A Market-Driven Approach to Retaining Talent

A Market-Driven Approach to Retaining Talent

Traditional strategies for employee retention are unsuited to a world where talent runs free. It's time for some fresh thinking

If YOU'RE LIKE MOST EXECUTIVES TODAY, YOU'RE A poacher. You regularly look outside your organization to find talented individuals to fill key posts. And when you spot attractive candidates, you do what it takes to lure them away from their current employers. You offer big signing bonuses, you buy out stock options, and you provide rich compensation packages of your own. All the while, you know that other companies are busily rifling through your own organization, hoping to poach your best talent.

The open competition for other companies' people, once a rarity in business, is now an accepted fact. Executives know that fast-moving markets require fast-moving organizations that are continually refreshed with new talent, and they've become adept at outside hiring. (See the sidebar "Strategic Poaching.") But if they're comfortable bringing talent in, they remain distinctly uncomfortable about seeing talent leave. To poach is fine; to be poached is not. One reason for the discomfort is emotional. Executives tend to judge themselves on their ability to instill loyalty in their people, and the departure of a talented employee can feel like a personal affront. Another reason is rational. In a time of tight labor markets, talent can be very hard-and very expensive- to replace. When a good employee walks, the business takes a hit.

In trying to stop people from jumping ship, many companies have fallen back on traditional retention programs. I recently attended a talk by a senior manager from DuPont who was telling of a corporate initiative to "re-engage" with employees. By designing and promoting new, long-term career paths and investing heavily in employee development, the company hoped to win back the loyalty of its workforce. When a member of the audience asked him if he really thought the company could stop the outflow of talent, the speaker replied, in a moment of unexpected candor, that he did not- the competition was simply too intense. But, he went on, the company's executives saw no alternative. They had to make the effort.

The speaker was right about one thing. It is futile to hope that by tinkering with compensation programs, career paths, training efforts, and the like, a company can insulate itself from today's freewheeling labor market. That doesn't mean, however, that companies should just go through the motions. There is an alternative: a market-driven retention strategy that begins with the assumption that long-term, across-the-board employee loyalty is neither possible nor desirable. The focus shifts from broad retention programs to highly targeted efforts aimed at particular employees or groups of employees. Moving to a market-driven strategy is not easy. It requires executives to take a hard-headed, analytical approach to what has long been viewed as a "soft" side of business- the management of people. But it is necessary. The clock can't be turned back.

Rethinking Retention

To adopt the new strategy, you first have to accept the new reality: the market, not your company, will ultimately determine the movement of your employees. Yes, you can make your organization as pleasant and rewarding a place to work in as possible- you can fix problems that may push people toward the exits. But you can't counter the pull of the market; you can't shield your people from attractive opportunities and aggressive recruiters. The old goal of HR management- to minimize overall employee turnover- needs to be replaced by a new goal: to influence who leaves and when. If managing employee retention in the past was akin to tending a dam that keeps a reservoir in place, today it is more like managing a river. The object is not to prevent water from flowing out but to control its direction and its speed.

Prudential is one company that has begun to adopt this market-driven perspective. Its "Building Management Capability" program, which integrates recruiting, retention, and training efforts, is geared toward an increasingly mobile workforce. "Gone is the notion that employees are going to stay with one company for life," says Kurt Metzger, a human resources executive at the company. The Prudential program is anchored by a sophisticated planning model that projects talent requirements and attrition rates. The model enables business-unit managers to develop highly targeted retention programs and create cost-effective contingency plans for filling potential gaps in skills. The model also provides a mechanism for constantly measuring the impact of human resources decisions, a capability crucial to managing people in this rapidly shifting labor market.

Prudential has begun doing what most companies avoid: making a truly honest assessment of how long the organization would like employees to stay on board. Such an analysis inevitably reveals that different groups of employees warrant very different retention efforts. There will always be some people a company will want to keep indefinitely-an engineering genius, an inspiring business head, a creative product designer, or a front-line worker deeply respected by customers. Another set of people will be important to retain for shorter, well-defined periods -employees with specific skills that are currently in short supply, for instance, or members of a team creating a new product or installing a new information system. And finally there will be people for whom investments in retention don't make sense employees in easy-to-fill jobs that require little training or employees whose skills aren't in demand in the market.

Once you know which employees you need to retain and for how long, you can use a number of mechanisms to encourage them to stay. The key is to resist the temptation to use the mechanisms across the board. Tailor your programs to your retention requirements for various employees and to the level of demand for them in the marketplace. Let's look at some of the mechanisms and their strengths and shortcomings.

Compensation. The most popular retention mechanism today is compensation. Most companies try to lock in their most valuable employees with "golden handcuffs"-pay packages weighted heavily toward unvested options or other forms of deferred compensation. The problem with pay-based incentives is that they're easy for outsiders to match. Recruiters routinely buy out golden handcuffs with signing bonuses- "golden hellos." Retention incentives end up becoming just another element of compensation, contributing more to wage inflation than to long-term retention. (See the sidebar "The Futility of Golden Handcuffs.")

But compensation can help shape who leaves and when. Some companies now pay special "hot skills" premiums to employees whose expertise is crucial and in short supply. The payments are an effective way to keep talent in place for critical periods, such as through the late stages of the design of an important product. The premiums cease the minute the skills become more readily available on the market or the employer decides that the skills are no longer as important to its business. Andersen Consulting, for example, recently eliminated its hot-skills premium for SAP programmers.

Paying signing bonuses in stages, rather than as lump sums, can also help to keep new employees in place, at least in the short run. Deferred signing bonuses are becoming the norm for executive-level hires. When Associated Communications (now Teligent) gave Alex Mandl, AT&T's heir apparent, a $20 million signing bonus to become its new CEO, it paid out the money over five years. Such bonuses are proving useful in retaining lower-level employees as well. Burger King, for example, offers workers a signing bonus but withholds payment until they've been on the job for three months. Three months may not seem like a long time, but in the fast-food business, where annual turnover averages 300%, it's an eternity.

A deferred bonus doesn't guarantee that a new employee will stay for the deferment period, of course. Such incentives are, after all, just a form of golden handcuffs. Another company can always come along with a big golden hello.

Job Design. To retain people with critical skills for longer periods, companies need better mechanisms than compensation. One is job design. By thinking carefully about which tasks to include in which jobs, companies can exert considerable influence over retention rates.

Consider what United Parcel Service did to improve its retention of drivers. UPS recognized that drivers have some of the most important skills in the delivery business. They know the idiosyncrasies of the routes and they have direct relationships with customers. Finding, screening, and training a replacement driver is time consuming; it may take a new hire months to learn the details of a particular route. When UPS studied the reasons its drivers left, it discovered that much of the turnover could be traced to the tedious and exhausting task of loading packages at the beginning of a run. It therefore unbundled the loading task from the drivers' job and assigned it to a new group of workers. The turnover rate for drivers fell dramatically.

Of course, turnover in the new loading jobs averages an eye-popping 400% per year. But that doesn't matter. With high hourly wages and low skill requirements, the loading jobs are fairly easy for UPS to fill, typically with students or other part-timers, and fairly simple for new employees to learn. A high turnover rate in the loading jobs is expected and manageable. In using job design to improve retention, UPS didn't attempt to decrease overall turnover; instead, it targeted the specific skills it wanted to retain. For employees without those skills, it allowed the revolving door to spin freely.

Jobs can also be defined in such a way as to influence when people will leave. Wall Street investment firms were once plagued by erratic, unplanned turnover among junior analysts. The companies addressed the problem by requiring the analysts to leave after three years. Forcing people to quit may seem like an odd way to solve a turnover problem, but it makes a lot of sense. The real issue, after all, was not that the junior analysts were leaving-it was expected that many would go on to business school but that the firms could not predict who would leave or when. As a result, project teams were often left understaffed, leading to delays and quality problems. Now that they know junior analysts will depart at the end of their third year, the firms can design projects to coincide with analysts' tenures. Having clear termination dates also creates large, well-defined employee cohorts, making training and development easier. The emergence of the three-year stint as an industry standard helps ensure that employees stay for the full period because a junior-analyst job lasting less than three years looks bad on a resume.

Job Customization. In addition to tailoring jobs to particular categories of employees, companies can also tailor them to the needs of individuals. Prudential is experimenting with such a program. It provides workers with a variety of tools to help them assess their own interests, values, and skills, and it encourages managers to tailor rewards, benefits, and assignments to individual requirements. A part-time arrangement might satisfy one employee's desire to pursue outside interests or meet a parenting need, while tuition reimbursement might be the key to keeping another employee happy.

Prudential's program draws on an array of employment options, most of which are available to all workers. It's easy to imagine, however, programs that would go even further in customizing jobs. Key employees might undertake a formal self assessment of their work and nonwork goals and of how those goals could best be achieved in the con- text of the company's operations. The assess- merits would form the basis for individual employment agreements, which might be created using cafeteria-style programs similar to those used in allocating employee benefits. Each employee would be able to allocate a set amount of money to "purchase" options in such areas as career development and balancing work and personal life. The amount available to allocate would depend on the importance of the employee to the company.

