Harvard Business Review

Executive Summaries

Cover Feature

Leadership's Online Labs

Byron Reeves, Thomas W. Malone, and Tony O'Driscoll Reprint: R0805C

Multiplayer online role-playing games are sprawling cybercommunities that offer a sneak preview of tomorrow's business environment. Players who lead teams in these online worlds hone the skills that they will need as business leaders in the future. Games also provide an environment that makes being an effective leader easier and that today's businesses might try to replicate selectively in their own organizations.

Those are the principal findings by Reeves, of Stanford University; Malone, of MIT's Sloan School; and O'Driscoll, of North Carolina State. As part of an analysis conducted by Seriosity, a company that develops game-inspired enterprise software, the authors studied people who headed up teams in online games. They also sought the insights of gamers who have led real-world business teams at IBM.

The authors identified three distinctive characteristics of leadership in online games that, as workplaces and the overall business climate become more dynamic and gamelike, will be essential for tomorrow's leaders: speed, risk taking, and acceptance of leadership roles as temporary.

The most important finding, say the authors, is that getting the leadership environment right can be as important as choosing the right leader. They point out two aspects of game environments that companies might consider adopting: One, nonmonetary incentives built into a game economy strongly motivate individuals to accomplish group aims. Two, hypertransparency of information about, for example, team members' capabilities and teams' real-time performance makes it simpler to match people with tasks and to empower individuals to manage themselves. Forethought

Where Is Advertising Going? Into `Stitials

Jeffrey F. Rayport Reprint: F0805A

It's high time to shift our focus from the "avails" in media content to those in consumers' lives--that is, where, when, and how people might be receptive to relevant commercial messages.

Working in the Gray Zone

Michel Anteby Reprint: F0805B

Supervisors often turn a blind eye when employees use company resources and time to work on personal projects. They realize that stamping out such behavior may do more harm than good.

Don't Just Capture Knowledge--Put It to Work

Katrina Pugh and Nancy M. Dixon Reprint: F0805C

By engaging employees in the systematic, facilitated collection and circulation of organizational knowledge, you can make sure innovations reach those who need them. Intel Solution Services offers a case in point.

The Best Advice I Ever Got

William P. Lauder Reprint: F0805D

The president and CEO of Estée Lauder Companies learned the importance of time management back when he worked under U.S. treasury secretary Donald Regan. The lesson sounds simple, but it has shaped his approach to strategy and his philosophy on motivating people.

The Blue-Collar Green-Building Boom

Charles Lockwood Reprint: F0805E

Most of the green buildings constructed over the past decade in the United States have been white-collar workplaces such as office buildings and schools. The greening of blue-collar workplaces, from warehouses to factories, has lagged far behind--but that's beginning to change.

To Get Value from a Merger, Grow Sales

Juergen Rothenbuecher and Joerg Schrottke Reprint: F0805F

After a merger, managers should strengthen sales and marketing before cutting costs. That's because revenue growth generates long-term shareholder value.

Even the Poorest Can Be a Thriving Market

Jean-Louis Warnholz Reprint: F0805G

Though many multinationals see little business potential in serving the poorest people in developing countries, two innovative mobile phone companies are making profits and improving local economic conditions by doing exactly that.

A Conversation with Christina Domecq Reprint: F0805H

The CEO of SpinVox, a start-up that provides voice-to-text service, says her relationships with her mentors have been the key to her entrepreneurial success.

Halting the Exodus After a Layoff Reprint: F0805J

A new study shows that downsizing often prompts demoralized survivors to quit, which hinders efficiency and costs companies money. To add insult to irony, career-development programs are associated with even higher turnover after the ax falls. The researchers say that certain types of HR practices may help.

Emerging Graphic Tool Gets People Talking

Martin Wattenberg and Fernanda B. Viégas Reprint: F0805K

Research by IBM's Visual Communication Lab shows that the sophisticated graphing of data generates impassioned conversations and collaboration. Executives seeking to engage employees in solving company problems should take note.

Reviews

Featuring Why Women Mean Business, by Avivah Wittenberg-Cox and Alison Maitland. HBR Case Study

Will Our Customers Bail Us Out?

David Silverman Reprint: R0805A

The Clarinda Company is struggling and may go under. Its main customer has just pulled out, taking a quarter of the typesetting firm's revenues with it, and David, Clarinda's president, has had to lay off 20% of his workforce. What's more, he's also had to fire Dan, his longtime partner and mentor, for drunkenness. How much of this should David share with the others in his small customer base? He'd like to get their help in selling the firm--but will they all bail if they know just how leaky the ship has become? Four experts weigh in on this case study, based on the author's real-life entrepreneurial travails, recounted in his recent book, Typo: The Last American Typesetter or How I Made and Lost 4 Million Dollars.

Tell all, says Jim Marsh, drawing from his own experience as CEO of a division of Cable & Wireless. By sharing both financial information and a turnaround plan, Marsh gained the trust and support of major customers, who gave his unit a second chance.

Don't panic, says Rick Rickertsen, a principal at a private equity firm, and don't necessarily sell the business. Share only what will become common knowledge, ask no favors, and don't mention Dan. Instead, get back on your horse and try to save the company.

Don't lie, says British publishing executive Richard Charkin, but don't tell the whole truth, either. Sell, by all means: Tell customers you're looking to refinance the company (selling is a form of refinancing) and invite them to take a stake.

Don't tell customers everything, says professor Kimberly D. Elsbach, who has taught Silverman's book in her first-year MBA classes at the University of California, Davis. Elsbach suggests that David is misguidedly inclined to reveal all because he is confusing competency-based trust with interpersonal trust. Features

The Science of Thinking Smarter

A Conversation with Brain Expert John J. Medina Reprint: R0805B

Advances in neurobiology have demonstrated that the brain is so sensitive to external experiences that it can be rewired through exposure to cultural influences. Experiments have shown that in some people, parts of the brain light up only when they are presented with an image of Bill Clinton. In others, it's Jennifer Aniston. Or Halle Berry. What other stimuli could rewire the brain? Is there a Boeing brain? A Goldman Sachs brain?

No one really knows yet, says Medina, a developmental molecular biologist, who has spent much of his career exploring the mysteries of neuroscience with laypeople. As tempting as it is to try to translate the growing advances to the workplace, he warns, it's just too early to tell how the revolution in neurobiology is going to affect the way executives run their organizations. "If we understood how the brain knew how to pick up a glass of water and drink it, that would represent a major achievement," he says.

Still, neuroscientists are learning much that can be put to practical use. For instance, exercise is good for the brain, and long-term stress is harmful, inevitably hurting productivity in the workplace. Stressed people don't do math very well, they don't process language very efficiently, and their ability to remember--in both the short and long terms--declines. In fact, the brain wasn't built to remember with anything like analytic precision and shouldn't be counted on to do so. True memory is a very rare thing on this planet, Medina says. That's because the brain isn't really interested in reality; it's interested in survival.

What's more, and contrary to what many twentieth-century educators believed, the brain can keep learning at any age. "We are lifelong learners," Medina says. "That's very good news indeed."

Rebuilding the R&D Engine in Big Pharma

Jean-Pierre Garnier Reprint: R0805D

From December 2000 to February 2008, the top 15 companies in the pharmaceutical industry lost roughly $850 billion in shareholder value. Although a number of factors--including the rise of generics, pricing pressures, regulatory requirements, and legal entanglements--are to blame, Garnier, the CEO of GlaxoSmithKline, believes that declining R&D productivity is his industry's primary problem. The way to solve it, he says, is to return power to the scientists--by reorganizing R&D into highly focused groups headed by inspirational leaders, seeking the best science outside as well as inside a company, fixing broken processes, and promoting a strong culture of innovation and passion for excellence.

GSK has replaced its organizational pyramid with 12 "centers of excellence." The company has worked to untangle the quest for breakthrough drugs from the effort to develop best-in-class offerings and has overhauled incentives for the scientists who actually make discoveries. It has also pursued contractual relationships with academia and biotech companies in a bid to secure the best science, wherever it may reside.

When the company began a sweeping reengineering of its R&D, it had only two products in late-stage development. Today it has 34--the most in the industry. But much more remains to be done, the author says. Significant cost efficiencies could be achieved by offshoring clinical trials. Development of new blockbuster drugs could be simplified and accelerated if researchers targeted only a limited segment of the potential patient population and then expanded to others over time.

The innovation malaise in pharmaceuticals is not unique, Garnier says. Many other industries face the same challenges. A cultural revolution and a broad transformation of the organization are necessary first steps to rebuilding the R&D engine.

When Winning Is Everything

Deepak Malhotra, Gillian Ku, and J. Keith Murnighan Reprint: R0805E

In the heat of competition, executives can easily become obsessed with beating their rivals. This adrenaline-fueled emotional state, which the authors call competitive arousal, often leads to bad decisions. Managers can minimize the potential for competitive arousal and the harm it can inflict by avoiding certain types of interaction and targeting the causes of a win-at-all-costs approach to decision making.

Through an examination of companies such as Boston Scientific and Paramount, and through research on auctions, the authors identified three principal drivers of competitive arousal: intense rivalry, especially in the form of one-on-one competitions; time pressure, found in auctions and other bidding situations, for example; and being in the spotlight--that is, working in the presence of an audience. Individually, these factors can seriously impair managerial decision making; together, their consequences can be dire, as evidenced by many high-profile business disasters.

It's not possible to avoid destructive competitions and bidding wars completely. But managers can help prevent competitive arousal by anticipating potentially harmful competitive dynamics and then restructuring the deal-making process. They can also stop irrational competitive behavior from escalating by addressing the causes of competitive arousal. When rivalry is intense, for instance, managers can limit the roles of those who feel it most. They can reduce time pressure by extending or eliminating arbitrary deadlines. And they can deflect the spotlight by spreading the responsibility for critical competitive decisions among team members.

Decision makers will be most successful when they focus on winning contests in which they have a real advantage--and take a step back from those in which winning exacts too high a cost.

How to Sell Services More Profitably

Werner Reinartz and Wolfgang Ulaga Reprint: R0805F

When products become commodities, manufacturing companies may seek to differentiate themselves with value-added services--a potentially profitable strategy. Unfortunately, companies often stumble in the effort. Reinartz and Ulaga conducted in-depth studies of 18 leading companies in a broad variety of product markets to learn what distinguished the successes from the rest. They discovered four steps to developing a profitable services capability.

Recognize that you already have a service company. You can identify and charge for simple services--as Merck did when it stopped quietly absorbing shipping costs. Switching services from free to fee clarifies their value for managers as well as for customers.

Industrialize the back office. To prevent delivery costs from eating up service-offering margins, build flexible service platforms, closely monitor process costs, and exploit new technologies that enable process innovations. The Swedish bearings manufacturer SKF provided off-site access to an online monitoring tool that could warn of potential failure in customers' machines.

Create a service-savvy sales force. Services require longer sales cycles and, often, decisions from high up in a customer's hierarchy; what's more, product salespeople may be inimical to change. Schneider-Electric did a major overhaul of its sales organization and trained its people to switch from cost-plus pricing to value-based pricing.

