Developing Leaders: How Winning Companies Keep On Winning

Reading Time: 28 min 



Last June, as the business world watched expectantly, General Electric (GE) promoted three key executives (David Calhoun, Joseph Hogan and John Rice). Each would report to one of the three potential candidates to succeed CEO Jack Welch (James McNerney, Jeffrey Immelt and Robert Nardelli). Thus, if one of the latter moves into Welch’s office, another seasoned GE professional will be ready to assume his role. Where do GE leaders come from? They do not spring up out of the earth overnight. For many years the company has worked hard to develop ongoing sources of leadership talent — not only to prepare for Welch’s retirement next year, but also to enrich every level of the organization with strong leaders. When Ronald Reagan was spokesperson for GE, their slogan was, “Progress is our most important product.” Today the mantra could be, “Leaders are our most important product.”

Leaders who keep learning may be the ultimate source of sustainable competitive advantage. With that understanding, many companies are investing in leadership development (programs that help key executives learn leadership skills). As early as 1993, Business Week estimated that $17 billion was being spent annually on helping managers develop the thought processes and company-specific skills that could enable them to move up and lead their business areas. Training magazine estimates that in 1998 U.S. companies spent $60.7 billion on training.1 But spending isn’t the only commitment. World-class executives are investing significant amounts of their time personally guiding and mentoring future leaders.2 To them, leadership development is not a luxury but a strategic necessity.

What processes transform managers into strong leaders ready for strategic action? How do the best leadership-development organizations design, manage and deliver world-class programs?

In January 1998, the nonprofit research group American Productivity and Quality Center, based in Houston, the American Society for Training and Development, based in Alexandria, Virginia and author Robert Fulmer set out to find the answers. The group developed a study to investigate best practices in leadership development; in 1999, they expanded the study to explore the challenge of developing leaders at all levels of an organization.3

Thirty-five organizations participated as sponsors. (See “Benchmarking Methodology.”) They sent representatives to a planning session, completed data-gathering surveys and attended or hosted on-site interviews. (See “Study Sponsors.”) The consortium identified six companies as having a strong or innovative leadership-development process. Those six agreed to participate in the study as best-practice partners. (See “And the Winners Are…Best-Practice Partner Organizations.”) They were chosen because they exhibited commitment to developing leaders — but with marked diversity of approach, emphasis and culture.

Benchmarking Methodology »

Study Sponsors

View Exhibit

And the Winners Are… Best-Practice Partner Organizations

View Exhibit

The Strategic Perspective

Most significant: Best-practice partners reported that they tied leadership development closely to business strategy and that they invested financial resources in it. CEOs did not support the programs out of a respect for education but from a conviction that such programs can assist in aligning functional areas with corporate strategy. Johnson & Johnson, for one, revised succession planning and performance-management systems to reflect the qualities anticipated by a leader-of-the-future exercise in a leadership-development conference.

Increasingly, programs that focus on developing future leaders are seen as a source of competitive advantage. GE’s CEO, Jack Welch, described the company’s Leadership Development Institute in Crotonville, New York, as a “staging ground for corporate revolutions.” In fact, innovative ideas such as the Six Sigma quality-improvement program and GE’s expansion into emerging economies have come from presentations made at leadership-development events. Tremendous growth, reductions in the number of GE employees worldwide and significant delayering of the organization in the 1980s and 1990s caused an enormous cultural shift. With fewer layers of management, individuals received fewer vertical promotions and hence fewer opportunities to practice being leaders. A new approach was called for. Today in the human-resource department’s “Session C” meetings, senior executives assess key GE personnel. After an initial meeting in March, there are two or three additional meetings and a wrap-up session in June or July to select employees who will attend the executive-development courses at Crotonville. At the end of the year, corporate leadership development, like all corporate functions, is measured by whether it was able to support GE initiatives. Steve Kerr, GE’s chief learning officer, says jokingly, “Crotonville is GE’s only unbudgeted and unmeasured cost center.” Then he adds seriously, “Everyone would know if we weren’t delivering strategic value.”

Ralph Larsen, chairman and CEO of Johnson & Johnson, champions J&J’s Executive Conference. Faithful to the company’s decentralization tenets, he leaves the program’s details to subordinates around the world but takes the time to suggest program themes.

