by Clayton M. Christensen, Matt Marx, and Howard H. Stevenson YEL MAG CYAN BLACK Managers can use a variety of carrots and sticks to encourage people to work together and accomplish change. Their ability to get results depends on selecting tools that match the circumstances they face. the primary task of management is to get people to work together in a systematic way. Like orchestra conductors, managers direct the talents and actions of various players to produce a desired result. It's a complicated job, and it becomes much more so when managers are trying to get people to change, rather than continue with the status quo. Even the best CEOs can stumble in their attempts to encourage people to work together toward a new corporate goal. In 1999, for example, Procter & Gamble's Durk Jager, a highly regarded insider who had recently been promoted to CEO, announced Organization 2005, a restructuring JIM FRAZIER T october 2006 73 T h e To o l s o f Co o p e rat i o n a n d C h a n g e Over our many years observing management successes and failures up close, we've found that the rst step in any change initiative must be to assess the level of agreement in the organization along two critical dimensions. The rst is the extent to which people agree on what they want: the results they seek from their participation in the enterprise; their values and priorities; and which tradeoffs they are willing to make in order to achieve those results. Employees at Microsoft, for instance, have historically been united around a common goal: to dominate the desktop. While of course there will always be pockets of employees who are an exception, this theme has dened the company's culture. The second dimension is the extent to which people agree on cause and effect: which actions will lead to the desired outcome. When people have a shared understanding of cause and effect, they will probably agree about which processes to adopt- an alignment that was clearly absent at P&G as Jager attempted to transform the company. The exhibit \"The Agreement Matrix\" depicts these dimensions. The vertical axis shows agreement by an organization's members on what they want; the horizontal axis shows their agreement on cause and effect. Employees in organizations in the upper-left quadrant share hopes for what they will gain from being part of the orga- Leaders who want to move their organizations in a new direction must rst understand the degree to which employees agree on two dimensions: what they want out of working at the company and cause and effect, or how to achieve what they want. A high level of agreement on both dimensions, such as exists at Apple Computer, requires a completely different set of change tools than leaders will need in, for instance, low- Broad consensus agreement environments. Extent to which people agree on what they want Assessing the Existing Level of Agreement The Agreement Matrix No consensus program that promised to change P&G's culture. However, not everyone at P&G agreed that such sweeping change was necessary or that the way to achieve it was to reduce investments in the company's core brands in order to fund radical, new products. The organization rebelled, and Jager was forced to resign only 17 months after taking the helm. The root cause of Jager's very public failure was that he didn't induce P&G employees to cooperate-a requirement of all change campaigns. To achieve such cooperation, managers have a wide variety of tools at their disposal, such as nancial incentives, motivational speeches, training programs, and outright threats. But although most competent managers have a good grasp of what cooperation tools are available, we've observed that they may be less sure about which to use. The effectiveness of a given tool depends on the organization's situation. In this article, which employs some ideas from Do Lunch or Be Lunch, by Howard Stevenson and Jeffrey Cruikshank, we explain how to choose the right tools and offer advice for managers contemplating change. Microsoft in 1995 Apple Computer Balkan Peninsula Companies employing independent contractors No consensus Broad consensus Extent to which people agree on cause and effect nization, even though each might have a different view of what actions will be required to fulll those hopes. Microsoft found itself in this situation in 1995, when Netscape was threatening to become the primary \"window\" through which people would use their computers. Everyone in the company wanted the same thing - to preserve Microsoft's domination of the desktop - but initially there was little consensus about how to do that. Many companies that employ independent contractors and unionized workers, in contrast, are in the lower-right corner. These employees may have little passion for the goals of the company but are willing to follow prescribed procedures if they agree that those actions will produce the needed results. In the upper-right quadrant are companies whose employees agree on what they want and how to get there. Clear consensus on both dimensions makes these organizations' cultures highly resistant to change: People are generally satised with what they get out of working in Clayton M. Christensen (cchristensen@hbs.edu) is the Robert and Jane Cizik Professor of Business Administration at Harvard Business School in Boston. Matt Marx (mmarx@hbs.edu) is a doctoral student at Harvard Business School. Howard H. Stevenson (hstevenson@hbs.edu) is the Sarom-Rock Professor of Business Administration at Harvard Business School and the vice provost for Harvard University Resources and Planning. He is also the chairman of the board for Harvard Business School Publishing. 