Focus on customers' processes and the opportunities they afford for new service offerings. You may need to acquire new capabilities to take advantage of those opportunities: The industrial coatings specialist PPG had to learn how painting robots function after it offered to take over Fiat's Torino paint shop.

Services can both lock in customers and help acquire new accounts. They should be developed with care and attention.

Strategy as a Wicked Problem

John C. Camillus Reprint: R0805G

In today's complex world, companies often find themselves facing confounding strategy problems. These issues are not just tough or persistent; they're "wicked"--a label used by urban planners for problems that cannot be definitively resolved. Poverty and terrorism are classic examples. A wicked problem has innumerable causes, morphs constantly, and has no correct answer. It can be tamed, however, with the right approach.

In this article, Camillus, a professor at the Katz Graduate School of Business, explains how executives can tell if they're dealing with a wicked strategy problem. In a 15-year study involving 22 companies, he identified five key criteria. If a problem involves many stakeholders with conflicting priorities; if its roots are tangled; if it changes with every attempt to address it; if you've never faced it before; and if there's no way to evaluate whether a remedy will work, chances are good that it's wicked. According to the author, the need for faster growth is, in all likelihood, a wicked issue for Wal-Mart.

Traditional linear processes--identifying the issue, gathering data, studying all the options, choosing one strategy--don't work with wicked problems. They instead demand social processes that constantly engage stakeholders, explore related issues, reevaluate the problem's definition, and reconsider the assumptions of stakeholders. A strong corporate identity is essential: It serves as a rudder that helps the enterprise navigate a sea of choices. Because it's impossible to tell which options are appropriate, executives should stop analyzing them and start experimenting with actions. Eventually they will make progress by muddling through. Envisioning possible futures and identifying moves that will realize the one the company hopes for will uncover promising remedies. That's how PPG Industries, a 100-year-old manufacturer, has successfully coped with its wicked strategy issues.

The Customer-Centered Innovation Map

Lance A. Bettencourt and Anthony W. Ulwick Reprint: R0805H

We all know that people "hire" products and services to get a job done. Surgeons hire scalpels to dissect soft tissue. Janitors hire soap dispensers and paper towels to remove grime from their hands. To find ways to innovate, it's critical to deconstruct the job the customer is trying to get done from beginning to end, to gain a complete view of all the points at which a customer might desire more help from a product or service.

A methodology called job mapping helps companies analyze the biggest drawbacks of the products and services customers currently use and discover opportunities for innovation. It involves breaking down the task the customer wants to accomplish into the eight universal steps of a job: (1) defining the objectives, (2) locating the necessary inputs, (3) preparing the physical environment, (4) confirming that everything is ready, (5) executing the task, (6) monitoring its progress, (7) making modifications as necessary, and (8) concluding the job.

Job mapping differs substantively from process mapping in that the goal is to identify what customers are trying to get done at every step, not what they are doing currently. For example, when an anesthesiologist checks a monitor during a surgical procedure, the action taken is just a means to the end. Detecting a change in patient vital signs is the job the doctor is trying to get done.

Within each of the discrete steps lie multiple opportunities for making the job simpler, easier, or faster. By mapping out every step of the job and locating those opportunities, companies can discover new ways to differentiate their offerings.

Marketing When Customer Equity Matters

Dominique M. Hanssens, Daniel Thorpe, and Carl Finkbeiner Reprint: R0805J

Measuring the effectiveness of marketing investments can be frustrating--especially if a company focuses on a long-term outcome like increasing customer equity. Though there are decision-support models to help marketers allocate their budgets, until recently such models aimed to maximize near-term sales, which aren't always consistent with long-term brand health. Now, however, Wachovia has created a model that helps it make customer-equity-driven marketing-mix decisions.

The model's architects--Hanssens, a professor at UCLA; Thorpe, a senior vice president at Wachovia; and Finkbeiner, an executive VP at TNS--began by parsing customer equity, or the lifetime value of customers, into three measurable components: customer acquisition, customer retention, and share of wallet. They then assembled data on past marketing activities and changes in the components--a daunting task, given the decentralized nature of previous investments. Once the database was created, a model of how investments affected outcomes could be built and used to design equations focusing on marketing-mix questions. Wachovia now has, for example, an equation to estimate how an increase in cable TV advertising will affect customer acquisition in a market. Just as important, the model highlights how much factors beyond the control of marketing, such as economic swings, affect outcomes.

After Wachovia began using the model, it found, as suspected, that a marketing mix with maximum short-term impact would not achieve the biggest increase in customer equity. Wachovia learned that it was overinvesting in traditional channels and underinvesting in emerging ones. The introduction of this decision-making tool has also supported a broader cultural shift at the company toward competing on analytics.

Leadership's Online Labs

Tens of millions of people are honing their leadership skills in multiplayer online games. The tools and techniques they're using will change how leaders function tomorrow--and could make them more effective today.

by Byron Reeves, Thomas W. Malone, and Tony O'Driscoll

Tomorrow's business landscape could well be alien territory for today's business leaders. At many companies, important decision making will be distributed throughout the organization to enable people to respond rapidly to change. A lot of work will be done by global teams--partly composed of people from outside the institution, over whom a leader has no formal authority--that are assembled for a single project and then disbanded. Collaboration within these geographically diverse groups will, by necessity, occur mainly through digital rather than face-to-face interaction.

What on earth will leadership look like in such a world--a world whose features have already begun to transform business?

Suspend your skepticism for a moment when we say that the answers may be found among the exploding space stations, grotesque monsters, and spiky-armored warriors of games such as Eve Online, EverQuest, and World of Warcraft. Despite their fantasy settings, these online play worlds--sometimes given the infelicitous moniker MMORPGs (for "massively multiplayer online role-playing games")--in many ways resemble the coming environment we have described and thus open a window onto the future of real-world business leadership.

True, leading 25 guild members in a six-hour raid on Illidan the Betrayer's temple fortress is hardly the same as running a complex global organization. For starters, the stakes are just a bit higher in business. But don't dismiss online games as mere play. The best ones differ from traditional video games as much as universities do from one-room schoolhouses. In fact, these enterprises are actually sprawling online communities in which thousands of players collaborate with and compete against one another in real time within a visually three-dimensional virtual world--one that persists and evolves even while a player is away.

The organizational and strategic challenges facing players who serve as game leaders are familiar ones: recruiting, assessing, motivating, rewarding, and retaining talented and culturally diverse team members; identifying and capitalizing on the organization's competitive advantage; analyzing multiple streams of constantly changing and often incomplete data in order to make quick decisions that have wide-ranging and sometimes long-lasting effects. But these management challenges are heightened in online games because an organization must be built and sustained with a volunteer workforce in a fluid and digitally mediated environment.

Getting a look at how leadership works in online games isn't easy. To see the best players in action, you need skills that allow you to participate at the highest levels of play, and those can take 400 or 500 hours to acquire. When IBM commissioned Seriosity to study leadership in games, Seriosity used a team of a half-dozen veteran players, with more than 50,000 hours of cumulative experience, to observe and record the actions of leaders in this rarefied setting. The eight-month study also included interviews with more than a dozen prominent gamers about their leadership endeavors in this arena. A follow-up survey at IBM of people with both gaming and business leadership experience helped validate some of our findings and suggested how they might be translated to fit real-world corporate contexts.

A number of our conclusions about the future of business leadership were unanticipated. For one, individuals you'd never expect to identify--and who'd never expect to be identified--as "high potentials" for real-world management training end up taking on significant leadership roles in games. Even more provocative was our finding that successful leadership in online games has less to do with the attributes of individual leaders than with the game environment, as created by the developer and enhanced by the gamers themselves. Furthermore, some characteristics of that environment--for example, immediate compensation for successful completion of a project with nonmonetary incentives, such as points for commitment and game performance--represent more than mere foreshadowing of how leadership might evolve.

Adopting some of these signature qualities of the game environment could actually make it easier to lead people in today's real-world companies. The startling implication: Getting the leadership environment right may be at least as important to an organization as choosing the right people to lead.

An Online Preview of Tomorrow's Leadership

Multiplayer online games are an increasingly popular and particularly compelling form of entertainment. Some estimates put the current number of registered players worldwide at more than 50 million; World of Warcraft alone claims 10 million players, who pay a subscription fee of roughly $15 a month. Participants play for an average of 22 hours a week, according to researcher Nick Yee of the Palo Alto Research Center; their average age is 27, and about 85% are men. (For descriptions of some popular games, see the exhibit "Online Fantasy Games.")

Sidebar Icon Online Fantasy Games (Located at the end of this article)

Although the games vary in theme and setting, many are similar in structure: Roughly 40 to 200 players form teams, or guilds, that undertake increasingly difficult tasks, whereby individuals acquire skills and tools that allow them to advance to the next level of play. Sometimes team members know one another in real life, but typically they form their relationships in the game world. Within teams, members adopt different roles and responsibilities to get things done on behalf of the group. Guild membership is often in flux, as players get fed up with colleagues or seek more attractive opportunities elsewhere.

Invariably, certain individuals emerge to set the team's direction and lead activities, although the leadership often changes. The leader or leaders of a guild, who specialize in ongoing recruitment, creation of incentive systems, and player evaluation, don't necessarily act as the leaders of a particular mission or raid. Raids involve a subset of the guild's members and take place in a single session.

The challenges undertaken by guilds can be complex, both organizationally and strategically: A raid on a dungeon in World of Warcraft, for example, may require the participation of dozens of players and go on for many hours. That competitive and goal-oriented environment makes these games very different from other immersive online worlds, such as Second Life. (For a comparison between these two types of online environments, see the sidebar "Online Games Versus Virtual Social Worlds.")

Sidebar Icon Online Games Versus Virtual Social Worlds (Located at the end of this article)

Online games are, of course, an imperfect analog to real-world business. Not only do they involve lower stakes, but the problems teams face, difficult though they may be, are also sharply defined and structured. Instead of having to identify and frame challenges--a central element of real-world business leadership--gaming leaders primarily plan and execute tactics to achieve goals specified by the game. In this sense, multiplayer games are more akin to warfare than to business.

Another difference between play and reality is that gamers operate through characters, or avatars, that they have adopted and personalized for the game and that easily disguise their real-world identities. This surrogacy gives gaming relationships a distinctive wrinkle. It's clear from laboratory research that players are highly invested psychologically in their avatars. That passionate attachment, coupled with the ability to act through an alter ego, makes heated disagreements both common and accepted. An atmosphere of intense honesty develops that many players say gradually makes them less averse to group conflict.

Nevertheless, our findings reinforced our basic premise that leadership in online games offers a sneak preview of tomorrow's business world. In broad terms, that environment can be expected to feature the fluid workforces, the self-organized and collaborative work activities, and the decentralized, nonhierarchical leadership that typify games. In more specific terms, we found several distinctive characteristics of leadership in online games that suggest some of the qualities tomorrow's business leaders will need in order to achieve success. (Online games offer an opportunity to develop not only these new skills but also many present-day leadership qualities. See the sidebar "A Leadership Simulator.")

Sidebar Icon A Leadership Simulator (Located at the end of this article)

Leadership demands speed.