Individualized deals always raise fairness concerns, of course. Basing rewards on skills, rather than just on performance, is something new, and it's sure to rub some people the wrong way. But there are plenty of precedents. Salaries have long been based on the labor market- those in hot fields get paid more. Relative compensation routinely hinges on criteria outside an employee's control, such as the performance of a division or the state of the stock market. And most companies have always had a fast-track career path for employees deemed more valuable than their peers on measures other than current job performance. Giving greater benefits to those with critical skills that are difficult to replace seems in tune with these established practices.

The bigger issue may lie with the form of the rewards rather than in how they are distributed. Few companies allow employees to design their own jobs, and those that do usually offer such programs across the board rather than selectively. That's the case, for example, with most flextime arrangements. Companies will need to consider carefully the effects on morale as well as the legal implications of selective programs, but they should not reject them simply because they're unusual and raise tough questions. The market is very creative in providing individualized rewards. Companies should be equally creative.

Social Ties. Loyalty to companies may be disappearing, but loyalty to colleagues is not. By encouraging the development of social ties among key employees, companies can often significantly reduce turnover among workers whose skills are in high demand. Carl Glaeser, general manager of Ingage Solutions, a Phoenix-based division of AG Communication Systems, has held the turnover of software engineers to 7%, mainly by developing programs that create a social community in the workplace. Golf leagues, investment clubs, and softball squads create social ties and bind workers to their current jobs. Leaving the company means leaving your social network of company-sponsored activities.

Arrangements that help create community within an organization have one big potential drawback: they make the trauma of any eventual restructuring all the more intense. Creating strong social ties is therefore inappropriate for employees who are likely to become less vital to a company in the near future. But you can achieve a similar bonding effect, minus the long-term complications, with teams. By creating closely knit teams to carry out particular projects, companies can increase the likelihood that the teams will remain intact for the length of the initiatives. People who would hardly think twice about abandoning a company find it very difficult to walk out on their teammates. Teams also have an added benefit: studies have shown that they increase employees' commitment to their work. (See the sidebar "Commitment Without Loyalty.")

Location. Large businesses have another good mechanism for managing retention: location. By carefully choosing the sites for various groups of employees, they can influence turnover rates. A high-tech company, for example, might find it useful to have a research and development operation in Silicon Valley in order to tap into cutting-edge thinking. The inevitable high turnover rate will be an advantage: the company will be exposed to a broad array of ideas. But an R&D project with a long lead time could be doomed by such high turnover. The company would be wise to set up a long-term R&D operation in a place where the skills of the development team are not in high demand, such as a rural community. People will still leave from time to time, but overall turnover will be much lower. At a Harris Semiconductor (now Intersil) facility in rural Pennsylvania, turnover recently averaged just 2% a year, far below the 20% average for the semiconductor industry.

Of course, trying to get people to relocate to remote regions poses its own challenges. But here again it makes sense to think about the individual circumstances and needs of your people. Employees who have young families, for instance, may be very interested in moving to a smaller, more rural community. Once they're there, it will be hard for them to pull up stakes and leave.

Hiring. When companies go out recruiting, they often focus on attracting precisely those people who will be the most difficult to retain. By shifting their sights to workers who can do the job but are not in high demand, organizations may be able to shelter themselves from market forces. Microboard Processing, a Connecticut-based assembler of electronic components, hires one-third of its assemblers from high-risk applicants, including welfare recipients, former drug addicts, and people with criminal records. The company often starts the new recruits on simple landscaping jobs to see how they do before moving them inside to the assembly operation. It also gives them lots of slack during the first few months on the job while they are growing used to the discipline of factory work. In return, the company says it is getting a hardworking pool of employees who are grateful and loyal to Microboard for giving them a chance.

Architectural Support Services, a computer-aided-design company providing technical support for architects, also uses hiring to bolster retention. In its early days, the company followed textbook HR practices, hiring the best and brightest professionals available. Yet it found its operations in shambles because of poor morale and high turnover- all caused by infighting among the high-powered staff. The company thought hard about its workforce and realized that it did not need to fill all its positions with gifted workers. It started recruiting from community colleges instead of elite four-year institutions. The company has been rewarded with a much more loyal and committed workforce-and its results have not suffered in the least.

Adapting to Attrition

Sometimes there will be no effective way to ensure the retention of a particular employee or group of employees. The market forces will be too strong. Look at the trouble companies have holding on to their information technologists. The extremely tight labor market gives talented techies a wealth of opportunities, and the field's rapidly changing skill requirements give them an incentive to seek new projects that will advance their expertise. Companies with a strong need to retain particular IT skills- for maintaining legacy systems, say are in a bind. Their best course is often to avoid the retention issue altogether by outsourcing the required skills. J.P. Morgan is among the many companies that have taken the outsourcing route. It collaborated with several IT companies to establish Pinnacle Alliance, which now manages Morgan's global IT operations. Morgan found that the best way to deal with an intractable skill shortage is to let somebody else deal with it.

In other cases, companies have found that high turnover isn't as big a problem as it appears. Just because a business is dependent on engineering skills, for example, doesn't mean that it has to go to great lengths to retain its engineers. If there's a large pool of engineers available, it might want to focus on recruitment rather than retention. That's exactly what a number of electronics companies in Ireland have been doing. Irish universities are producing a steady supply of talented engineers trained in the latest technologies. The electronics companies recruit aggressively at these schools, but they make relatively little effort to retain their current engineers. That way, they continually infuse their organizations with the most up-to-date skills. Moreover, since new hires have lower salaries than longer-term employees, the companies are able to keep a lid on compensation levels.

There are also ways to adapt organizations and operations to high turnover. Simplifying and standardizing jobs and cross-training workers in multiple jobs make companies less dependent on any one individual. Many semiconductor companies, for example, have responded to high turnover rates among machine operators by certifying operators on more machines and rotating them to new positions every three months or so. Moving from legacy systems, even if they suit the organization's needs, to more common, off-the-shelf systems helps ensure that needed skills will be readily available in the marketplace. And organizing work around short-term projects with clear end points can make turnover easier to manage. Companies can focus their retention efforts on keeping employees just until a project is completed- a much easier task than building long-term loyalty.

Information technology can also help employers cope with turnover by preserving some of the institutional memory that employees would otherwise take with them. Customer relationship software automates sales and gives clerks access to client histories, including prior orders and complaints, allowing the clerks to sound familiar with accounts they know nothing about. Groupware applications like Lotus Notes can standardize interactions and keep records of decisions and crucial contextual information, providing something like an electronic record of employee knowledge. Other programs, such as Open Text's Livelink, enable all employees to track and share documents on an intranet. New simulation software for team-based project management, such as Thinking Tools' Project Challenge, helps new teams learn how to work together much more quickly than on-the-job experience would allow.

Even a technology as simple as e-mail can prove to be a godsend when key employees are lost, as Pamela Hirshman, a project manager at Young & Rubicam, recently found out. She was called in to take over a project after the entire original project team had left. "The project file had a record of all the e-mails between the team and the client," she says, "and after reviewing about So of these, I was up to speed on the problems of the client and where the project was heading."

Cooperating with Competitors

Because of the intensity of the talent war, companies instinctively view retention and recruitment as competitive exercises a perspective that has kept them from seeking help from one another. But history shows that cooperation, even among competitors, can be one of the most effective ways of dealing with talent shortages. In the 1950s, the big aircraft companies like Lockheed, McDonnell-Douglas, and Northrop competed fiercely for the government contracts that were their lifeblood. When a company won a new contract, it faced the challenge of quickly hiring skilled staff to carry out the work. When a company lost a contract or simply finished a project, it had the problem of excess staff.

A solution emerged in southern California, where many of the companies had operations. They began to "lend" teams of employees to one another. A company that lost a fighter contract, for example, would hire out a team of experienced employees to the company that won it. The team members would remain employees of the first company. Lockheed reported that its program, known as Lending Employees for National Development (LEND), had a wide range of benefits. In addition to avoiding layoffs, the company retained its investment in key employees, maintained its capability to bid on future contracts, and broadened the experience of its leased employees.

The aerospace industry saw another type of cooperation, one between prime contractors and subcontractors. Complicated components for large projects would be created at a subcontractor, then moved to the prime contractor, where they were assembled into a larger module and passed on to a final assembler before being delivered to a client like NASA-a process that could take years. Key employees of the subcontractor followed the component to the prime contractor, becoming employees of the prime contractor and working alongside its staff.

Perhaps the most expansive current example of cooperation in hiring is the Talent Alliance, which began at AT&T and has grown to include about 3o large companies. It started as a kind of sophisticated job bank during the era of downsizing and high unemployment. Companies that had to lay off skilled workers could market them to other employers that might be looking for such skills. The Talent Alliance has since expanded its charter. It now provides standardized tests for screening and evaluating people and for matching them with jobs at member firms.

Other, more ad hoc, collaborations are appearing among noncompeting companies. Cascade Engineering, a plastic parts manufacturer in Grand Rapids, Michigan, has teamed up with a local Burger King to coordinate recruiting. Applicants who do not have the skills necessary for Cascade's production positions but who otherwise seem like good workers are offered jobs at Burger King. Successful Burger King employees who begin looking for more skilled positions are offered vocational counseling at Cascade. The prospect of moving to Cascade provides an incentive for people to join and stick with Burger King, and the Burger King employees become a dependable labor pool for Cascade. The career development that individuals in the past would have experienced within a single company now takes place across two companies.

