Skip main navigation

Insights & Publications

Article|McKinsey Quarterly

The irrational side of change management

Most change programs fail, but the odds of success can be greatly improved by taking into account these counterintuitive insights about how employees interpret their environment and choose to act.

April 2009 | byCarolyn Aiken and Scott Keller

In 1996, John Kotter published Leading Change. Considered by many to be the seminal work in the field of change management, Kotter’s research revealed that only 30 percent of change programs succeed. Since the book’s release, literally thousands of books and journal articles have been published on the topic, and courses dedicated to managing change are now part of many major MBA programs. Yet in 2008, a McKinsey survey of 3,199 executives around the world found, as Kotter did, that only one transformation in three succeeds. Other studies over the past ten years reveal remarkably similar results. It seems that, despite prolific output, the field of change management hasn’t led to more successful change programs.

It also hasn’t helped that most academics and practitioners now agree on the building blocks for influencing employee attitudes and management behavior. McKinsey’s Emily Lawson and Colin Price provided a holistic perspective in “The psychology of change management,”1 which suggests that four basic conditions are necessary before employees will change their behavior: a) a compelling story, because employees must see the point of the change and agree with it; b) role modeling, because they must also see the CEO and colleagues they admire behaving in the new way; c) reinforcing mechanisms, because systems, processes, and incentives must be in line with the new behavior; and d) capability building, because employees must have the skills required to make the desired changes.

This prescription is well grounded in the field of psychology and is entirely rational. One of its merits is its intuitive appeal: many managers feel that, once revealed, it is simply good common sense. And this, we believe, is precisely where things go wrong. The prescription is right, but rational managers who attempt to put the four conditions in place by applying “common sense” typically misdirect time and energy, create messages that miss the mark, and experience frustrating unintended consequences from their efforts to influence change. Why? Because when they implement the prescription, they disregard certain, sometimes irrational—but predictable—elements of human nature.

In our research and by working with companies attempting change, we have identified nine insights into how human nature gets in the way of successfully applying the four conditions required for behavioral change. As we describe these insights, we’ll show how various companies have, either by conscious awareness or simple luck, overcome or leveraged counterintuitive sides of human behavior in making change happen.

Creating a compelling story

Change-management thinking extols the virtues of creating a compelling change story, communicating it to employees, and following it up with ongoing communications and involvement. This is good advice, but in practice there are three pitfalls to achieving the desired impact.

1. What motivates you doesn’t motivate most of your employees. We see two types of change stories consistently told in organizations. The first is the “good to great” story: something along the lines of, “Our historical advantage has been eroded by intense competition and changing customer needs; if we change, we can regain our leadership position.” The second is the turnaround story: “We’re performing below industry standard and must change dramatically to survive. We can become a top-quartile performer in our industry by exploiting our current assets and earning the right to grow.”

These stories both seem intuitively rational, yet they too often fail to have the impact that change leaders desire. Research by a number of leading thinkers in the social sciences, such as Danah Zohar, has shown that when managers and employees are asked what motivates them the most in their work they are equally split among five forms of impact—impact on society (for instance, building the community and stewarding resources), impact on the customer (for example, providing superior service), impact on the company and its shareholders, impact on the working team (for example, creating a caring environment), and impact on “me” personally (my development, paycheck, and bonus).

This finding has profound implications for leaders. What the leader cares about (and typically bases at least 80 percent of his or her message to others on) does not tap into roughly 80 percent of the workforce’s primary motivators for putting extra energy into the change program. Change leaders need to be able to tell a change story that covers all five things that motivate employees. In doing so, they can unleash tremendous amounts of energy that would otherwise remain latent in the organization.

Consider a cost reduction program at a large US financial-services company. The program started with a change story that ticked the conventional boxes related to the company’s competitive position and future. Three months into the program, management was frustrated with employee resistance. The change team worked together to recast the story to include an element related to society (to deliver affordable housing, for example), customers (fewer errors, more competitive prices), the company (expenses are growing faster than revenues, which is not sustainable), working teams (less duplication, more delegation), and individuals (more attractive jobs).

This relatively simple shift in approach lifted employee motivation measures from 35.4 percent to 57.1 percent in a month, and the program went on to achieve 10 percent efficiency improvements in the first year—a run rate far above initial expectations.

2. You’re better off letting them write their own story. Well-intentioned leaders invest significant time in communicating their change story. Road shows, town halls, and Web sites are but a few of the many approaches typically used. Certainly the story (told in five ways) needs to get out there, but the insight we are offering is that much of the energy invested in communicating it would be better spent listening, not telling.

In a famous behavioral experiment, half the participants are randomly assigned a lottery ticket number while the others are asked to write down any number they would like on a blank ticket. Just before drawing the winning number, the researchers offer to buy back the tickets from their holders. The result: no matter what geography or demographic environment the experiment has taken place in, researchers have always found that they have to pay at least five times more to those who came up with their own number.

