Chapter Six

April 21, 1994, was one of those mornings when Southern California lives up to its hype: the sky was clear over Santa Monica, the air pleasantly cool; a breeze that barely ruffled the blue Pacific was enough to toss the hair of the blondes who glided by on roller blades. Warren Bennis arrived to meet me for breakfast, perpetually youthful, preternaturally suntanned. Fortune was doing a story about how e-mail networks were beginning to change the style and content of management, and I wanted to talk about it with him as he was a professor at the University of Southern California, an expert on both the style and substance of leadership. I particularly wanted his perspective as a student (and heir to the intellectual mantle) of Douglas McGregor, the author of The Human Side of Enterprise.

I had eggs. He had fruit and, I think, toast. We chatted a bit about this and that; then I broached my topic. He began his reply by saying he wasn't sure he was the best person to talk on the subject— he himself wasn't all that familiar with electronic networks. He didn't quite say it, but I guessed he meant he'd never sent an e-mail message in his life. No shame in that, then: April 1994 was the month in which Mark Andreessen filed the papers incorporating Netscape, the company that made the World Wide Web navigable and popular; that month, America Online—which has more than 21 million subscribers as this is written—had just 712,000 members. Bennis had celebrated his sixty-ninth birthday just six weeks before, and might reasonably have figured that by the time e-mail became omnipresent, he wouldn't be around.

Having warned me of his ignorance, Bennis went on to offer a set of observations and hypotheses more expert than any others I heard while reporting that story—all of them, moreover, amply and accurately borne out as electronic networks have become ubiquitous and data traffic (much of it e-mail) has grown till its volume is greater than voice traffic. Bennis himself, now seventy-five, has been known to contribute some to that torrent of bits.

Here's what he said: Networks, by definition, connect everyone to everyone. Hierarchical organizations, by definition, don't do that—they create formal channels of communication, and you're expected to follow them. He didn't use the ecological metaphors that have since become fashionable, but if he had he might have said that hierarchies are like concrete-lined irrigation ditches, where the water flows along clearly laid out, prescribed lines, whereas networks are flat, rich, mysterious Okeefenokees of every-which-way communication.

A hierarchy, Bennis went on—this was the key point—acts as a “prosthesis for trust.” Organization charts—showing who reports to whom, who owes fealty to whom—define more than reporting relationships. They are the trellises on which trust's fragile vine twines and blooms. Indeed, as he reminded me, bureaucratic rules and procedures came into being in part as safeguards against untrustworthy behavior like nepotism, favoritism, and corruption. Bennis said, “That organizational armature reinforces or replaces interpersonal trust.”

Reinforcement and replacement: Both are important. People in organizations act from many motives and are acted on by many pressures. Personal trustworthiness might need reinforcement by organizational strictures; if the rules are clear enough and the hierarchy strong enough, personal trustworthiness might not even be an issue. In The Organization Man, William H. Whyte showed how in large, hierarchical corporations, the Protestant ethic, with its notion of individual responsibility for one's actions, mutated into an organizational ethos.¹ In that regime, real interpersonal trust is unnecessary: its doppleganger will do. You can count on me because you're my boss; I can count on you for the same reason; together we can count on others because the boss of bosses has told us what he wants. Everyone has his place, and everyone else knows what that place is. Position substitutes for persuasion. In theological terms, it's a rule of law, not grace—and it works.

When a network becomes the main means by which information is conveyed and work gets done in a corporation, those hierarchical crutches are knocked away. Networked organizations have few promotions to give out, and rank is unclear. Colleagues might be thousands of miles away. Rewards may go to teams, not individuals. Those teams are likely to be interdepartmental—so that hierarchical power, position power, isn't around to guarantee that work gets done. More and more often these teams are temporary—like floating crap games, Bennis said that morning—which disband when the project is done; today's team leader might be tomorrow's underling. Networks encourage people to operate informally, outside the rule of law. Relationships therefore depend much more on cooperation than on control. Cooperation, in turn, depends on trust.

Flattened hierarchies and networked relationships change the sources and uses of power—a subject about which there's a fair amount of scholarship. Few businesspeople I meet are mystified by power. Yes, bosses can't throw their weight around as they once did, but power is still easy to recognize, and when the boss says “Jump!” the reply is still “How high?” more often than it's “Why should I?”