Cooperating with other companies to develop employees and lay out possible career paths goes against the grain of traditional HR management, which is based on the assumption that employees are captive and proprietary assets. But it is in tune with the current reality of the market-driven workforce. One thing is for sure: as the early years of our new century unfold, executives will be challenged to abandon their old ways of thinking and adopt ever more creative ways of managing, retaining, and, yes, releasing their talent. Those who begin that difficult process now will be one step ahead of the game.

~~~~~~~~

by Peter Cappelli

Peter Cappelli is the George W. Taylor Professor of Management Studies at the Wharton School in Philadelphia sources. He is author of The New Deal at Work: Managing the Market-Driven Workforce.

Strategic Poaching

The signs of an explosion in outside hiring are everywhere. The executive-search arm of A.T. Kearney reported that the number of searches undertaken by its clients in 1997 was 15% higher than in 1996, which was itself a record year, and that CEO searches were up 28%. When one St. Louis headhunter, John R. Sibbald, tracked the careers of 150 up-and-coming executives, he found that within two years, 8o% had changed employers. Job fairs, once mainly nonprofit, informal events, have turned into highly profitable extravaganzas run by specialized companies that charge up to $5,000 per recruiting company. And electronic job markets, unknown a couple of years ago, are overflowing with recruiters and resumes.

One reason poaching has spread so rapidly is that companies have learned to use outside hiring strategically as well as tactically. They bring in experienced people not just to fill open positions but also to get the expertise they need to quickly expand into new markets or even to launch new businesses. As the consumer electronics and computer industries have moved into each other's markets, for example, they have begun to recruit each other's employees as a way to tap into each other's expertise. The U.S. unit of Mitsubishi's consumer electronics group recently hired 20 engineers from computer companies for its research staff in one swoop. "What you have is the NFL draft of electronics executives," says Bob Lee, head of Manpower Staffing Services in San Jose, California.

The strategic use of outside hiring is not limited to the volatile high-tech industry. Today when an oil company wants to expand the sales of products at its service stations, it hires managers from Pepsi and Frito-Lay with expertise in retailing. When an airline wants to get better at managing customer relationships, it recruits executives from Marriott with experience in customer service. When a power company prepares for deregulation, it hires people from a phone company that has already gone through the transition. Businesses have found that it's quicker to steal new competencies than to develop them from scratch.

Poaching can also provide a relatively easy way for companies to move into new regions. In 1995, Ernst & Young built up its presence in Spain by recruiting nearly the entire Madrid office of its competitor Coopers & Lybrand-9o people in all. When Allegheny Health Systems moved into the Philadelphia market in the mid-1990s, it raided the cardiology and cardiac surgery departments of Presbyterian Medical Center to develop an instant presence in those specialties. Presbyterian then turned around and hired all but one of the cardiologists from Cooper Health Systems in nearby Camden, New Jersey.

Hiring outside executives is now even seen as an effective, and frequently less risky, alternative to acquiring entire companies. A few years ago, AT&T wanted to enter the computer-systems-integration business, but it worried about whether it could incorporate an entire company into its culture. Instead, it asked recruiters to find the top 5o system integrators in the country. AT&T hired them and started its own systems-integration operations.

It would be a mistake to look at all this activity and think it's a passing phenomenon-a symptom of the booming economy and the tight labor market. The underlying changes in business are fundamental, not transient, and the use of outside hiring as a strategic tool will only grow in the future. While the overall demand for labor will rise and fall, the war for talent will rage on.

The Futility of Golden Handcuffs

Today virtually every company offers its key people some form of deferred compensation in hopes of buying their loyalty. When such golden-handcuffs programs were new, they effectively kept corporate recruiters at bay. Just as burglar alarms redistribute burglary to unprotected homes, golden handcuffs redistributed poaching to unprotected businesses. Some companies, like Emerson Electric, became famous for the elegance and elaborateness of their handcuffs.

But now that almost all companies offer their key people some form of deferred compensation, golden handcuffs are no longer a deterrent. With no unprotected companies left to pilfer from, recruiters have been forced to unlock the handcuffs by offering huge signing bonuses. In 1996, Alex Mandl, who appeared to be next in line for the CEO job at AT&T, was lured away to Associated Communications, a small operation that offered him a signing bonus of more than $20 million, much of which simply offset the $10 million in AT&T stock options he forfeited on leaving. No matter how generous the deferred compensation package you offer to one of your stars, there will always be another, more desperate company willing to pay even more.

In addition to being ineffective, golden-handcuffs programs sometimes backfire. When Lou Gerstner arrived at IBM in 1993, he discovered that despite all the unvested stock options the company offered, his key employees were being picked off by competitors right and left. The reason was the decline in IBM's stock price over the preceding years. The options had been rendered worthless, to the disgruntlement of employees. As Gerstner put it in a Fortune interview, "What I've got is an employee group with absolutely no incentive to stay here because every one of their options is under water."

Researchers have identified other perverse effects of golden handcuffs. A study of the semiconductor industry found that when large profit-sharing bonuses were distributed or when a company's stock price was booming, engineers often cashed in their profits and left to start their own businesses. That finding dovetails with a conclusion reached by labor economists: when employees get big windfalls, they tend to work less, essentially buying more leisure by retiring early or shifting to easier jobs with shorter hours.

At best ineffective, at worst counterproductive, deferred compensation programs are nonetheless a necessity in business today. Since all companies offer them, your company will have to as well. The market demands it, and the market rules.

Commitment Without Loyalty

Executives have long seen loyalty and commitment as two sides of the same coin, believing that employees who lack loyalty to a company must also lack commitment to their work. From that perspective, the erosion of employee loyalty looks very scary. Research, after all, has shown a strong correlation between commitment and performance. If your employees lack commitment, you're in big trouble.

But there are many ways to engender commitment to the work without requiring loyalty to the company. Organizing work around projects is one such method. Studies have shown that when employees have control over a piece of work, they have greater commitment to seeing it done well. If it goes well, they get the credit, which increases their prestige (and helps build their resumes); if it fails, their reputations take a hit.

Creating teams is another way to build commitment. After all, commitment is far easier to establish among individuals than between an individual and an abstract entity such as a corporation. Team members work hard because they do not want to let the rest of the team down. The more accountable a team is for its performance, the greater the peer pressure on members to make sacrifices for the team. Team-based compensation, in particular, helps create the sense that the fate of the community relies on the performance of its members. Even in industries long characterized by hostile relations between employees and employers, such as the U.S. auto industry, the redesign of production work around teams has contributed to sharp improvements in quality and overall performance, at least in part by engendering greater worker commitment.

The confusion of loyalty and commitment underlies another widely held but false belief: that commitment can exist only in a long-term relationship. We know this is not true from our experiences in other arenas. Many people are extraordinarily committed to their alma maters, for example, donating time and money to them years after graduation. Many people have similar relationships with former employers, particularly those where they held their first jobs. McKinsey & Company, for example, is famous for the level of commitment it enjoys from its former consultants -even those who were pushed out of the firm. Indeed, short-term relationships often create a higher level of commitment than long-term relationships. One of the midterm exams given every year at Wharton asks members of the first-year class to explain how they were managed in their previous jobs. Almost without exception, the students coming from "terminal" jobs-those with fixed departure dates, such as junior analyst positions at investment banks- are the most positive about their former employers. The advantage of such temporary relationships is that the people in them have a clearer idea of what's expected of them and what they'll gain. They understand going in that they will have to work hard and that after a certain period they will have to leave. Not only are such employees committed to the companies during their tenure, but 'their positive feelings after they leave pay additional benefits to the companies- influencing word-of-mouth reputation, facilitating future business deals (why not deal with companies you know and like?), and creating a pool of potential future recruits.

THE LEADERSHIP TEAM

Title: THE LEADERSHIP TEAM., By: Miles, Stephen A., Watkins, Michael D.

THE LEADERSHIP TEAM

Complementary Strengths or Conflicting Agendas?

SENIOR LEADERSHIP TEAMS whose members play complementary roles have been chronicled as far back as Homer's oral history of the Trojan War. Though the Greeks were led in their quest for retribution against Troy by the powerful King Agamemnon, their victory would not have been possible without Achilles, the mighty warrior; Odysseus, the wily tactician; and Nestor, the wise elder. Each had a crucial, distinct role to play in the Greek high command. Achilles rallied the troops in the heat of battle. Odysseus provided sound strategic advice during and between engagements. Nestor was a source of cool-headed counsel and diplomacy, mediating between the titanic egos of Agamemnon and Achilles. No one of them could have played all the varied roles necessary to guide the enterprise to victory; collectively they prevailed and won their place in history.

Not much has changed at the top of large organizations in the past 3,000 or so years. Today, complementary-leadership structures are common and, in some cases, even institutionalized. Think, for example, of the chief executive officer and the chief operating officer, a pairing in which one leader is concerned mainly with external issues and the other focuses on internal matters.

Even though most complex organizations are run, formally or informally, by teams of two or more, far more attention is paid to CEO performance and succession than to such issues as how complementary teams should be designed and what happens when their membership changes, especially during a succession process. While acknowledging the symbolic and actual importance of an organization's ultimate leader, we need to expand our focus beyond this unitary position.