A game hour is unlike 60 minutes at your desk or in a meeting. Actions that might take weeks or months to unfold in real life are often compressed into hours or even minutes online. For example, in a World of Warcraft battle that we recorded, a hastily formed team of 10 players decided who would lead the assault, assessed the strengths and weaknesses of its rivals from another team, formulated an attack plan, and coordinated battle assignments--all before the game clock had counted down one minute.

The fast pace of leadership has some interesting consequences. For example, the need for ultraquick decision making may occasionally trump team consensus--a tension the leader must carefully manage because of the need to constantly motivate people who are free to leave the team on a whim. Another implication of speed: Decisions are nearly always based on incomplete information and then modified as more data become available.

The lightning pace of games is unlikely to become widespread anytime soon in the business world, except perhaps in selected contexts such as high-velocity financial trading. However, business decision making is accelerating, driven in part by the almost instant, if not always complete, availability of certain kinds of data. To keep up with rivals, real-world leaders will increasingly need to be willing and able to act on such information without pausing for long periods to weigh options. They'll need to be comfortable with--and operate in a corporate culture that readily accepts--modifying decisions in response to contingencies and adopting iterative strategies marked by repeated course corrections.

Risk taking is encouraged.

Trial and error play a big role in accomplishing game tasks. Failure, instead of being viewed as a career killer, is accepted as a frequent and necessary antecedent to success.

In one incident that we recorded from EverQuest, seven guild members prepared for a brand-new quest that required them to get their team across a large lake protected by a gruesome and hostile creature. Although they had formulated a strategy based on information gathered in advance, everyone seemed comfortable with the high likelihood of failure, at least initially. After a first attempt, in which the whole team nearly drowned and was forced to retreat, members quickly began plotting a new strategy in the spirit of a fundamental gamer maxim (one not heard very often in business): "Let's try that again."

Obviously, the team members took the initial plunge because they knew they'd get another chance: Neither millions of dollars in shareholder value nor the livelihoods of thousands of employees were at stake. But it would be wrong to assume that nothing is ventured in such an effort. Games can exact severe penalties for failure, ranging from wasted hours of planning to the loss of hard-earned privileges and reputation. Although hard-core gamers don't mind failure in the short term, don't tell a hard-core gamer that failure in the long term doesn't matter. In the words of one veteran player, "No one wants to be a member of a guild that always wipes out."

Frequent risk taking allows players to practice the art of weighing odds calmly in uncertain environments. Confronting risk routinely and with a level head will be an important leadership skill as the real-world business environment becomes more uncertain and as success comes to depend more on innovation than on execution. Organizations can help prepare leaders by fostering a culture in which failure is tolerated. They can expose leaders to risk by mimicking the structure of games, breaking down big challenges into small projects. Failure, after all, is clearly more palatable for the individual and more affordable for the organization when it happens at the project level rather than on a larger scale.

Leadership roles are often temporary.

Perhaps the most striking aspect of leadership in online games is the way in which leaders naturally switch roles, directing others one minute and taking orders the next. Put another way, leadership in games is a task, not an identity--a state that a player enters and exits rather than a personal trait that emerges and thereafter defines the individual.

Don't get us wrong: Leadership stars do exist in games. Some guild leaders have successfully led 100-strong teams for a year or more--an eternity in this new medium. As in business, players with exceptional relationship skills are particularly good at forming effective teams, delegating responsibility, and keeping groups motivated and moving forward. However, games do not foster the expectation that leadership roles last forever. Someone leading a guild today may grow weary of the stress and hand over the reins after a month or two. The leader of a raid knows that someone else's skills and experience may be better suited to commanding the next effort. Even during the frenzied activity of a raid, the leadership role can be transferred as conditions change or because the person in charge doesn't happen to be around when the need for a decision arises. Notably, choices about who will lead and who will follow are often made organically by the group--frequently because someone volunteers to take over--not by some higher authority.

The expectation that leadership is temporary has benefits and consequences. Leaders clearly must be--and often are--good followers because their experience equips them to understand what the person in charge is trying to accomplish. Frequent swapping of roles also helps leaders avoid burnout--a very real problem in the hyperintense environment of online games. One noteworthy consequence is that people who wouldn't normally seek out or be chosen for leadership end up assuming it, sometimes simply to ensure that a specific task gets done. "My guild was struggling to merge with another," one experienced and successful guild leader, a 27-year-old man, recalled in an interview. "When things didn't work out to plan, our guild leader called it quits. No one volunteered to take over, so I stepped up to the plate. It wasn't my desire to lead, but I knew if I didn't, everything we had worked so hard to build would crumble."

When otherwise reserved players find themselves acting as leaders, they can surprise even themselves with their capabilities. A 46-year-old woman, unsure of her fitness to lead a guild when friends recruited her, said, "Follow-up and assertiveness now feel more natural to me, even in real life. It has been an amazing opportunity to push myself beyond my boundaries."

The idea of temporary leadership is alien to most business organizations. Companies usually identify people as leaders early in their careers. The selected few carry that designation with them through different jobs, each typically lasting several years, as they move up the corporate hierarchy. That model may not work well in the future. The growing complexity of the business environment means that no single leader will be an expert in every area. Beyond the obvious benefit of matching an individual's expertise to a challenge, treating leadership as a temporary state can empower employees to volunteer to lead and, thereby, can unearth previously overlooked talent among the ranks.

Game Elements to Make Leadership Easier Today

Most writing about leader selection and development focuses on people's backgrounds and natural talents. Whether leadership ability is inborn or acquired through training, the assumption is that expertise resides within the individual.

Our study provided us with an arrestingly different view: Perhaps the right environment is what really matters, whoever the leader happens to be. This concept, which as far as we know is absent from the academic and professional literature about leadership, wasn't something that we set out to prove. The notion arose from the experienced gamers on our research team, who were puzzled by our initial preoccupation with the individual qualities of game leaders. "If you want better leadership," they asked, "why not change the game instead of trying to change the leaders?"

So we began to focus on identifying distinctive aspects of online game environments that could improve leadership in business and other real-world settings. We pinpointed at least two properties of games that we believe facilitate and enhance leadership: nonmonetary incentives rooted in a virtual game economy; and hypertransparency of a wide range of information, including data about individual players' capabilities and performance. These two elements--along with the rich mix of text, audio, and visual communication in games--make it easier for leaders to be effective. Players know exactly what they should be doing and, to a large degree, have the tools they need to manage themselves. This suggests that organizations can benefit by selectively "gamifying" their work environments in order to improve the quality of leadership--not in the future but right away.

Nonmonetary incentives.

A game leader faces important motivational challenges. Consider, for instance, having to persuade dozens of team members from around the world to leave their real-life activities and show up online at a specific time ready to participate in a raid that will last for hours. How do you motivate these players to contribute their time and skills to a coordinated activity that benefits the entire group?

The game leader is aided by an array of sophisticated incentive mechanisms--some of them built into a game by its developers, many created and refined by the leaders themselves--that reward the individual performance of players and their contribution and commitment to the team. The incentives can be either short-term (sometimes almost immediate) or long-term.

We studied one guild leader who offered bonuses--known in many games as dragon kill points, or DKPs--to raiding groups that were able to push performance to new levels. He was often able to make such offers on the spot--"I have an additional four DKPs for those that can clear the dungeon in under three hours!"--because of the capacity games give players to track individual achievement continuously during a mission. The immediacy of the incentive made it particularly compelling. The leader also used incentive systems to encourage sustained effort. For instance, he announced that players would receive additional DKPs for mining raw materials needed by the group in advance of a raid, thereby encouraging enterprising miners in the guild to perform an otherwise boring task over the course of days.

Some of the most pervasive and sophisticated incentive mechanisms are those used to distribute the loot that a team accumulates during a successful raid. A leader needs a way to dole out to raid participants what may be only a handful of plundered items, themselves not always divisible--say, the Axe of the Gronn Lords or the Helm of the Fallen Defender. One method involves an auction in which team members bid using DKPs they have earned for their participation and performance in the current and past raids.

Listening in on a DKP manager conducting an auction can provoke snickers if you're not a gamer: "OK, bidding is now open for the Salamander Scale Pants." But the reward system may involve sophisticated algorithms that measure an individual's input, ranging from attendance to quality of play, as part of a group process. Each guild member's DKP totals are posted and kept current--often automatically, using game-provided or plug-in software--on a guild's website.

A point system like DKPs, used by leaders to motivate team members, is also part of a broader game economy. Players use synthetic currencies, such as virtual gold pieces, to buy and sell items of value to one another--everything from weapons to information to an agreement to collaborate on a particular task. (Players can also use real-world currency to purchase valuable items, such as skills or tools that others have earned in the game world, at numerous online auction sites. One of a leader's tasks when putting together a team is to sniff out players who have tried to buy their way to a certain level of accomplishment.)

Incentive systems used by leaders affect motivation in several ways. Dividing up the winnings from a quest immediately after it's completed--or, occasionally, awarding loot to someone even as the battle rages--creates a strong connection between effort and reward. Furthermore, DKP systems enable individuals to see in advance what they are likely to get when their team succeeds (more DKPs equal greater access to team booty), sharpening the incentive to join the effort. Even when it's clear they're unlikely to share in the spoils of a raid, players know that their participation will earn them points for future use. Finally, because individual compensation is based on objective performance data that can be automatically gathered and processed, and then publicly posted in real time, the reward system is generally viewed as fair.

The sorts of contributions people make to a corporate cross-functional team aren't, of course, as easy to precisely quantify, track, and reward as are contributions that game players make to their guilds. Still, we believe that game-inspired incentives have the potential to dramatically improve leadership effectiveness in business organizations. Companies might devise ways to shorten the lag time between successful outcomes and the monetary compensation for those who contribute to them. For instance, instead of getting an end-of-year bonus, people in certain businesses could be rewarded for their contributions to a project as soon as it was completed--a prospect likely to galvanize their efforts. Also, before the launch of a group project such as a prolonged cross-functional sales effort, people might be given a breakdown of how rewards for a successful outcome will be divvied up.

One of the most powerful takeaway messages about game incentives is that those built around synthetic currencies have tremendous value. As Indiana University economist Edward Castronova has noted, economies have always been a game, with currencies the mechanism for keeping track of success and failure. Online games give us the further insight that people care a great deal about virtual gains and losses, even if the currency that records them can't be exchanged for dollars. That reality opens up entirely new ways of thinking about business incentives, including how we document and value people's contributions to group efforts.

One major impediment to group collaboration in business is uncertainty about whether an individual will get credit for contributing useful information, especially digital work that can be easily forwarded or repurposed, after it is passed around the organization. Take the case of someone's e-mail notes that end up in a widely circulated internal document. A virtual currency system that identified the source of digital information and tagged its subsequent use could ensure that the originator would receive credit--formal acknowledgment or some more tangible reward--when those data were forwarded, reused, or cited. That would create a strong incentive to share.

We have seen virtual incentives affect digital communication in other ways. In a study done at a Fortune 100 company on methods for reducing information overload, employees received an allotment of a virtual currency, which they could use to indicate the relative importance of e-mail messages they sent. Attaching a large amount of the scarce currency to a particular message would draw attention to it or even serve as a feedback mechanism: You send me an e-mail you value at 100 units, and I respond with one valued at 200, giving you a credit of 100 units to validate the usefulness of the information you sent. One experiment showed that the currency, as a marker of information importance, in fact influenced how quickly colleagues opened and read different messages in their inboxes. Other gamelike elements were also tested: For example, the ability to win publicly visible digital badges in return for efficient communication helped reduce unnecessary e-mail.

