Statistical Leadership

by Ronald D. Snee and Roger W. Hoerl

Statisticians, as we have known them for almost half a century, may become an endangered species. Although there is a great deal of discussion about how new technologies can be used to perform the traditional role of the statistician efficiently, there is little discussion of whether this old role still makes sense in today’s business environment. We need to fundamentally rethink our 50-year-old paradigm for the statistics profession; otherwise, we are likely to decline in societal influence.

Consider the environment in which the profession evolved up to the 1970s: Much of today’s common vocabulary—website, internet and URL—didn’t exist. Few people had access to computers, and there was virtually no commercial statistical software on the market. Only extremely limited statistics courses were available to students, and statistics was not well integrated into other disciplines. In the large, functionally oriented organizations that were common in government, business and industry, statisticians owned statistical methods because no one else was adequately prepared to use them.

Statisticians focused on data analysis because there was no infrastructure in place for others to do this. Data analysis was considered a separate job, removed from the process that created the data. Statisticians simply provided information they had extracted from the data to others, such as engineers, business leaders, government officials and social scientists, who used it to guide their decisions. Both the public and private sectors were willing to fund the
creation of separate consultative statistical groups, and the role of statisticians evolved as passive and consultative.

During the past 10 years, however, a series of far-reaching upheavals has rendered the traditional consultative role of the statistician out-of-date. Today, most employees in academia, government and the private sector have a personal computer connected to the Web, which makes virtually any information available to anyone in the world at any time. Commercial statistical software is also ubiquitous. For example, Minitab is loaded on the PCs of all General Electric (GE) employees.

Statistics classes are now commonplace in academia. They are included in most engineering, business, economics and social science curricula and even some liberal arts programs. According to Curt Hinrichs of Duxbury Press, nearly half a million U.S. students will take a general introductory statistics course in college this year, a figure that doesn’t include the many students enrolled in introductory courses targeted for business, engineering and other disciplines.

Similarly, the private sector has been providing mass statistical training through initiatives such as Six Sigma. At the same time, the private sector has been downsizing, and specialized consultative organizations have been hit hard. The internet has also made it relatively easy to move white-collar jobs to low cost locations, such as India, Mexico and China.

As the market for data analysis to support a client’s decision making rapidly erodes, statisticians must adapt by expanding their role from consultation to leadership, from reactive to proactive and from value neutral to value adding. How might such a change look in practice?

In the traditional consultative role, statisticians analyze data and design experiments, teach statistical tools, work with technical people, consult on other people’s projects and are seen to have narrow expertise and little accountability. In the new leadership role, they also determine appropriate tool sets for organizations, design training systems, work with managers, lead cross functional projects and possess broad expertise and significant accountability.

GE, for example, was forced into the new paradigm as a result of its across the board commitment to Six Sigma. Statistical tools are no longer the exclusive property of statisticians; everyone uses them. Statistical thinking has surpassed statistical methods in importance. It’s more important to know how to apply statistics to attacking unstructured problems than to know how to do a regression or design an experiment. GE still needs statisticians, but not just to analyze data because employees often do that for themselves.

The statisticians and nonstatisticians at GE know tools don’t make improvements—leaders do. This isn’t unique to GE. We see similar trends in other companies implementing Six Sigma.

To become leaders, statisticians must first understand the basic change that has taken place in the way we work and grasp how that change demands a clear understanding of the difference between managing work and leading people. They must understand what leaders do, and they must take up the task of learning how to do it themselves.

The Way We Work Now

The litany of change is so familiar that we’re likely to put our hands over our ears when we hear it repeated: globalization, more demanding customers and stockholders, tougher competition, accelerating technology, a wired world and far less room for error. We listen more closely, however, when we reflect on how these mega trends have impacted our daily activity. In the past, we went to work and did our job. Today, however, we have two jobs: doing our work and improving how we do our work.