The two of us have studied numerous complementary-leadership structures, often at very close range in consulting relationships we've each had with companies (including several mentioned in this article). We've sought to understand why these structures emerge, what purposes they serve, and what challenges they create. Our study supports the case for complementarity: Such teams by their very nature are able to do things that individuals and noncomplementary teams can't. At the same time, we've seen that with the benefits come risks. Because of members' different strengths and styles, they may pursue incompatible ends or employ inconsistent means to achieve their goals. Succession also presents particular challenges. What happens when members of complementary teams move on, as they inevitably do? When succession looms, the logic of complementarity can get turned on its head – that is, the more complementary the team, the greater the potential for difficulties when it comes to a change of command.

Fortunately, the board of directors and the CEO can avoid problems like these. With a good understanding of complementary leadership, they can create an effective team and ensure a stable and productive succession when the membership changes. While this article applies broadly to leadership teams, including relatively large groups such as the senior executive committee, it is particularly relevant to relationships involving the top two or three people in an organization.

The Promise of Complementary Leadership

The pervasiveness of complementary leadership in large organizations results in part from the obvious differences between various roles – the CEO and the COO, for example, or the heads of different functional areas. Sometimes complementary-leadership teams are designed into an organization. (For a description of one firm's experience with co-leaders, see the sidebar "A Commitment to Complementarity.") More often, however, such teams emerge from a process akin to natural selection, in which leaders – through a mixture of succession planning, unexpected opportunities, personal capabilities, and self-selection – over time come to perform complementary functions. These are distinct because of the fundamental limits on a single person's ability to focus attention, acquire new capabilities, process information, and play diverse social roles. Indeed, the limitations of people's information-processing capacity, which are well documented, make it impossible for one individual to manage a large and complex enterprise. Bruce Chizen, CEO of the software and technology company Adobe Systems, says of his own position, "The job is simply too big for any one person." Bringing together two or more people with complementary strengths not only compensates for the shortcomings of each but also results in a team in which the whole is much greater than the sum of the parts.

Complementary leadership generally manifests itself in four ways. Increasing demands on the CEO's attention and time have led to the fairly straightforward task complementarity. Faced with unmanageable levels of complexity and uncertainty, leaders divide management responsibilities into coherent blocks of tasks. As we noted above, one familiar split assigns one leader (usually the CEO) the job of managing the external environment, while her counterpart (often the COO) concentrates on internal management issues. Another way to divvy up tasks is to designate executives to take primary responsibility for different businesses or groups of businesses.

A second fairly clear-cut division of responsibilities is expertise complementarity. Before becoming general managers, executives usually get their training and experience in one or at most a few business functions, such as sales, marketing, finance, operations, or R&D. Although effective general managers acquire a good working knowledge of other functions, their original areas leave indelible imprints on them. The resulting differences in expertise among general managers naturally lead to the formation of teams with complementary expertise. One common configuration in consumer brands companies is a CEO with a sales and marketing background and a COO with expertise in finance or operations. In technology companies, the COO often has deep technology expertise. For example, Chizen, Adobe's CEO, has a background in sales and marketing, and Shantanu Narayen, the company's president and COO, came up through the engineering and product ranks.

A third, less sharply delineated type of synergy –what we call cognitive complementarity – involves differences in how individuals process information. It is extraordinarily rare, for example, to find leaders who are equally good at dealing with the big picture – creating and communicating compelling visions and crafting breakthrough strategies – and at driving execution through an intense focus on tactics, details, and follow-through. Both Aart de Geus, the founder and CEO of Synopsys, and Chi-Foon Chan, the company's president and COO, have immense technical expertise – which is crucial to their credibility in their highest of high-tech industries, electronic design automation. Despite this overlap, however, they play very different roles: De Geus is the idea-a-minute visionary, while Chan stays grounded in realistic goals and the details of achieving them.

Finally, leaders often play discrete and complementary social roles in organizations – a phenomenon we call role complementarity. One person can rarely assume more than one social role; it is difficult, for example, for a leader to be both feared and loved. Agamemnon could not inspire the troops as Achilles could, nor could Achilles formulate an overarching battle strategy as the king could. Rather than sacrifice one or the other of these powerful motivating forces, senior teams naturally evolve into configurations in which one leader provides the "pull" through rewards and inspiration and another provides the "push" through disciplined goal setting and sanctions.

A common example of role complementarity is when one leader plays "diplomat" (similar to the traditional "good cop" role) and another plays "warrior" (the "bad cop") in dealing with external constituencies. When Roberto Goizueta and Douglas Ivester were, respectively, CEO and COO of Coca-Cola, Goizueta was the urbane diplomat championing the Coca-Cola brand with such stakeholders as regulators and bottlers. Ivester, by contrast, described the company as a "wolf seeking its prey"–in this case, a sheep called Pepsi. According to Warren Buffett, then a member of Coca-Cola's board, Ivester would take it as a personal insult "anytime anyone swigs any beverage other than a Coca-Cola product." Another division of roles occurs during times of wrenching internal change: One leader plays the "guardian," the preserver of what is good in the organization, and another plays the "entrepreneur," the destroyer of what is bad. (For a description of how these complementary roles have played out conspicuously in a political context, see the sidebar "Complementary Leadership and the Quest for Peace.")

Mapping Team Complementarity

The extent to which senior teams exhibit complementarity varies considerably. In some cases, leaders play complementary roles along all or most of the four dimensions we have described. In extreme instances, a deeply complementary relationship effectively becomes a kind of shared leadership.

There are few examples of this so-called "two in a box" leadership structure. Bill Gates and Steve Ballmer at Microsoft are a well-known pair. After more than 20 years working together, plus a stint as hall mates in a Harvard undergraduate dorm, the chairman and the CEO augment their complementary areas of expertise (the creation of visionary technical architectures for Gates, hard-driving sales and marketing for Ballmer) with a shared passion for the company. The relationship will be tested as Gates continues his retreat from day-to-day duties and the recently named COO Kevin Turner – in the tradition of technology company COOs, an operationally focused former chief information officer (in his case, at Wal-Mart)–tries to create a similar if perhaps less intense dynamic with Ballmer.

A complementary-leadership relationship that has survived the stress of several changes exists at Starbucks. In 2002, when Jim Donald joined the company from Pathmark Stores (where he had been president and CEO), he became the newest member of a "three in a box" leadership team that had previously been known as H2O– a reference to the first names of Starbucks founder and chairman Howard Schultz, head of international operations Howard Behar, and CEO Orin Smith. Sustaining this successful setup wasn't easy. Partly through his efforts to assure rank-and-file employees that they could count on continuity in the company's leadership approach, Donald helped ensure the success first of the new three-person team (he took Behar's place, creating a group in which Schultz was the visionary, Smith the administrator, and Donald the merchant) and then of a two-person version (he became CEO when Smith retired).

Deeply complementary relationships like these are unusual; teams with some complementary aspects are quite common. Steven Reinemund, formerly the CEO and now the chairman of PepsiCo, had a strong partnership as COO with his CEO predecessor, Roger Enrico. When Reinemund became the chief executive, he fostered a similarly productive relationship with his ultimate successor, CFO Indra Nooyi, based mainly on task complementarity. Giving Nooyi the additional title of president while she was CFO, he let her take an increasingly public role with external constituencies, which was good preparation for her ascension to CEO but also a way to balance Reinemund's operational strengths.

Disk drive maker Seagate is run by a four-person team that successfully exploits expertise complementarity. CEO Bill Watkins, president and COO David Wickersham, and two executive vice presidents – Brian Dexheimer, who is in charge of marketing and strategy, and Charles Pope, who heads up finance– work well together because they draw on one another's different functional strengths. As we will discuss, however, overlap in the expertise of Watkins and Wickersham initially threatened the team's success.

The Perils of Complementary Leadership

The advantages of complementarity – which, after all, are rooted in differences among leaders – don't come for free. For one thing, teams risk squandering the potential benefits through the confusion that complementary leadership can spawn. "People, especially those lower down in the organization, would sometimes wonder when they should talk to me and when they should talk to him," recalls Adobe CEO Chizen, speaking of his close complementary relationship with Narayen. Over time, employees came to understand the semiformal division of labor: Chizen was solely responsible for presenting financial proposals to the board and for strategy decisions, but beyond that the two shared duties, with Chizen focusing on marketing, brand, and customer issues and Narayen tackling product and operational matters.

Another challenge involves achieving and sustaining agreement about organizational priorities. What happens, for instance, if Agamemnon decides to attack the city and Achilles stays in his tent? Given that the team will intentionally include people with significant differences – in terms of their assigned tasks, areas of expertise, mind-sets, or social roles–how does a company avoid ending up with a group of leaders who, in the words of the adage, sleep in the same bed but dream different dreams? The risk that team members will head off in different directions is all too real, especially as organizational environments become increasingly complex and ambiguous.

A less obvious problem is that team members' similarities in certain areas may interfere with the team's overall complementarity. That is, if there is too much overlap in the Venn diagram depicting members' tasks, areas of expertise, mind-sets, or social roles, the problem may be more than redundancy: In the overlapping areas, people may compete to do things their own way. This need not happen, of course. Recall how Aart de Geus and Chi-Foon Chan of Synopsys complement each other, even though they share deep technical expertise, because they enjoy the breathing room provided by their different cognitive roles: De Geus's visionary approach and Chan's down-to-earth way of processing information.