Networks also change the sources and uses of trust. That's a subject far less studied. A visit to Amazon.com and a simple search on the word “power” turns up 28,003 possible titles, versus just 1,819 for “trust”; most of the books in each case fall into categories like self-help that are not pertinent to management of organizations—but the fifteen-to-one ratio of study feels about right. Yet, Bennis argues, trust is more important. His newest book Douglas McGregor, Revisited (written with Gary Heil and Deborah C. Stephens) says: “Gathering information, and above all developing trust, have become the key source of sustainable competitive advantage.”²

Trust, unlike power, baffles people. How do we create a climate of trust in the company? How do I know if we can trust our suppliers? How can I make sure that people will do what they say they will do, when they don't report to me? How can we reconcile the need to protect confidential information with the desire to be open? People find these questions much harder than questions about power. When Owens-Corning moved into its airy, elegant new office building on the banks of the Maumee River in Toledo, Ohio, the CEO, confident of his authority, had no qualms about having a glass-walled office; indeed, he insisted on it. But the legal and accounting departments worried about whether their secrets would be safe behind glass, same as they did at Alcoa, when it moved to new space across the Allegheny from its old tower in downtown Pittsburgh. In these days of hacker attacks and hyper-competition, information systems managers are rightly zealous about security. Whom can we trust? Without pretending to answer that question—or indeed to do more than raise questions—I hope in these pages to suggest some of the sources and uses of trust in a postmodern corporation.

Real trust is hard even between people who have chosen to be together and have years to work on it, like spouses. It's harder still where they have little or no say in selecting their colleagues and where time is short. Impossible where an organization is large. The goal of real interpersonal trust might be misguided as well unattainable: Certainly there are limits to the trust between colleagues or between boss and subordinate, since everyone retains the option to end the relationship. Trust at work therefore needs support—forces that create incentives for trustworthy behavior and reassurance for people who rely on others. In the absence of hierarchy, what organizational strictures will keep someone in line?

Trust's first truss is competence: I can trust you if I believe you're good at what you do, and cannot trust you if I doubt your skill. We trust competence all the time, with automobile mechanics, physicians, computer technicians, chefs.

Life at work demands that kind of entrustment more and more often. Traditionally industrial tasks were handed from one department to another—from research to development to design to manufacturing to distribution, sales, and service. At each stage, department heads vouched for the competence of their staffs. In each functional department, the boss became boss by virtue of being the best—at least, that was the idea. In an earlier life I worked for a large publishing company. One day the chairman walked by while I was typing something, poked his head into my office, and asked, “How fast can you type?”

“About fifty words per minute,” I answered, looking up from the manual machine I ostentatiously preferred to use.

He said: “If you ever want to be the head of a Fortune 500 company, you have to be able to do everything better than everyone else. I type sixty.” Leave his arrogance aside (not to mention the fact that few Fortune 500 CEOs could type at all in those days)—what's interesting about his quip is the assumption that the boss has to be best at everything, a notion almost as quaint as typewriters and carbon sets.

“Smarter Than My Boss” says a button I keep in my office. I won't say whether that's true in fact (power being something recognizable), but it's true in theory. The boss today isn't the most talented specialist in a functional department but is instead Peter Drucker's conductor-CEO, a coordinator of specialists. The conductor knows the score best, but the trumpeter knows how to blow the horn. When musicians have trouble with a passage, conductors say, “Take it to your teacher”—who is not the boss. The twenty-eight-year-old CEO of a dot-com company told me, “The role of ‘manager’ or ‘boss’ never existed for me. My role is to look for new and emerging markets and help us get there.” The boss expects the specialists—the finance people, the techies—to work more or less unsupervised.

So the boss trusts in their competence—but what about judging it? In the functional organization, that wasn't a problem. The old-style chief engineer hired, trained, evaluated, and promoted other engineers, assigning people to jobs he could do himself, on the basis of his expert knowledge of their ability—he chose to delegate, rather than to entrust. The leader of a team consisting of a butcher, a baker, and a candlestick maker has less ability to evaluate and no choice but to entrust. “I leave it in your hands,” the generalist tells the butcher, because the details of meat-cutting are a mystery.

So trust needs a second crutch: community. With or without computers, networked organizations naturally spawn informal groups of like-minded souls. When these communities emerge around a common discipline or problem—a work-related subject like graphic design or the behavior of derivative financial instruments—they become “communities of practice.” The term, coined in 1987 by Etienne Wenger and Jean Lave of the Institute for Research on Learning, has come into such widespread use that a search on the Web turns up about 4,400 hits, four times as many hits as for the phrase “on becoming a leader.”³ These communities are where work and (particularly) learning occur. I elsewhere described them as “the shop floor of human capital, the place where the stuff gets made.”

Communities of practice support trust because they create and validate competence, a role performed by functions in hierarchical companies. The boss may not know which butcher is best, but the other butchers do. And when butchers get together, they kibitz, teach, and form groups to work on unsolved problems. General Electric demonstrates the role of informal communities in creating trust, or at least something that substitutes for it.