This reveals something about human nature: when we choose for ourselves, we are far more committed to the outcome (almost by a factor of five to one). Conventional approaches to change management underestimate this impact. The rational thinker sees it as a waste of time to let others discover for themselves what he or she already knows—why not just tell them and be done with it? Unfortunately this approach steals from others the energy needed to drive change that comes through a sense of ownership of the answer.

At BP, to develop a comprehensive training program for frontline leaders, a decision was made to involve every key constituency in the design of the program, giving them a sense of “writing their own lottery ticket.” It took a year and a half to complete the design using this model but was well worth it: now in implementation, the program is the highest rated of its kind at BP. More than 250 active senior managers from across the business willingly teach the course, and, most important, managers who have been through the training program are consistently ranked higher in perfor-mance than those who haven’t, both by their bosses and by the employees who report to them.

3. It takes a story with both + and – to create real energy. The “deficit based” approach—which identifies the problem, analyzes what’s wrong and how to fix it, plans, and then takes action—has become the model predominantly taught in business schools and is presumably the default change model in most organizations. Research has shown, however, that a story focused on what’s wrong invokes blame and creates fatigue and resistance, doing little to engage people’s passion and experience.

This has led to the rise of the “constructionist based” approach to change, where the change process is based on discovery (discovering the best of what is), dreaming (imagining what might be), designing (talking about what should be), and destiny (creating what will be). The problem with this approach is that an overemphasis on the positive can lead to watered-down aspirations and impact. The reason is that, as humans, we are more willing to take risks to avoid losing what we’ve got than we are to gain something more. Some anxiety is useful when it comes to spurring behavioral change.

We believe the field of change management has drawn an artificial divide between deficit-based and constructionist-based approaches and stories. While it is impossible to prescribe generally how the divide should be split between positive and negative messages (as it will be specific to the context of any given change program), we strongly advise managers not to swing the pendulum too far in one direction or another. Consider Jack Welch, former CEO at GE, who took questions of “what’s wrong here?” (poorly performing businesses, silo-driven behavior, and so forth) head-on, as well as “imagining what might be” (number one or two in every business, openness, and accountability).

Role modeling

Conventional change management suggests leaders should take actions that role model the desired change and mobilize a group of influence leaders to drive change deep into the organization. Unfortunately, this does not necessarily deliver the desired impact.

4. Leaders believe mistakenly that they already “are the change.” Most senior executives understand and generally buy into Gandhi’s famous aphorism, “Be the change you want to see in the world.” They commit themselves to personally role modeling the desired behaviors. And then, in practice, nothing significant changes.

The reason for this is that most executives don’t count themselves among the ones who need to change. How many executives when asked privately will say no to the question, “Are you customer focused?” and yes to the question “Are you a bureaucrat?” Of course, none. The fact is that human beings consistently think they are better than they are—a phenomenon referred to in psychology as a self-serving bias. Consider that 94 percent of men rank themselves in the top half according to male athletic ability. Whereas conventional change-management approaches surmise that top team role modeling is a matter of will or skill, the truth is that the real bottleneck to role modeling is knowing what to change at a personal level.

Typically, insight into what to change can be created by concrete 360-degree feedback techniques, either via surveys, conversations, or both. Look at Amgen CEO Kevin Sharer’s approach of asking each of his top 75, “What should I do differently?” and then sharing his development needs and commitment publicly with them. Consider the top team of a national insurance company who routinely employed what they called the circle of fire during their change program: every participant receives feedback live—directly from their colleagues—in relation to being the change, such as “What makes you great?” and “What holds you back?”

5. “Influence leaders” aren’t a panacea for making change happen. Almost all change-management literature places importance on identifying and mobilizing those in the organization who either by role or personality (or both) have disproportionate influence over how others think and behave. We believe this is sound and timeless advice. However, we have observed that the role of influence leaders has gradually shifted—from being perceived as a helpful element of a broader set of interventions, to a panacea for making change happen.

Our experiences working with change programs suggest that success depends less on how persuasive a few selected leaders are and more on how receptive the “society” is to the idea. In practice it is often unexpected members of the rank and file who feel compelled to step up and make a difference in driving change. That’s why we warn against overinvesting in influence leaders and advocate that change leaders’ attention should be balanced across the right application of all four conditions for change, to ensure they reinforce each other in ways that maximize the probability of the change spark taking off like wildfire across the organization.

Reinforcing mechanisms

Conventional change management emphasizes the importance of reinforcing and embedding desired changes in structures, processes, systems, target setting, and incentives. We agree. To be effective, however, these mechanisms must take into account that people don’t always behave rationally.

6. Money is the most expensive way to motivate people. Companies that try to link the objectives of change programs to the compensation of staff find that it rarely enhances their motivation for change to the extent desired. The reason for this is as practical as it is psychological in nature. The reality is that in the vast majority of companies, it is exceedingly difficult to incorporate a meaningful link to the change program within compensation systems that are based on a vast array of metrics. Moreover, many studies have found that for human beings satisfaction equals perception minus expectation (an equation often accompanied by the commentary, “reality has nothing to do with it”).