74 harvard business review | hbr.org T h e To o l s o f Co o p e rat i o n a n d C h a n g e Moving from Agreement to Cooperation The tools of cooperation can be grouped into four major categories: power, management, leadership, and culture. In the exhibit \"The Four Types of Cooperation Tools,\" we've matched each category with a quadrant of the agreement matrix. While the boundaries are not rigid, the broad labels can give managers a sense of which tools are likely to be effective in various situations. Power tools. When members of an organization share little consensus on either dimension of agreement, the only tools that will elicit cooperation are \"power tools\" such as at, force, coercion, and threats. Marshal Josip Broz Tito, the leader of Yugoslavia during most of the Cold War, wielded power tools effectively. He herded the disparate and antagonistic ethnic groups of the CEO of JPMorgan Chase, used these tools during the bank's integration with his previous company, Bank One. Convinced that pay had gotten out of control (the head of HR at Bank One was paid more than $5 million), Dimon met with executives individually to tell them they were vastly overpaid and slashed hundreds of salaries by 20% to 50%. He drove a replacement of the rm's myriad IT systems with a single platform, threatening to make all the decisions himself if the IT staff didn't reach any decisions in six weeks. He yanked hundreds of unvisited small-to-midsize businesses from the investment bank's \"prospects\" list so that the commercial bank could have the chance to work with them. Dimon also recongured control systems so that retail branch managers, who had received modest bonuses for meeting sales quotas on mortgages and other products, now stood to lose their jobs for missing quotas. We have included three tools in the exhibit - negotiation, strategic planning, and nancial incentives-to make a point. These tools will work only when there is a modicum of agreement on both dimensions of the matrix. In environments of antagonistic disagreement - whether in the Middle East or in the infamous clashes between Eastern Air Lines' management and its machinist union - negotiation generally doesn't work. A leader might use strategic planning to gure out where the organization ought to go next, but in the absence of the requisite degree of agreement on both dimensions, the strategic plan itself won't elicit the cooperative behavior required to get there. And using nancial incentives - essentially paying employees to want what management wants - may backre in an environment of low consensus. Consider, for example, the world of K-12 public education, which is decidedly in the lower-left quadrant of the agreement matrix. A WISE MANAGER IN A LOW-CONSENSUS ENVIRONMENT would not agree to lead a change program without the authority to wield the right power tools. Balkan Peninsula into a more or less articial nation and said, in effect,\"I don't care whether you agree with me or with one another about what you want out of life or about how to get it. What I want is for you to look down this gun barrel and cooperate.\" His approach worked, and the Balkan nations lived in relative peace for several decades. This is not to suggest, of course, that managers bring rearms to the ofce. But when organizational factions can't agree on what they want or what to do, power tools are the only ones that work. Jamie Dimon, currently the october 2006 Teachers, taxpayers, administrators, parents, students, and politicians have divergent priorities and disagree strongly about how to improve. Most pay-for-performance schemes have failed miserably in producing enduring change in schools, because nancial incentives are a tool that just won't work in this situation. Power tools can be extremely effective in low-agreement situations. The key is having the authority to use them. Managers sometimes nd themselves in balkanized circumstances without the power to wield the only tools 75 YEL MAG CYAN BLACK the organization and agree strongly about how to maintain that status quo. The nal scenario is the lower-left quadrant of the agreement matrix, where participants do not agree either on what they want or on how the world works. The perpetually warring nation-states of the Balkan Peninsula exemplify this lack of agreement. We will return to each situation in the following pages. It's important to note that there is no \"best\" position for managers to aspire to in the agreement matrix. To choose the right tools for fostering cooperation among employees, however, managers must assess where their organization lies. The tools that will induce employees in one quadrant to cooperate with a change program may well misre with employees in a different quadrant. In fact, in any given situation, most tools for eliciting cooperation will not work. T h e To o l s o f Co o p e rat i o n a n d C h a n g e that will induce cooperation under those conditions. If managers are asked to lead a matrixed or \"lightweight\" project team whose members' loyalties are in conict with the objectives of the project, for instance, the road to success will be tortuous. Just as a carpenter would never undertake a job without having the requisite tools in his or her toolbox, a wise manager in a low-consensus environment would not agree to lead a change program without the authority to wield the right power tools. Management tools. The tools of cooperation that drive change in the lower-right quadrant of the agreement matrix focus on coordination and processes. These \"management tools\" include training, standard operating procedures, and measurement systems. For such tools to work, group members need to agree on cause and