenes function as a bridge between strategy and operational design. In product development, prototypes make abstract requirements testable; a work scene does something similar for organizational plans. People can challenge a concrete sequence, spot missing information, and imagine their own role in it. The method also improves communication because a memorable episode carries more context than a list of values. Its limitation is that a vivid scenario can make one future feel inevitable. Multiple scenes, including plausible failure cases, can broaden discussion and keep a persuasive story from being mistaken for evidence.
Change at least four parts of the work system substantially.
Look across the whole system. The Work Systems Model connects behavior to eight parts of a work environment: reporting structures, workspace and tools, task flows, people and skills, rewards, measures, information access, and decision authority. Each part sends cues about what matters and how work gets done. Changing only one may leave the others reinforcing the old routine. The authors recommend significantly altering at least four parts, enough for employees to experience a genuinely different environment.
Treat the number as a guide. A new software tool, for instance, may not help if employees lack training, managers still reward the old process, and decision rights remain unchanged. Leaders should compare the desired scene with current conditions, identify which elements help or obstruct it, and coordinate the needed changes. The framework also works as a go or no-go test: if the organization cannot change enough of the conditions, the initiative may not be ready to proceed.
This systems view is stronger than treating change as a communication campaign because it asks whether the organization’s operating parts reinforce one another. Sociotechnical design makes the same broad point: technology, social roles, and work processes need to be considered together. The four-part threshold is useful as a planning prompt, but it is a practitioner’s rule of thumb in this book, not a demonstrated universal cutoff. A small change may succeed if one lever is decisive; a sweeping redesign may fail despite touching many. The key test is coherent influence on the target behavior, not a mechanical count.
Treat resistance as a clue that workplace signals may contradict the change.
Inspect the local logic. People tend to adapt to the conditions in which they work. If a company asks employees to collaborate across units while rewarding each unit only for its own results, the request competes with the system’s stronger instruction. A leader may label the response resistance, but it can be a rational attempt to succeed under existing rules. The authors caution against assuming that hiring different people or offering more training will fix a problem produced by the work design.
Change the conditions too. Hiring, coaching, and skill development matter, but newly trained employees can quickly fall into the habits the environment rewards. The book connects this pattern to the fundamental attribution error, the tendency to explain others’ behavior by their character while neglecting the situation. In Halloran’s supply-chain project, local agents had reasons to protect authority and trusted supplier ties. Those realities needed to be addressed, not dismissed as reluctance.
This framing can make change efforts less punitive and more empirically curious. In safety science, investigators often look beyond an individual’s mistake to the procedures, tools, and pressures that made it likely. The same question helps leaders distinguish unwillingness from incompatible demands. It also protects against a common organizational habit: using training to compensate for incentives or authority structures that remain untouched. Still, not all resistance is a design defect. Employees may identify real risks, inequities, or losses. Their objections are useful data, but they deserve deliberation rather than automatic acceptance or dismissal.
Reward the innovation process, not merely the people who suggest ideas.
Make the desired work pay. Whirlpool wanted innovation to come from across its workforce, not only from engineering and marketing. It tied capital allocation to innovation projects, requiring divisions to devote 10 percent of investment funds to them, then increasing the requirement to 20 percent. A third of senior executives’ pay was linked to innovation results. The company also trained hundreds of employees as innovation mentors and created funds for ideas that the usual annual budget cycle might otherwise block.
Make ideas actionable. These moves linked expectations to resources, skills, recognition, and authority. The reported innovation pipeline grew from $1.3 billion to $3.3 billion by 2006. That result matters less as proof of any one intervention than as a picture of coordinated design: asking for ideas is not enough if no one can develop or fund them. A strategy becomes credible when employees can see the time, support, decisions, and rewards that make its associated behavior worthwhile.
Whirlpool’s case illustrates a valuable distinction between exhortation and infrastructure. Innovation depends not just on creativity but on a path from proposal to experiment, funding, and adoption. Mentors and discretionary funds can lower organizational friction and distribute problem-solving capacity. Incentives need care, however. Once a metric becomes a target, people may optimize the reported number rather than the underlying value. A large pipeline does not alone establish profitable or useful innovation. Measures should track learning and implementation quality as well as the volume of ideas, with room for prudent experiments that do not succeed.
Treat new technology as a work redesign, not a plug-in upgrade.
Technology changes the job. The book’s example is IDx-DR, an artificial-intelligence system approved by the FDA in 2018 to detect diabetic retinopathy from eye images without a clinician interpreting the result. The opportunity is significant because the authors report that only 15 percent of people with type 2 diabetes received the recommended annual or biennial eye screening. But a device alone cannot overcome limited access, outdated clinical workflows, or reimbursement rules that do not recognize its output.
Redesign around the tool. A clinic must decide where screening occurs, who operates the equipment, how results enter medical records, who handles referrals, and how care is paid for. IDx-DR required a new reimbursement code, and its autonomous decision-making raised questions about responsibility for errors. The example shows why implementation must include roles, tasks, information, decisions, and incentives, rather than treating technology as a standalone purchase.
The example anticipates a central challenge in automation: a technical system can perform a task while the surrounding institution remains organized for human performance of that task. Implementation therefore involves professional boundaries, accountability, and payment, not just usability. Similar questions arise whenever software makes or recommends consequential decisions. Who can contest the result, and who bears the cost of an error? These details shape whether access improves equitably. The book’s system lens is particularly useful here, though a full evaluation must also test accuracy across populations and monitor unintended effects after deployment.
Design public policy around the behavior and support people actually need.