Hypertransparency of information.

Game environments make a broad array of information, conveniently organized on data-rich dashboards, immediately visible not just to leaders but to the entire team. The information includes detailed statistics on individual and group performance, real-time status reports on operations, and relevant facts about players' capabilities and performance histories. All of these make leading easier.

Real-time updates about a team's mission help a leader choose a strategy in the heat of the action. Data about individual players allow a leader to quickly locate guild members who can bring needed skills and weapons to a raid and then to assign them to suitable roles. As we noted when discussing incentives, transparent and quantitative score keeping, which governs the distribution of rewards and the assignment of roles, encourages players to see the system as fair and to buy in to a leader's goals. In fact, most games are indeed meritocracies: The chances of getting preferential treatment simply because you're a friend of the boss are relatively low.

Dashboards, or cockpits, display both status and communications functions on the same densely populated user interface and often on a single computer screen, eliminating the need to open and close different software applications. Constantly visible during play, the cockpits allow a leader to stay within the narrative of the game while acquiring necessary information about players and communicating instructions to the group. Unlike a corporate dashboard that is located on a handful of computers at headquarters, with access limited to the senior executive team, these personal, view-as-you-go game cockpits give people in the field access to information as soon as it is available. That, in turn, allows game players to act on it without waiting for instructions from a guild leader. What's more, the information allows players to assume impromptu leadership roles as needed. In many of our video clips, we see three or four people barking orders to team members during a raid, briefly taking the lead in the improvisational style of a jazz ensemble.

Most real-world companies are already working on capturing and integrating real-time information about people, activities, and results. Certainly, the concept of an überdashboard that would synthesize and display all current company metrics is something CEOs have long sought, although fitting them all on one screen might be difficult. A more relevant issue is whether leaders might benefit from relinquishing control of some of that information, in order to provide employees with better tools for making their own decisions and to spur group insights that would never occur to a single leader.

The kind of game information that may be particularly applicable to business is the detailed data about players. Unlike a relatively static employee file, which provides a snapshot of someone's past experience and training, player data are constantly and automatically updated. They thus provide a kind of streaming video of a player's résumé, including information about what he or she is like right now. This approach to employee information could transform how managers and their subordinates collaborate. An employee's profile could include a wide variety of voluntarily provided information about the person's informal skills and personal passions, all of which could be put in a searchable database. Someone looking for a sales manager in Thailand might unearth a worker with, say, a Thai spouse (whose local knowledge could help the manager be more effective there) or a longtime dream of living in Thailand (which could enliven a person's work assignment).

But even this tool would lack the dynamism of a gamer's constantly changing profile. Envision, for instance, a system in which all kinds of actions and interactions are, with the employee's permission, automatically tagged and fed into a personal "tag cloud"--a visual schematic of the employee's contacts with people, activities, and ideas. To a leader who is assembling a team in an evolving business environment, such a real-time, composite view of a person is likely to be more relevant than formal certification of an employee's mastery of rigidly defined and probably outdated skill sets.

The Future Is Here

Even if they buy into the argument that game elements can make leadership easier, most business leaders will remain skeptical that a business can adopt them--unless, that is, these leaders themselves have spent time playing multiplayer online games.

To get the reactions of managers who regularly visit these online leadership labs and then return to the world of business, IBM surveyed 135 of its employees who had led business teams and had also been a leader or member of a guild in a multiplayer online game. For the most part, they found games to be surprisingly relevant to their day-to-day work. Three-quarters of the respondents said that environmental factors within multiplayer games could be applied to enhance leadership effectiveness in a global enterprise. Nearly half said that game playing had already improved their real-world leadership capabilities, particularly for managing teams whose members didn't fall under their formal authority.

Many said, however, that widespread adoption of the leadership approaches found in online games would require a change in most organizations' cultures. Failure to achieve a goal on the first try is generally viewed as a learning experience in multiplayer games, after which you "reattempt with new knowledge," according to one respondent. That's in contrast to the corporate world, where, he acknowledged, "reattempting is hard."

But games, and the generation that has grown up steeped in the game environment, may end up being catalysts for change in business leadership. This new crop of workers will bring with them--first as followers, then as leaders--game-informed notions about the best methods for leading.

Ultimately, the entire workplace may begin to feel more gamelike--with game-inspired interfaces becoming 3-D operating systems for serious work--which could enhance not just leadership but all sorts of collaboration and innovation. At the very least, digitally enabled environments and techniques could increase productivity by making many aspects of work simpler, less tedious, and--dare we say it?--more fun. That wouldn't necessarily be a bad thing. Online Fantasy Games

Here's a sampling of multiplayer online role-playing games that might help you and other leaders in your organization hone leadership skills--and have fun in the process.

World of Warcraft

Blizzard Entertainment

Paying members: 10 million

The Basics: Gamers create their avatars by choosing among 10 "races" (including dwarves, orcs, and humans) and nine roles (such as hunter, mage, and rogue).

How Players Advance: They join guilds and collaborate to explore new destinations and complete complex quests.

Eve Online

CCP

Paying members: 225,000

The Basics: Players participate in a hypercapitalistic competition among corporations that battle in outer space for galactic domination.

How Players Advance: Gamers progress by making money rather than killing monsters.

EverQuest

Sony Online Entertainment

Paying members: 375,000 (includes EverQuest II)

The Basics: This medieval fantasy game involves three types of play: adventuring (gaining experience and loot), trading with other players, and social interaction.

How Players Advance: Gamers must cooperate with rather than hinder one another.

Lineage

NCsoft

Paying members: 2.1 million (includes Lineage II)

The Basics: This game is based on a castle siege system, in which castle owners set tax rates in cities and collect taxes on items purchased in city stores.

How Players Advance: Victory in player-versus-player fighting is the route to success, but gamers are penalized for killing players who don't fight back.

Star Wars Galaxies

Sony Online Entertainment

Paying members: 100,000

The Basics: Based on the Star Wars characters and narrative, this game requires that players either train for professions that provide game services or make useful products--and then market those services or products on game planets.

How Players Advance: They must keep the supply chain filled and satisfy customer demand.

Numbers of paying members are taken, as of January 2008, either from reports by the game makers or from estimates by mmogchart.com. Online Games Versus Virtual Social Worlds

Visually three-dimensional online social worlds--such as Second Life, for adults, and Club Penguin, Webkinz, and Habbo, for children--are becoming increasingly popular. Like multiplayer online games, these virtual environments are characterized by real-time interaction among avatars (online personae that participants create for themselves) and virtual economies in which participants buy and sometimes sell virtual assets.

However, unlike online games, virtual social worlds lack structured, mission-oriented narratives; defined character roles; and explicit goals. Instead of collaborating to slay monsters, people simply do things together--go to a club with friends to listen to music or invite people over to a home they've decorated with items purchased from virtual stores. Yes, people in an internet social world can play games, many of which offer prizes or currency to buy more virtual stuff--but the world itself isn't created around game objectives.

Still, virtual social worlds offer significant opportunities for collaboration--in Second Life, virtual residents can work together to build things or even establish businesses that offer an array of virtual products and services--and for leadership and other training. That's why real-world businesses of all kinds have set up shop in Second Life--to experiment with the possibilities it offers. A Leadership Simulator

Online game leaders operate in a context that may well foreshadow the business environment of the future, but unlike the dragons they sometimes battle, they are not strange creatures who are nothing like us. They actually exhibit many of the skills of today's successful real-world leaders--making sense of ambiguous situations, transforming strategy into action, managing diverse teams collaboratively, and so on.

Put simply, online games can be informal but realistic simulators for contemporary leadership training. In fact, companies could explicitly integrate these games into their leadership development programs in order to teach the "soft" aspects of leadership, complementing simulation tools that emphasize hard, analytic skills. The benefits would extend not only to individuals but also to business teams, which might use the games to try out various leadership structures for the group.

Opportunities for practicing leadership are plentiful in games. Their pace means that leaders often have to make hundreds of strategic decisions during an hour of game play. The relatively mild consequences of failure allow players to easily test out a variety of leadership techniques. And the temporary nature of many leadership roles in games provides people who are followers in the real world with opportunities to see what it's like to lead.

Game leaders often comment on the parallels between play and reality. "The closest thing I can liken the leadership of an 80-person modern raiding guild to is the management of a medium-size business," says a World of Warcraft game leader, a former U.S. Army officer with a master's degree in human resource management. "You need to allocate resources, construct balanced compensation for your employees, stay ahead of the competition, ensure growth, and keep everyone happy and productive while handling many other day-to-day details. In the end, I worked at our guild's success for close to 60 hours a week--leading raids but also answering e-mails and questions from team members, refining guild policies, and updating the roster and raid points for everyone. I came away from it proud of my ability to manage so many diverse people and accomplish the endgame, but I was drained emotionally and physically. It is tough work."

Rebuilding the R&D Engine in Big Pharma

To save themselves, pharmaceutical companies will have to break up their giant R&D organizations, overhaul core processes, and put passionate scientists back in charge.

by Jean-Pierre Garnier

Historically, the pharmaceutical industry has been a leader in financial performance and value creation. In recent years, however, its stock-market record has raised doubts about the sustainability of that history along with fundamental questions about the industry's health. From December 2000 to February 2008 the top 15 companies in the industry lost roughly $850 billion in shareholder value, and the price of their shares fell from 32 times earnings, on average, to 13.

Sidebar Icon From Leaders to Laggards (Located at the end of this article)

The common explanation for investors' loss of faith is the well-known perfect storm of trends--pricing pressures, regulatory requirements, legal entanglements, inroads by generics, and declining R&D productivity--that have increased the industry's costs enormously and reduced its revenue and profit potential. I certainly agree that all of these trends are problems for the industry. But I believe that declining R&D productivity is at the center of its malaise.

Some critics question whether so-called Big Pharma can fix its R&D engine. They predict that more-nimble new enterprises like those in the biotech sector will supplant the lumbering dinosaurs. I strongly disagree. There are benefits to size: It provides the critical mass needed for global clinical development and acquiring crucial technology platforms.

The way to solve the productivity problem is not to break up the pharmaceutical giants into smaller companies. It is to return power to the scientists by reorganizing R&D into small, highly focused groups headed by people who are leaders in their scientific fields and can guide and inspire their teams to achieve greatness. It is to seek the best science wherever it resides, inside or outside a company. It is to fix broken processes and promote a strong culture of innovation marked by a passion for excellence and an awareness that results matter. The basic philosophy for modern R&D should be to morph big into small in recognition of the fact that critical mass in fundamental research is the size of one human brain.

We have been striving to do all these things at GlaxoSmithKline since 2000, when we began a sweeping reengineering of R&D. Our results to date suggest that we are on the right track. When we began our effort, the company had only two products in late-stage development--one of the smallest numbers in the industry--despite a decade of high R&D spending. Today we have 34 drugs and vaccines--the largest number in the industry, according to Cowen equity research. A study by CMR International, a respected pharmaceutical R&D benchmarking agency, compared eight Big Pharma players on key R&D metrics (pipeline fill and flow); it shows that our productivity is now two or three times as high as the average of our competitors.

