We can deal with the framing bias, for example, by recasting our language and by using a technique called causal mapping. A causal map (which can take a visual or verbal form—the term map is a metaphor) requires one to identify the true antecedents, actions, and consequences associated with a situation.^[149] Reframing calls for a revised description of those facts. Here, for example, is one way of describing a research project:

Here's another way to look at it:

The first framing leads to a no-go decision. The second calls for a different conclusion, particularly if the project manager has dreams of following the relationship manager's path to partnership.

If the firm should decide to proceed, the project manager and her team can then go through the causal mapping exercise. Doing so would help them determine how best to minimize the anticipated problems and increase the probability of a mutually satisfactory outcome. After reflection, for example, the team might conclude that the client organization's evolving, unstable, post-IPO culture poses the risk that internal politics could disrupt the project. The response might be to staff the project with consultants who have worked with this client on similar projects before. They know the methodologies required, understand the organization's cultural landscape, and have credibility with the client-side project manager. This combination should help the consulting team anticipate and deal with client politics that threaten to interfere with the project. Similarly, mapping the project would help the consulting team identify where in the project timeline lie the greatest risks that an undisciplined client could fundamentally redefine the project's requirements. Armed with this information—a map of virtual landmines, so to speak—the team might decide to price the project flexibly (time and expense) rather than rigidly (fixed fee) to protect their firm from unexpected scope changes. The team could also plot the key decision-making contingencies on the project time line. By doing this, they could highlight the financial and project quality implications of hitting, or missing, major project milestones.

Reframing and causal mapping take managers and their teams through a clarifying exercise that gets beneath the surface of the framing bias. Approaches like these help managers and their teams to identify the true risks associated with planned work. But even a carefully crafted causal map won't be worth the pixels it's printed with if it fails to produce a realistic estimate of the time and resources required to get work done. Managers also need a way to force reality into the project outcome estimates. One approach, called reference class forecasting, can help managers overcome the unjustified optimism that underlies the planning bias.

Reference class forecasting uses the actual outcomes of a group of similar actions to provide a context for a project under consideration. In effect, it provides an outside view that helps prevent some of the misestimation that often characterizes purely inside views. Reference class forecasting does this by taking the planner through three steps:

The first use of reference class forecasting to challenge the planning bias occurred in 2004 as part of the budgeting process for a public transportation project in Edinburgh, Scotland. The promoter of the project, Transportation Initiatives Edinburgh, had submitted a business case estimating the project would come in at £255 million. With a contingency allowance of £64 million, the total project estimate came to about £320 million in capital cost. After considering a reference class of similar projects, a firm of consulting engineers recommended that the transportation authority raise its estimate to £400 million—a 25 percent increase—if it wanted to have an 80 percent probability of staying within the budget.^[151]

Admittedly, most managers aren't dealing with financial numbers this big. Still, the need to budget accurately and then meet the budget remains central to managerial success. Avoiding an £80 million overrun just might save some politician his job. Techniques like reference class forecasting can enable a manager to suspend her natural (and often cognitively flawed) judgments and look for outside evidence that provides a more accurate context for planning.

In applying reference class forecasting to the research project, the team would begin by identifying the consulting costs to complete a group of similar past projects. They might discover, for example, a tendency to overrun consulting budgets by an average of 12 percent on those projects. Therefore, they might decide to raise their cost estimate by 15 percent, factoring in the vicissitudes of this particular organization's unpredictable politics. They could incorporate even more aggressive adjustments if they wanted to make doubly sure not to underestimate costs or overestimate profits.

Techniques like causal mapping and reference class forecasting can help managers overcome framing and planning biases. But what should they do to deal with the illusion of control? Having a carefully crafted project map, complete with all the up-and-down topography of a challenging project, might actually heighten a manager's feeling of influence over circumstances and outcomes. But that confidence might be an illusion, just like thinking that throwing the dice harder will produce more elevens. Managers also need a way to reflect back on real events and to imagine what might have changed the outcomes. One way to do this is to engage in a mind experiment called counterfactual thinking.