Change the response system. Portugal faced severe drug-related harms in the 1990s, including high rates of addiction and drug-related AIDS. In 2001, it shifted possession of small amounts of drugs from a criminal offense to an administrative one, while keeping trafficking punishable. It also redirected attention and resources toward prevention, treatment, and public health. Mobile teams met drug users where they lived, offering services such as medical care, syringe exchange, and referrals to treatment.
Make recovery workable. The approach also supported former users’ return to employment and society. The book reports that heroin use fell from about 100,000 people before the policy to 25,000 in 2019, while HIV diagnoses connected to injection drug use declined by more than 90 percent. These results followed a broad redesign of institutions and services, not a change in messaging alone. The case demonstrates how policy can alter the choices available to people by changing access, incentives, and the system’s response.
Portugal’s case broadens the book’s argument beyond business: behavior is shaped by the institutions that classify, treat, and support people. Harm reduction offers a practical insight here. Services can reach people before they are ready or able to meet demanding conditions, creating a route toward greater stability. The reported outcomes are compelling, but they do not by themselves isolate which policy component produced which result, nor guarantee that the model transfers unchanged to another country. The case is best read as evidence for coordinated public-health design, alongside careful attention to local institutions and outcome measurement.
Sustain change by making the new behavior the easiest way to work.
Build cues into the day. Disney designs customer service as a complete experience, from the physical setting to employee training and daily procedures. Its analysis found that visitors tended to carry a discarded wrapper about 27 paces before dropping it, so the company placed trash cans at similar intervals. That small design choice makes the desired action convenient without requiring a reminder from a manager. Disney also gives employees some discretion to make a guest’s experience better, rather than scripting every interaction.
Keep testing the system. The authors advise leaders to revisit the work environment as change unfolds, checking whether roles, procedures, measures, and decisions still support the intended behavior. If leaders cannot materially alter enough of those conditions, they should reconsider whether to launch. No system makes change permanent, but repeated cues and practical support can make the new routine easier to sustain than the old one.
Disney’s mix of reliable standards and frontline discretion shows that consistency need not mean robotic service. Good design can remove predictable friction while leaving room for judgment in unusual situations. This resembles the distinction between a protocol and a principle: a protocol handles recurring cases, while a principle guides exceptions. The balance is delicate. Excessive scripting can suppress initiative, while unrestricted discretion can make service uneven. Continual observation helps reveal whether the environment is genuinely supporting the intended behavior, rather than merely producing compliance when leaders are watching.
Analysis
The book’s core contribution is to treat organizational change as a design problem with behavioral consequences. It relocates the central question from how to persuade employees to how to arrange work so the desired action is practical, informed, supported, and rewarded. That move connects strategy to execution at a level leaders can inspect: reporting lines, tools, workflows, skills, incentives, measures, information, and decision rights. Its scene-making method adds a useful middle layer between an abstract aspiration and an implementation plan. A future episode of work can be discussed, challenged, and translated into concrete design choices.
The case studies make the argument vivid across very different contexts. Halloran shows how costs and benefits assigned to different groups can quietly sabotage a plan. Whirlpool shows that inviting ideas without funding and development capacity is mostly theater. The AI example reveals that technology adoption touches professional authority, reimbursement, and liability. Portugal demonstrates that changing a social problem’s institutional response can change the options available to people. Together, the cases argue for coherent intervention rather than isolated slogans or tools.
The framework is strongest as a diagnostic and planning discipline, not as a causal law. The recommendation to alter at least four levers is a useful forcing question, but the book does not establish that four is a scientifically validated threshold. Its success stories offer practical detail, while providing less basis for comparison with failed or partially successful implementations. Leaders should therefore use the model to form explicit hypotheses, then measure whether behaviors and outcomes actually change.
There is also a human dimension worth making explicit. Designing environments can enable employees, but it can also become a method of control if leaders define the desired behavior without meaningful input from the people doing the work. Scenes are most credible when employees help write them, surface trade-offs, and identify risks. Finally, measures and rewards require scrutiny: people can optimize a metric in ways that undermine the purpose it represents. The model’s enduring value lies in making these choices visible, connected, and revisable, not in promising a frictionless transformation.
People Also Read
Glossary
Work Systems Model
Map of work conditionsA framework for understanding how organizational behavior is influenced by connected features of the work environment. It examines structure, workspace and tools, task processes, people and skills, rewards, measurement, information distribution, and decision allocation. Leaders use it to identify which conditions support or obstruct a desired pattern of behavior, then redesign those conditions together.
8 Levers of Change
Eight environmental design choicesThe eight parts of the work environment that leaders can alter to encourage desired behavior: organization, workplace design, task, people, rewards, measurement, information distribution, and decision allocation. Each can send a cue about how work should be done. The authors emphasize coordinating changes across several levers so that the environment does not reward one behavior while asking for another.
Scene
A concrete future-work storyA detailed account of how work should look after a change succeeds. A scene follows people in particular roles through actions, decisions, information exchanges, and interactions with tools. It translates broad ambitions into recognizable behavior and helps leaders identify what the work environment must support. Scenes can also expose contradictions between different parts of a proposed future.
RACI
Decision-role clarification toolA responsibility-assignment method that identifies who is Responsible for doing work, Accountable for the result, Consulted for input, and Informed about decisions. In the book, it is used to clarify decision-making roles and reduce confusion about who participates, who has authority, and who needs updates. The authors describe it as one practical aid within broader work-system redesign.
Download PDF
Download EPUB
.epub digital book format is ideal for reading ebooks on phones, tablets, and e-readers.