These jobs are not synonymous. Work in organizations is done through processes, whether they are business, manufacturing, operational or administrative processes. When we are operating within those processes, we are doing job one. When we’re trying to improve those processes by reducing manufacturing errors, streamlining logistics or reengineering the capital budgeting process, we are doing job two. As anyone who has participated in an improvement initiative knows, balancing those two jobs can be challenging.

Managing vs. Leading

The dual nature of our work roles is reflected in the dual nature of the activities required to run the organization: managing and leading. Both activities are valuable and necessary, but they differ significantly.

Managers make sure the work gets done; leaders make sure the work gets done differently and better. Managers work within an existing paradigm; leaders help a group of people move from one paradigm to another. Managers focus on doing things right; leaders focus on doing the right things. Managers work hard to maintain improvements; leaders work hard to create improvements. Managers manage processes; leaders lead and develop people.

Organizations need excellent managers as well as excellent leaders. Improvements that aren’t sustained become worthless, and excellent processes that never improve become mediocre over time. In fact, given the dual nature of work today, most organizations need people who can both manage and lead. In many organizations the percentage of time spent on each activity, according to one’s role, loosely follows the pattern in Table 1.

Although there is a vast how-to literature on both management and leadership, a great mystique and a number of myths have sprung up around the latter. It is assumed leadership is a rare skill, inborn and restricted to the people at the top of an organization. Many assume leaders must be charismatic, eloquent speakers who enjoy controlling, directing, prodding and manipulating. Statisticians who wish to move from a consultative role to an expanded role must look past these myths and understand what leaders really do.

What Leaders Do

As the distinction between managing and leading suggests, leaders lead change. Change takes place on both the business front and the personnel front. It requires not only substantive changes to a process, practice or strategy, but also shifts in the knowledge, skills and attitudes among the people being led. Unfortunately, many people resist change because it takes them out of their comfort zone into new, uncharted territory. As ancient maps warned of unknown territory—there be monsters.

To succeed you must practice leadership in a way that simultaneously changes the business and helps people change themselves. That is what real leaders do, and they do it through following these four main practices:

  1. Create the direction. Don’t as-sume all matters of strategy and vision must be left to top management. Certainly, top leadership is primarily responsible for the overall direction of the enterprise, but as Joseph Juran made clear, “All improvement takes place project by project and in no other way.”1 Projects, no less than grand strategy, need objectives, vision and direction.

    At this level, creating the direction often begins with identifying the specific projects that need to be undertaken to achieve some larger strategy or goal. Once you have determined what project to undertake or have been given a project, you must determine what success will look like, how you will achieve it, what you will focus on and the specific initiatives you will undertake within the overall framework. It is at the helm of such individual projects that leaders are made.

  2. Communicate the direction. You must provide people with a clear understanding of where you intend to lead them, what benefits going in that particular direction will produce and the landmarks that will let them know they are progressing. That much is obvious, yet it is surprising how often grandiose plans fail to get translated into easily understandable goals with clear metrics and milestones.

    Communication should be clear, concise and continuous. It should also take place in a variety of media because not everyone absorbs and processes information the same way. Above all, it should provide not only understanding but hope, based on concrete progress, that the change can be achieved.

  3. Set people up for success. Many of us are painfully familiar with organizations that invest just enough in a project to fail. Such failures, especially the failure to allocate sufficient resources, suggest little confidence existed in the project in the first place. You must enable success and create its expectation by providing enough people, time and money from the outset.

    You also need to provide training to build the skills that are needed for success. Letting employees sink or swim on their own is the mark of an incompetent leader. A real leader re-moves barriers to success by coaching and counseling, not cajoling or criticizing, to help people through the inevitable rough patches.

  4. Catch people doing things right. To achieve progress and maintain momentum in a change, you must recognize accomplishments and results and reward them. Rewards can be psychological (praise, formal recognition, awards), financial (bonuses, raises, benefits), professional (promotions, expanded responsibility, special training) or some combination of all three.

    That doesn’t mean you shouldn’t candidly assess how people are doing through management reviews because people want and need honest feedback. But frequent and public recognition of desired behavior is likely to get you closer to your goal than steely silence. Psychologists call it positive reinforcement. Leaders call it leadership.

