John P. Kotter on What Leaders Really Do: A Harvard Business Review Book

Read an Excerpt

Chapter One

Leadership at the Turn of the Century

For 30 years I have been studying the actions of those who run organizations, trying to record and clarify what they do, why they behave as they do, and what effect their choices have on other individuals and enterprises. I did not begin this work with an explicit focus on "leadership" in mind—indeed, the word can be found rarely if at all in my early writings. Nevertheless, it is toward leadership that my efforts have ultimately been directed, a fact that says something important about the evolving nature of what we call "managerial" work.

    After conducting fourteen formal studies and more than a thousand interviews, directly observing dozens of executives in action, and compiling innumerable surveys, I am completely convinced that most organizations today lack the leadership they need. And the shortfall is often large. I'm not talking about a deficit of 10% but of 200%, 400%, or more in positions up and down the hierarchy. This is not to say that untalented, unenergetic people occupy managerial positions. The typical case is just the opposite, with bright, experienced, and hardworking individuals, some quite extraordinary, almost all trying to do what they believe is right. The problem is that far too few of these people are providing the leadership that is increasingly needed in business, government, everywhere.

    Some say that there has been a dearth of quality leadership throughout history. This may or may not be true. What is clear is that the increasingly fast-moving and competitiveenvironment we will face in the twenty-first century demands more leadership from more people to make enterprises prosper. Without that leadership, organizations stagnate, lose their way, and eventually suffer the consequences.

    The central issue here is not one of style. I often hear people say that we need a "new leadership style" for the new century. In a globalizing world with a better-educated workforce that is no longer inclined to be seen and not heard, a new leadership style is in fact called for, but style is not the key leadership issue. Substance is. It is about core behavior on the job, not surface detail and tactics, a core that changes little over time, across different cultures, or in different industries.

    There are those who say that the consequences of inadequate leadership are minimal because so many factors inside and outside enterprises affect performance. While I would agree in general that many elements contribute to an organization's results, most of these factors can be influenced by good or bad leadership. When that influence points in the wrong direction, in no direction despite rapid change, or along a reasonable trajectory at inadequate speed, the consequences can be tragic. In extreme cases, enterprises fail, jobs are lost, customers and communities and financial interests are hurt, careers are derailed. In less extreme situations, firms underperform, essentially dealing with the same consequences as their ailing counterparts, but to a lesser degree. These failures show up in numbers that are sometimes easy to interpret but are often subtle, as when an enterprise turns out a performance similar to that of its competitors but actually should have done much better, given its assets and starting position.

    More dramatic than any numbers are individual cases of real human beings who suffer under tyrants or incompetents or even well-meaning bosses whose failure to lead helps bring down the ship. The pain, broadcast loudly or suffered silently, can be huge, as people lose their jobs to incompetent reengineering or strain under the pressure of propping up a shaky bottom line.

    Confront those at the very top of organizations with these facts and many will argue, or at least quietly disagree. "Yes, we could be doing better, but ..." "Competition forces us to do ..." "What do you expect in an industry where ..." This simple confrontational test often separates those executives who understand effective leadership from those who do not.

    The leadership gap exists for many reasons, and correcting the problem is made difficult for many additional reasons, not the least being the complexity of the issues. But the problem is not that there is only a limited number of people out there with leadership potential. Even if only one person in a hundred had any potential at all, we would have tens of millions of individuals around the globe providing leadership. This is far from the case today, a fact that should tell us something important about a failure on the part of educational and work organizations, and probably on the part of many families, too.

    Instead of nurturing talent, encouraging people to lead and to learn from mistakes and successes, organizations all too often ignore leadership potential, offer no relevant training or role models, and punish those who make small errors while trying to lead. Individuals, too, get in their own way by failing to assess their developmental needs realistically and to proactively seek means of meeting those needs.

    The confusion around some of these points occasionally strikes me as staggering. People say "leadership" but describe "management," talk only of a commanding style, serve up speeches about how more than one leader creates chaos, or talk in mystical terms. I have witnessed this cluttered thinking endless times in intelligent people. When capable individuals make such remarks, we have a clear indication of the need for a better understanding of what leaders really do.

TEN OBSERVATIONS

The body of my work on these subjects has been reported in more than a dozen books and a series of articles. The most important of these articles are six published in the Harvard Business Review between 1979 and 1997. These six pieces, along with my integrative commentary, form the mass of this book.

    The material collected here deals with the challenges inherent in what is called managerial work and with what differentiates effective from ineffective responses to these challenges. The articles in the first part of the book focus explicitly on leadership and change, that part of organizational activity which I think has become more and more important over the past few decades. Part 2 shows how managerial work today is less about wielding power than about coping with dependence, how managers are put into a far more complex web of interaction with influential others than any organization chart can suggest, and how from these observations important implications follow. The two parts connect through the notion, much clearer to me now than when I wrote the articles, that more change demands more leadership, which places managers in more complex webs of interaction.

