McKinsey Quarterly

The psychology of change management

| Article

Over the past 15 or so years, programs to improve corporate organizational performance have become increasingly common. Yet they are notoriously difficult to carry out. Success depends on persuading hundreds or thousands of groups and individuals to change the way they work, a transformation people will accept only if they can be persuaded to think differently about their jobs. In effect, CEOs must alter the mind-sets of their employees—no easy task.

CEOs could make things easier for themselves if, before embarking on complex performance-improvement programs, they determined the extent of the change required to achieve the business outcomes they seek. Broadly speaking, they can choose among three levels of change. On the most straightforward level, companies act directly to achieve outcomes, without having to change the way people work; one example would be divesting noncore assets to focus on the core business. On the next level of complexity, employees may need to adjust their practices or to adopt new ones in line with their existing mind-sets in order to reach, say, a new bottom-line target. An already "lean" company might, for instance, encourage its staff to look for new ways to reduce waste, or a company committed to innovation might form relationships with academics to increase the flow of ideas into the organization and hence the flow of new products into the market.

But what if the only way a business can reach its higher performance goals is to change the way its people behave across the board? Suppose that it can become more competitive only by changing its culture fundamentally—from being reactive to proactive, hierarchical to collegial, or introspective to externally focused, for instance. Since the collective culture of an organization, strictly speaking, is an aggregate of what is common to all of its group and individual mind-sets, such a transformation entails changing the minds of hundreds or thousands of people. This is the third and deepest level: cultural change.

In such cases, CEOs will likely turn for help to psychology. Although breakthroughs have been made in explaining why people think and behave as they do, these insights have in general been applied to business only piecemeal and haven’t had a widespread effect. Recently, however, several companies have found that linking all of the major discoveries together in programs to improve performance has brought about startling changes in the behavior of employees—changes rooted in new mind-sets. Performance-improvement programs that apply all of these ideas in combination can be just as chaotic and hard to lead as those that don’t. But they have a stronger chance of effecting long-term changes in business practice and thus of sustaining better outcomes.

Four conditions for changing mind-sets

Employees will alter their mind-sets only if they see the point of the change and agree with it—at least enough to give it a try. The surrounding structures (reward and recognition systems, for example) must be in tune with the new behavior. Employees must have the skills to do what it requires. Finally, they must see people they respect modeling it actively. Each of these conditions is realized independently; together they add up to a way of changing the behavior of people in organizations by changing attitudes about what can and should happen at work.

A purpose to believe in

In 1957 the Stanford social psychologist Leon Festinger published his theory of cognitive dissonance, the distressing mental state that arises when people find that their beliefs are inconsistent with their actions—agnostic priests would be an extreme example. Festinger observed in the subjects of his experimentation a deep-seated need to eliminate cognitive dissonance by changing either their actions or their beliefs.

The implication of this finding for an organization is that if its people believe in its overall purpose, they will be happy to change their individual behavior to serve that purpose—indeed, they will suffer from cognitive dissonance if they don’t. But to feel comfortable about change and to carry it out with enthusiasm, people must understand the role of their actions in the unfolding drama of the company’s fortunes and believe that it is worthwhile for them to play a part. It isn’t enough to tell employees that they will have to do things differently. Anyone leading a major change program must take the time to think through its "story"—what makes it worth undertaking—and to explain that story to all of the people involved in making change happen, so that their contributions make sense to them as individuals.

Reinforcement systems

B. F. Skinner is best known for his experiments with rats during the late 1920s and the 1930s. He found that he could motivate a rat to complete the boring task of negotiating a maze by providing the right incentive—corn at the maze’s center—and by punishing the rat with an electric shock each time it took a wrong turn.

Skinner’s theories of conditioning and positive reinforcement were taken up by psychologists interested in what motivates people in organizations. Organizational designers broadly agree that reporting structures, management and operational processes, and measurement procedures—setting targets, measuring performance, and granting financial and nonfinancial rewards—must be consistent with the behavior that people are asked to embrace. When a company’s goals for new behavior are not reinforced, employees are less likely to adopt it consistently; if managers are urged to spend more time coaching junior staff, for instance, but coaching doesn’t figure in the performance scorecards of managers, they are not likely to bother.

Some disciples of Skinner suggest that positive-reinforcement "loops" have a constant effect: once established, you can leave them be. Over time, however, Skinner’s rats became bored with corn and began to ignore the electric shocks. In our experience, a similar phenomenon often prevents organizations from sustaining higher performance: structures and processes that initially reinforce or condition the new behavior do not guarantee that it will endure. They need to be supported by changes that complement the other three conditions for changing mind-sets.

The skills required for change

Many change programs make the error of exhorting employees to behave differently without teaching them how to adapt general instructions to their individual situation. The company may urge them to be "customer-centric," for example, but if it paid little attention to customers in the past, they will have no idea how to interpret this principle or won’t know what a successful outcome would look like.