Learning To Lead

Despite the myth that leaders are born, not made, you can certainly learn to be a leader. However, don’t make the mistake of thinking of leadership as a narrow technical skill. Effective leaders not only know how to lead, they also have substantial business knowledge. They understand how business works in general, how it works for a specific industry and how it works for their company. They understand strategic planning (the process for arriving at change objectives) and strategic deployment (the process of implementing strategy).

Leaders are also comfortable with process and systems thinking. After all, improving the way work gets done inevitably entails improving processes. Statisticians, steeped in analytic rigor, are uniquely positioned in that regard. Leaders are also familiar with the proven, structured improvement methods such as Six Sigma or lean that provide the practical means for improving processes.

Leaders also need the so-called soft skills. They must know how to create stakeholders in a project, including those executive stakeholders and champions who are necessary for success. They must be adept at reviewing, coaching and communicating. They understand group dynamics, know how to lead teams, are adept at designing and facilitating meetings and excel at project planning and management.

That’s a lot to learn. But there are almost as many ways to learn as there are things to learn. Read books and articles (see bibliography), attend courses, discuss issues with colleagues, find a mentor, observe good and bad examples of leadership and, most importantly, be a leader every chance you can.

Most organizations are continuously contemplating change programs of some kind. Your company is likely no exception. Find out where the action is and do your best to get involved. For example, Six Sigma programs, now ever-present in organizations of all types, simultaneously teach people how to improve processes and how to lead change. Such programs, which often include change projects of various sizes, offer an ideal opportunity for you to learn to lead by leading at a level with which you are comfortable.

Find these opportunities and seize them. Start small but think big. You have much more to gain than you have to lose.


  1. Joseph M. Juran, Leadership for Quality: An Executive Handbook, Free Press, 1989, p. 35.


  1. Bennis, Warren G., and Burt Nanus, Leaders: The Strategies for Taking Charge, Harper-Row, 1985. Bennis has also written several other books on leadership.
  2. Bens, Ingrid, Facilitation at a Glance, GOAL/QPC, 1999.
  3. Brassard, Michael, The Memory Jogger Plus, Goal/QPC, 1989.
  4. Covey, Stephen R., The Seven Habits of Highly Effective People, Fireside, Simon and Schuster, 1989.
  5. Dalton, Gene W., and Paul H. Thompson, Novations: Strategies for Career Manage-ment, Novations Group, 1993.
  6. Doyle, Michael, and David Straus, Making Meetings Work, Jove Books, 1982.
  7. Project Management Memory Jogger, GOAL/QPC, 1997.
  8. Hayword, Steven F., Churchill on Leadership, Prima Publishing, 1997.
  9. Kotter, John P., Leading Change, Harvard Business School Press, 1997.
  10. Kotter, John P., “Leading Change: Why Transformation Efforts Fail,” Harvard Business Review, March-April 1995.
  11. Kouzes, J.M., and B.Z. Posner, The Leadership Challenge, Jossey-Bass, 1995.
  12. Phillips, D.T., Lincoln on Leadership, Warner Books, 1992.
  13. Scholtes, P.R., B.L. Joiner and B.J. Streibel, The Team Handbook, Oriel, 1996.
  14. Snee, R.D., “Non-Statistical Skills That Can Help Statisticians Become More Effective,” Total Quality Management Journal, Vol. 9., No. 8, pp. 711-722.

RONALD D. SNEE is principal of performance excellence at Tunnell Consulting in King of Prussia, PA. He has a doctorate in applied and mathematical statistics from Rutgers Univer-sity in New Jersey. Snee is a Shewhart Medalist and an ASQ Fellow.

ROGER W. HOERL is manager of General Electric’s Corporate Applied Statistics Lab. He has a doctorate in applied statistics from the University of Delaware. He is a Fellow of ASQ and a Brumbaugh Award recipient.

Average Rating


Out of 0 Ratings
Rate this article

Add Comments

View comments
Comments FAQ

Featured advertisers