    The evolution of these ideas began when I discovered that the most effective people in managerial jobs seemed to act in ways that defied stereotypes of managerial behavior. Later I decided that this "unusual" behavior was mostly related to leadership, which in turn was related to change, one of the central themes of our times.

    Reflecting on the six articles now, in 1998, I think hindsight and a bit of finagling put all the big ideas into a set of ten central, interrelated observations. (See Exhibit 1-1.) Each of these observations reflects important changes that continue to occur in the contexts in which managers work, changes driven by powerful forces associated with technology, the globalization of competition and markets, and workforce demographics. These forces have been destroying the mid-twentieth-century stability and pushing up the speed of so much, demanding from managers both incremental change and bigger leaps.

    In the mid-twentieth-century world of oligopolies, monopolies, and many barriers to global competition, the longer jumps were not as necessary. Industries changed more slowly, demanding in turn less organizational change. Incremental shifts were for the most part sufficient, perhaps accompanied by a leap every decade or two. In an increasing set of industries today, this behavior is insufficient and can lead to disaster. Thus the number of efforts to transform organizations has increased dramatically in the past few decades, efforts that go by the names of reengineering, restructuring, restrategizing, quality programs, cultural change, and mergers and acquisitions.

Observation #1

When managers today produce successful change of any significance in organizations, regardless of the specific approach, the process is time consuming and highly complex, never a one-two-three, hit-and-run affair.

    In the most successful change efforts, people move through eight complicated stages in which they (1) create a sense of urgency, (2) put together a strong enough team to direct the process, (3) create an appropriate vision, (4) communicate that new vision broadly, (5) empower employees to act on the vision, (6) produce sufficient short-term results to give their efforts credibility and to disempower the cynics, (7) build momentum and use that momentum to tackle the tougher change problems, and (8) anchor the new behavior in organizational culture.

    Each of these eight actions associated with successful leaps takes time. Step 1 alone, pushing up the level of urgency, may require many months in a complacent organization of any size. Formulating an appropriate vision and strategies might be done in weeks, but it often requires 12, even 24 months. Creating all the conditions necessary to make a new set of actions stick, not evaporate over time, can take years. These time frames are incomprehensible to short-term, reactive managers. Partially because they are not proactive—indeed, are always trying to catch up with the best competitors—they find the temptation to skip a step or two, or to run through the process too quickly, irresistible. So the reengineering, restructuring, etc., only pushes the organization incrementally, often at great cost.

    People do create the appearance of successful change both more quickly and more easily. The big acquisition is made, the structures and systems rationalized, and the process seems to be completed in five months. But appearances deceive. The biggest chore associated with an acquisition of any size is to merge the two (or perhaps more) different cultures. Five months into the process, cultural integration typically has barely begun. Yet if this part of the transformation is ignored or handled poorly, problems will surface for years, maybe decades. Two different ways of operating and two different teams will clash in subtle and obvious ways, diverting attention from the real business of the enterprise.

Observation #2

Although change generally involves this complex eight-step process, regardless of the setting, some essential actions taken by effective managers with transformational goals always vary from case to case to fit key contingencies in their situations.

    Differences appropriately arise around the degree to which the vision should be set in stone early on, to which large groups of people should be involved in determining and / or implementing the vision, and to which resistance from various quarters should be heeded.

    Many factors affect the choices that need to be made in any specific situation, including the amount of resistance anticipated (the more resistance, the harder it is to push through it); the stakes involved (the bigger the stakes, the more important it is that the vision be right, even if getting it right takes more time and involves many revisions); and the extent to which lower levels in the organization are needed to construct or implement the vision (more dependence on lower levels means allowing them more involvement and participation).

    People often get into trouble when they try to apply the tactics that worked in their last change experience without considering how the new situation is different. If, for example, an approach heavy on training has succeeded again and again in helping to empower employees, managers are likely to try it again, even if the central problem is resistance from supervisors and managers, not uneducated subordinates. Or if an approach based on maximum speed has worked well historically, executives continue to charge ahead even though more time is needed to deal with a much more powerful group of resisting employees. In general, the longer a set of tactics has worked, the more oblivious we become to new contingencies. The better a hammer has served in the past, the more all new problems look like nails.

Observation #3

For a number of reasons, even very capable and well-intentioned people can make a predictable set of mistakes when they are attempting significant, nonincremental change.

    Failure is associated not only with the untalented. The very capable sometimes allow too much complacency up front. They put together too weak a guiding coalition. They fail to create enough vision, or undercommunicate the vision, or fail to remove sufficient obstacles to change. Their plan to create short-term, credibility-building results is insufficient, they declare victory too soon once initial results are in. They fail to sufficiently connect new approaches to the culture or to create new cultures that can support these approaches. More tactically, they may educate when they need to apply pressure, negotiate over details when better communication is needed, manipulate when being supportive would work much better.