effect but not necessarily on what they want from their participation in the organization. For example, in many companies the reasons unionized manufacturing workers come to work are very different from the reasons senior marketing managers do. But if both groups agree that certain manufacturing procedures will result in products with targeted levels of quality and cost, they will cooperate to follow those procedures. Measurement systems can also elicit cooperation in such situations. During Intel's rst two decades, grossmargin-per-wafer-start was the widely agreed-upon metric for protability. In the 1980s, the company's DRAM products, which had enjoyed high gross margins in the 1970s, were withering under Japanese competition. Focused on the accepted metric - and even without an explicit execu- The Four Types of Cooperation Tools When people in an organization disagree on what they want and on how to achieve desired results, the only tools that induce cooperation are \"power tools,\" which are essentially variations on coercion and at. If people want the same thing but disagree on how to achieve it,\"leadership tools\" such as role modeling and charisma can move them toward a consensus. If people agree strongly on cause and effect but little on what they want, leaders can employ \"management tools\" such as training and measurement systems. Companies where employees agree on both dimensions of the matrix, and so are generally happy with the status quo, have very strong cultures that are difcult to change. In such circumstances, it is possible only to tweak direction, using such \"culture tools\" as rituals and folklore. Managers do have other tools at their disposal - such as negotiation and nancial incentives - but these will work only when there is a certain level of Broad consensus Charisma Salesmanship Folklore Leadership Tools Rituals Culture Tools Role modeling Tradition Negotiation Role denition Power Tools Fiat Coercion Threats Religion Vision No consensus Extent to which people agree on what they want agreement on both dimensions of the matrix. Democracy Apprenticeship Strategic planning Financial Transfer pricing incentives Management Hiring and promotion Tools Measurement systems Control systems Training Standard operating procedures No consensus Broad consensus Extent to which people agree on cause and effect 76 harvard business review | hbr.org tive mandate-middle managers in disparate parts of the organization cooperated to shift manufacturing emphasis from DRAMs to microprocessors, which had become higher-margin products. Leadership tools. The tools useful in the upper-left quadrant of the agreement matrix tend to be results oriented rather than process oriented. Such \"leadership tools\" can elicit cooperation as long as there is a high level of consensus that a change is consistent with the reason employees have chosen to work in the enterprise-even if consensus is low on how to achieve the change. Charismatic leaders respected by employees, for example, often do not address how to get things done. Instead, they motivate people to \"just go out and do it.\" Good sales managers employ these tools skillfully. Bill Gates used the leadership tool we call vision in his 1995 Internet Tidal Wave memo, which helped Microsoft's employees see that maintaining the company's dominance in the software industry (what they wanted) required an aggressive acknowledgment that the nascent World Wide Web would become an integral part of computing rather than a sideshow to the thendominant desktop applications - an acknowledgment that ran counter to most employees' deeply held beliefs. The erce response of the company's Internet Explorer team crippled Netscape and won Microsoft a more than 90% share of the browser market. Faced with stiff competition from Google in late 2005, Gates reemployed this technique in his memo regarding a \"services wave,\" calling for a shift from sales of shrink-wrapped software to sales of subscriptions. The same actions viewed as inspiring and visionary among employees in the upper-left corner of the matrix can be regarded with indifference or disdain by those in the lower quadrants. Consider vision statements. When members of a group agree on what they want to achieve, statements that articulate where the organization needs to go can be energizing and inspiring. But if employees don't agree about what they want, vision statements won't help much in changing their behavior - aside from inducing a collective rolling of eyes. Culture tools. In organizations located in the upperright quadrant of the matrix, employees will cooperate almost automatically to continue in the same direction. Their deep consensus on priorities, and on what set of actions will allow the company to achieve those priorities, is the essence of a strong culture. As MIT's Edgar Schein wrote in Organizational Culture and Leadership, culture is \"a pattern of shared basic assumptions that was learned by a group as it solved its problems of external adaptation and internal integration, that has worked well enough october 2006 to be considered valid and, therefore, to be taught to new members as the correct way to perceive, think, and feel in relation to those problems.