At Arthur Andersen, the mission of the Partner Development Program (PDP) is “to help partners worldwide acquire and build the knowledge, skills and behaviors required to be valued and trusted business advisors in an ever-changing marketplace.” To meet the needs of a business that continues to diversify and globalize, managers aim to keep the program closely linked with Arthur Andersen’s evolving business strategy.

Hewlett-Packard, under the leadership of CEO Carly Fiorina, is rushing to reclaim its status as a top high-tech innovator. Fiorina must convince the public and HP employees that HP is the hottest new company of the Internet era — without losing the old-time commitment to quality and integrity. Past HP glory led many excellent engineers to focus on what used to be important, instead of on the future. Once HP started to improve leadership development, the company could make better business decisions.

Today HP’s senior executives actively participate in leadership development. Fiorina uses management meetings and leadership-development programs to articulate her vision of making the company “represent the next decade rather than the past one.” Her predecessor, Lewis E. Platt, showed his support for leadership development by making personal appearances at all HP Accelerated Development Programs, opening and closing them with an opportunity for participants to have a dialogue with him. And Bob Wayman, HP’s CFO, was the internal champion for a worldwide broadcast on closed-circuit television. During the broadcast, Wayman played an active role as facilitator of the panel discussion “Challenging the Growth Barrier.” Senior HP executives have served as faculty in part of every core program. Fiorina’s early commitment to communication with her management team has led to an expectation within the company that she will continue using leadership development as a “bully pulpit” for strategic change.

Cor Herkstroter, the former chairman of the Royal Dutch/Shell Group (which has dual headquarters in the Netherlands and the United Kingdom), once asked his top 50 to 60 leaders to suggest improvements in the company’s financial performance. Shell’s committee of managing directors decided that a new leadership-development process could be a catalyst for organizational change, and Shell’s Leadership and Performance (LEAP) program was created. After the program showed measurable returns in the United States and elsewhere, Shell made it corporatewide.

When James Wolfensohn joined the World Bank as president in 1995, he created a mission statement that continued the bank’s longstanding commitment to dispersing knowledge and financial resources but placed a stronger emphasis on the goal of reducing poverty worldwide. The new focus required change. Recognition of that need led to the Executive Development Program (EDP) for managers — a unique collaboration among Harvard Business School, the Kennedy School of Government, Stanford University, INSEAD and IESE (the graduate business school of the University of Navarra, Spain) — which offers five weeks of classroom training and a project to help the bank become more of a world leader.4

The Five Critical Steps

Pronouncing a strategic vision is not enough to bring about change or to tie leadership development to the company’s goals. Our data suggest that there are five critical steps to achieving those ends. (See “Making Leadership Development Strategic.”) Examples of each step can be found in the corporate leadership programs of benchmark companies.

Making Leadership Development Strategic

View Exhibit


The need for a process to build leadership skills has best-practice organizations looking both inside and outside their organizations for approaches that work. The foundation of such companies’ leadership development is awareness — awareness of external challenges, emerging business opportunities and strategies, internal developmental needs and the ways other leading organizations handle development.

Arthur Andersen uses both internal and external data to determine the learning and development needs of the partners in the firm. Internal data come from client-satisfaction and employee-satisfaction surveys, upward communication and analysis of what the firm calls 450-degree feedback (360-degree feedback plus client evaluations).5 Arthur Andersen wants to know how its partners are perceived in terms of their technical competence and their responsiveness to customers.

External data about new financial and managerial tools or about challenges in the business environment come from market research, business trends and leading-edge thinkers. The partner-development program also exploits the research Arthur Andersen does while serving clients — and the conversations on emerging trends partners routinely have with leading thinkers in management education and business practice. The PDP also makes use of the literature on new concepts in leadership development.

To ensure that its leadership-and-performance program does not simply react to the immediate needs of the business, Shell’s LEAP team has an ongoing conversation with the committee of managing directors (representing all of Shell’s geographic and functional areas) about corporate transformation. LEAP staff members negotiate an agreement with the executive of the business unit and the critical players who will go through the program. Together they create budgets for the team project and set time expectations and goals.

To gather external perspectives, Shell has joined the Global Research Consortium, a group of transnational companies that sponsors research. The consortium gives its members the opportunity to hear and discuss the latest on leadership and learning. Like other best-practice companies, Shell also works with consultants and professors to stay abreast of the latest in leadership research.