    If we all had more experience dealing with major change these errors would occur much less often. But too many people have been trained for and raised in a more stable world, a world that, for the most part, no longer exists. Too many people have been trained only to manage the current system or to make incremental shifts. They have not been shown how to provide the leadership necessary to make bigger leaps.

Observation #4

The issue of leadership is centrally important here because leadership is different from management, and the primary force behind successful change is the former, not the latter. Without sufficient leadership, the probability of mistakes increases greatly and the probability of success decreases accordingly.

    Here I'm talking about leadership as the development of vision and strategies, the alignment of relevant people behind those strategies, and the empowerment of individuals to make the vision happen, despite obstacles. This stands in contrast with management, which involves keeping the current system operating through planning, budgeting, organizing, staffing, controlling, and problem solving. Leadership works through people and culture. It's soft and hot. Management works through hierarchy and systems. It's harder and cooler.

    This distinction between management and leadership is neither arbitrary nor semantic. It is, instead, enormously important and a source of great confusion. The person who thinks management is leadership will manage change, hence keeping it under control, but he or she will be unable to provide the stuff required to make larger and more difficult leaps.

    Executives who do not lead are almost programmed to fall into all sorts of traps. They rarely push urgency high enough. They underestimate the need for a strong coalition to guide the change. They create only plans and budgets, not the visions and strategies needed to accomplish the vision. They undercommunicate almost any new direction. They fail to eliminate sufficient obstacles so that employees can act on the vision. They declare victory too soon. They do not institutionalize new approaches in organizational culture.

    The point here is not that leadership is good and management is bad. They are simply different and serve different purposes. The fundamental purpose of management is to keep the current system functioning. The fundamental purpose of leadership is to produce useful change, especially nonincremental change. It is possible to have too much or too little of either. Strong leadership with no management risks chaos; the organization might walk right off a cliff. Strong management with no leadership tends to entrench an organization in deadly bureaucracy.

Observation #5

Leadership is a growing part of managerial work because the rate of change has been growing.

    A few decades ago a product life cycle might have been 15 years, while today it is 4 for the very same type of product. Thirty years ago, the response time to a customer complaint may have been in days, while now it must be hours. Forty years ago organizations replaced computers within a decade or two, not every 36 months. As a result, effective top executives now might spend up to 80% of their time leading, up from 40% not that long ago. Even those at the bottom of the management hierarchy might spend at least 20% of their time on leadership, and in dynamic industries, even more.

    This shift in the nature of "managerial" work is of fundamental importance because leadership and management are so different. It is as if the work needed has suddenly shifted from accounting to accounting and piano playing. Or, for those who play piano at a small restaurant as a second job, it's as if they've been asked to take as their main job membership in a philharmonic orchestra while doing accounting on the side. Under any circumstances, such a shift would be difficult for most people to make and impossible for some. But confusion over what is accounting and what is piano playing would significantly confound the problem.

    In a similar way, the confusion over what is leadership and what is management significantly confounds the problem in enterprises today. So people are asked to lead, yet they manage more intensely, becoming frustrated when their actions are not rewarded by superiors or customers. Managers and leaders are not put into teams—one solution to the problem—because the categories are not clear. So managers are teamed with managers. Leaders—what few there are—continue to be seen as potentially dangerous renegades.

    Because those helping to run organizations are now being pushed to both manage and lead, the complexity of their position has increased, requiring some new vocabulary.

Observation #6

Increasingly, those in managerial jobs can be usefully thought of as people who create agendas with both plans and budgets (the management part) and visions and strategies (the leadership part), as people who develop implementation networks both through hierarchy (management) and a complex web of aligned relationships (leadership), and who execute both through controls (management) and inspiration (leadership).

    The proportion of planning versus creating vision, organizing the hierarchy versus aligning the web of relations, controlling versus inspiring, varies with the inclinations of the individual and the demands of the job. Generally, the higher a person goes in an organization, the more he or she needs to be engaged in the work of leadership.

    The broad concept of agendas for action and networks of relationships can be useful when we are talking about managerial behavior because they cut across the categories of leadership and management, thereby providing a degree of unification. Pure managers create agendas and develop networks, as do pure leaders, although the two groups do so in very different ways. But today there are few situations that call for only one or the other; there is usually a need for both types of actions.

    The temptation on the part of leadership enthusiasts is to abandon the older vocabulary. But even in jobs demanding much leadership there is typically a managerial component. Agendas and networks can help us evolve from talking only about planning and controls, to talking only about vision and inspiration, to making use of the best of both sets of ideas.