Counterfactual thinking simply means comparing what actually happened in a situation with what might have been had past circumstances unfolded differently.^[152] It involves reviewing a chain of events, mentally altering something in the action sequence, and imaging how the outcome would have differed. Often, reflecting back and creating "if-only" scenarios can compel a realistic reassessment that challenges an individual's illusion of control. In reviewing the outcomes of similar efforts in the reference class, the research project team might reaffirm a tendency for similar projects to produce cost overruns. A causal map might reveal that falling behind in producing the final analysis and report, and having to scramble to hit the deadline, consistently caused those overruns. Our project manager might start by looking for things she could have done differently. She might order a round of lattes, convene her team, and run through a set of if-only scenarios like these:

The manager—and probably the team as well—would likely conclude that the last scenario provides the best solution to the problem. In the process, the project manager would realize that she never had complete control over the outcomes. And so the illusion of control bites the dust. In the future, she would have to work more closely and effectively with others who have influence over project results (specifically, the client's project manager).

Causal mapping, reference class forecasting, and counterfactual thinking complement each other. Reframing and causal mapping require us to challenge the way we look at facts and get beyond the labels that bias our judgment. Reference class assessments force us to look at relevant past outcomes and push us to revise our typically optimistic initial projections. Counterfactual thinking supports our causal mapping by providing scenarios that lead to outcomes different from those actually experienced. Each technique, in its way, reinforces the balanced processing element of manager authenticity—the requirement that a manager objectively analyze all available information, especially facts and insights that challenge assumptions. These approaches require a manager to have sufficient self-awareness (another authenticity element) to recognize and address his own cognitive weaknesses. They also call for no small amount of humility; a successful offstage manager recognizes that overcoming cognitive biases usually requires help from others.

By themselves, techniques like these won't magically turn managers into thoroughly rational thinkers. It will take a few million more years of evolution to accomplish that. Moreover, overcoming judgment biases is a necessary but not sufficient requirement for successful work planning. A manager's approach to planning must be part of a broader philosophy for how work should get done. That approach should acknowledge that the best (that is, the least cognitively biased) planning, like other aspects of the manager's job, happens when managers have a light hand and when they involve others. In the time allocation structure laid out in Chapter Four, planning activities falls into the work process oversight category. Of course, work planning efforts must dovetail with the manager's other responsibilities, especially those associated with people focus. Ensuring that these role elements mesh effectively provides a platform for unit and individual success.

A customized role never leaps fully formed from the pages of a job description. In fact, just the opposite. Job descriptions are often works of fiction—outdated, oversimplified, sometimes wildly inaccurate snapshots of work realities. Human Resources should provide managers with the formal job specifications for the positions in the unit, but only as a point of departure. Manager and employee must look well beyond the words on the page and consider a variety of dimensions that meet at the intersection of required tasks and individual interests and abilities. At the most fundamental level, the basic elements of a job must align with the interests of the job holder. The several typologies of job components tend to agree that people's work-related interests will fall into the general categories shown in Table 5.1.

By listening to employees talk about what they like and don't like, observing how they go about their work, and perhaps running some experiments (for instance, letting people try small changes in their jobs), an astute manager can help people achieve a fit between their jobs and their work interests. With attention, most managers can figure out which employees display quantitative talent, which are conceptual thinkers, and which are most creative. The same goes for the social elements of work. Whether a person seeks or avoids social contact, displays an agreeable nature or likes to work alone, or has a calm demeanor or is prone to anxiety are all empirical questions. Careful observation and manager-employee cooperation can provide answers and help managers overcome the actor-observer and transparency biases.

In addition to these personal building blocks, managers must consider important dimensions of jobs themselves. One dimension is job demand, the magnitude of sustained mental or physical effort required to execute the basic job components. Job demands are high when the job holder faces frequent challenges and problems: pressure to perform, long task lists, complex role requirements, and emotional stress. Job demand is complemented by job resources: support from teammates, problem-solving assistance from a supervisor, performance feedback (especially recognition for success), and autonomy in how they go about their work. These complement the tangible resources noted in the prior chapter's discussion of the interplay between manager and employee roles. Access to job resources helps people manage large workloads and juggle disparate assignments. They do a better job of handling emotional interactions with customers and reconciling work and home responsibilities. Said another way, job resources help to buffer the effects of work demands, reduce obstacles to performance, and bolster employee engagement in the face of stressful work requirements.^[156]

Fostering autonomy requires managers to recognize and support individuals' discretion and control in deciding how to do their work. Autonomy differs in subtle but important ways from empowerment. Empowerment has become so common in the business lexicon that it hardly needs a definition, and so cries out for one. When people speak of empowerment, they generally refer to the transfer of power from a manager to an employee. The root words are em (from the Latin preposition meaning into) and power (from the Latin potere, to be able, from which we get words like potent). Empowerment refers to the transfer of a quantity of power from manager to employee. It's a zero-sum power game, where the manager's power diminishes by the amount given to the employee. Jonathan Gosling and Henry Mintzberg recommend that managers go beyond empowerment, "a word implying that the people who know the work best must somehow receive the blessing of their managers to do it."^[157]

Autonomy refers to something subtly but importantly different. Essentially, it means self-rule. Its roots are the Greek words auto (self) and nomy (law). Rather than subdividing a given quantity of power, autonomy adds to the total amount of power available to do work. Self-determination is a good synonym. Think of it this way—a royal authority can empower, whereas autonomy requires self-government.