Some of the unique actions we have taken in the past eight years have already proved their value. The most significant is the breakup of our formerly mammoth R&D organization into small cross-disciplinary groups, each of which is focused on a family of related diseases. Other potentially transformational initiatives are still works in progress. Our effort to untangle the process for pursuing breakthrough therapies from that used to develop the best medicine in an already-discovered class of drugs is the most noteworthy.

This article focuses on the pharmaceutical industry. But I think many if not all of the lessons we have learned are applicable to other industries whose long-term survival depends on true breakthrough discoveries, not just incremental improvements.

The Crux of the Problem

The business model of Big Pharma is straightforward. New products are discovered, developed, launched, and protected by various patents. Initially the products benefit from monopolistic--or at least oligopolistic--pricing. After 10 or 12 years, in general, patents expire and lower-priced generics come in, wiping out the revenues of blockbuster drugs in a matter of weeks. R&D must continually replace older products with new ones to stop the revenue base from shrinking. The problem, of course, is that doing this has proved increasingly difficult.

The decrease in the pharmaceutical industry's R&D productivity--as measured by the average R&D cost per new approved drug, which includes the amount spent on failures--has been fully documented. Although the industry's collective investment in R&D from 1980 to 2006 mushroomed from $2 billion to $43 billion, the number of drugs approved by the FDA in 1980 and in 2006 was roughly the same.

One consequence of the slide in R&D productivity is the alarming decline in average sales exclusivity: the remaining period of time drugs will be protected from competition by patents. Since 1999 the average exclusivity of patent-protected drugs (weighted by sales) has declined from five and a half years to less than four, the lowest level ever. A number of blockbuster drugs will lose their monopolies within the next four years, making it a watershed period for the industry.

A variety of forces are collectively responsible for the drop in the number of blockbuster drugs launched every year and the overall slide in R&D productivity. Those commonly cited include tougher challenges (the diseases easiest to treat or cure have been tackled), greater regulatory requirements, and the skyrocketing costs of every aspect of R&D, from producing a case report on a patient in a clinical trial to building a chemical pilot plant. Discovering and developing a new medicine takes at least 12 years, and the average cost is now more than $1 billion--higher than NASA's budget for sending a rocket to the moon. (See the sidebar "An Old Model Under Attack.")

Sidebar Icon An Old Model Under Attack (Located at the end of this article)

Another culprit is the enormous size and complexity of the traditional pharmaceutical R&D organization. In drugs, electronics, software, and other industries where fundamental discovery (as opposed to continuous improvement) is the key to success, size has become an impediment. Two classic company examples are Bell Labs and IBM, which in their heyday innovated and won recognition again and again. Then, for a variety of reasons--most of them rooted in increasing size and bureaucratization--they stopped being the true innovators, the breakthrough discoverers. Top scientists with a low tolerance for bureaucracy left; with no clear mission, those who stayed often aimlessly pursued their own interests.

If not creatively managed, complexity can cause passionate engagement and courageous risk taking to give way to risk aversion, promises with no obligation to deliver, and bureaucratic inertia. To make matters worse, the leaders of major corporations in some industries, including pharmaceuticals and electronics, have incorrectly assumed that R&D was scalable, could be industrialized, and could be driven by detailed metrics (scorecards) and automation. The grand result: a loss of personal accountability, transparency, and the passion of scientists in discovery and development.

In the pharmaceutical business, one of the biggest disappointments of the past decade is that the sequencing of the human genome and the industrialization of techniques employed in the early discovery process have not become miracle cures for sliding R&D productivity. Indeed, the share of experimental drugs that fail in the clinical stage (when testing on people occurs) has actually risen, hitting a record 93% in 2006, according to CMR International.

The hope was that the huge amount of data from sequencing the genome would provide shortcuts in pinpointing targets for attacking diseases, and that new techniques (including combinatorial chemistry for creating new compounds and high-throughput screening for testing the medicinal potential of compounds) would allow researchers to produce safe and effective therapies much faster. But for the most part, the sequencing of the genome has provided only an alphabet; we still do not understand how to assemble the letters to form words and sentences. And the tools themselves are no substitute for first-rate scientific minds; a fool with a tool is still a fool.

How can Big Pharma's R&D problems be solved? Here are some ideas.

Redesign the Organization

By and large, pharmaceutical companies have maintained an organizational setup that worked well in the 1960s: a pyramid with functional silos (chemistry, pharmacology, clinical development, and others) all joining at the top. However, back in the 1960s the largest players employed only about 1,000 scientists each; the pyramid contained only a few management layers; there were fewer projects; scientists worked together on one campus; and resource allocation was relatively straightforward. Something happened on the way to the new century. Employment multiplied by 20. The pyramid became a monster and everything suffered. Silos could not communicate seamlessly. Overly complex matrix teams were created to try to overcome the rigidities. The decision-making process slowed. And middle and upper managers lost their command of the fast-evolving science.

In the late 1990s I concluded that the organizational pyramid had become obsolete and decided to break it into a constellation of highly focused centers of excellence designed to improve transparency, increase the speed of decision making, and restore freedom of action to the scientists actually conducting the research. GSK now has 12 centers. Each is focused on a family of related diseases (for example, Alzheimer's and other neurological diseases, or diabetes and obesity), has a CEO with the authority to initiate and kill projects, and contains a few hundred scientists from all the crucial disciplines. There are only two or three management layers between each center's CEO and key bench scientists.

The lessons learned in breaking the R&D pyramid into a constellation are many. The centers must be built around two things: a specific mission (such as discovering the most effective therapies for Alzheimer's and other neurological diseases) and the stage of the R&D process required to perform that mission (such as the choice of a particular target for attacking the disease). Anything that is not critical to the core mission and R&D process must occur outside the center of excellence. This means that all other functions--toxicology, drug metabolism, formulation, and so on--indispensable though they might be, must become service units. Leaders of these units must be selected for their ability to deliver an effective service at the lowest possible cost, which should be regularly benchmarked. We compare their performance to that of outside organizations selling the same services for a profit. This new construct has enabled us to adopt a make-versus-buy approach. Finally, the roof of the old pyramid--the functional senior vice presidents--and any other vestiges of the silo organization should be eliminated. Their responsibilities should be distributed throughout the decentralized organization.

Improve the Quality of Leadership

The R&D process is grueling and discouraging. Most projects fail, and few scientists ever experience the thrill of producing a successful new drug. Drug discovery and development is a team sport, with moments of brilliant thinking and hours of painstakingly detailed work. In such a difficult environment, inspiring and nurturing leaders who are also accomplished scientists is indispensable.

R&D organizations, however, have traditionally promoted their best scientists to management positions--sometimes without paying enough attention to leadership abilities. Complexity and the leadership void have given rise to teams that focus too much on process and too little on producing meaningful results, and have allowed sleepwalkers and nine-to-fivers to hide in the pyramid.

Recognizing that every project needs a strong leader, we have aggressively embarked on a program to build a cadre of exceptionally gifted individuals. Such leaders are fairly easy to spot. They love the science, show passion in their desire to win, have the resilience to soldier on in the face of multiple setbacks, and genuinely care about the members of their teams. These inspiring product finders typically represent less than 1% of the entire R&D population, but their value is exponentially greater. They must be identified, protected, and supported.

Launch a Cultural Revolution

When I was a graduate student in pharmacology at the University of Louis Pasteur in the late 1960s and early 1970s, my thesis directors were two eminent scientists who would often review my work on a Monday. Come most Fridays, I descended into a panic because I was convinced that I had not produced anything worthy that week to show them. I then literally lived in the lab through the weekend, desperate to come up with something that might impress them. I even had a bed in the lab. That is the kind of engagement and passion I want all our scientists to share. We need every one of them to be consumed by the desire to win against all odds, to make a difference.

Such a culture existed in drug companies 50 or 60 years ago, before they became giants. But restoring it is going to take a revolution. Frankly, I think it's our biggest challenge, because scientists are a unique breed: They are nonconformists; they care much more about external recognition than money; and they tend to tune out the rest of the world. So even though their awareness that not all is well has been raised by numerous staff reductions in R&D and beyond and by questions about Big Pharma's long-term survival, that's not nearly enough.

At GlaxoSmithKline we've overhauled our incentives in R&D. We've tailored the bonus system to reward scientists for what they do, not what the rest of the company does. For example, when a potential drug reaches the proof-of-concept stage (its efficacy and safety have been proved), it triggers a significant payout to the core team of discoverers. We also give handsome rewards to scientists for solving major problems--such as figuring out how to make a previously insoluble drug soluble. Why do this if scientists aren't especially driven by money? Because I'll do anything that makes even a little difference. If someone convinces me that serving beer in the parking lot at five o'clock every afternoon will help attract and retain good scientists and get them to work harder, I'll do it.

My belief is that only when the right leaders are in place will the right culture emerge. R&D leaders must restore a sense of purpose to every project team, while requiring engagement, accountability, and transparency. They must establish inspiring objectives (it's all about the patient and conquering the disease) that motivate people every day, every hour, in every cubicle, in every lab. Project teams must be empowered, and their progress (or lack thereof) must be made visible to all. Success must be celebrated and generously rewarded. Poor performers should no longer be shuffled from project to project. To make our scientists feel that they have skin in the game, resource allocation must be reinvented so that teams have to compete for funding.

In short, we've got an enormous amount of work to do. At GSK we have planted a lot of seeds, but the flowers have yet to bloom.

Overhaul R&D Processes

I have two radical ideas for changing R&D processes. One--separating R&D for first-in-class drugs from that for best-in-class drugs--is in progress at GlaxoSmithKline. The other--something I call "the progressive blockbuster"--would require a reinvention of the clinical stage of R&D for breakthrough drugs and is just a notion at this point.

Separating first in class from best in class.

The R&D organization of a major pharmaceutical company typically pursues two objectives: to be first in class by discovering new targets and disease mechanisms and producing a breakthrough medicine, and to come up with the best-in-class compound for attacking a validated disease target. Currently, the pursuits of these objectives are intertwined in most R&D organizations. That is a mistake, because the requirements and risks of the first objective differ radically from those of the second. A significant means of improving R&D productivity in the short term is the intelligent separation and optimization of these two distinct activities.

To be first in class means to go where no one has ever gone before. Truly understanding the fundamental pathology of a disease requires resources, concentration, and the building of institutional knowledge and competence over a long period of time--as long as 25 years. From an economic standpoint this is by far our riskiest venture, because it involves pioneering work in biology (to discover a new target), in chemistry (to find the right compound), and in drug development and approval (to make the chemical safe and effective and to gain approval in an area where regulators lack established guidelines or rules). Despite considerable efforts, Big Pharma has had limited success in the first-in-class arena. When we do succeed, however, it enhances our standing with payers, patients, and external constituencies. When all is said and done, curing disease is our raison d'être. Without the discovery of first-in-class drugs, Big Pharma would lose its soul.

Once a target for attacking a disease has been validated, the race is on to discover and develop the best-in-class molecule to attack it. Time is of the essence; chemistry is king; experimental biology is irrelevant. The objective is to optimize the molecule--to develop a better mousetrap. Pfizer's Lipitor, which for years has been the uncontested best statin, was the fifth statin discovered. Because the search for a best-in-class medicine does not involve breaking new ground in biology or in drug development and approval, it is much less risky than a first-in-class program but has the same--or sometimes even greater--financial rewards.