Observation #7

Because management tends to work through formal hierarchy and leadership does not, as change demands more leadership in organizations, managerial jobs are placing people in ever more complex webs of relationships.

    In a more stable world, in which management is the prime activity, jobs of consequence operate through hierarchy. So people look down to their subordinates and up to a boss, as shown on the company organization chart. In a world that is continually changing, where additional leadership is necessary, more individuals outside one's chain of command take on added importance, as do intangibles not on the organization chart, intangibles like corporate culture.

    Because a new product strategy may require new information systems, new performance appraisals, new jobs, new attitudes, and more, executives launching the change are forced to concern themselves with many more people than they would in simply pursuing the existing product strategy. Managers fail if they focus only on their subordinates, ignoring for the most part the human resources staff, the information technology (IT) staff, and the layers of other managers who are trying to implement their own visions of change.

Observation #8

Because managerial work is increasingly a leadership task, and because leaders operate through a complex web of dependent relationships, managerial work is increasingly becoming a game of informal dependence on others instead of just formal power over others.

    The wily employee near the bottom of the hierarchy who is seemingly either irrelevant or impotent can make life difficult (or easier) for the "important" manager through any number of strategies. The prototype is Radar in the movie and TV series M.A.S.H., a lowly soldier who could stop a colonel in his tracks. In a world that creates many Radars, an increasing part of managerial work involves actively dealing with dependence on others who are above or below in the hierarchy, peers inside the organization, and even people outside.

    The dependence issue is not entirely new. It has been discussed at least since the 1930s, but for the most part quietly and indirectly, at least in serious circles. Less quiet have been tomes on authority, responsibility, span of control, organizational structure, etc., pieces that can be most useful when done well, but which tend to put the spotlight on formal authority, hierarchy, and management. One of my first significant insights as an assistant professor was that because managerial positions make people increasingly dependent on others, including others outside their chain of command, a focus on the dependencies was superior to a traditional emphasis on only formal powers. This idea, which will seem obvious to some, was nevertheless counterintuitive to all who thought of superiors first and foremost as the ones with the power.

    Some of my MBA students, especially those with little experience in managerial jobs, find the notion of dependence on others more than confusing. For them managerial work is attractive in the first place because of its attendant perceived power and control. The idea that their prospective careers might push them into jobs of dependence, not just power, is unnerving.

Observation #9

When you start to think in terms of networks and dependence and leadership, not just hierarchy and formal authority and management, all sorts of interesting implications follow.

    Giving orders becomes a less important part of the job. Developing good working relationships with people in the network becomes a bigger part of the challenge. Having focus beyond direct subordinates is obviously necessary. Actively managing relations with the boss is a necessity for the good of the enterprise.

    This last point struck many as odd when I wrote about it in 1979. With a managerial mindset that looks mostly down the hierarchy, down the chain of command, the concept of managing the boss (MTB) makes little sense outside of a political context. Yet good managers MTB when they have poor bosses to get those superiors to do their managerial tasks well. Also, good leaders always pull everyone relevant along, and the boss is always relevant.

    Managing one's superior well, as a part of good management and leadership, especially the latter, means understanding the boss and his or her context, assessing yourself and your needs, and developing and maintaining a relationship that fits both needs and styles, a relationship that is characterized by mutual expectations, that keeps the boss informed, that is based on dependability and honesty, and that makes selective use of the boss's time and resources. This behavior will help you do routine and transformational work, in the latter case helping to avoid the suggestion of being a cowboy. Most organizations put cowboys out to pasture.

Observation #10

What a manager / leader does on a minute-by-minute, hour-by-hour basis rarely jibes with any stereotype of a manager, a heroic leader, or an executive, a fact that can create considerable confusion for those new to managerial jobs. This behavior is nevertheless understandable if one takes into consideration the diverse tasks (leadership and management), the difficult work (maintenance and change), and the complex web of relationships (beyond formal hierarchy) that come with the manager's territory.

    A close examination of the day-in, day-out actions and responsibilities of a "manager" or "leader" will produce a picture that doesn't resemble anything like the "able manager" or "visionary leader" of our dreams. In "real life," effective executives spend a lot of time just talking to other people, including people who are not their subordinates. They deal in a broad sweep of topics rather than just their functional specialty, are much more likely to ask questions than to give orders, and actually make "big" policy decisions only rarely. They engage in the kind of chitchat and joking that cements relationships. And despite the fact that they do all of this in a very casual, often disjointed way, they are remarkably efficient, accomplishing diverse tasks (that is, the tasks of leadership and of management) in short periods of time.

    If you carefully study effective executives, you can see the logic of these methods, but the logic is subtle. Because these tactics have been little studied or taught, relative to other managerial or business subjects, even executives themselves have difficulty describing what they do and why they do it.

(Continues...)