Although the results of empowering and fostering autonomy may seem similar, the dialogue that precedes them is different. The empowerment exchange between manager and employee sounds like this:

The autonomy dialogue unfolds differently:

Creating the conditions for autonomy to thrive requires special attention from managers. People can be empowered in small increments, but autonomy is difficult to partition. In general, people either have autonomy in how they do their work, or they don't. Effective autonomy requires the characteristics of the job to complement those of the individual. The work and work environment specifications most critical to autonomy are:

In effect, autonomy serves two functions: it is a feature of a rewarding job and a reward unto itself. Although autonomy clearly has value to employees, it isn't free. It requires a cognitive investment and brings responsibility for outcomes. Conversely, it can moderate the relationship between the accountability burden that comes with self-determination and the emotional stress brought on by the weight of obligations.^[159] Effectively autonomous employees also confer a benefit on the organization. They enable managers to spend less time performing and closely overseeing work and more time thinking about how to improve work processes, build networks, and craft and deliver appealing individual deals for employees. In effect, employee autonomy moves managers away from the Widget Wizard role structure and toward the People Powermeister form.

The effects of job demands and job resources will naturally vary from person to person. This is where an additional element, individual disposition, enters the equation. In this context, disposition refers to the tendency for people to be either proactive or passive. Proactive people believe they can control circumstances and create change in the work environment. They adopt what psychologists call a "promotion focus," meaning they exhibit persistence, motivation, flexibility in the face of challenges, and, generally, satisfaction with their work. They tend to be extroverts, seeking social contacts and looking for excitement. Managers will recognize them as the optimists who concentrate on winning and want to receive rewards for their success. Passive employees, in contrast, tend to view opportunities for change as beyond their personal scope and ability. They respond to circumstances as they occur and hope for the best. They display vigilance, risk-aversion, and loss-avoidance.^[160]

So the full equation contains three factors: job demands, job resources, and personal disposition. The interaction among these three elements yields outcomes in five categories:

The chart in Exhibit 5.1 shows how these factors interact for people who have a strong inclination to influence their work environments (that is, people who exhibit high proactivity, flexibility, and motivation to succeed). Note the cluster of dark bullets in the Fulfilling Job column. For these employees, confidence in their ability to execute, ownership of outcomes, and engagement all peak when their managers take the time to ensure that competencies, roles, and resources come together to give them high control over a challenging job. Alternatively, given low control or less-demanding work, a naturally proactive person will generally reach lower levels of mastery, achieve lower levels of confidence in personal effectiveness, assume less ownership of results, and feel less engaged. Efforts to achieve self-actualization will be thwarted.

Figure 5.1. Proactive Employees Thrive with High Control and Challenging Work

Job content also plays a role in determining how well employees understand and respond to the link between their work and organizationally valued outcomes. For example, ensuring that people have periodic contact with the beneficiaries of their efforts can have a powerful effect on effort and performance. In one study, researchers introduced a group of university fundraisers (the folks who call during dinner to ask you to contribute to your alma mater) to one of the scholarship recipients who had benefited from their money-raising success. He told them how much he valued their effort and how helpful the scholarship money had been to him. A month later, the research team measured the performance of the fundraisers over a week's time. They looked for differences among three groups: a group that had met with the student, one that read a letter from the student, and one that had no contact at all. They found that the fundraisers who had spent time with the student more than doubled their per-caller minutes over what they had spent before meeting him. They also secured almost three times as much money as they previously had. The researchers concluded, "Merely interacting respectfully with a beneficiary of their work enabled them to maintain their motivation, as observed in their persistence behavior and objective job performance."^[161]

Exhibit 5.2 shows the same array of job factors for more passive employees, those who focus their attention on preventing downside events. They function best with low job demands and high control (the Manageable Job depicted in the left-hand column). Too much challenge causes discomfort and tension. Even given their relatively passive personalities, however, these employees learn more, take greater ownership of results, and exhibit higher engagement when they have direct influence over the circumstances of work.