But in such situations, conflict can overshadow what would otherwise be a productive relationship. Consider the recent history of the leadership at Seagate. CEO Watkins, who has a strong interest and expertise in operational matters, remembers working hand in glove as COO with former chief executive Steve Luczo, who focused primarily on corporate strategy. When Luczo became chairman and Watkins became CEO, Watkins filled the COO spot with Wickersham, an executive with a similar passion for operational excellence – a crucial element in a vertically integrated business that moves some 86 million parts in its supply chain every day. Initially, the relationship was somewhat strained. "When I first came into the CEO job, I wouldn't give him responsibility for the part of the organization that oversees product quality," Watkins recalls. "Then I realized I was messing up, using my ownership of quality to hammer operations. It wasn't so much an issue of his not being ready; it was an issue of my not being ready to give up responsibility."

As a result of Watkins's decision to pull back from day-today operations and give Wickersham some freedom to maneuver, Seagate has been able to benefit from their complementary social roles–Watkins as the emotional keeper of the company's culture, Wickersham as the data-driven executive who gets the job done.

The Four Pillars of Effective Complementarity

The risks inherent in complementary leadership can't be avoided. But organizations can manage them by heeding the four pillars of alignment in successful complementary teams: a common vision, common incentives, communication, and trust. As a team's complementarity increases, so does the importance of these pillars.

The failure to craft and commit to a shared vision and supporting strategy is at the root of some notable senior team collapses. Ed Zander, who is now chairman and CEO of Motorola, says he knew it was time to leave his position as president and COO of Sun Microsystems – thus dissolving what had been a successful complementary relationship with then CEO Scott McNealy – when the two began to diverge in their visions for the company. "Scott and I had a really good relationship," he recalls, "but I could see us start to philosophically move apart – not in a negative fashion but in an honest disagreement about the company's priorities." At that point, Zander says, it was clear the relationship would no longer be productive.

Commitment to a common vision and strategy, while necessary, will not be sufficient if members of the team have competing incentives. People won't work in concert and make sensible tradeoffs if their rewards depend on achieving different goals. For example, one member of the leadership team might be pursuing short-term results while another is focusing on the long term. Or one might be rewarded for driving top-line growth and another for driving earnings growth. Their differing incentives could lead them to take actions that, while advancing their individual goals, are suboptimal for the business as a whole.

Making complementarity work also requires outstanding coordination and communication among the team members. The deeper the complementarity, the denser this communication must be. The most effective complementary teams are in near constant communication and have well-established protocols so that potential divergences can't take root. "The danger is that you'll get into something like a parental situation, where people in the organization think they can go to one person to get an answer that the other one won't give them," says former PepsiCo CEO Reinemund, recalling his relationship with then CFO and president Nooyi. "I'm sure it happened occasionally. But on big issues it never happened because Indra and I were connected enough to avoid it. People knew we talked all the time." Jim Donald, the Starbucks CEO, puts it another way: "I bug the heck out of Howard Schultz just to stay in touch."

The fourth pillar, which may be the most crucial for a team's stability, is trust among team members. Common vision, aligned incentives, and close communication enable purposeful and powerful cooperative action, but they have no value unless team members know that their counterparts can and will further the best interests of the enterprise.

Members of successful complementary teams describe the phenomenon in different ways. Reinemund says it involves not questioning your teammates' motives and caring about their personal well-being in a way that "supersedes day-today results." Donald talks about giving others on the team "a free runway" to do what they need to do. Watkins characterizes trust as "knowing that when things hit the fan, we will try to do the right thing for Seagate – and that we'll stick together to get it done."

This kind of trust yields tremendous benefits. "If you have it, you can screw up, as we all do on a regular basis, and still make things work," says Reinemund. "Without that, when the first tough storm comes, things just fall apart." According to Watkins, his team's deep confidence that it ultimately won't make many mistakes "gives us the freedom to have some pretty good conflicts before agreeing on a plan."

The Challenges of Succession

Making complementarity work is a challenge even for teams that are intact; when the composition of the team is altered, especially in the case of leadership succession, the challenges become more substantial. Researchers have repeatedly identified succession as one of the most worrisome issues for corporate boards, partly because it's one of the areas in which companies most frequently fall short. With a complementary-leadership team, succession is particularly problematic and becomes more so the more extensive the team's complementarity. The classic case is when a COO or president who has worked in a complementary fashion with the CEO moves into that top role.

Consider Ivester's move from COO to chairman and CEO at Coca-Cola in 1997, when Goizueta died unexpectedly. As we have described, the two men had an unusually effective complementary relationship. Many predicted continued success for Ivester: Fortune dubbed him the "prototype boss for the 21st century." Coca-Cola's board confirmed his accession to CEO in a 15-minute meeting.

But the successful pairing of Goizueta and Ivester contained the seeds of Ivester's downfall. An accountant by training, Ivester had spent nearly 20 years rising through the ranks to become Goizueta's right-hand man. He had been named Coca-Cola's CFO in 1985, at age 37, and made his mark in 1986 by orchestrating the successful spin-off of the company's bottling operations. He also succeeded in his first operating role, as president of European operations, overseeing the company's expansion into Eastern Europe in 1989.

But Ivester failed as CEO. He refused to name a new COO, even when strongly pressed to do so by Coca-Cola's board. Instead, he continued to act as a "super-COO," maintaining daily contact with 16 direct reports. His extraordinary attention to detail, a virtue in finance and operations and an effective foil to Goizueta's big-picture approach, proved to be a hindrance in his new position, preventing him from taking on the strategic, visionary, and statesman roles of an effective CEO.

He made a series of mistakes, from the ham-handed treatment of European regulators ruling on Coca-Cola acquisitions to the belated acknowledgment of a festering racial discrimination suit in the company's Atlanta headquarters. He had little chance of recovering from such setbacks because by the end he had alienated nearly everyone who might have saved him.

Ivester's unfortunate story vividly illustrates why leadership succession can be so risky in complementary teams: The strong leadership traits that make someone a successful member of a complementary team may limit his ability to adapt to a different role–because of his inherent capabilities and because of what the organization, comfortable with him in his old role, will permit him to do in his new one.

For example, making the move from COO to CEO requires substantial changes in the four dimensions of complementarity we described earlier. To succeed in the chief executive role, Ivester needed to shift his focus to different tasks (by spending more time on external matters), draw on different areas of expertise (particularly sales and marketing), exercise different cognitive capacities (particularly the development of vision and strategies), and play a new social role in the organization (one that would build personal relationships and trust). He made a common but disastrous mistake in his failure to "promote himself" into the role of CEO, continuing instead in what was essentially a COO role camouflaged by a new title.

Of course, the magnitude of the leap to CEO would tax anyone's adaptive capacity. While hard work may help someone learn new tasks or acquire new expertise, the ability to adopt new cognitive capabilities or play different social roles may be limited by who that person is. To use an analogy from the theater, some actors have a narrow range–they are extremely effective in a few types of roles – while others can perform a broader repertoire. But even the most versatile actor can't change her inherent talents, and few can convincingly play all roles, just as few executives will be able to excel in all leadership positions.

Even if a leader has a tremendous ability to adapt to a new role, he still has to deal with the legacy of the roles he has previously played in the organization. Just as audiences may not like to see a typecast actor in a different kind of role, people in organizations may hold entrenched attitudes that make it extremely difficult for someone to undertake a major role shift. Having played the bad cop to Goizueta's good cop for so many years, Ivester was typecast as the hard-edged disciplinarian. Even if he had been capable of playing a softer, more diplomatic role at Coca-Cola, it is far from clear that people in the organization would have accepted the transformation.

Ensuring a Smooth Transition

Companies that come to understand the potential pitfalls of complementary leadership teams should not, of course, give up on complementarity and stock their leadership pipelines with clones. Rather, the challenge is to enjoy the advantages of complementarity without sowing the seeds for disaster during succession, particularly to the top job. An organization's board of directors and CEO can take various steps to manage the short-term benefits and longer-term risks of complementarity.

The first step is to recognize the inherent tension between creating complementary leadership teams and laying the foundation for smooth succession processes. This can mean, for example, deciding to put in place a COO whose experience and talents partially overlap with those of the CEO and who therefore might make the transition into the top role more easily. Another option is to extend the life of the complementary CEO-COO relationship by appointing as COO someone who will take time to grow into the number two job. Giving that person a stretch assignment will ensure that he has time not only to learn his own job but to prepare, with the CEO's help, for the top position.

More-formal procedures can also help. The board can, for example, broker an explicit agreement between the CEO and the COO outlining the gradual transfer of responsibilities. The COO can take increasing responsibility for dealing with external constituencies such as analysts, shareholders, and the press; become more involved in strategy setting or meetings with key customers; or assume greater responsibility through a seat on the board or a new set of reporting relationships. Many of today's successful complementary leaders – like Jim Donald at Starbucks, Bill Watkins at Seagate, and Indra Nooyi at PepsiCo – have benefited from that kind of preparation as part of an earlier complementary team.