How can adults best be equipped with the skills they need to make relevant changes in behavior? First, give them time. During the 1980s, David Kolb, a specialist in adult learning, developed his four-phase adult-learning cycle. Kolb showed that adults can’t learn merely by listening to instructions; they must also absorb the new information, use it experimentally, and integrate it with their existing knowledge. In practice, this means that you can’t teach everything there is to know about a subject in one session. Much better to break down the formal teaching into chunks, with time in between for the learners to reflect, experiment, and apply the new principles. Large-scale change happens only in steps.

Second, as the organizational psychologist Chris Argyris showed, people assimilate information more thoroughly if they go on to describe to others how they will apply what they have learned to their own circumstances. The reason, in part, is that human beings use different areas of the brain for learning and for teaching.1

Consistent role models

Most clinical work confirms the idea that consistent role models, whom the famous pediatrician Benjamin Spock regarded as decisive for the development of children, are as important in changing the behavior of adults as the three other conditions combined. In any organization, people model their behavior on "significant others": those they see in positions of influence. Within a single organization, people in different functions or levels choose different role models—a founding partner, perhaps, or a trade union representative, or the highest-earning sales rep. So to change behavior consistently throughout an organization, it isn’t enough to ensure that people at the top are in line with the new ways of working; role models at every level must "walk the talk."

The way role models deal with their tasks can vary, but the underlying values informing their behavior must be consistent. In a company that encourages entrepreneurial decision making at low levels, one middle manager might try to coach junior employees to know how to spot a promising new venture; another might leave this up to them. Both, however, would be acting in line with the entrepreneurial principle, whereas a boss who demanded a lengthy business case to justify each $50 expenditure would not be. But organizations trying to change their value systems can’t tolerate as much variance in their role models’ behavior. If entrepreneurial decision making were a new value, both of these middle managers might have to act in roughly the same way in order to encourage their subordinates to make bold decisions.

Behavior in organizations is deeply affected not only by role models but also by the groups with which people identify. Role modeling by individuals must therefore be confirmed by the groups that surround them if it is to have a permanent or deep influence. (Most teenagers could tell you a lot about this.) Say that a well-respected senior leader is waxing lyrical about making the culture less bureaucratic and even conforming to the new regime by making fewer requests for information. If the sales reps in the company canteen spend every lunchtime complaining that "we’ve heard this a thousand times before and nothing happened," individuals will feel less pressure to change their behavior. Change must be meaningful to key groups at each level of the organization.

Putting the approach into practice

The case of a retail bank shows how these four conditions can coalesce to change mind-sets and behavior and thereby improve performance. But though we have grouped the actions of the bank under the four conditions, it didn’t apply them in a neat sequence. As in any change program, there was much disruption and risk. Nonetheless, basing the program on four proven principles gave the CEO confidence that it would eventually succeed.

A few years ago, this CEO took the helm of a large European retail bank that employed more than 30,000 people. He set several targets: doubling the economic profit of the bank, reducing its cost-to-income ratio to 49 percent (from 56), and increasing its annual revenue growth from the current 1 to 2 percent to 5 to 7 percent—all within four years. But retail banking is almost a commodity business. No financial-engineering shortcuts or superficial changes in practice could win a competitive edge for the bank. It could meet these performance goals, the CEO realized, only by galvanizing its people to deliver far better customer outcomes at a much lower cost. That meant changing the culture of the bank by transforming it from a bureaucracy into a federation of entrepreneurs: managers would be rewarded for taking charge of problems and deciding, quickly, how to fix them.

The story of change

First, the CEO developed these insights into a story that would make sense to all of the bank’s employees, top to bottom, and would persuade them to change their behavior in line with the new principles. His principal technique was dialogue-based planning, a refinement of double-loop learning (see sidebar, "People want to develop," for a different technique). First, he drafted a top-level story of the way he perceived the bank’s position and refined the story with the help of his executive directors. Each of them in turn developed a chapter of the story relevant to his or her direct reports; the human-resources director, for example, explained how she would improve the system for identifying potential highfliers and redraw their career paths so that they would spend less time in low-impact jobs. Every director assigned responsibility for each "deliverable" in the story to one member of his or her team. Each team member then had to develop a performance scorecard setting out what he or she would do differently to meet the new goals.

The directors and the CEO then met again to retell their chapters and to get feedback from one another. Each director shared the amended version with his or her subordinates, who in turn retold the relevant part of the story to their own direct reports, and so on down five levels of the organization to the branch managers. At each retelling, the emphasis was on making the story meaningful to the people listening to it and to the groups to which they belonged.