General Electric's core competence is leadership; leadership development, therefore, is the company's most important business process. GE executives say they spend between a quarter and half their time on these issues. With 340,000 employees, GE might easily be an impersonal, difficult-to-navigate company; with a highly competitive culture, it might be a place where ideas are hoarded and clever politicians squash good businesspeople; with businesses ranging from medical equipment manufacturing to freight-car leasing, employees might be expected to have few skills and know few people outside their specialties. Instead, largely because the place is threaded through and through with communities of practice, the company is remarkably informal, so successful at creating topflight executives that it consistently produces more than it needs and exports its “trade surplus” in talent to companies around the world, and so networked that everybody at GE, it seems, knows people in every other GE business everywhere. In 1999 I visited GE businesses in eight European cities accompanied by an American-born, London-based media relations manager.⁴ Unsurprisingly—it was her job—she knew the senior leaders of the businesses. It surprised us both, however, that in every city every time we went to a meeting with a dozen or so employees, she ran to embrace someone she'd met and worked with at one of GE's many training and networking events. Ask any GE person about the value of attending courses at Crotonville, its fabled leadership development institute. The answer is always, as it is with any great school, “The people I met were more important than the courses I took.”

These communities play a key role in leadership development at GE. The company has elaborate formal leadership programs. There's training—not only at Crotonville but in many other locations—and every candidate for a leadership position in the company undergoes extensive training in Six Sigma quality methodologies as well as in traditional subjects. To evaluate talent, GE uses a forced ranking of employees into “A, B, and C players,” a second ranking in which every manager rates all direct reports on a strict bell curve regardless of letter grade, and—most important—an annual staffing review, called Session C, for which all GE professional employees submit self-assessments and career-development plans and during which they are evaluated by squads of senior managers, with the top people and those singled out for high potential also reviewed at headquarters in Fairfield, Connecticut. CEO Jack Welch himself takes part in Session C evaluations of several hundred people.

The formal processes are so rigorous they would amount to hazing were it not for the role of GE's many communities of practice. The place is full of them—manufacturing councils, finance councils, technology councils—literally hundreds of interdisciplinary and interbusiness affinity groups. Through them (and through action-learning projects in training), GE's young leaders form the networks of friendships they will use during the rest of their careers with the company. They're expected to bring ideas to share at these meetings, where their friends and equals debate them, improve them, and take them home to implement in their own businesses. It is here that they get noticed, and it is from these communities that managers learn who's really good, who's really up-and-coming. Leadership, like any art, is easier to recognize than it is to define. Good “grades” in reviews and accomplishments in training, however searching the tests or superb the school, cannot create, show, or anneal leadership talent—and cannot produce trust—the way communities of practice can. Without these communities, GE's leadership development processes would be more competitive and political, and less cooperative and effective. It's essential to see the conclusions of the development processes verified by the communities in which a candidate participates.

Commitment, a third source of and support for trustworthiness, is an adjunct to both competence and community, neither of which necessarily implies loyalty to the organization. Indeed, communities of practice create a rival allegiance, where the interests of a community (for example, advanced research in cardiology) might conflict with the goals of an employer (such as a managed care company).

Trust obviously depends on the degree to which people are willing to support the organization's purposes. This is not a question of motivation. As Douglas McGregor argues, people are intrinsically motivated—but to do what? The convicted spies Kim Philby and Aldritch Ames were highly motivated men, but their goals were diametrically opposed to those of the government intelligence organizations they claimed to serve. As we ask people to be more entrepreneurial, as we flatten hierarchies so that bosses supervise fifteen or twenty people instead of six or eight, as we empower people, it's vital that there be a shared commitment to the same mission and values. Moreover, in a knowledge economy, the nature of work has changed. Repetitive, unthinking jobs—adding spreadsheets, filing sales reports, running routine tests—have been automated; in general, we are asking all workers—and especially managers and knowledge workers—to think and to make decisions. We need their inner gyroscopes aligned with the corporate compass.

That can't happen unless people know what they are committing to. Statements of vision and mission are notoriously vacuous; they breed cynicism, not trust. George Bailey, a partner in the consulting business of PriceWaterhouseCoopers, once printed up half a dozen companies' vision statements, then challenged their CEOs to identify which one was theirs. Half failed. And who wouldn't? Most companies could get better mission statements if they used Mad-Libs, or tried the Dilbert Web site, which will generate both mission statements and performance reviews by randomly combining buzzwords and bromides. It just gave me, “Our mission is to completely negotiate enterprise-wide materials while continuing to proactively leverage existing error-free solutions to exceed customer expectations.” I couldn't agree more—or care less.

