Is There a Best Way To Lead Scientists and Engineers?

By Michael Maccoby

Research Technology Management; Volume 49. No. 1. January-February, 2006. pp 60-61.

You lead an R&D organization, either a department, team or project. How do you go about getting the best possible performance from your people? How are you organizing their work? Do you think there are processes you could design to improve organizational effectiveness. Should you give your people training, new information tools? Are you satisfied with the way the people you manage share their knowledge? Do you worry about how to motivate people to achieve a higher performance?

Thomas H. Davenport addresses these questions in his useful book, Thinking for a Living, How to Get Better Performance and Results from Knowledge Workers (Boston: Harvard Business School Press, 2005). Davenport shows why research/technology managers, like all managers of highly qualified knowledge workers, are right to concern themselves with these questions. No one seems to have found the perfect way to manage knowledge workers, particularly those expected, to create new knowledge.

I would put it that there is no analogue for knowledge work to the Toyota model is to car manufacturing (1). Davenport writes “We’re in the early days of thinking and knowing about how to improve knowledge work... Every attempt to make it better is—or should be—an experiment. The problem is that we don’t treat them as experiments. We adopt new ways of officing, organizing and operating based not on rational experimentation and learning, but rather on faddish inclinations and gut feel.”

To be sure, less complex knowledge work can be dramatically improved with new tools. Less complex transactional work in call centers, retail, and banks can be formatted and in many cases automated. Increasingly, we go to the Internet for advice on buying cars, getting loans, even for legal and medical advice.

But as Davenport reports, attempts to design artificial intelligence (AI) and expert systems to make important decisions or judgments have invariably failed. The reasons why:

“The knowledge intended for the system was difficult to extract from the expert’s brain.
“The knowledge in a system generally needed to change more rapidly than the system designers anticipated, and constantly revising for such change was difficult and expensive.
“The best systems proved to be those that augmented human experts rather than replacing them—which lowered the potential economic returns from the expert systems.”

A good example of “the best systems” is the order-entry technology increasingly used by physicians in leading hospitals to dramatically improve medical performance and lower the shockingly high rate of deadly errors caused by prescribing the wrong drugs or failing to follow best practice.

The more complex the knowledge work, the more productivity depends on the qualities of individual workers. Not surprisingly, Davensport has found that the most effective knowledge workers are the best learners. They network both inside and outside the organization, not for the purpose of seeking better job opportunities but to increase their knowledge. Superior performers need time to read, talk to people, reflect and think. (I know I do.)

Distinguishing Dreamers from Doers

But how do you tell the difference between someone who’s dreaming vs. thinking up a great solution? Do you, as a manager, consider that a researcher is working while reading a business book? You would probably be wise to encourage this particularly from excellent performers. With the stars, it’s a no-brainer, they produce. But what about an employee who has not reached his or her potential? How do you help them increase their competence.

In the past, management experts maintained that employees should be responsible for their own development. But Davenport describes ways some companies have provided useful education. He cites a Hay Group survey of over three hundred companies where knowledge workers said that the ability to learn new skills was far more important in their willingness to stay with a company than money or any other factor.

Whatever you do, knowledge workers perform better when they understand the business. An R&D organization is part of a socio-technical system that has a purpose and a business strategy. The more the system’s parts, including people, are aligned with this purpose, the more effective the system will be.

Davenport recognizes that knowledge workers, especially the more qualified, are hard to manage. They’re anti-bureaucratic. They resent top-down organizational changes. They want to work in a communal culture.(2) The best way to get them to buy-in to new policies and processes is to involve them in designing their own work. But they may also need education for organizational effectiveness: how to run a meeting to reach good decisions, how to use personal technology, how to organize information efficiently, how to teach others what they know.

Exceptional Leadership

To a large extent, I’ve found that knowledge workers can manage themselves. But the most creative groups of knowledge workers have had exceptional leadership.(3) Although management is a collection of processes and functions that can be carried out either by individuals or teams, leadership always implies a relationship. People may or may not comply with management directives; they will enthusiastically follow an effective leader. Ideally, all managers will be leaders. When they are not, knowledge workers make their lives hell.

Davenport only touches the subject of leadership, but it is clear from all he writes that traditional assumptions about management from the industrial world don’t fit knowledge work. For example, industrial bureaucracies workers usually know more about their jobs. That is not the case for most managers reading this article.

What then is needed from leaders of engineers and scientists? My answer is that research/technology organizations need different types of leaders: strategic, operational, project bridge-builders. People will follow them because they communicate meaningful goals and build trust.

Davenport agrees, “Knowledge workers don’t want to work toward a goal because someone else has set it, but rather because they believe that it’s right.”

Understand Your People

Leaders build trust by practicing what they preach, being ethically consistent and transparent about the reasoning underlying decisions. But leaders who understand individual personality and take account of differences are trusted more than those who act as though everyone is the same.

Understanding should not make leaders go soft, however. A common weakness of the most people– oriented leaders is failing to fire people close to them who are not adding value and sometimes even undermining their authority. Part of the reason for this is a lack of what I call Personality IQ, a combination of knowledge about personality types and a heart that listens to another person’s emotional attitudes.(4) Another part is oversentsitivity to being thought a tyrant.

In short, althought there is no one best way to manage knowledge workers, there are clearly ways you should avoid. And you should not confuse the need for participation in a collaborative community with a neglect for standards of behavior, attitude and achievement.

(1)Maccoby, Michael, “Is there a Best Way to Build a Car.” Harvard Business Review 75, no. 6 (November-December 1997): pp. 161-171.
(2)For examples of collaborative communites of knowledge workers see Charles Heckscher and Paul Adler, editors, The Firm as Collaborative Community: Reconstructing Trust in the Knowledge Economy. Oxford, 2006.
(3)See, for example, Warren Bennis and Pat Biederman, Organizing Genius. Addison-Wesley, 1997.
(4)Maccoby, Michael. “Understanding the People You Manage,” Research • Technology Management, May-June 2005. pp. 58-60.