By intertwining the processes for pursuing first in class with those for pursuing best in class, we have been demanding that R&D try to perform as a ballet dancer and a football player at the same time. We end up with low productivity in both pursuits, which is a major reason why Big Pharma organizations rarely discover truly novel molecules and have become slow followers in finding best-in-class compounds on a consistent basis.

The clear separation of these activities where possible (complete separation is impossible) will require a significant retooling and redesign of industry efforts. For first-in-class R&D we must select many fewer endeavors, concentrate our resources, staff the effort with true discoverers, intensify our partnerships with leading academic centers, place bets on novel approaches through venture-capital-like investments, and stay on course decades or longer to succeed.

To optimize best-in-class R&D, several activities need to be strengthened. The ranks of the chemists should be reinforced, and they should be supplied with cutting-edge tools such as structure relation analysis, a software program that can help find the best candidate for a medicine from among thousands of compounds. The capability to conduct rigorous patent reviews should be improved to help teams navigate their fields of research.

Untangling the two processes will require R&D to rethink its organizational setup, allocation of resources, recruitment, culture, scorecard, and incentive systems. I fully understand that all this is much easier said than done. One especially big challenge will be deciding what to separate and what to keep together in order to minimize the duplication of functions. Although it's clear that processes and resources supporting first-in-class and best-in-class projects should be separate in the start-up phase, that's not the case further along. For example, the point at which a new target becomes a validated target is often ambiguous. And to make the best possible use of the thousands of people participating in a clinical program, first-in-class and best-in-class projects should share them. Such realities illustrate the execution challenges related to this process transformation.

As difficult as the untangling will be, it will be worth the effort. As we've seen in those centers of excellence where GSK has taken this step, the successful execution of this new model will produce a significant uplift in R&D productivity.

The progressive blockbuster.

This idea would involve abandoning pursuit of the instant blockbuster in order to first target a limited segment of potential patients and then expand to other segments over time.

Traditionally, new drugs with potential for treating a common disease or condition such as diabetes or high cholesterol have been tested on a large, diverse patient population in clinical trials. If the trials produced the expected results and the FDA approved the drug, the result was an "instant blockbuster." However, it is common for a drug to cause side effects in a small segment of the target population. Years ago these side effects would often go undetected until after the drug was on the market and millions of patients had used it. Today the FDA requires companies to try to identify rare or unexpected side effects before a product launch; this has led to a record number of initial rejections and, as a result, delays in introducing new drugs.

A solution would be to begin by limiting the clinical trials to a highly uniform segment of patients--those who, for example, have a specific genetic profile or suffer from exactly the same concomitant conditions, such as diabetes and cardiovascular disease. If testing produced good results, the company could seek the FDA's approval for marketing the drug to that segment only. This approach would reduce the likelihood that rare or unexpected side effects went undetected during clinical trials and would make it easier to monitor the drug's performance after launch. Development of the drug should continue in the same way: A second uniform patient segment should be chosen. And so on.

Such systematic clinical development would address society's low tolerance for surprises (side effects in even a tiny subset of patients) and should simplify and speed up product development. Ultimately, everyone would win.

Capture Cost Efficiencies

Companies can leverage their huge R&D budgets by reducing waste and improving the quality of decision making. One area in which the opportunity for savings is huge is Phase II and Phase III clinical trials, which test the safety and efficacy of drugs on the targeted patient groups. By switching 50% of its trials from high-cost places such as the United States and Western Europe to low-cost places such as India and South America, a midsize pharmaceutical company with 60,000 patients in clinical trials could save $600 million annually. (A top-notch academic medical center in India charges $1,500 to $2,000 per patient case report, while a second-rate center in the United States charges $20,000.)

Offshoring management of clinical trial data and using IT to automate parts of the clinical trial process would also generate significant savings, and there are many other ways to cut costs as well. The key to identifying and capturing the biggest opportunities is to strengthen R&D organizations in the way I've described above: by devolving power to strong project teams that are staffed by people who know the most about the work, are supported by a culture that empowers leaders, and are guided by a transparent and science-based framework for making decisions. Decisions tainted by wishful thinking or unduly influenced by bureaucratic or political considerations constitute one of the most wasteful elements of the R&D process and must be vigorously attacked.

Place More Bets

Improving R&D productivity alone won't be enough to save Big Pharma. Given their enormous need for new products, companies are going to have to place more bets. This will require them to change their business model so that they can increase R&D funding. The industry has been investing about 16% of revenues in R&D for the past 10 years. In view of the "all or nothing" consequences of R&D success, it seems prudent (and competitively advantageous) to raise the stakes gradually. The funding of this expansion should be made possible by shrinking an oversize selling and marketing machine. The best companies have already started on this quest.

At the same time, companies must forge alliances with academia and biotech companies. The model of a company concentrating its R&D resources inside its four walls is obsolete. Big Pharma players can no longer hope to generate the absolutely best science in all areas on their own. Therefore, standard operating procedure should be to decide on a scientific bet (for example, kinases in oncology), shop around among all the external players that are pursuing such research, and establish a contractual relationship with the best. The CEOs of our centers of excellence have the power to decide what to do inside and outside. At the end of the day we judge them on what's in their pipelines, regardless of the source.

An open architecture for R&D projects has many advantages. It forces competition between internal and external scientific teams as well as between different approaches to a common therapeutic solution. It makes the enterprise more flexible, allowing it to cancel programs without having to go through painful restructuring. In order to operate in this fashion, however, companies will need to strengthen and expand their ability to evaluate opportunities, negotiate deals, and nurture external scientific bets on a large scale--in other words, to act like a venture capitalist with in-depth understanding of the science at stake.

The innovation malaise in the pharmaceutical industry is not unique. Many other industries face the same challenges. The scientific revolution is visible to us all, but its translation into significant goods and services remains a distant goal for most large, established corporations. So instead of waiting for the scientific revolution to save us, we must rebuild our R&D engines. In most cases a cultural shift and a broad transformation of the organization are necessary first steps. This is an enormous task. In pharmaceuticals only the very best players will be able to meet the challenge and rebuild their R&D engines. The approaching wave of patent expirations leaves them little time. From Leaders to Laggards

The graph below compares the average annual return to shareholders (stock appreciation plus dividends) of various industries, weighted by market capitalization and expressed in percentages. The pharmaceutical industry's significant lead in creating shareholder value evaporated between the two time periods shown. This occurred even though the industry's gross and EBITDA margins rose steadily for 20 years--to 78% and 32%, respectively, in 2005.

An Old Model Under Attack

Big Pharma's P&L structure has long been heavily weighted toward sales and promotion and, to a lesser extent, R&D. Until the 1980s this model was extremely successful. The industry's revenues soared, and its profit margins and total return to shareholders were among the highest relative to other industries. Since the early 1990s, however, its world has dramatically changed, threatening that model.

Shorter product monopolies have fueled marketing wars. As companies expanded their research capabilities and raced to move into any promising new opportunity that emerged, the duration of product monopolies within a given therapeutic class shrank from several years in the 1970s (for, say, Tagamet, an H2-receptor antagonist for treating gastric ailments) to a few months in the 1990s (for, say, saquinavir, a protease inhibitor for treating HIV). This trend intensified marketing wars and led to enormous sales forces. Rule changes in 1997 that allowed companies to broadcast advertising in the United States only escalated the battle. In 2006 the top seven pharmaceutical companies spent twice as much on SG&A (about 33% of revenues) as on R&D (about 16% of revenues).

A slowdown in new-product introductions has left companies with oversize sales and marketing machines. Starting in the 1980s, R&D productivity (as measured by the average cost of discovering and developing new drugs) steadily declined. Meanwhile, the number of new blockbuster drugs (those with annual global sales of $500 million to $1 billion) has diminished. Some major drug companies have trimmed their sales and marketing functions or consolidated them in mergers, but at most companies these functions are still too large relative to the products they have to sell.

Pricing pressure is increasing globally. Private and public payers around the world are willing to spend a lot more for truly innovative medicines: Daily treatment reimbursement rates for the new generation of oncology drugs are nearly a thousand times as high as for past oncology treatments. But such drugs are a tiny minority.

Developed countries lack a standardized system for determining the value of new products or line extensions that offer incremental or hard-to-discern improvements, making it difficult for pharmaceutical companies to win price increases for such products. Their success has heavily depended on the type of product (new category versus multiple competitors), the status of a country's health care budget, and ancillary factors such as R&D and manufacturing investments.

Overall, U.S. prices continue to surge ahead of those in developed countries with single-payer systems, and developed countries pay more, on average, than emerging or poor nations. A few pharmaceutical companies, including GlaxoSmithKline, now take a country's standard of living into account when setting prices. They charge higher prices in the United States, Europe, and Japan; intermediate prices in developing countries; and not-for-profit prices in poor countries (as defined by the United Nations). Chances are good that pricing pressure will only increase.

Revenues plunge when patents expire. Until the mid-1980s the revenues of branded drugs declined gradually after their patents expired. No longer: A U.S. law requiring pharmacists to substitute generic versions of drugs when available, together with pressure on consumers from payers to buy lower-cost generics that differ somewhat from branded drugs, has accelerated a shift to generics. As a result, the soft landing for branded drugs has disappeared. Today multibillion-dollar blockbusters lose 90% of their revenues and profits within a few weeks of losing their patent protection.

Meanwhile, with generics the benchmark, payers are less and less willing to reimburse for line extensions at a premium. This has made it tough for pharmaceutical companies to extend the life cycle of existing medicines.

All costs are continuing to climb. Various third-party studies have assessed the cost of discovering and developing one new product at more than $1 billion--a dramatic increase from less than $100 million 15 years ago. Significant factors include regulatory requirements; larger, more complex clinical trials; the growing public debate about drug safety; and lawsuits.

Trends that will significantly increase demand for innovative drugs over the long term won't have a major impact on overall global volume anytime soon. Developed nations will continue to consume innovative medicines, albeit at a slower pace. But generics now account for 70% of all prescriptions dispensed in the United States, and will continue to gain market share as multiple blockbusters lose their patent protection.

Several trends will ultimately provide a major opportunity for the pharmaceutical industry, but their impact on total global demand for branded drugs will be gradual. They include:

• Demand from emerging economies. Countries with high standards of living that collectively make up less than 20% of the world population (the United States, European nations, Japan) now account for 80% of the industry's global revenues and profits. This will change as rapidly developing countries such as China, India, Turkey, Russia, and Vietnam gain economic power and gradually improve their health care systems.
• Aging populations. The number of Alzheimer's patients in the United States, for example, is expected to grow from 5 million to 20 million within 30 years.
• New diseases. Although impossible to predict, the surfacing of new diseases comparable to AIDS and avian flu will trigger demand for new therapies.

When Winning Is Everything

The primal urge to win often overwhelms rational decision making. Here's how to tame competitive arousal, head off emotionally charged competitions, or manage them to your advantage.

by Deepak Malhotra, Gillian Ku, and J. Keith Murnighan

Have you ever made a decision in the heat of competition only to wonder, when faced with the consequences, "What was I thinking?" Such charged decision making is driven by an adrenaline-fueled emotional state we call competitive arousal. It's all too common in business--and all too often leads to costly mistakes.