Such a grooming process requires a CEO who is willing to transfer power in both real and symbolic ways. "It became clear to me that certain aspects of the CEO job were not my strong suit," says Chizen at Adobe. "Yet these were things that Shantanu was really good at. He wanted more responsibility, and I was more than happy to give it to him. He continued to do a good job as COO, which confirmed for me that I could safely give him even greater responsibility." Jim Donald, speaking of the responsibility he was given as Starbucks COO when Orin Smith held the top spot, says, "If you don't have a great number one, you are not going to have the latitude and longitude to do what it takes to become a great number one yourself." Of course, handing over the reins is not easy: Consider Watkins's reluctance to turn over responsibility for quality at Seagate when Wickersham became COO.

One way to sidestep the challenges of a complementary COO's transforming himself into a CEO is to ignore the traditional definitions of the two positions–that is, don't worry about which roles the CEO and the COO play as long as they maintain the logic of complementarity. Such a "role-swapping" strategy could mean, for example, that a COO who is strong at playing the "insider" role would continue to do so after becoming CEO, complemented by a new COO who would play the "outsider" role. We saw a version of this strategy at Pepsi-Co, where CFO Nooyi continued to deal with external constituencies, such as analysts, when she took on the role of president, while CEO Reinemund continued to play to his operational strengths. But that kind of approach won't work in all situations. The CEO must be comfortable sharing the spotlight with the second in command, and important stakeholders – investors, analysts, the media –must accept the division of responsibilities.

Instead of trying to manage the sometimes difficult ascension of the number two executive to the top position, a board can orchestrate a complete changing of the guard, selecting a highly complementary CEO-COO pairing–or even a trio of leaders – and then replacing both people when the time comes. That was the pattern for several generations of top executives at Johnson & Johnson. From 1976 to 1989, chairman and CEO James Burke shared leadership responsibilities with president David Clare. In 1989, the two passed the baton to Ralph Larsen, who became chairman and CEO, and Robert Wilson, who became president. They, in turn, passed it in 2002 to William Weldon, who became CEO and was also the worldwide chairman of the pharmaceuticals group, and James Lenehan, who was worldwide chairman of the medical devices and diagnostics group. When Lenehan interrupted this orderly succession process by resigning suddenly in 2004, the company expanded its complementary-leadership structure, appointing in the next year both CFO Robert Darretta and global chairman of pharmaceuticals Christine Poon to the position of vice chairman in the office of the chairman, a move that made them part of a new complementary team headed by Weldon.

Whatever the mechanism, the aim is to realize the enormous potential of complementarity in senior leadership teams while avoiding the perils that come with it. The stakes in getting this right could not be higher.

============================================================================================

A Commitment to Complementarity

The benefits and challenges of running an organization with leaders who play complementary roles can be seen at Goldman Sachs, where for decades many parts of the business – and sometimes the firm itself – have been headed by teams of two co-leaders.

The practice emerged almost by chance. In 1976, when the senior managing director died, the firm decided to fill his position with two partners and members of the management committee, John Weinberg and John Whitehead, who had worked closely together for years. "As friends, they were able to collaborate in a noncompetitive way," recalls Jonathan Cohen, a Goldman Sachs advisory director who started at the firm in 1969. "It was natural for them to come together." Weinberg and Whitehead ran the firm for eight years, and a precedent was set.

Over time, the notion of co-leadership became ingrained in the firm's culture. Although no formal policy mandates that certain businesses be run by more than one person, when a position opens, Cohen says, "you look over the best people for the job, and often there are two with complementary strengths." The practice has extended to the top. Before leaving to become U.S. treasury secretary in July 2006, chief executive officer Henry Paulson, Jr., worked in a close complementary fashion with then president and chief operating officer Lloyd Blankfein, who is now CEO. Earlier, Paulson headed a three-person team comprising himself and co-presidents and COOs John Thornton and John Thain.

The benefits of such arrangements are several. Co-leadership can act as a restraint on the naturally strong egos found at a top-tier investment bank. It can help assimilate senior hires into the organization's culture by pairing the newcomers with veterans of the firm. It also allows the leadership to be in two or more places at once – something that proved beneficial after the attacks of September 11, according to Steven Kerr, a former chief learning officer at Goldman Sachs who is now a senior adviser to the firm (and a director of Harvard Business School Publishing). At the time of the attacks, Kerr says, Paulson and Thornton were out of the country, but Thain was in New York and could thus oversee efforts to restore order at the firm.

Perhaps the greatest benefit of co-leadership is diversity of thought and talent. Decisions, while they might take slightly longer to reach, often are better because two different minds have been at work on them. Co-leaders can play to their individual strengths. When Paulson and Blankfein worked together, Paulson, who had spent his career building client relationships, was Mr. Outside; Blankfein, who had a background in the technical intricacies of financial instruments, was Mr. Inside. Faced with the succession challenge that is often embedded in such a complementary relationship, Blankfein has had to work to raise his public profile and increase his involvement with clients since becoming CEO.

The success of co-leadership at Goldman Sachs reflects both personal and institutional commitment to the concept, Kerr says. He notes that when he reported to both Thain and Thornton, "I would say something to one, and ten minutes later the other would know it. If they had disagreements, and I'm sure they did, I saw none of that." Cohen recalls that Weinberg and Whitehead set ground rules for the relationship early on, including an agreement that if one person felt very strongly about something, they both would head in that direction.

Kerr says such semiformal agreements continue to this day. He and others often interview both members of a new leadership team, map each one's view about who will be responsible for what, and then bring the two together to work out differences.

Of course, not all relationships between co-leaders succeed. After all, Kerr says, "people marry for love and half of them don't make it – and these guys aren't always in love when they are put together." But the successes outweigh the failures, in large part because the firm has learned, sometimes painfully, from its mistakes over the past 30 years. "If we had known what the full cost would be, maybe the practice would never have started," Kerr says. "But that cost is now sunk. We've paid our dues."

Complementary Leadership and the Quest for Peace

Senior leadership teams whose members have complementary strengths abound in government as well as in business. The partnership between Israeli politicians Yitzhak Rabin and Shimon Peres – and the way their complementary roles led to the 1993 signing of the Oslo Accords – provides a classic example of complementarity in a governmental context.

In the early 1990s, Rabin was prime minister and Peres was foreign minister of Israel's Labour government. The two politicians were longtime rivals, representing the hawk and dove wings of the Labour Party. In the partnership, Rabin, a former general in the Israeli Defense Forces, played the role of "guardian," preserving national identity and ensuring the country's security. Peres, who lacked the credibility associated with military service and was viewed as soft on security, acted as the "entrepreneur," developing new approaches to engaging with the Palestinians and aggressively pushing those ideas forward. Together, the two men did what neither could have done alone, pursuing and completing negotiations that laid the foundation for the Oslo Accords, for which the two, along with Palestinian leader Yasser Arafat, received the Nobel Peace Prize.

After the assassination of Rabin in 1995, Peres's history of playing the entrepreneur role hindered his candidacy for prime minister when he ran against Benjamin Netanyahu: Many Israelis worried that he couldn't step into Rabin's role as trusted guardian of the nation's future.

Promise-Based Management: The Essence of Execution

Promise-Based Management: The Essence of Execution

By examining the commitments people make to colleagues and customers, executives can figure out why work stalls and how to get it moving again.

by Donald N. Sull and Charles Spinosa

Managers have a full set of tools for translating strategy into action. They can redraw their organization charts, redesign their business processes, realign employee incentives, or build sophisticated IT systems to track performance. Nevertheless, critical initiatives stall, and important work goes undone. Emerging business opportunities fall by the wayside or, even worse, into the hands of more agile competitors.

Execution fails for a variety of depressingly familiar reasons: Employees disengage because they don’t buy in to the company’s priorities; they become dissatisfied and unproductive. Functional silos hinder the coordination necessary for companies to seize new business opportunities. Matrix organizational structures obscure accountability for projects and initiatives. Indeed, execution becomes especially difficult when executives are charged with managing the activities not only of their direct reports but also of a far-flung network of suppliers, partners, knowledge workers, and colleagues in different time zones around the world.

Managers cannot overcome these and other obstacles to execution by doing more of the same; instead, they must fundamentally rethink how work gets done. Specifically, they must acknowledge that a company is more than a bundle of processes or a set of boxes and lines on an org chart. At its heart, every company is a dynamic network of promises. Employees up and down the corporate hierarchy make pledges to one another—the typical management by objectives. Employees also make commitments to colleagues in other divisions and to customers, outsourcing partners, and other stakeholders. Promises are the strands that weave together coordinated activity in organizations.

Most of the vexing challenges leaders face—improperly executed strategy, lack of organizational agility, disengaged employees, and so on—stem from broken or poorly crafted commitments. Executives can overcome some of their thorniest problems in the short term and foster productive, reliable workforces for the long term by practicing what we call “promise-based management”: cultivating and coordinating commitments in a systematic way.

Why Promises, and Why Now?

Promise-based management builds on a tradition that extends back at least to the emergence of contract law in the Roman Empire . It draws on the tenets of speech act theory, a branch of linguistic philosophy that explores how people commit themselves to action through assertions, questions, requests, promises, declarations, and other speech acts. (See the sidebar “A Primer on Speech Act Theory.”) Promise-based management is particularly relevant to today’s executives as they increasingly specialize in their core businesses, divest noncore units, and outsource peripheral activities. It also helps executives to capitalize on business opportunities outside their core competencies and to engage and retain employees within a highly mobile workforce. Let’s examine each of these business challenges in turn.

Increase coordination and collaboration.