There is greater awareness today that best-practice organizations’ corporate leadership-development function is specifically for strategic issues; more-tactical management skills and business-specific challenges are usually left to business units. That seems to work well — corporate leadership efforts in best-practice organizations complement learning experiences within the business units and don’t compete with them. Business operations are better equipped to handle their own management-skill training; corporate leadership programs concentrate on helping decision makers become more effective at using those skills.

All the best-practice leadership programs tap leaders with extensive line experience. At Arthur Andersen, Johnson & Johnson and Shell, the heads of the leadership-development process have senior-level business experience. The use of business leaders is based on a belief that participation from executives will help ensure buy-in from the businesses and will keep the programs practical.

GE and Shell International bring in high-potential individuals on two-year rotational assignments to oversee leadership development. HP recruits key people from line positions for the same purpose. In addition to drawing on the business units, best-practice organizations access the experience of individuals in corporate education, human resources and academia. The director of GE’s Crotonville center came from a university setting, and the head of World Bank’s EDP has a background in corporate education.


Although business cases traditionally focus on the past and best-practice reviews focus on the present, the best leadership-development programs emphasize the future. Top leadership-development companies use anticipatory learning tools: focus groups that explore potential challenges or the impact of emerging technologies; decentralized strategic planning (planning that builds on many organizational levels’ imagining of the future); analysis of future scenarios; and the Delphi method (successive rounds of composite predictions used to build awareness and consensus).6

Strategy guru Gary Hamel recommends decentralized planning because revolutions are beneficial and they “seldom start with the monarchy.”7 The participative and future-centered Merlin Process is an example of decentralized planning.8 Managers imagine the organization a decade from the present and describe what it would look like if totally successful. In contrast to more conventional, top-down strategic planning, the Merlin Process has groups throughout the organization describe their ideal. The resulting presentations provide insight and input for senior executives and lead to more-formal planning sessions.

From 1993 to 1996, J&J followed that pattern. During its second set of Executive Conferences, executives from around the world worked together for a week with outside consultants to create a vision for a decade later. Participants challenged conventional wisdom about the evolution of the health-care industry and focused on actions their divisions could take to create their future. J&J 2002, an extended scenario developed from future-focused interviews with more than 100 executives in six countries — and from published predictions about the future of health care — projected multiple trends and discontinuities. Using a modified Delphi approach, participants assessed the probability and impact of 14 hypothetical developments. An integrative exercise called the Merlin Exercise was used to tie the various aspects of the program together. (A Merlin Process asks people to create a future vision of the company; the Merlin Exercise has participants apply course concepts to that vision.) Participant groups made formal presentations of the desired future to the CEO or the vice chairman.

The Merlin Exercise gets results. Upon being congratulated about a significant promotion to head a new J&J venture, one program graduate responded, “Well, it took almost a year, but I finally got our Merlin.” The first step in the envisioned future had become a reality because of his commitment to the potential.

For some organizations, anticipation involves developing a list of the competencies that the company will need. (See “About Competencies.”)

About Competencies »


Action, not knowledge, is the goal of best-practice leadership-development processes. Best-practice groups bring the world into the classroom, applying real-time business issues to skill development.9 The answers to tough questions are not in the instructor’s head; learners must discover them on the spot. And with program participants implementing their own recommendations, the learning experience benefits both the organization and the learner. Such action learning can be complicated and costly, however. That is why Arthur Andersen, for example, uses a modified approach that still includes prework and postwork. Before the course begins, the firm gives participants criteria for selecting a client with a business problem and a protocol for interviewing that client. Learners work in a team to develop client recommendations. After the course, the team must make a presentation to the actual client or the program sponsor.

At GE, Welch himself has been the one to choose the action-learning topics for each of the three annual business-management courses and for the annual executive-development course. Participants in both courses are highly motivated to carry out projects, important as they are to the company’s direction.

Recommendations made by the participant teams are usually implemented. Students in one management course went to Russia and developed proposals for GE’s operations there. A quality report from an individual in another course led to corporatewide adoption of the Six Sigma initiative, a quality-assurance program designed to eliminate defects from all products.