Consider Boston Scientific's disastrous acquisition of medical-device maker Guidant. In December 2004, Johnson & Johnson announced plans to acquire Guidant for $25.4 billion. Soon after, Guidant recalled 170,000 pacemakers, 56% of its total production. Not surprisingly, J&J threatened to pull out. Guidant responded by suing J&J to complete the deal, and J&J countered with a reduced offer of $21.5 billion.

Suddenly, Boston Scientific--J&J's longtime rival--offered $24.7 billion for Guidant. This triggered a bidding war, even as Guidant's financial and public relations woes worsened. The bidding finally ended in January 2006 with Boston Scientific's offer of $27.2 billion--$1.8 billion more than J&J's initial bid.

Was this a good deal? In June 2006, Boston Scientific had to recall 23,000 Guidant pacemakers and advise 27,000 patients who'd already had them implanted to consult their doctors. Boston Scientific's share price, which was near $25 at the start of bidding, fell below $17. Fortune later called the Guidant acquisition "arguably the second-worst ever"--behind only AOL's infamous purchase of Time Warner.

The Guidant case exemplifies the decision errors that can result when managers and executives, overcome by competitive arousal, shift their goals from maximizing value to beating an opponent at almost any cost. Fortune's investigation of the Guidant acquisition reveals the role that competitive arousal played in the bidding war. "What emerges is a roller-coaster tale of bet-the-franchise corporate brinkmanship, miscalculation and overreaching," Fortune's writer noted. "It is a stark lesson on how the single-minded pursuit of victory can blind even brilliant execs to the true costs of a deal....it sheds new light on

• is marked by deep personal animosity and ferocious combat in the marketplace and the courtroom...."

As we'll show, there is strong evidence that competitive arousal has fueled many other high-profile business mistakes. In a variety of contexts--be they auctions, negotiations, legal disputes, mergers and acquisitions, employee promotion contests, or the pursuit of hot managerial talent--decision makers can easily become fixated on beating their competitors. There is nothing necessarily irrational about incurring a cost to win; people enjoy winning--especially against their rivals--even at a price. There may actually be strategic benefits in doing so: If winning a contract will damage a competitor for the long term, it may make sense to pay more than fair value for that victory. But cases like that require a clear, upfront analysis of the limits of acceptable losses and the benefits that winning will yield. When analyses are conducted in the heat of the moment, competitive arousal crowds out clarity. The result is often a Pyrrhic victory.

In this article, we describe the causes of competitive arousal, when it is most likely to derail strategy and destroy value, and how managers can avoid or reduce its pernicious effects. We identify three principal drivers of competitive arousal in business settings: rivalry, time pressure, and audience scrutiny--what we call the "spotlight." Individually, these factors can seriously impair managerial decision making. Together, their consequences can be all the more dire.

Sidebar Icon The High Cost of High Stakes (Located at the end of this article)

Rivalry

As the Guidant case suggests, competitive arousal is most common--and most dangerous--when rivalry is intense. In our research, we theorized that head-to-head rivalry would interfere with rational decision making more than any other kind of rivalry. Our first test of the theory was a large field study of auctions of artist-designed, life-size fiberglass cows. The proceeds of the auctions would go to charitable causes. In November 1999, 140 cows were auctioned live in Chicago and on the internet, generating almost $3.5 million--seven times initial estimates. Other cities soon followed suit, auctioning fiberglass cows as well as pigs, moose, fish, and bears. We collected bidding data from these auctions (including chronological listings of how much was bid and by whom) and conducted pre- and postauction surveys. We found that people were more likely to bid past their preset limits when they were competing against a few rather than many other bidders: A small field--particularly a field of just two--created intense feelings of rivalry, more bidding, and more overbidding. We've seen this same phenomenon in laboratory experiments: People tend to overbid more when they face only one bidder (even when we control for their perceived chances of winning); they also report more psychological arousal (in the form of excitement and anxiety), which fuels their overbidding.

Rivalry can derail one-on-one negotiations, too. Consider a large Asian conglomerate that recently contemplated selling its telecommunications holdings. Because one potential buyer had been a competitor, there was personal animosity between the two companies' executives. As the conglomerate considered the sale, one of its owners told us, "We don't care if they are willing to pay $100 million more than everyone else. We won't sell to them. We don't want to give them the satisfaction....Now we have what they want, and we're not going to let them score one last win." It did not seem to matter that the seller and the buyer would no longer be competitors once the assets were sold; what did matter was their past enmity and the chance to thwart an archrival.

Time Pressure

A ticking clock--in auctions, negotiations, disputes, and other competitions--can overwhelm people with the desire to win. In live auctions, bidders must make decisions rapidly as the auctioneer calls out, "Going once! Going twice!..." Job seekers often confront "exploding" offers. Home sellers get bids that can expire in a matter of hours. Potential acquirers often face bidding deadlines.

Years of research has provided compelling evidence that time pressure seriously impairs decision making by increasing psychological arousal, which decreases the ability to find and apply relevant information and leads to an overreliance on simple decision heuristics, such as pursuing a strategy simply because it has worked in the past. In our auction studies, we found that people were more likely to bid past their preset limits as the deadline approached (this finding held even after we controlled for the size of their bids and the number of other bidders). When time was running out, bidders seemed increasingly fixated and limited in their thinking, clinging to the hope that their next bid would clinch the deal. Although one more bid may indeed increase the chances of winning an item, it can also result in overpaying for it.

Sometimes the rules of the game include deadlines, so time pressure is unavoidable. Other times, though, decision makers create unnecessary pressure for themselves. One executive told us that he used to negotiate important deals over breakfast because this provided an informal atmosphere at a time of the day when he was at his best. After years of using this approach, he came to understand that a breakfast meeting did not give him enough time to seriously consider and respond to unexpected proposals and requests. During these breakfasts, he realized, he had often agreed to provisions and price concessions that he later regretted.

The Spotlight

Psychological research shows that the presence of an audience, particularly one that's highly engaged, increases psychological arousal and can reduce performance on physical tasks as well as on tasks that require problem solving or creativity. This means, for example, that live auctions incite considerably more competitive arousal than internet auctions. When an auctioneer or a bid spotter has singled you out with a challenge to increase your bid, and the audience is watching your every move, it's hard to avoid a rush of adrenaline and the urge to bid even higher. In contrast, online bidding is more private; the "spotlight" is dimmer. As a result, online bidding tends to be more temperate and rational. In the fiberglass-animal auctions, for example, internet overbidders exceeded their limits, on average, by $1,134; live-auction bidders overbid by an average of $5,609--almost five times as much. We also found that when these auctions garnered less media attention, which dimmed the spotlight further, fewer people overbid. This finding held even after we controlled for a host of macroeconomic variables, such as stock market performance and consumer confidence index.

In negotiations, bidding wars, and business disputes, the strength of the spotlight can vary considerably. Some disputes are public affairs; others are shielded by gag rules. The brighter the spotlight, the greater is the potential for competitive arousal and bad decisions. The Blackstone Group, a prominent private-equity and investment management firm, provides an example. In early 2007, under the glare of the business-media spotlight, Blackstone acquired Equity Office Properties Trust, the largest owner of office buildings in the United States. Blackstone's final offer of $23 billion in cash and an assumption of $16 billion in debt surpassed Vornado Realty Trust's final offer by $3 billion and was the largest private-equity deal ever. Would Blackstone have been willing to pay such a premium without the tremendous media attention? Our research on competitive arousal suggests that the spotlight may have been influential in this case. In its coverage of the deal, the Wall Street Journal highlighted the increasing tension created by the sometimes incompatible goals of wanting to score a much publicized win and wanting to make a sound economic decision. "The culture of private-equity giants like Blackstone," the Journal wrote, "is built on two competing foundations, the reluctance to lose any deal--particularly one as big as this--and an equal unwillingness to pay too much for any deal."

A Perilous Mix

Rivalry, time pressure, and a bright spotlight can each fuel competitive arousal. Collectively, they can lead to decision disasters. We demonstrated this in a recent experiment, in which participants who were outbid in an online charity auction received one of three messages: a default message informing them that they had been outbid and could continue bidding by returning to the auction's website; a charity message appealing to their sense of compassion ("We hope you will continue to support this charity"); or a competitive message fueling their desire to win ("The competition is heating up... are you up for the challenge?"). Bidders who received the competitive message were 50% more likely to bid again than those who received default or charity messages if two conditions were simultaneously met: high time pressure because it was the last day of the auction and high rivalry because only one other bidder remained.

In another case, involving a company we consult for, we saw rivalry, time pressure, and a spotlight conspicuously reinforcing one another. The company had hired us to advise it in its first business-to-business reverse auction. As with many such auctions, our client and its unidentified competitors were prequalified to bid in a computer-based auction in which the lowest bidder was contractually bound to provide services at that price. A company vice president was in charge of making the key decisions, but 10 of his colleagues were watching, placing him under intense scrutiny. The format of the auction gave him precious little time to make his bids, even though the stakes were many millions of dollars. Before the bidding ended, the VP had bid well below his previously calculated, rational limit. Luckily, the company lost the auction. The VP later admitted that he had stopped when he did only because he was unsure who the remaining rival was; had he known it was his company's biggest competitor, he said, he probably would have continued to bid.

The pitched battle for Paramount Pictures offers yet another example of how the combination of these potent factors can distort decision making. In late 1993, longtime rivals Sumner Redstone (then the CEO of Viacom) and Barry Diller (then the CEO of QVC) locked horns in a public pursuit of Paramount. After a series of bids and counterbids, Viacom triumphed. Research by Pekka Hietala, Steven Kaplan, and David Robinson suggests that Viacom overpaid by $2 billion and that QVC would have overpaid by $688 million had it won. Indeed, Redstone himself admitted to Time that as a result of the bidding war he paid about $1.5 billion more than he had intended to. The magazine described how the decision-making process looked behind the scenes: "Time was running out on the Viacom executives and advisers who hunkered down to a Sunday-afternoon skull session....Unless Viacom came back fast and hard, everyone present knew, the fight would soon be over....one thought dominated all those at the meeting: how to throw a knockout punch that would be, as one of them put it, a `Diller-killer.'" This hardly reads like a description of rational decision making. Rather, a long-standing rivalry, mounting time pressure, and a glaring media spotlight seem to have converged to drive the overpayment. The root of the problem can perhaps be found in the title of Sumner Redstone's autobiography, published a few years after the Paramount acquisition: A Passion to Win.

Sidebar Icon How Lawyers Can Fuel Competitive Arousal (Located at the end of this article)

Managing Competitive Arousal

The risk factors for competitive arousal are ever present: Managers and executives must constantly deal with rivals, make quick decisions, and operate in the public eye. They can minimize the potential for competitive arousal as well as the harm it can inflict by following two broad strategies: avoiding certain types of competitive interaction and mitigating the risk factors.

Circumventing competition.