Figure 5.2. Passive Employees Seek Lower Demands—But Still Value Control

The common factor for the most rewarding work, regardless of individual personality, is control over results. Jobs structured to incorporate this element, and the other elements depicted in the exhibits, can provide the skill development, decision input, career advancement, and work opportunities that people crave. Research in the workplace (as with one analysis conducted in the call centers of a British financial institution) confirms that this combination of factors also produces high performance.^[162] Call center jobs, the research confirmed, can be stressful and unfulfilling. Astutely managed companies, however, can bolster the resource side of the equation and improve both individual performance and customer satisfaction.

For example, consider the case of Zappos, the quirky online retailer. Zappos has probably moved as far from the call-center-as-electronic-sweatshop as it's possible to go. In fact, Zappos is such an appealing place to work that it debuted at number 23 on the 2008 Fortune list of the best companies to work for, the highest newcomer ranking in the history of the list.

The Zappos approach shows how job design can support a customer focus strategy. The call center rep's job structure lies at the center of the company's service model. The organization begins by taking potential hires through a careful screening process that eliminates the shy, the egotistical, and the unfunny. They want people from the proactive end of the personality spectrum only. "We do our best to hire positive people and put them in an environment where the positive thinking is reinforced," says Tony Hsieh, Zappos chief executive.^[163] That environment is script-free and freedom-rich. The principal job of front-line reps is not just to close a sale, but also to please the people who call in to order.

They are trained to encourage callers to order more than one size or color, because Zappos pays for shipping both ways. They refer callers to competitors when Zappos is out of stock. But most important, their jobs are constructed to let them use their creativity and their imaginations to delight customers. In one case overheard by a magazine reporter, a customer complained that her boots from Zappos had begun leaking after almost a year of use. The rep not only sent out a new pair, in spite of a policy that used shoes can't be returned, but also mailed the customer a handwritten thank-you note.^[164] Rumor has it that a Zappos call center rep once spent six hours on the telephone with a single customer. This kind of initiative, combined with a job structure that affords the freedom to exercise it, captures what Zappos means by WOW: "To WOW, you must differentiate yourself, which means doing something a little unconventional and innovative.... We are not an average company, our service is not average, and we don't want our people to be average. We expect every employee to deliver WOW."^[165]

Zappos managers are also encouraged to spend 10 to 20 percent of their time with team members outside the office. This intensive interaction enables managers to get to know their team members far better than they can in the context of work alone. CEO Hsieh says that when he confers with managers after they've taken their team members on an outing, they talk about improved communication, higher trust, and budding friendships. "Then we ask, 'How much more efficient do you think your team is now?'" Hsieh says. The answer? "Anywhere from 20 percent to 100 percent."^[166]

As the Zappos example suggests, designing engaging jobs doesn't have to be limited to high-end knowledge work. Granted, research into engineering positions suggests that knowledge workers sometimes have more latitude than others to alter the boundaries of their roles by taking on unassigned tasks, teaching new skills to others, adding coworkers to their networks, and participating voluntarily on teams. But a study of hospital cleaning staff showed that, despite having a relatively prescribed set of tasks, even custodial workers can craft their jobs individually. One of the groups analyzed stayed with the standard job description, which involved little interaction with patients and other staff. They reported that they disliked cleaning in general, judged the required skill level to be low, and had little willingness to engage with others or alter job tasks. They saw their job as cleaning—nothing more. Members of a second group went out of their way to interact with patients, visitors, and others in their unit. They made an effort to add tasks and to time their work to coordinate with others. Cleaners who crafted their jobs saw themselves as playing an important role in the experience of patients.^[167]

Or take the case of nurses, who can craft their jobs by creating expanded opportunities for interaction with other care providers and with patients' families. To quote the researchers from one ethnographic study, "The nurses acted as job crafters by actively managing the task boundary of the job to deliver the best possible patient care. By paying attention to the patient's world and conveying seemingly unimportant information to others on the care team, nurses re-created their job to be about patient advocacy, rather than the sole delivery of high-quality technical care."^[168]

Employee and manager both have roles to play in customizing jobs. In the words of one pair of psychologists, "People are not passive receivers of information from their work environment, but rather active in interpreting their jobs, and consequently in shaping their jobs."^[169] Another researcher says that "Not all employees are motivated to fulfill needs for control, positive image, and connection at work. Individuals who look to fulfill these needs at work likely will look for opportunities to craft their jobs in ways that allow them to meet their needs. Others may find that these needs are met elsewhere in their lives."^[170] Only a direct manager—by empathizing with an employee's individual work needs, mediating within a work group to craft a set of personalized job demands, and ensuring that job roles support the successful execution of unit and enterprise strategy—can help an individual make the multidimensional matches reflected in Exhibits 5.1 and 5.2.