GE also supports what it calls the Change Acceleration Process (CAP), a systematic attempt to turn managers into professional change agents by disseminating GE’s accumulated knowledge about how to initiate, accelerate and secure change. If CAP is successful, says Welch, “people who are comfortable as coaches and facilitators will be the norm at GE. And the other people won’t get promoted.”10

At Johnson & Johnson, the purpose of the third group of Executive Conferences, which started in 1997, was to emphasize J&J’s Standards of Leadership (a model developed by key executives working with McKinsey consultants) and to tie the standards to specific business issues through action learning. The principal session lasted five days, with prework and follow-up extending the experience. Before the core session, each operating unit discussed the business topic it would focus on.

Different J&J executives in the various businesses “sponsor” each conference session. Those who choose the topic are asked to pick one that can have significant or transformational impact. Past program topics have included top-line growth, product-development cycles, new-market entries and leadership development. Once the topic is defined, the executive sponsor chooses 50 to 130 program participants, who do additional preparation, such as gathering data and interviewing people in the company who might have some relevant insight.

Participants go through the program and return later for a day to report on implementation results. Typically, the process takes six to nine months. J&J’s Executive-Conference approach includes work teams from the business area that is experiencing the problem being studied. The company’s action-learning approach at the middle-management level, however, brings together high-potential individuals from all parts of J&J to tackle a more broad-based issue. The Executive-Conference issues aim more at organizational development, whereas middle-management programs focus more on development of individuals’ skills.


Because best-practice organizations recognize the importance of alignment between leadership development and other corporate functions, they often tie educational efforts to formal succession planning. At a few of our best-practice partners, the leadership-development function and the succession-planning function report to the same executive; other enterprises merely emphasize that a natural link exists.

At J&J, all development functions use 360-degree-feedback evaluations as a part of leadership development. Facilitators assess a multiple-choice, behavioral questionnaire, in which participants rate their performance in many areas and get ratings from supervisors, peers and subordinates. Plans may be made for participants to be coached later or to engage in activities to strengthen weak areas as part of the program, but the facilitators’ assessments are not typically fed directly into succession planning.

Although the data were not conclusive, we believe the best companies are beginning to integrate and align assessment, development, feedback, coaching and succession planning. In the integrated model, leadership development becomes an important part of maintaining a steady flow of information throughout an organization and ensuring that top talent is tracked and continues to grow.

GE openly ties leadership development to succession planning. All employees are rated in a nine-block system for the annual Session C review. The review includes discussion about people’s performance and their adherence to the values in GE’s value statement. The system is an approximation of a typical competency model but was created quickly, simply and with GE self-confidence from a comment by Welch and elaboration by his HR team. It features a chart on which an employee’s bottom-line performance is rated on one axis, with adherence to GE values on the other axis. Those who don’t make their performance numbers but do adhere to GE values are given a chance to improve those numbers and get a higher rating. Those who make their numbers but don’t demonstrate the GE values are rated low in the four-level model, which gauges promotion suitability. Those who do neither are rated lowest.

Senior managers spend most of their resources developing their best and brightest. They see that approach as delivering the most mileage: creating both role models and alignment on what is valued. Crotonville’s leadership-development offerings are targeted at high-potential individuals, people the organization refers to as its “A Players.” Each year the institute trains about 10,000 of GE’s approximately 300,000 employees.

It is commonly said that, at GE, the corporate headquarters owns the top 500 people in the company and just rents them out to the businesses. To encourage the sharing of business talent, GE includes a negative variable in its performance appraisals for managers who hold back talented employees. Outstanding business performance and development of leaders go hand in hand.

Hewlett-Packard provides myriad opportunities for emerging leaders to develop and grow. Platt, the former CEO, recognized that many people who grew up with the founders were retiring and that their immediate successors looked a little too much like one another. He saw that as the company became more global, it would need more diversity of ethnicity and gender. Having a female CEO now may help change perceptions about who is leadership material. And HP’s leadership-development process is clearly supportive of diversity goals, providing stretch assignments for the most promising people and making accelerated programs available for individual contributors and first-level managers.

Best-practice organizations use the goals of their leadership-development program as guides to putting the right people in the right programs. The goal of Shell’s LEAP program is to create leaders at all levels, so the programs are open to anyone within the organization (although certain programs are targeted for those with the highest potential). GE and HP are more selective about entrance because they want to focus only on those individuals with the potential to move quickly through the ranks.