Managers can prevent competitive arousal altogether by anticipating potentially harmful dynamics and then creatively restructuring the deal-making process. Consider how NBC revised its approach to deal making after a protracted and costly negotiation with Paramount over the TV rights for Paramount's hit comedy series Frasier. The show had aired on NBC for eight years and had done spectacularly well. As the contract period neared its end, Paramount demanded a three-year contract with a 10% price-per-episode increase, despite the show's (arguably) waning popularity. NBC executives wanted a lower price and a one-year contract (or a provision allowing them to forgo the third year if ratings faltered) and were confident that no other network could profitably pay even that amount for Frasier.

But Paramount was in a peculiar position: Its parent company, Viacom, had recently acquired NBC's chief rival for the Frasier deal, CBS. This created a difficult (and unexpected) competitive dynamic for NBC. Even though NBC executives were sure that CBS could not profitably outbid NBC for Frasier, CBS might persuade Viacom to force Paramount to sell to CBS just to score a win against NBC. This implied threat may have been sufficient to stimulate competitive arousal among NBC's executives, but several other risk factors were also present: The media were all over the story. The lead negotiator for NBC, Mark Graboff, had recently been hired away from CBS, and Frasier was his first negotiation on the job, putting him under both external and internal spotlights. And time pressure was intense--negotiations continued not only beyond the contract's expiration date but also past the deadline set for an exclusive negotiation period between Paramount and NBC.

NBC got the show but appears to have paid more than it was worth. Frasier did well for one more year but then, consistent with the network's initial analysis (and biggest fears), had its two worst years. Alas, NBC had not managed to negotiate the provision allowing it to forgo those disastrous years.

As a result of this experience, NBC restructured its deal-making process. In particular, the company began to include a "no self-dealing" clause, barring producers negotiating with NBC from shopping their shows to networks they own (or to those owned by their parent company). We've seen other companies avoid damaging bidding wars and escalating competition by using rolling contracts, which can be renewed prior to expiration without lengthy renegotiation, limiting the possibility that competing bidders will enter and trigger irrational price wars. Similarly, noncompete clauses can help firms avoid hypercompetition with rivals that might otherwise threaten to steal their star employees. And, having learned from bitter experience, some companies avoid B2B reverse auctions altogether because bidding wars can drive competitors to make irrationally low bids.

Mitigating the risk factors.

It's not possible to avoid potentially destructive competitions and bidding wars completely. However, by targeting the underlying causes of competitive arousal, organizations can head off the escalation of irrational behavior.

Sidebar Icon Defusing Competitive Arousal (Located at the end of this article)

Defusing rivalry.

The desire to win at any cost is most powerful when the competitor is an erstwhile nemesis or is seen as evil. In such instances, it is helpful to remember that competitors are simply parties with their own interests. Like you, they are probably smart, reasonably rational, and somewhat emotional, and they probably see you as the nemesis. Of course, it is not always easy to control your feelings about rivals, but adopting your competitor's perspective can promote cool, rational decision making, even in tough competitions.

When rivalry is intense, organizations should consider limiting the roles of those who feel it most. For instance, the COO of a large family-owned business, one of our clients, is a seasoned manager who was once a top political administrator in his country. His role in the company is to act as a devil's advocate and thereby temper the CEO's competitive arousal. His stature allows him to speak truth to power even when no other executive or board member feels comfortable doing so. In more than one instance, the COO has persuaded the CEO to step aside as the lead decision maker or negotiator when the CEO's feelings of rivalry seem to be undermining his judgment. As this COO understands, when an arousal-prone negotiator can't control the competitive fire in an intense rivalry, it may be in everyone's interest to remove him or her from the bargaining table.

Managers for the National Hockey League saw the wisdom in this approach, albeit belatedly, during a highly publicized labor dispute in 2005 and 2006. Most observers realized that the rivalry between commissioner Gary Bettman and Players' Association executive director Bob Goodenow made effective negotiating impossible. The two had a history of contentious, win-lose negotiations and even disagreed on details as trivial as how many meetings they had conducted. As a result, their seconds-in-command (chief legal officer Bill Daly for the owners and senior director Ted Saskin for the players) took over many of the substantive discussions, preventing the conflict from escalating further.

Another way to dilute the impact of rivalry is to quantify early on--before competitive arousal kicks in--how much you are willing to lose in order to "win." In negotiations, auctions, legal disputes, and the like, this involves identifying not only the benefits that can accrue from the deal (assets, synergies, reputational advantages, and so on) but also the costs that you are willing to incur in the name of pride and ego.

We recently worked with an executive to do exactly this. He was planning to sell his 50% stake in a company to his partner. They had come to hate each other and could no longer work together effectively. When we asked how much he would accept for his share in the company, he declared that he would not sell for less than $8 million. After some financial analysis and a consideration of his alternatives (whether he could sell to an outside party), we calculated that his bottom line should be closer to $7 million. "He doesn't deserve that good a price," the seller responded. After explaining that he was implicitly pricing his pride and his ego at $1 million, we asked him to explicitly put a price on beating his partner and to enter it into the spreadsheet we had created. After thinking this through overnight, he priced his pride at $200,000; he acknowledged that his original, emotional response had led him to overprice this element of the deal and that he was comfortable with the new figure (as large as it seemed to us).

Once you have carefully assessed your limit in an auction, a negotiation, an acquisition, or a dispute, publicizing it to colleagues or to your executive board can provide additional safeguards against competitive arousal. Public commitments make it harder to violate your limits.

Reducing time pressure.

Who can stop a bidding war that has spiraled out of control? Let's reexamine the battle over Paramount. Time reported that Barry Diller came to his senses at the end of 1993, realizing that the acquisition was strategic and financial rather than a competition with Sumner Redstone. His first step was to defuse his dangerous feelings of rivalry. He then removed himself from the time-pressured deal-making environment. "Diller had packed up 10 lbs. of Paramount documents and hauled them along on a year-end Caribbean vacation," Time reported. "Running the numbers while onboard the rented yacht Midnight Saga as he cruised off St. Barts, Diller decided that Paramount was not worth a penny more than the $10 billion in cash and stock that QVC was bidding." When Diller returned from the trip, he held to his original offer--and lost the deal. Although time pressure can be imposed from outside, as with deadlines, it is essentially perceptual, shaped by how one feels about time ticking away. By shifting contexts, from his office to his yacht, Diller reduced perceived time pressure, slowing the ticking clock so that he could make a calculated, rational decision.

Effective decision makers create time to reevaluate the bases of their value calculations, to discuss substantive issues, and to negotiate. Because individuals consistently underestimate the time needed for complicated tasks--and consistently overestimate their ability to make wise decisions, particularly under time pressure--it is easy to set ill-considered deadlines that lead to bad decisions.

Extending or eliminating arbitrary deadlines may serve the purposes of both parties. Recent negotiations between Comair and its flight attendants and pilots are an excellent example of this kind of temporal flexibility. The airline's union and management agreed to extend negotiation deadlines twice, in both cases to prevent time pressure from leading to value-destroying competitive behaviors. As Captain J.C. Lawson, chairman of the Comair pilots' union, said, "The pilots and flight attendants will not have contract terms imposed upon them, regardless of some artificial deadline." Comair spokeswoman Kate Marx conveyed the same message, albeit in different terms: "We have agreed to the deferral as a way to continue our work toward a consensual agreement, which has been our goal since we began this process over a year ago." Given the simplicity of deferral as a solution, executives need only ask themselves in negotiations and in other high-stakes decisions, "Do I really need to make this decision today?" If the answer is no, a short deferral can be tremendously effective. Asking this simple question, however, may not always be easy--after all, it took our negotiate-over-breakfast executive several years to recognize the time trap he had set for himself.

Deflecting the spotlight.

A spotlight can originate outside an organization--in the media, for example. It can also shine from within, in the form of observant colleagues. To counter the effects of an internal spotlight on competitive arousal, organizations should consider spreading the responsibility for critical, competitive decisions across team members. That way, no one manager or executive will stand alone in the spotlight. For instance, one of our clients, a financial services company, recently adopted a policy requiring sales managers to report any large push by a customer for price concessions to a team of peers and bosses. The team then formulates a strategy and takes collective responsibility for the consequences. In the past, if clients threatened to switch to a competitor, sales managers feared that they would be held personally responsible for the loss and that their failure would be highly visible throughout the organization. This anxiety often pushed them to agree to extreme and unnecessary concessions. The new policy deflects the spotlight, allowing them to feel more comfortable holding the line on prices.

Another way to deflect the spotlight is to put individual managers in charge of multiple accounts and judge them on their overall--rather than on account-specific--performance. Editors at publishing companies, for instance, are often caught in emotionally charged bidding wars as they try to woo attractive authors. A big deal may come along rarely for an individual editor, and so the spotlight shines brightly. It's not hard to understand why he or she may have a hard time resisting competitive arousal. To address this problem, some book publishers allow only one person in the organization--the CEO, who is familiar with the company's entire portfolio--to approve advances above a set amount (for example, $250,000). The CEO won't suffer the glare of the spotlight because each deal represents just one of many.

Many of the acquisition battles we have discussed here have had widespread media coverage, with analysts dissecting the competitors' every action and reaction. The media spotlight can be difficult to deflect, but advance work can limit its effects on decision making. If negotiators anticipate that an acquisition will make headlines, they should carefully calculate their reservation prices before word of their potential bid hits the newsstands. These calculations should include the premiums that an acquirer is willing to pay in a variety of scenarios (for example, if competitor X enters the bidding or if the media begin to play up the rivalry between bidders). Although this prescription is not difficult to follow, firms routinely wait until after a new competitor has outbid them to reevaluate the premium they're willing to pay. By then, competitive arousal may be clouding their judgment.

In some contexts, competitors are shielded from the spotlight as a matter of policy: Dispute mediations often have gag rules that prevent parties from discussing the process or the outcome. This policy has a dark side, of course, which must be considered before implementing it, especially in out-of-court settlements involving harmful behavior that might be repeated. For example, a company that has created products that can harm consumers may be free to continue manufacturing these products if guilt cannot be disclosed.

Whenever competitive arousal can be anticipated--whether because of rivalry, time pressure, or the spotlight, or because all three are likely to emerge--mental preparation can be an important defense. A simple and effective way to avoid unwise competitive behavior is to consider not simply prior mistakes but potential mistakes that may occur in competitive interactions. A number of our former students and clients role-play to prepare for major negotiations. By simulating upcoming deals, negotiators can anticipate the emotional reactions of competitive arousal and avoid behaviors that might derail otherwise sound strategy. Our research has shown that if managers do not have time to simulate an entire deal, they can still help avoid missteps by imagining future regrets about overpaying.

• o o

Research clearly demonstrates that we tend to overestimate how rational, careful, and logical we are. We are also prone to believe that others are more susceptible than we are to irrational decision making. Both of these biases make it easy to ignore or underestimate the harm that can befall us as a result of competitive arousal. Executives and managers will be most successful in competitions when they not only prevent or mitigate competitive arousal, but also set in place organizational processes that help focus their competitive energies toward efficiently winning contests in which they have a real advantage--and away from those in which winning comes at too high a cost. The High Cost of High Stakes

Many of the examples of competitive arousal we cite in this article entail high-stakes decisions, suggesting that even when millions or billions of dollars are on the line, the overwhelming desire to win can derail sound strategy. O