When Paul Fitts stepped into the cockpit of a fighter jet at Wright-Patterson Air Force Base near Dayton, Ohio, he wasn’t dressed for flight. No helmet, no parachute. Just a notebook, a pencil and an abiding fascination with human nature. Indeed, he didn’t even know how to fly a plane.

In 1942, just after the United States had entered World War II, Fitts was therefore an unusual choice to be entrusted with recommending how to get the best performance from an aircraft worth roughly $2 million in today’s terms, never mind every twin engine fighter the United States had in operation. Fitts was on a mission, though not the kind the Air Force typically dispatched. He wanted to know why some of America's most experienced  pilots—many of whom who had flawlessly logged thousands of hours in the  sky— were suddenly making catastrophic errors in judgment at the most critical moments.

As the war intensified, Fitts was chosen for this special mission because he possessed a skill America desperately needed: He specialized in understanding human behavior under stress. The Air Force had recruited him from his Ph.D. program at the University of Michigan, seeing in him a rare combination of practical intuition and scientific precision. Unlike many academics of his time, he wasn’t content to theorize from behind a desk. He insisted that to truly understand human behavior, he needed to observe it firsthand, in  real- world settings.

Fitts soon found himself squeezed into the cockpit beside seasoned pilots, observing them at work. He noted every hesitation, every misplaced reach, every flicker of confusion. And soon, a troubling pattern emerged. Pilots were crashing into trees, ending up in riverbeds and landing without lowering the wheels. As planes were getting better, pilots seemed to be getting worse.

Fitts saw clearly what others had missed. The pilots weren't failing the plane; the plane was failing the pilots. Unwittingly, the military had set good pilots up to fail. Pilots weren't randomly making mistakes; they were consistently mixing up similar-looking controls. Important instruments were oddly placed, almost deliberately contrary to intuition. Under the intense stress of combat or tricky maneuvers, these design flaws became deadly traps. His report summed it up in a blunt phrase: “A great many accidents result directly from the manner in which equipment is designed and where it is placed in the cockpit, and therefore can be eliminated by attention to human requirements in the design of equipment.”

Fitts and his colleagues’ research findings guided engineers and designers to reshape cockpits with an eye toward human instinct. Controls were given unique shapes and textures so that pilots could distinguish them by touch alone. Instruments were grouped logically according to their purpose and urgency, and vital switches were repositioned exactly where a pilot’s hand would naturally reach. Implementing Fitts’s recommendations transformed aviation safety, significantly reducing pilot errors and fatal plane crashes.

Our former colleague Yves Morieux has a maxim that encapsulates Paul Fitts’s spirit: “Context drives behavior, which drives performance.”

Most transformations fail not because the strategy is wrong, but because the cockpit is. The issue is not what to do, but how to do it. Very often, employees know what needs to be done, but feel the system is fighting them. Goals and deadlines are established by executives, sometimes without an appreciation for how complex and time-consuming they will be to achieve. The organization’s structure doesn’t encourage timely decision-making or productive collaboration. The metrics and incentives for different teams can contradict each other. In meetings, people devote more time to reviewing what has been done than to preparing for upcoming work or providing help to achieve the next level of performance.

The lessons of Paul Fitts reach far beyond the cockpit. New technology or different tools are a constant feature of almost all organizations and jobs. Yet the challenge of shifting human behavior in new directions remains. The ever-increasing improvements in analytics, digitization and artificial intelligence have all been accompanied by the reality that as technology progresses, it is human beings who will remain at the center of the change. As Harvard psychologist Stanley Smith Stevens put it almost 80 years ago in “Machines Cannot Fight Alone”:

Although the human operator proved to be a major weakness, it was realized that much of the fault was in the original design of the equipment, which was often poorly matched to the physical and intellectual capabilities of the men and women who had to use it. … Research was needed to find designs which were more compatible with human capabilities.

Successful change isn't about forcing new behaviors on people. It’s about building environments that through a combination of design, incentives and nudges naturally lead to the desired behaviors. It requires leaders to go where the work happens and understand what prevents people from behaving in ways that leaders expect. Companies thrive when they design workplaces for the ways people behave, feel and think. Just as Paul Fitts turned confusing cockpits into safer ones by respecting human nature, change leaders can increase their odds of success by redesigning their organizations with an understanding of real human behavior.

To put your organization among the minority that run successful change programs:

Get true agreement, not false alignment. When executives decide to embark on large-  scale change programs, they often fail to truly agree on why things need to change, what will be changing and how that change will occur. When this happens, the rest of the organization can’t make progress. In successful change programs, leaders have a process for bringing together their peers into a true agreement that creates the conditions for successful, sustained change.

Increase agency, not just involvement. People are more open toward change when they feel they have had some agency in designing it. In successful change programs, leaders help employees develop high-agency mindsets through specific experiences: making deci-sions, influencing someone else’s decisions and being represented by trusted peers.

Expect take up to be earned, not automatic. Organizations are full of barriers that prevent people from doing what you expect of them, or what they expect of themselves. In successful change programs, leaders are curious about situations where employees don’t have the capability, opportunity or motivation to make a change. Then they use the organization’s resources to lower those barriers until none are left. They inspire, encourage and enable.

Understand emotions through feedback, not instinct. Companies don’t guess how customers feel about their products. They measure it and use what they learn to make improvements. In successful change programs, leaders do the same for their employees by continuously measuring emotions, confidence and capacity. When concerning results appear, they take action. When they observe a behavior they would like to see more of, they encourage it.

Use a process with rituals, not reactions. The more decisions people need to make, the harder it becomes for them to make those decisions well. Change programs require intensive organization-wide coordination, which creates risk of decision fatigue for everyone. In successful change programs, leaders reduce the risk by designing ritualized meetings for every team, including structured celebration, reflection and debate. These rituals give people back the cognitive capacity to do the work that matters most.

Share stories and symbols, not just dollars. In successful change programs, leaders understand their employees’ goals and desires. They use that knowledge to craft and tell a compelling, emotional story about their change and reinforce it through symbolic objects, actions and events.

Create momentum, not just at the start. Change demands tenacity. Early wins matter but don’t create sustained momentum on their own. In successful change programs, leaders continuously predict when a program is likely to lose momentum, and adjust their plans to encourage successful teams—and reset struggling ones.

Reprinted by permission of Harvard Business Review Press. Excerpt from “How Change Really Works: Seven Science-Based Principles for Transforming Your Organization” by Julia Dhar, Kristy Ellmer, Philip Jameson. Copyright 2026. All rights reserved.