\" In organizations with strong cultures, people instinctively prioritize similar options, and their common view of how the world works means that little debate is necessary about the best way to achieve those priorities. Companies with strong cultures in many ways can be self-managing. But this very strength can make such organizations highly resistant to change. So-called culture tools-such as rituals and folklore - only facilitate cooperation to preserve the status quo; they are not tools of change. Leadership and management tools can also be used in this quadrant to foster cooperation, but only in order to reinforce or enhance the existing culture. A manager of such a company might see herself as a visionary leader wanting to chart a new course for the organization. She may want to use a vision statement as a tool for analyzing and rening the vision in her mind. But as a tool of change? Employees in the upper-right strong-culture quadrant are unlikely to cooperate with any strategy that is at odds with their deeply shared beliefs about what they want and what must be done. Hewlett-Packard's Carly Fiorina 77 YEL MAG CYAN BLACK T h e To o l s o f Co o p e rat i o n a n d C h a n g e T h e To o l s o f Co o p e rat i o n a n d C h a n g e learned this the hard way when she tried to challenge the so-called HP Way. Her very public clashes with HP's employees and board led to her ouster in 2005, following the company's controversial merger with Compaq. Essentially, as P&G's Durk Jager needed to recognize, the only tools that can be wielded are those that are effective in the domain where the employees are-and in strong cultures, the tools in the upper-right quadrant lead to cooperation in gradual change, at best. What Managers Can - and Cannot - Do We noted earlier that there is no \"best\" position in the matrix of agreement; each quadrant carries its own challenges. A company's position may reect where it is in its life cycle and is largely determined by how successful it has been. Most organizations start at the left and often at the bottom of the matrix, where the founder's ats drive much of what gets prioritized and how it gets done. If employees develop effective methods that result in success, consensus will begin to coalesce on the horizontal dimension of agreement - what actions yield the desired results. As the company succeeds, employees who t with these ways of working, and who want what senior management wants, tend to be promoted. Those who don't tend to leave. Hence, success is the mechanism that builds consensus around what people want and how they can get it. Success shifts the organization toward the upperright quadrant. Crisis and failure, in contrast, can destroy that consensus, plunging the organization toward the lower-left quadrant. Employees in crisis are no longer certain or unanimous in their beliefs about what actions are necessary. Managers who are able and willing to use power tools during crises can get employees to cooperate in a remedial course of action, provided those managers know where the organization needs to go and what must be done to get there. Indeed, scholars of organizational change frequently prescribe \"creating a crisis\" because it forces employees into a situation where they can be compelled to cooperate. While there is merit to the create-a-crisis strategy, there's a rub to this simple solution: What if the CEO sees the need to change direction while the business is still healthy - when the crisis is in the future, not the present? And what if this healthy company also has an extremely strong culture? That was the situation facing John Sculley, CEO of Apple Computer from 1983 until 1993. Fresh from a triumphant career at PepsiCo, Sculley was an exceptional executive. During his rst several years at Apple, the company continued to prosper. By the late 1980s, however, Sculley sensed trouble over the horizon and saw the need to change strategy in three specic ways. First, he saw edgling low-cost computer makers, such as Dell, menacingly exploring how to make higher-performance computers within their low-cost business models. Sculley declared that Apple needed to move down-market aggressively, reducing its prices by as much as 75% in order to blunt this disruptive attack. Second, before Microsoft introduced its Windows operating system, Sculley urged The Tools of Politics In institutions with well-established cultures (those in the upper-right portion of the exhibit \"The Agreement Matrix\"), democracy can be used as a tool to encourage cooperation. An important insight from this model is that democracy will not work except where people agree strongly on both dimensions of the matrix: what they want and the rules of cause and effect. The very functioning of democracy depends upon the existence of strong cultural beliefs that are often rooted in the teachings of certain religions. The religious institutions at the root of these cultures have taught that people are meant to be free and that they should voluntarily be honest and respect the life, property, and equal opportunity of 78 others - because even if the police don't catch and punish them, they will be rewarded or punished in some way in the afterlife. The successful practice of these beliefs - together with a shared value that every person should be allowed to worship God in his or her own way - has created successful societies in places such as India, Japan, the United States, and Western Europe. The practices have become so deeply embedded over so many years that almost all people in these societies, regardless of religious belief, now strongly share these values and are ensconced in the upper-right quadrant of the agreement matrix. The vast majority of people living in these cultures obey the law voluntarily - and, as a result, democracy works. On occasion, Americans in particular have tried to impose democracy on countries whose populations are not in the upper-right corner of the agreement matrix - where religious or other institutions have not built the type of cultural consensus that is consistent with democratic principles. When America has essentially