What's needed is a clear understanding of what makes the difference between success and failure, and how that translates into behavior and decisions. No more hiding the business model behind high-sounding nonsense. One company makes money by being the low-cost producer: We drive hard bargains, are a no-nonsense kind of place, are fussy about expense reports and impatient with slow learners, and if that makes you uncomfortable, don't work here. Another makes money by being the leader in innovation: We'd rather see half-baked ideas than fully cooked ones, and if you haven't failed around here, you haven't tried. A third makes money by coddling customers—making sure, however, to deal only with customers who will pay for pampering. It's crucial to link the mission to the business model; crucial, too, that personal success—career advancement—comes to people who commit to the behavior you ask of them. The company that asks for innovation and rewards punctiliousness should not be surprised if its creative people seem alienated.

Some companies have strong-flavored cultures. GE is an example, a culture strong enough that I can frequently recognize GE people without being told that's where they work. The Pentagon's culture is so clear that Hollywood can dress someone in mufti yet let you know he's a military officer with just a few words and facial gestures. Hewlett-Packard is another strong culture. I can't feel it, but Debra Engel, an HP alumna who went on to 3Com and now is a venture capitalist, can. People who worked at HP can almost sniff the presence of others who did, she told me once, by way of explaining how it was she had ended up in a con versation with the only other HP alumnus in a room of three hundred people. GE, the Pentagon, and HP are high-commitment organizations. You know you're there, and you know what you're supposed to do.

Beyond competence, community, and commitment, trust of course depends on communication, which can be its best friend or its worst enemy. That morning in Santa Monica, Bennis said that communication “will take a hell of a lot more time than it used to. And it will take a lot of emotional labor on the part of the leader.” He understated the case.

Hierarchies can lie, and get away with it pretty well. Naked emperors go unchallenged. Incoming CEOs rewrite history with an avidity Orwell would recognize—and for reasons he would understand. Their newest trick is to take a big restructuring charge as quickly as possible after taking office, thereby reducing current earnings so that, a year later, they can boast the improved results while polishing their résumés in preparation for their next gig.

A revolutionary way to build trust: tell the truth. A few years ago the corporate communications head at AlliedSignal asked me: “What news travels faster than any other news through a factory?” He answered himself: News that a competitor won an order the company was bidding for. “And what news,” he went on, “is never, never, never even mentioned in any plant newspaper?” It was a struggle, he said, to get the editors of those newspapers to understand that credibility mattered more than cheerleading.

“If you can't say something nice, don't say anything at all” might be good etiquette, but it's bad management, certainly in the age of networks. Rick Levine, Christopher Locke, Doc Searls, and David Weinberger, the authors of The Cluetrain Manifesto, a rabble-rousing, best-selling credo of the post-hierarchical age, exaggerate only slightly when they say: “There are no secrets. The networked market knows more than companies do about their own products. And whether the news is good or bad, they tell everyone…. As with networked markets, people are also talking to each other directly inside the company—and not just about rules and regulations, boardroom directives, bottom lines…. We are immune to advertising. Just forget it.”⁵ Open-book management, an outgrowth of McGregor's (and Bennis's) thinking, turns out to be inevitable as well as desirable. “Is democracy inevitable?” Bennis asked in 1964. He was talking about geopolitics, but his answer—it is—was also correct when it comes to the management of organizations.

One of trust's important, little-noticed allies—and the last I'll mention—is cupidity. (I'd have preferred another word—reward, perhaps; but after competence, community, and the rest, I needed one that begins with “C.”) The point here is simple and obvious: If trust is a source of competitive advantage, it should pay. Failure always breeds mistrust—backbiting, toxic politics. “I get the willies when I see closed doors,” says Nametk, the protagonist of Joseph Heller's novel Something Happened, and we all know why—some thing's not right, and “they” don't trust us to know what it is. Trust needs to be seen to be good business. Bosses should display it in stormy times as well as in balmy, palmy ones. Instead, when the going gets tough, managers dust off their old command-and-control hats, destroying the comity that's their best chance of getting out of the mess. That's got to change.

Business begins with trust. It begins with a deal: If you pay me X, I will give you Y. As companies abandon bureaucratic mechanisms, their leaders need to understand that trust is as important to management as it is to relationships with customers. Trust is hard, and it should be “hard stuff,” not “soft stuff”—that is, it should be a virtue that can be documented and even measured, and not just in employee-attitude surveys. For that to happen, leaders will have to understand that trust isn't magic. It doesn't occur by itself. It can be created—and of course destroyed. Managers need to use the tools of trust as deftly as they do the tools of power.