At every level, information flowed upstream as well as down. Part of the story told by the director of retail operations, for example, was the customers’ desire for faster banking processes. One thing slowing them down, according to the staff of the branches, was the document imagers, which broke down, on average, every three days. Ordering a new imager thus became a detail in each branch manager’s story, and the branch staff could translate the top-level story—"our customers want faster operations"—into a practical result that also made their lives easier. At each level of the organization, an employee heard the relevant version of the proposed changes from his or her immediate boss, the person widely regarded as the most effective communications channel.2

How could the CEO know that people really bought into his story? The secret, he felt, was to ensure that it described how life would be better for all of the bank’s stakeholders, not just investors and analysts.

Reinforcing systems

The most dramatic structural change at the bank was eliminating 20 percent of its managerial jobs. The hypothesis, later proved correct, was that doing so would remove a swath of useless activity, without any falloff in performance. All of the bank’s managerial jobs were terminated, and managers were invited to apply for the remaining 80 percent. Applicants knew that they had succeeded if they were invited to a dialogue-based planning session—another way of signaling the importance of the process. Unsuccessful candidates left the bank. The goal was not, primarily, to improve the bank’s cost-to-income ratio; on the contrary, the cost of laying them off was quite high. Rather, since fewer managers now had to make the same number of decisions, this move was intended to force the survivors to make them more quickly.

Simultaneously, the bank’s performance-management process was sharpened. Under the old system, managers were rated from 1 to 5 each year and remunerated accordingly. On average, 84 percent of them got a rating of 3 or more, though the performance of the bank was hardly as good as those results would imply. It injected reality into the process by introducing rankings within cohorts. To reveal the true relative performance of the bank’s employees, a manager assessing ten people, say, could rank no more than three as top performers and had to put at least one person in the lowest level. The ten directors evaluated the top 50 managers in meetings chaired by the CEO. The bonus for gaining the first rank was increased to 20 percent, from 10. Managers in the lowest rank, who would formerly have received a bonus of 5 percent, got none at all. Those who consistently ranked in the lowest level were asked to leave.

Skills for change

There was more drama to come. After four months of developing the new strategy with the ten directors, the CEO realized that only five of them were committed to change and equipped to see it through. To ensure that his bank had the right skills to change its practices and culture, he replaced the other five with new directors, three of them outsiders.

Meanwhile, the top 50 managers spent two days at a skill-development center where their leadership abilities—in coaching and decision making, for example—were assessed, and each drew up a personal plan to develop those talents. The company began to assess the performance of its people not just on whether they "made the numbers" but also on the leadership dimension. One manager who had consistently won high bonuses was known to be hell to work for, a fact acknowledged by the new measurement scheme: he was paid the lowest sum appropriate to his post. This news, which traveled fast on the grapevine, underlined the message that leadership really counted.

Consistent role models

Dialogue-based planning ensured that leaders at each level of the organization were "singing from the same song sheet." Their planning sessions were high-profile events where they themselves started modeling the new type of behavior that the bank wanted its staff to adopt. The CEO’s enthusiasm also inspired employees to behave differently. He convinced them that although change would take a long time and would be very hard to achieve, his passion for improving the life of everyone involved with the bank was heartfelt.

Both messages came through strongly in the way he reshaped his executive team. The five departing directors left just as the most disruptive changes were starting, and the work of the remaining five became even more intense during the six months it took to find replacements. It would have caused far less chaos to search for them while leaving the old team in place—and in the dark—but the CEO’s conscience told him not to do so. Besides showing other managers that there was nothing soft about the change program, his approach demonstrated his integrity and his respect for the needs of all of the bank’s people, even those he didn’t want to keep in the long term. In such a large-scale change in behavior, the leader’s character and integrity matter enormously.

The outcome

The bank, which is now two years into its four-year improvement timetable, is about halfway toward meeting its targets for reducing its cost-to-income ratio and increasing its revenue and economic profit. This achievement is a sure sign that behavior is heading in the intended direction throughout the bank. Does it prove that mind-sets too are changing? No numerical evidence is available, but from close observation we can see that the culture really has evolved. The bank isn’t a comfortable place to work, but the focus on performance is far stronger, functional silos are being broken down, and people treat every task with far more urgency. A small but indicative example: average queuing times in branches have dropped by over 30 percent, largely because branch managers can count on their employees to work a more flexible shift system by making the most of part-time work and temporary cover. The imagers are working as well.

It is neither easy nor straightforward to improve a company’s performance through a comprehensive program to change the behavior of employees by changing their mind-sets. No company should try to do so without first exhausting less disruptive alternatives for attaining the business outcome it desires. Sometimes tactical moves will be enough; sometimes new practices can be introduced without completely rethinking the corporate culture. But if the only way for a company to reach a higher plane of performance is to alter the way its people think and act, it will need to create the four conditions for achieving sustained change.

Explore a career with us