In real life, few people sit squarely at one end or the other of the proactive–passive spectrum. As we saw in Chapter Two, most people experience some level of conflict as they choose between dominating and deferring. We want to lean far over to grab the brass ring, all the while holding tightly to the carousel horse's neck. Likewise, jobs rarely fall cleanly into either the "easy" or "impossible" categories. Managers must deal with ranges of factors for both people and their jobs. How can a manager get an accurate understanding of an individual's disposition and overcome the actor-observer and transparency biases? Research suggests that good judges of people do three things well. First, they create situations in which people manifest relevant clues to their personalities. Then, the skilled judges interpret those clues accurately and put them to use.^[172] Table 5.2 summarizes how a manager might perform this sequence.

Good judges of personality have predictable traits. They tend to display:

In short, they act authentically, observe astutely, and interpret accurately. Mostly, they are good listeners, and much more.

Sometimes, a team structure best meets the job-related needs of individuals and the strategic requirements of organizations. Structuring work in teams has become a ubiquitous artifact of organizational life. Effective team collaboration, says organizational and evolutionary psychologist Nigel Nicholson, is "widely attributed to be the single most important predictor of successful enterprise, and group breakdowns one of the most common causes of failure."^[173]

Rarely do organizations dedicate themselves as thoroughly to a collaborative way of working as does high-tech icon Cisco Systems. Cisco, which makes the routers and switches that link networks and power the Internet, pursues a strategy relying on product differentiation bolstered by operational discipline. Connecting people and communities isn't just the competitive goal—it's also how the company does its work. Cisco pursues this strategy by arraying its units functionally. In the words of Randy Pond, Cisco's executive vice president of Operations, Processes, and Systems, "Cisco is the largest functionally organized company that I know of."^[174] Functional organizations, which separate themselves into discrete areas like R&D, engineering, manufacturing, sales, and product development, sometimes have difficulty getting new products developed and delivered to the market. The organizational stovepipes are often so self-contained and impermeable that cooperation in the interest of a broader initiative—such as combining research expertise with product development insights—becomes next to impossible. That would be fatal for a company like Cisco, whose business relies on collaboration that drives innovation and new product introduction. So, rather than change to a fundamentally different structure (one based on lines of business, for example), the company has instituted an intricate system of committees made up of executives and managers from various functions. At the top of this structure are twelve councils, with an average of fourteen people each. Councils take charge of markets that could reach $10 billon in revenue. The small business segment, comprising companies with fewer than a hundred employees, is an example. Below the councils are some forty-seven boards, also with about a dozen members each. A board concentrates on market opportunities in the $1 billion neighborhood. Finally, working groups, small temporary teams that concentrate on individual projects, support the board and council structure.^[175]

If this arrangement sounds like a matrix—point taken. Matrix structures are hardly new, and they don't come without complications. A matrix arrangement brings reporting ambiguities, as people try to figure out whether they answer mainly to their nominal manager or to a project manager for the team they've joined. It can also cause fights between managers as they duel for resources and jockey to position their projects at the head of the organizational queue. According to Nigel Nicholson, "The injection of matrix forms into traditional hierarchies typically fails because the pull of functional identity overrides the requirements of interdisciplinary collaboration."^[176]

Interdisciplinary collaboration is precisely what Cisco needs to be the innovator it wants to be. So what does this all mean for managers and employees and their work? Managers must begin by deploying assets exceptionally well. The key assets under management are information and technology. Moreover, given a matrix structure, managers must do their asset deployment, as well as their coordinating and cultivating, in a focused but infrequent way. You might call it punctuated management—intense and vital contact between managers and employees, in an episodic form.