snapped its ngers at these countries, ordering them to establish stable democracies - and quickly - chaos typically has ensued. The crime, corruption, and tax evasion that characterize much of Russia; the collapse of civil order that torments Haiti; and the costly, tragic dilemma that America now faces in Iraq - all are testaments to the fact that democracy doesn't work when the enabling preconditions don't exist. harvard business review | hbr.org T h e To o l s o f Co o p e rat i o n a n d C h a n g e DEEP CONSENSUS ON PRIORITIES, and on what set of actions will allow the company to achieve those priorities, is the essence of a strong culture. The Tool of Disaggregation All is not lost for managers who see the need to change a successful company before the onset of a crisis. They can wield the tool of disaggregation-the separation of organizations into units. This allows managers at the new october 2006 unit to build a different consensus among its employees regarding what they want and how to get there, while the prior culture continues to thrive in the original unit. Disaggregation works by eliminating the need for cooperation between groups with opposing goals. This is how Hewlett-Packard succeeded in the disruptive ink-jet printer business even while its laser-jet printer business was prospering with a very different prot model. HP disaggregated the printer business, leaving the laser-jet unit in Boise, Idaho, and setting up the ink-jet unit in Vancouver, Washington. Likewise, IBM stayed strong in computers for many years, whereas all its mainframe and minicomputer rivals failed, because it used the tool of disaggregation. When minicomputers began disrupting mainframes, IBM created a separate business unit in Rochester, Minnesota, to focus on minicomputers, which had to be designed, built, and sold within a very different economic model than mainframes. When personal computers disrupted minicomputers, IBM disaggregated again, setting up in Boca Raton, Florida, another freestanding unit, which developed a business model tailored to PCs. Had IBM executives tried to convince the managers and employees of the original computer business to cooperate on a strategy, economic model, and culture to succeed simultaneously in mainframes, minicomputers, and PCs, the company would have failed. Mastering the Tools of Cooperation at Continental Airlines It would be rare, of course, for all employees in a company to be in one place in the agreement matrix at a given time or across time. While the founding group of senior managers may be in the upper-right quadrant, manufacturing employees may be in the lower-right. Those in sales and creative design might be in the upper-left, sharing an understanding of what is important but unwilling to subject themselves to the sorts of standards and processes that are effective in the lower-right quadrant. Most managers, unfortunately, have a limited tool kit and thus can successfully manage only in certain types of situations. One of the rarest managerial skills is the ability to understand which tools will work in a given situation - and not to waste energy or risk credibility using tools that won't. Gordon Bethune, CEO of Continental Airlines from 1994 until 2004, was such a manager. Bethune was the airline's 79 YEL MAG CYAN BLACK Apple to open its proprietary product architecture and begin selling its vaunted operating system. Third, he saw that portable, handheld devices would become an important growth market. In retrospect, Sculley saw the future of his industry with remarkable clarity. But being a visionary leader isn't all it's cracked up to be. When leaders like Sculley conclude that their organization's course must change, they need to consider where the rest of the employees are in the agreement matrix. At Apple, they were decidedly in the upper-right quadrant-some said that Apple put the \"cult\"in \"culture.\" Sculley tried reorganization, rings, control systems, nancial incentives, training, measurement systems, standard procedures, vision statements, salesmanship, strategic planning, and many more tools to elicit cooperation behind the changes he envisioned. But none worked. The Apple employees wouldn't listen. Sculley gradually lost credibility with his board and employees as tool after tool failed to produce the changes he desired, and he was ousted in 1993. Apple's board then appointed Michael Spindler, head of the company's successful European operations, as CEO. Spindler also found that the only tools of cooperation at his disposal were those that reinforced Apple's culture, and he was dismissed after three years. The board then brought in Gil Amelio, who had turned around the deeply troubled National Semiconductor - expecting that he could do the same at Apple. He couldn't and was gone in 18 months. Unable to recruit another qualied CEO, Apple's board turned in desperation to ousted Apple founder Steve Jobs as interim CEO. Jobs essentially stopped trying to change the company and instead encouraged the troops to resume designing cool, innovative, high-end products such as the iMac and iPod. Apple now dominates the digital music industry. But if there had been any tools to wield within this strong culture to elicit cooperation behind the new direction Sculley foresaw, Apple might have captured much of the fruit that ultimately fell into the hands of Compaq, Dell, and Microsoft. T h e To o l s o f Co o p e rat i o n a n d C h a n g e tenth CEO in ten years, following a disastrous run including industry worsts in lost baggage, customer complaints, overbooking, and on-time departures. Moreover, Continental had declared bankruptcy twice during the previous decade and