It’s fairly straightforward for managers to get things done when all the relevant people and resources fall within the same P&L or functional division. There is a clear hierarchy, and positional power motivates people to honor their promises—bosses wield carrots and sticks. But the drift toward corporate specialization has been steady in recent years. Executives struggle to make things happen in matrix organizations or networks of loosely allied firms when the people they’re relying on don’t share their assumptions or objectives. A researcher in a pharmaceutical company, for instance, may define success as a breakthrough drug developed over decades. Meanwhile, an outside sales rep may focus on units sold in the short term, rarely looking beyond the next quarter’s quota. Well-made promises can help bridge the gap between such individuals, who may be literally and figuratively miles apart. The dialogues that are central to promise-based management allow people from disparate backgrounds to achieve a common understanding of what needs to be done. Promises also foster a mutual sense of personal obligation to deliver the goods.

Increase agility.

Companies with well-honed business processes usually do a good job of executing on high-volume, routine activities. However, those same processes can prevent firms from taking advantage of opportunities that fall outside their core capabilities—say, entering an emerging market, rolling out a large-scale IT system, or managing an ecosystem of partners to create and capture value. The very standardization that generates continuous improvements in traditional business processes limits companies’ flexibility—and agility matters. In a recent McKinsey survey on building nimble organizations, 89% of the more than 1,500 executives polled worldwide ranked agility as “very” or “extremely” important to their business success. And 91% said it had become more important for their companies over the past five years. Promise-based management can help organizations act more quickly and flexibly. When putting out a request for help with a project or an initiative, for instance, employees can cast their nets wide, within the organization and beyond, to find the right person for the job. Each party to the promise can establish terms to suit his or her specific circumstances and can renegotiate as new information comes to light or as priorities shift—and that’s much less cumbersome than reengineering a well-oiled business process. Because both sides have voluntarily agreed to the commitment—and have put their reputations on the line—they are likely to act with urgency and discipline.

Increase employee engagement.

Many managers attempt to rein in today’s fragmented workforce by creating rigid processes that dampen employees’ initiative and engagement. But organizations that engender well-made, reliable promises create a sense of community among workers—that is, people promise to do things because they buy in to the company’s overall mission and priorities and see their part in making things happen. Promise-based management empowers individuals to act like true entrepreneurs within the organization—to spot opportunities, assemble the resources required to seize those opportunities, and adjust on the fly. Within the bounds of the firm’s objectives, employees can own and run their own personal networks of promises. This sense of ownership, when coupled with wide latitude in managing the negotiations around individual promises, dramatically increases employees’ engagement and therefore boosts overall performance.

Although promises are critical to business success, too often they fail in practice. To a large extent, these breakdowns result from managers’ and employees’ imperfect understanding of how to make effective commitments.

Conversations for Commitment

A promise is a pledge a provider makes to satisfy the concerns of a customer within or outside an organization. For our purposes, “customer” and “provider” refer to roles, not individuals, and these roles can vary depending on the situation. The CIO, for example, is a customer when requesting financial data from the CFO or soliciting a commitment from a subordinate. But she is a provider when supplying technical support to the finance department or making promises to her boss.

A promise rarely occurs in isolation: In order to deliver on a promise, a provider must solicit and oversee a network of supporting commitments from colleagues, subordinates, partners, vendors, and so on. Having to weave this web of pledges makes it that much more complex for the provider to deliver the goods and underscores the importance of managing commitments effectively.

People often take a legalistic view of promises, defining them according to the terms of a deal, much as lawyers might focus on specific clauses in a contract. More important than the actual content of a promise, however, are the discussions that give it life. Both sides must explicitly thrash out what the customer wants and why, how the provider would go about satisfying the request, and any constraints or competing priorities that could derail fulfillment of the promise.

Specifically, the customer and the provider should rigorously go through three phases of conversation to develop and execute an effective promise. The first is achieving a meeting of minds, which is easier said than done. This phase typically starts when the customer requests something from the provider. The two parties will have different takes on what should be done to fulfill the request, why, how quickly it can be done, and which resources should be used. Because of divergent worldviews—across divisions, companies, countries, and languages—people often end up talking past one another. The customer and the provider must therefore sit down and explore the fundamental questions of coordinated effort: What do you mean? Do you understand what I mean? What should I do? What will you do? Who else should we talk to?

The customer and the provider should strive to obtain a common and realistic understanding of what it will take to satisfy the customer, possible obstacles to delivery, and what the customer can do to help if difficulties arise or other priorities compete for the provider’s time and attention. This phase of discussion concludes when the provider makes a promise that the customer accepts.

In the next phase—making it happen—the provider executes on the promise. Regardless of what the provider may think, now is not the time to take the phone off the hook. Conversation is more critical than ever. Even well-crafted promises remain fragile, susceptible to shifts within the organization or in the broader business environment that prompt executives to reshuffle priorities and reallocate resources. In light of such shifts, the customer and the provider will need to continue interpreting and reinterpreting the promise. Indeed, if the provider realizes he cannot satisfy the promise he made to the customer, he should immediately renegotiate the terms of delivery. Likewise, the customer is obliged to initiate renegotiations if her priorities or circumstances change in ways that affect what she has asked the provider to do. This phase ends when the provider declares the task complete and submits it to the customer for evaluation.

In the final phase—closing the loop—the customer publicly declares that the provider has delivered the goods (or failed to do so). Closing the loop gives the customer and provider a chance to offer each other feedback on how they could work more effectively in the future, thereby building continuous improvement into the quality of other promises they make.

Note that the customer and the provider must come not only to a meeting of minds but also to a common purpose. A provider may be reluctant to enter into a commitment for good reasons—such as keeping her options open and protecting her reputation for delivering the goods. It’s critical that conversations about what to do go hand in hand with discussions about why it matters for both sides. In their haste to get things done, many managers rush through these important dialogues or skip them altogether.

The Five Characteristics of a Good Promise

In more than a decade of research on commitments, we’ve asked hundreds of managers to evaluate the quality of promises made within their organizations. We’ve asked them what percentage of all commitments made to them they could actually rely on. The typical response is about 50%. When promises are unreliable, managers waste a lot of time checking progress, exerting political pressure, or duplicating work. Organizational efficiency and effectiveness suffer.

If managers and employees understand how to solicit and make good promises, they can minimize this kind of friction. More important, they’ll be able to overcome the execution challenges thrown at them. We’ve found that well-made promises share the following five characteristics.

Good promises are public.

Promises that are made, monitored, and completed in public are more binding—and therefore more desirable—than side deals hammered out in private. When employees make promises out in the open, in front of their peers and bosses, they can’t conveniently forget what they said they would do, recall only a few conditions of a promise, or back out of an uncomfortable commitment entirely. Nor will they want to, in all likelihood: Psychologists have found that most people strive to make good on declarations they’ve made in public. After all, their reputations for competence and trustworthiness are on the line.

A good example of the power of public promises comes from Royal Bank of Scotland . In the past decade, RBS has moved from the number two bank in Scotland to one of the top ten banks in the world. It broke into the big leagues through its 2000 acquisition of England ’s NatWest, a bank three times its size. RBS did not make the first or the highest bid for NatWest, but it won the prize by promising to improve the target company’s operating performance. RBS didn’t make vague statements about projected synergies or scale efficiencies; instead, its leaders publicly promised to deliver on 154 specific initiatives that, combined, would grow revenues by £390 million and cut costs by £1.2 billion. Moreover, RBS pledged that its managers would take personal responsibility for delivering on those initiatives.

A promise made in public should remain public throughout the life of the commitment. The managers at the Brazilian brewer AmBev each year publicly promise to accomplish five individual goals, all of which are linked to the company’s overall objectives. They pledge to hit target numbers for, say, increasing company margins, improving service levels, or cutting costs. The managers’ performance against these stated objectives is tracked weekly, and the data are posted in the office for all to see. The resulting culture of transparency and execution has helped propel AmBev from the number two brewer in Brazil to the largest brewer in the world (by volume) through InBev, its joint venture with the Belgian company Interbrew.

Good promises are active.

In many organizations, customers hurl requests at providers like paperboys cycling through a neighborhood chucking newspapers onto doorsteps. Providers catch the requests, throw them on a pile, and go back to work. Requests like these rarely elicit good promises. As we discussed earlier, negotiating a commitment should instead be an active, collaborative process.

Misunderstandings will inevitably occur when providers and customers come together from different disciplines, business units, organizations, or countries, or when they are pursuing a novel initiative. Even worse, when an organizational promise is broken, people often believe that the other party has acted in bad faith. Business unit managers complain about the idiots in IT, while software engineers grumble about managers who don’t know what they want. No one gets the benefit of the doubt, and every miscommunication is interpreted as further evidence of evil intentions. This downward spiral of distrust poisons relationships and impedes performance. In such situations, probing discussions can unearth the different assumptions customers and providers are making.