In another sense, the Cisco structure calls for a redefinition of line management. In this way of working, project managers oversee the execution of tasks by paying attention to:

This approach to configuring work calls for an extraordinary level of trust among managers and between managers and employees. Managers must have confidence that employees who are out of sight (except in a teleconference) are nevertheless working to their productive maximums. In effect, managers must fight against the illusion of control, since there's no way they can closely direct people who have many bosses. Employees, in turn, must trust their managers to perform a conscientious and comprehensive assessment of their work results, because those results will often be delivered to someone other than an employee's formal supervisor. Managers who share resources, including the human capital of employees in common, must negotiate for the use of those resources in good faith, with the greater good of the organization always forefront in their minds. As Cisco CEO John Chambers says, leading only the people reporting directly to them is not how Cisco managers should behave: "That's not what cross-functional leadership is about.... Whoever serves on each of these councils and boards and working groups, from each functional group, has to be able to speak for the whole group. Not go back and ask permission, but has to be able to speak for the group."^[177]

Cisco would probably admit that the roles of supervisors and managers in its technology turbocharged matrix are still under construction. After all, laying a matrix on top of a functionally oriented pyramid calls for substantial focus and energy. For this reason, the council and board structure has taken from 2001 to 2009 to become fully effective. No doubt managers still struggle with player-coach responsibilities, and spans of control probably need to be worked out for various managerial positions. But one thing is clear: Cisco has decided that the collaborative, punctuated leadership and management encouraged (in fact, required) by its matrix structure are critical to acquiring and defending a differentiated product position. The competition is no pushover—companies like Hewlett-Packard and IBM aren't likely to let Cisco win without a fight. An innovative organizational architecture, with all it implies for leaders and managers at all levels, represents a key weapon in Cisco's competitive arsenal.

Many organizations want to foster team-based collaboration without the complexities of a matrix approach. Some choose instead to institute self-managing teams. Like matrix configurations, self-managing teams are not new. And, like organizational matrices, they are easy to get wrong. Especially in organizations with cultures that encourage heavy manager direction or the opposite, unfettered individual initiative, autonomous teams can prove more frustrating than effective. The paradox of managing effectively from the sidelines has its quintessence in how supervisors build and work with self-managing teams.

A study done in the customer services division of Xerox Corporation explored the kind of manager behaviors that enhance or hinder the performance of self-managing work groups. Like Cisco, the modern incarnation of Xerox has chosen to pursue an innovation-intensive focus on product differentiation. Anne Mulcahy, former Xerox CEO, ensured that research and development would remain a continuing priority of the company. She is on record as emphasizing that Xerox is also dedicated to delivering a great customer experience. She came by her customer focus naturally, having started with Xerox as a sales representative in 1976.^[178] If anything, her successor, Ursula Burns, is working to accelerate the company's progress toward higher revenue through tighter customer ties and more profitable products and services.^[179]

At the time of the study, the Xerox service organization was divided into nine geographical areas that were in turn subdivided into districts and subdistricts. A manager headed up each subdistrict, which had twenty to thirty technicians apiece. Technicians worked in teams of between three and nine members. Their jobs were to respond to customer calls about machine breakdowns and visit customer sites to perform preventive maintenance. The study looked specifically at two forms of manager involvement with the teams: team design and manager coaching.

Thoughtfully designing a self-managing team involves a broad array of dimensions. The Xerox research team, headed by Dartmouth professor Ruth Wageman, looked at such design elements as:

Different styles of manager coaching encompassed such elements as:

The research team classified the first two forms of manager coaching—providing autonomy-reinforcing rewards and building the group's toolkit of skills—as "positive coaching." Conversely, two other elements—intervening in tasks and pointing out the team's problems—tended to reduce the team's sense of collective responsibility. The researchers labeled these "negative coaching."^[182]

The Dartmouth group assessed team performance by looking at such factors as customer satisfaction with service; parts expense; time to respond to customer calls and complete repairs; and machine reliability. The team came to a conclusion that at first seems surprising but makes sense on further reflection: no form of manager coaching improved team performance. The coaching elements that the researchers called positive coaching were associated with overall levels of successful team self-management and quality of group process. But they had no significant correlation with achievement of objective performance standards. In contrast, teams that were designed well (had a clear statement of purpose, consistently worked interdependently on tasks, executed a strategy of sharing successful practices, and benefited from a team-oriented reward system) did perform better.^[183] The takeaway for managers: create the architecture for self-managing teams carefully and then honor the concept of self-management. Be a good manager by providing information and material resources. Be a good leader by providing direction, establishing clear and challenging goals, helping when the team needs to get over process hurdles, and rewarding the team when it succeeds—but only as a group. Otherwise stay out of the way.