was losing $55 million per month despite years of cost cutting. Bethune turned down the top job twice even though he was already serving as Continental's COO. The rst offer was to be acting CEO during the existing CEO's six-month leave of absence, and the second was to serve in the ofce of the CEO after that executive decided to retire. Although board members respected Bethune, they believed that the only way to restore protability was through further upper-right quadrant, working to reinforce what has become a very productive culture. Bethune's well-timed choice of tools mirrored that of Jack Welch at General Electric, who started out as Neutron Jack, using power tools when the company was a collection of businesses with vastly different cultures, operating procedures, and expectations about growth and protability. As he oriented the company around the mantra of being rst or second in each of the conglomerate's businesses, GE moved from the lower-left corner of the matrix toward the upper-right, and Welch shifted his focus to culture-reinforcing activities, teaching up-andcoming managers at the company's Crotonville campus. ONE OF THE RAREST MANAGERIAL SKILLS is the ability to understand which tools will work in a given situation - and not to waste energy or risk credibility using tools that won't. cost cutting-a path Bethune was convinced would lead to disaster, not deliverance. Given the signicant disagreement about how to restore protability, Bethune knew he could do nothing without the full authority that came with the top job, without the qualiers of \"acting\" or \"ofce of.\" Even after the board approved Bethune as CEO, few within the company agreed with his unconventional view that Continental needed to be less restrictive of its employees and spend more in order to get out of bankruptcy. As Bethune wrote in his book From Worst to First, when the operations staff rebuffed his instruction to repaint all of the carrier's more than 200 airplanes, he threatened to shoot them unless they complied. Concerned that customerservice employees were micromanaging customers by relying too heavily on a very thick instruction manual, he set re to a stack of manuals in the parking lot. Having won some initial battles by sheer force, Bethune achieved preliminary success and began to move the company out of the lower-left quadrant toward the upperright. As the company started to recover, Bethune began employing more traditional management tools, including nancial incentives. After he offered each employee a $65 bonus every month that Continental placed among the top ve for on-time departures, Continental jumped to fourth the subsequent month and rst thereafter. Our model suggests that this incentive would not have worked in the environment of distrust and disagreement that characterized the company when Bethune began his work. By 1998, the company had posted 11 straight quarters of improved prots and had won two consecutive J.D. Power and Associates' awards. Bethune spent the nal years of his career using the tools in the 80 The success of Bethune and Welch, of course, is both good news and bad news for their successors. As long as the shared purposes and unied view of how to achieve them are appropriate for their companies' challenges, Larry Kellner and Jeffrey Immelt ought to be able to preside over continued success using the cooperation tools handed to them on their arrival. However, if there are shifts in the competitive environment that mandate signicant changes either to what people want or to the required actions, the two CEOs may nd that the tools their predecessors used to turn their organizations around cannot be wielded effectively in the strong-culture quadrant. For example, much has been written about former CEO Lou Gerstner's success in refashioning IBM from a \"big iron\" company to one built on services. Managing change is always hard. But our model suggests that because he took IBM's helm when the company was in genuine crisis, losing billions of dollars, Gerstner was fortunate. The situation demanded power tools. As IBM's service businesses mature, his successor, Sam Palmisano, may face the tougher challenge. There is no current crisis that enables the effective use of power tools to marshal a cooperative march in a new direction. He faces a cultural challenge that will likely prove more difcult than the crisis Gerstner faced. Bethune, Welch, and Gerstner were blessed with an instinct for choosing the right tools at the right time. Our hope is that by making the instincts of effective managers more explicit, even those of us who are not born knowing how to manage change can learn to do so more effectively. Reprint R0610D; HBR OnPoint 1458 To order, see page 151. harvard business review | hbr.org Harvard Business Review Notice of Use Restrictions, May 2009 Harvard Business Review and Harvard Business Publishing Newsletter content on EBSCOhost is licensed for the private individual use of authorized EBSCOhost users. It is not intended for use as assigned course material in academic institutions nor as corporate learning or training materials in businesses. Academic licensees may not use this content in electronic reserves, electronic course packs, persistent linking from syllabi or by any other means of incorporating the content into course resources. Business licensees may not host this content on learning management systems or use persistent linking or other means to incorporate the content into learning management systems. Harvard Business Publishing will be pleased to grant permission to make this content available through such means. For rates and permission, contact permissions@harvardbusiness.org