The beauty of this equation for change managers is that small, unexpected rewards can have disproportionate effects on employees’ satisfaction with a change program. Gordon M. Bethune, while turning around Continental Airlines, sent an unexpected $65 check to every employee when Continental made it to the top five for on-time airlines. John McFarlane, former CEO of ANZ Bank, sent a bottle of champagne to every employee for Christmas with a card thanking them for their work on the company’s “Perform, Grow, and Break-out” change program. Most change managers would refer to these as merely token gestures and argue that their impact is limited and shortlived. Employees on the receiving end beg to differ. Indeed, they consistently report back that the rewards have a disproportionately positive impact on change motivation that lasts for months, if not years.

7. The process and the outcome have got to be fair. Employees will go against their own self-interest if the situation violates other notions they have about fairness and justice. Consider a bank, which, as part of a major change program, created new risk-adjusted return on capital (RAROC) models and delivered the resulting new pricing schedules to the front line along with new and appropriate sales incentives. The result: customer attrition (not only of the unprofitable ones) and price overrides went through the roof and significant value was destroyed by the effort. What went wrong? Because the frontline bankers perceived the changes as unfair to the customer, a significant number of them vocally bad-mouthed the bank’s policies to customers and used price overrides to show their good faith, even though it meant they were less likely to achieve individual sales goals.

In making any changes to company structures, processes, systems, and incentives, change managers should pay what might strike them as an unreasonable amount of attention to employees’ sense of the fairness of the change process and its intended outcome. Particular care should be taken where changes affect how employees interact with one another (such as head count reductions and talent-management processes) and with customers (sales stimulation programs, call center redesigns, and pricing). Ironically, in the pricing example described above, the outcome was inherently fair (customers are being asked to pay commensurate to the risk the bank is taking on), and therefore the downward spiral described could have been avoided (and has been by other banks adopting RAROC-based pricing) by carefully tending to employees’ perceptions of fairness in the communications and training surrounding the changes.

Capability building

Change-management literature emphasizes the importance of building the skills and talent needed for the desired change. Though hard to argue with, in practice there are two insights that demand attention in order to succeed.

8. Employees are what they think, feel, and believe in. As managers attempt to drive performance by changing the way employees behave, they all too often neglect the thoughts, feelings, and beliefs that, in turn, drive behavior. Consider a bank that through a benchmarking exercise discovered that its sales per banker were lagging behind those of the competition. After finding that bankers spent too little time with customers and too much time on paperwork, the bank set about reengineering the loan-origination process in order to maximize customer-facing time. Unfortunately, six months later, the levels of improvement were far lower than envisioned.

A further investigation, with an eye to the bankers’ mind-sets rather than their behaviors, revealed that they simply found customer interactions uncomfortable and therefore preferred paperwork. This feeling was driven by a combination of introverted personalities, poor interpersonal skills, and a feeling of inferiority when dealing with customers who (by and large) have more money and education than the bankers do. Finally, most bankers were loath to think of themselves as salespeople—a notion they perceived as better suited to employees on used-car lots than in bank branches.

Armed with these root-cause insights, training for bankers was expanded to include elements related to personality types, emotional intelligence, and vocational identity (recasting “sales” as the more noble pursuit of “helping customers discover and fulfill their unarticulated needs”). This enhancement not only put the program back on track within six months but also ultimately delivered sustainable sales lifts in excess of original targets.

9. Good intentions aren’t enough. Good skill-building programs usually take into account that people learn better by doing than by listening. These programs are replete with interactive simulations and role plays, and commitments are made by participants regarding what they will “practice” back in the workplace. But come Monday morning, very few keep their commitments.

This lack of follow-through is usually not due to ill intent: it is because nothing formal has been done to lower the barriers to practicing new skills. The time and energy required to do something additional, or even to do something in a new way, simply don’t exist in the busy day-to-day schedules of most employees. This failure to create the space for practice back in the workplace dooms most training programs to deliver returns that are far below their potential.

We advocate a number of enhancements to traditional training approaches in order to hardwire day-to-day practice into capability-building processes. First, training should not be a one-off event. Instead, a “field and forum” approach should be taken, in which classroom training is spread over a series of learning forums and fieldwork is assigned in between. Second, we suggest creating fieldwork assignments that link directly to the day jobs of participants, requiring them to put into practice new mind-sets and skills in ways that are hardwired into their responsibilities. These assignments should have quantifiable, outcome-based measures that indicate levels of competence gained and certification that recognizes and rewards the skills attained.

In the same way that the field of economics has been transformed by an understanding of uniquely human social, cognitive, and emotional biases, so too is the practice of change management in need of a transformation through an improved understanding of how humans interpret their environment and choose to act. While sustained impact can be measured only over numbers of years, our early results when applying these insights give us the confidence to broadly share our thinking.

About the authors

Carolyn Aiken is a principal in McKinsey’s Toronto office, and Scott Keller is a principal in the Chicago office.

McKinsey Quarterly

McKinsey Quarterly 50th Anniversary art

The Quarterly has been shaping the senior-management agenda since 1964.more

About this content

The material on this page draws on the research and experience of McKinsey consultants and other sources. To learn more about our expertise, please visit the Organization Practice.