Best-practice organizations always assess the impact of their leadership-development process. To collect information on the perceived value, the best-practice partners use a number of tools and techniques. The Kirkpatrick Four-Level Model of Evaluation (participant reaction, knowledge acquired, behavioral change, business results) is typical.11 Participants, human-resource-development staff, consultants and, in some instances, financial staff, do the assessments — the latter weighing program expenditures’ return on investment. (See “An Executive Primer.”)

An Executive Primer: Five Steps Help Managers Get the Most From Leadership-Development Programs »

Most best-practice partners use an assessment method called the Kirkpatrick levels to quantify the effect of leadership programs on business results. But both the study sponsors and best-practice partners use other metrics, too — including corporate performance, customer satisfaction and employee satisfaction. In general, best-practice partners were more aggressive than sponsors about measuring and evaluating program effectiveness.

Of all the best-practice organizations, Arthur Andersen is probably the most dedicated to assessment — and has reaped the benefits. The vast amounts of data the firm collects not only demonstrate the partner-development program’s correlation with improved business results, but also show where the organization needs to head.

Although measurement is expensive and sometimes tricky, its benefits cannot be discounted. Arthur Andersen combines impact research with participants’ course evaluations. Program attendees fill out evaluation forms before they take the class, immediately after the class is finished and three months later. The forms contain questions about the knowledge participants believe they have gained.

The impact research consists of comparing, course by course, partners who have attended PDP with those who have not. The results show that attendance increases both client satisfaction and per-hour billings. Impact research is done in a two-year cycle, with information gathered on partners a year before the program and a year after.

The use of both participant-satisfaction and impact-research measures helps provide a balanced set of results. Arthur Andersen found that one of its programs was not getting a high participant-satisfaction rating, but an impact analysis showed that the program was having a greater effect than any of PDP’s other courses. (But, PDP staff members prefer to see both strong satisfaction and strong impact, so if partners have a low level of satisfaction the course will not be recommended.)

Regression analysis shows that course duration is an important factor in how satisfied participants are. The perception that a course is too long has a negative effect. (However, a belief that the program length is appropriate does not by itself improve overall satisfaction.)

Another key factor in determining satisfaction is whether program participants have similar levels of familiarity with the topic. What may be an exciting concept for one person could be old news to another. Arthur Andersen believes that moving to a problem-based course design will help it address different levels of participant knowledge.

To Shell’s LEAP staff, a program adds value only if the team project generates revenues at least 25 times greater than the project’s cost (a 25:1 ROI). During the initial contracting process, a LEAP staff member and the leader of the business unit determine the desired project outcomes, including financial targets. The business leader expresses his or her objectives in sending the candidate to the program; in many cases that defines the program and problem the team or individual will address.

You get what you pay for. Best-practice companies do consider costs, but their main focus is on the value the program can provide. When asked to rank the importance of various criteria in selecting an outside vendor, the companies put fees near the bottom of the list. Arthur Andersen invests approximately 6 percent of total revenues in education (more than $30 million). If course offerings achieve their objective to improve business results, support is likely to continue.

Cost for each participant in the executive-development program at the World Bank is $22,000, which includes travel, lodging and business school fees for three modules and for the Grass-Roots Immersion Program (which gives managers one-week stints in a developing-country village or urban neighborhood so they can acquire a firsthand understanding of poverty). The cost is not charged back to the business groups but funded centrally through the bank’s $12 million annual executive-education budget.

The New Strategic Reality of Leadership Development

Globalization, deregulation, e-commerce and rapid technological change are forcing companies to reevaluate the way they operate. Approaches that have worked for years are no longer effective. Development of leaders who think strategically is increasingly a source of sustainable competitive advantage. Hence observations of companies known for excellent leadership-development practices can be invaluable.

Leadership development has become too specialized to relegate to human-resource departments. In best-practice companies, top-level managers get involved. Without their support, leadership-development processes would founder. Of course, corporate leaders are more likely to offer support if programs are producing business results. By monitoring the effectiveness of the leadership-development processes, capitalizing on quick wins and communicating successes throughout the organization, the best-practice companies sustain a virtuous cycle.

Development groups such as Arthur Andersen’s PDP and the one at GE’s Crotonville site emphasize diligent crafting of programs, careful listening, constant monitoring and frequent communication. That helps senior executives understand how a leadership-development process can shape and disseminate an organization’s culture, overcome resistance to change and achieve strategic goals.