In many organizations, the active negotiation of a promise turns into an exploration of multiple assertions and scenarios, leading everyone to engage in time-consuming rebuttals and “gotcha” questions designed to demonstrate the inquisitor’s cleverness rather than get closer to a good promise. These discussions often start out productively but stall when the participants seek complete certainty before hammering out a deal. The top executives of one biotech firm were caught in exactly this trap. The senior team consisted of brilliant scientists who spent all their time trying to prove they were right. The discussion was insightful and erudite, but in the end nothing got done.

Active conversations should comprise offers, counteroffers, commitments, and refusals rather than endless assertions about the state of nature. In the biotech company, the senior executives eventually agreed to make clear requests of one another and provide only the background information necessary to flesh out those requests for potential providers. (One of the executives was tasked with interrupting those who went overboard explaining their rationales.) Potential providers were allowed two rounds of questioning for clarification. They were then required to get back to the customer within 48 hours with a refusal, a commitment to act, or a counteroffer. In a matter of weeks, the team’s discussions shifted from endless debates about reality to clear requests for action and promises to deliver.

Good promises are voluntary.

In many organizations, people feel compelled to comply with each and every request in order to be seen as team players, please their bosses, or avoid looking like jerks. For instance, in the past, employees at General Motors made liberal use of the “GM nod”—a polite yes to every request. But when the response to every request is yes, what does it really mean? It might mean “yes”—but it might mean “no” or “Is it time for lunch yet?” or “I have a pulse.”

The most effective promises are not coerced; they are voluntary. The provider has viable options for saying something other than yes. Contracts signed under duress are not binding in a court of law. Similarly, psychologists have found, people assume little personal responsibility for promises made under threat (although they may comply out of fear). By contrast, people feel deeply obliged to follow through on a promise if they exercised free will in making it.

While providers shouldn’t be expected to fulfill every request, they also cannot be allowed to avoid making promises. Instead of automatically saying no, a provider can respond to a customer’s request with a counteroffer—for instance, “What you’re asking is not possible, but this is what I can do for you.” A thoughtless yes and a reflexive no are both passive responses to a customer’s request, but a counteroffer signals the provider’s active interest and voluntary engagement in helping the customer succeed.

Senior executives must therefore give providers the space to decline customers’ requests or to make counteroffers. An executive in one information technology company we worked with gave his direct reports a set of cards in which most were marked “yes” or “counteroffer” and three were marked “no.” Using those cards, subordinates could decline three requests per quarter, provided they publicly offered a clear explanation why.

Of course, managers should recognize that some team members may abuse an opt-in philosophy toward making promises. Keeping commitment-phobic employees on the team degrades the power of promises for everyone else.

Good promises are explicit.

Customers and providers should clearly acknowledge who will do what for whom and by when. The need for explicit negotiation increases in situations in which a new party replaces an established one, a company’s employees are culturally diverse, or an abstract construct (optimization or innovation, for example) gives rise to multiple interpretations. Implicit promises are quick and easy to establish but often result in misunderstandings.

The customer and the provider must be explicit about their promise throughout its life cycle. Requests must be clear from the start, progress reports should accurately reflect how the promise is being executed, and success (or failure) should be outlined in detail at the time of delivery rather than after the fact, during a quarterly performance check-in or through annual 360-degree feedback.

A large hydroelectric engineering joint venture we worked with recognized the need for clarity of organizational promises and created a system for making sure the lines of communication stayed open between customers and providers. From its inception in 2000, Voith Siemens Hydro Power Generation battled upstart Chinese and Indian manufacturers at the low end of the market and established rivals, including GE and Alstom, at the high end. Voith Siemens decided that the best way for it to compete would be to offer its customers integrated solutions—entire power houses, including turbines, generators, and other components. To make this strategy work, however, managers needed to re-create the way employees in different disciplines, departments, and regions coordinated their activities. CEO Hubert Lienhard and his team initiated a program to improve the quality of commitments people in the organization were making. The engineers in the various disciplines, for instance, created and widely distributed a set of checklists and memo templates to be used as guides for making requests and promises. The checklists specified a half-dozen or so aspects of any request or promise that must be explicit to both parties. These included names, dates, underlying rationales for requests, the skills necessary to fulfill promises, and so on. The engineers also established periodic design freezes, during which design coordinators, referring to their checklists, would ensure that customers and providers maintain identical understandings of their requests, promises, and counteroffers.

Explicit promises foster coordination and execution across an organization. They keep customers satisfied and providers on point. But that doesn’t mean the terms of a promise should be etched in stone; they can and will evolve as circumstances change, priorities shift, or new information emerges. Renegotiation of promises may not always be pleasant—it can be risky, time-consuming, and resource intensive—but it is critical. Customers and providers must have the scope to recalibrate in order to seize emerging business opportunities.

Onset Ventures, in California, has cultivated more than 100 early-stage technology start-ups since 1984, and nearly 80% of them (compared with the industry average of about 20%) have gone on to higher rounds of financing. Like most venture capital companies, Onset stages its funding in rounds. At the start of each round, the entrepreneur and Onset negotiate a small set of explicit objectives to meet before moving on to the next round—for instance, develop a working beta product, sign five reference customers, and survey 100 potential customers to determine demand. The promises are ironclad within a round, but they are expected to change (and do) in subsequent rounds. Throughout the rounds, the entrepreneur and the VC firm reassess the situation and agree on new promises to accommodate shifting business needs.

Good promises are mission based.

Often, a customer will solicit a promise from a provider without offering any explanation for why the request matters. As a result, the provider infers that the request isn’t critical or that the customer doesn’t consider the provider important enough to deserve an explanation or smart enough to understand it. In any case, the outcome won’t be pretty. The most effective promises are mission based—that is, the customer explains the rationale for the request and invests time to ensure that the provider understands the mission. Sure, it can be cumbersome to explain where a division fits in the corporate strategy and where a particular request belongs within that division. But when providers understand why their promise matters, they are more likely to persist in executing even when they encounter conflicting demands and unforeseen roadblocks. They can also exercise creativity in addressing customers’ underlying concerns rather than blindly fulfilling the letter of the stated request.

The U.S. Marine Corps, for instance, uses what it calls mission-based orders. These requests clearly articulate what the commanding officer wants and why, while leaving the methods of implementation to the discretion of the subordinate officer closest to the situation on the ground. Each order includes an explanation—known as the commander’s intent—of why the objective matters to the commanding officer and to his superior as well. Business leaders can apply a similar discipline by explaining to providers why requests matter to them. They can gauge whether providers understand and support the overall rationales for a request by asking them to articulate in their own words why the request matters.

• • •

Promises are the fundamental units of interaction in businesses. They coordinate organizational activity and stoke the passions of employees, customers, suppliers, and other stakeholders. While they hold an organization together, they are as fragile as they are crucial. Individuals’ divergent worldviews and objectives tug constantly at the filaments of promises, and unexpected contingencies can tear precarious agreements. Leaders must therefore weave and manage their webs of promises with great care—encouraging iterative conversation to make sure commitments are fulfilled reliably. If they do, they can enhance coordination and cooperation among colleagues, build the agility required to seize new business opportunities, and tap employees’ entrepreneurial energies. If they don’t, they will lose out to rivals who do.

A Primer on Speech Act Theory

Most executives prefer doing to talking, but they also spend between two-thirds and three-quarters of the workday in formal or casual discussions. So how do they get things done with words? Speech act theory—a branch of linguistic philosophy that explores how people use words to coordinate action—says that talking is doing.

For centuries, philosophers viewed language as a tool for describing external reality. Sentences such as “It is raining” were considered true or false on the basis of how well they corresponded to real-world conditions. But in the 1950s, Oxford philosopher John L. Austin argued that many statements are intended to get things done rather than describe reality. When an umpire calls a strike, a military officer issues an order, or a supplier promises to provide a service, that individual is not describing reality but changing it through his or her utterances. Austin argued that speech always falls somewhere along a wide spectrum between purely descriptive statements, such as scientific equations, and purely active statements, such as a priest’s declaration that a couple is married.

University of California philosophy professor John Searle later introduced a taxonomy of speech acts based on the roles that different statements play in getting things done. Commissives bind the speaker to a future course of action and include not only promises but also offers (I will do this if you accept) and counteroffers (I can’t do that but could do this). Directives attempt to induce the listener to do something; they include entreaties, requests, and commands. Declarations are authorized pronouncements that change the state of affairs in the world, as when a boss fires a subordinate. Expressives commit the speaker to feeling a certain way about the current state of affairs, as when someone apologizes for doing something. Assertives commit the speaker to a truth and imply future actions consistent with it.

Searle’s student Fernando Flores argued that most corporate conversations are waylaid by attempts to unearth absolute truths that everyone can agree on and that will produce a clear agenda for all to follow. Like philosophers, Flores argued, managers have been seduced by the belief that talking is about describing rather than doing. Requests and promises are the basic units of coordination in commercial organizations, and assertives should be used primarily to clarify those requests and promises.

Consciously or not, managers (through their utterances) create an intricate web of requests, commitments, assertions, and declarations that affect how people in their organizations act.