At GE, the corporate leadership-development group endeavors to maintain buy-in. It interviews company leaders around the world on a regular basis to gauge future business needs and the characteristics future leaders should have. Additionally, the group at Crotonville identifies early adopters of a given development initiative and leverages their support.

Hewlett-Packard has garnered support for its leadership-development process by having both the CEO and senior managers participate in its programs. The executives serve as mentors, faculty and supporters in leadership-development design and programs.

Senior-level support for Johnson & Johnson’s executive conferences is evidenced by the fact that either the chairman or a member of the company’s executive committee participates in each session, articulating J&J’s credo and values — and the program’s link to business success.

Although the best-practice firms differ in their emphasis on making leadership development strategic, the development program of each includes elements of the five critical steps:

  • building awareness of external challenges, emerging strategies, organizational needs and what leading firms do to meet the needs;
  • employing anticipatory learning tools to recognize potential external events, envision the future and focus on action the organization can take to create its own future;
  • taking action by tying leadership-development programs to solving important, challenging business issues;
  • aligning leadership development with performance assessment, feedback, coaching and succession planning; and
  • assessing impact of the leadership-development process on individual behavioral changes and organizational success.

Most people, even if they have heard of the specific leadership-development practices of Johnson & Johnson or GE, have not grasped how to manage an integrated set of variables in order to achieve excellence in developing executives. Increasing the budget for education or changing the name of a training department to “corporate university” doesn’t guarantee improved performance. Our study shows that, despite the diversity of approaches to leadership development in the best-practice companies, all share common goals: anticipating, supporting and aligning the organization’s strategic initiatives with development, as well as gaining and sustaining competitive advantage. And increasingly, those companies choose an action-oriented, ongoing learning process closely linked to the strategic needs of the business.



1. “Industry Report 1998: Training Budgets,” Training, October 1998, 47. See also:

J. Reingold, M. Schneider and K. Capell, “Learning To Lead,” Business Week, Oct. 18, 1999: 76.

2. John A. Byrne, “PepsiCo's New Formula,” Business Week, Apr. 10, 2000, 172; and

T.A. Stewart, “How to Leave It All Behind,” Fortune, Dec. 6, 1999, 345–348.

3. G. Hollenback and W. Vestal, eds., “Developing Leaders At All Levels” (Houston:

American Productivity & Quality Center, 1999); and R.M. Fulmer, J. Camillus and J. McMorrow, eds., “Leadership Development: Building Executive Talent” (Houston: American Productivity & Quality Center and American Society for Training & Development, 1999).

4. A.A. Vicere and R.M. Fulmer, “Leadership by Design” (Boston: Harvard Business School Press, 1998), 218–223.

5. Arthur Andersen uses the term “450-degree feedback” to emphasize that feedback from clients is included in the process. Technically, 360-degree feedback is feedback from four quadrants of perspective: boss, subordinates, peers and customers. Respondents rate themselves.

6. For example, see:

M.A. Hammer, “The Power of Reflection,” Fortune, Nov. 24, 1997, 291–295.

7. G. Hamel, “Strategy as Revolution,” Harvard Business Review 74 (July–August 1996): 69–81; and G. Hamel, “Leading the Revolution” (Boston:

Harvard Business School Press, 2000).

8. R.M. Fulmer and S. Perret, “The Merlin Exercise: Future by Forecast or Future by Invention?” Journal of Management Development 12 (no. 6, 1993): 44–53.

9. For example, see:

Stratford Sherman, “How Tomorrow’s Leaders Are Learning Their Stuff,” Fortune, Nov. 27, 1995, 90–98; and

Thomas A. Stewart, “Telling Tales at BP Amoco,” Fortune, June 7, 1999, 220–223.

10. “General Electric: The House That Jack Built,” The Economist, Sept. 18, 1999, 23–27.

11. D. Kirkpatrick, “Evaluating Training Programs: The Four Levels” (San Francisco: Berrett Koehler, 1994).

Reprint #:


More Like This

Add a comment

You must to post a comment.

First time here? Sign up for a free account: Comment on articles and get access to many more articles.

Comment (1)
John Kjenner
Unfortunately, this didn't protect Arthur Anderson from Enron.