The Organized Mind

7 ORGANIZING THE BUSINESS WORLD

How We Create Value

At midday on September 30, 2006, the de la Concorde overpass at Laval outside of Montreal collapsed onto Quebec Autoroute 19, a major north-south artery. Five people were killed and six others were seriously injured when their cars were thrown over the edge. During the bridge’s construction, the contractors had installed the steel reinforcing bars in the concrete incorrectly and, to save money, unilaterally decided to use a lower-quality concrete, which didn’t meet design specifications. The ensuing government inquiry determined that this caused the bridge to collapse. Several other cases of low-quality concrete used in bridges, overpasses, and highways were identified in Quebec during a government inquiry into corruption in the construction industry. The history of shoddy construction practices is long—the wooden amphitheater in Fidenae near ancient Rome was built on a poor foundation, in addition to being improperly constructed, causing its collapse in 27 CE with 20,000 casualties. Similar disasters have occurred around the world, including the Teton Dam in Idaho in 1976, the collapse of Sichuan schools in the Chinese earthquake of 2008, and the failure of the Myllysilta Bridge in Turku, Finland, in 2010.

When they work properly, large civic projects like these involve many specialists and levels of checks and balances. The design, decision-making, and implementation are structured throughout an organization in a way that increases the chances for success and value. Ideally, what everyone is working for is a state in which both human and material resources are allocated to achieve maximum value. (When all the components of a complex system achieve maximum value, and when it is impossible to make any one component of the system better without making at least one other component worse, the system can be said to have reached the Pareto optimum.) The asphalt worker paving a city street shouldn’t normally make the decision about what quality of paving materials to use or how thick the layer should be—these decisions are made by higher-ups who must optimize, taking into account budgets, traffic flow, weather conditions, projected years of use, standard customs and practices, and potential lawsuits if potholes develop. These different aspects of information gathering and decision-making are typically distributed throughout an organization and may be assigned to different managers who then report to their higher-ups, who in turn balance the various factors to achieve the city’s long-term goals, to satisfice this particular decision. As Adam Smith wrote in The Wealth of Nations in 1776, one of the greatest advances in work productivity was the division of labor. Dividing up tasks in any large human enterprise has proved extremely influential and useful.

Up until the mid 1800s, businesses were primarily small and family-run, serving only a local market. The spread of telegraph and railroads beginning in the mid 1800s made it possible for more companies to reach national and international markets, building on progress in maritime trade that had been developing for centuries. The need for documentation and functional specialization or cross-training grew dramatically along with this burgeoning long-distance commerce. The aggregate of letters, contracts, accounting, inventory, and status reports presented a new organizational challenge: How do you find that piece of information you need this afternoon inside this new mountain of paper? The Industrial Revolution ushered in the Age of Paperwork.

A series of railroad collisions in the early 1840s provided an urgent push toward improved documentation and functional specialization. Investigators concluded that the accidents resulted from communications among engineers and operators of various lines being handled too loosely. No one was certain who had authority over operations, and acknowledging receipt of important messages was not common practice. The railroad company investigators recommended standardizing and documenting operating procedures and rules. The aim was to transcend dependence upon the skills, memory, or capacity of any single individual. This involved writing a precise definition of duties and responsibilities for each job, coupled with standardized ways of performing these duties.

Functional specialization within the workforce became increasingly profitable and necessary so that things wouldn’t grind to a halt if that lone worker who knew how to do this one particular thing was out sick. This led to functionally compartmentalized companies, and an even greater need for paperwork so workers could communicate with their bosses (who might be a continent away), and so that one division of a company could communicate with other divisions. The methods of record keeping and the management style that worked for a small family-owned company simply didn’t scale to these new, larger firms.

Because of these developments, managers suddenly had greater control over the workers, specifically, over who was doing the work. Processes and procedures that had been kept in workers’ heads were now recorded in handbooks and shared within the company, giving each worker an opportunity to learn from prior workers and to add improvements. Such a move follows the fundamental principle of the organized mind: externalizing memory. This involves taking the knowledge from the heads of a few individuals and putting it (such as in the form of written job descriptions) out-there-in-the-world where others can see and use it.

Once management obtained detailed task and job descriptions, it was possible to fire a lazy or careless employee and replace him or her with someone else without a great loss of productivity—management simply communicated the details of the job and where things had been left off. This was essential in building and repairing the railroads, where great distances existed between the company headquarters and the workers in the field. Yet soon the drive to systematize jobs extended to managers, so managers became as replaceable as workers, a development promoted by the English efficiency engineer Alexander Hamilton Church.

The trend toward systematizing jobs and increasing organization efficiency led the Scottish engineer Daniel McCallum to create the first organizational charts in 1854 as a way to easily visualize reporting relationships among employees. A typical org chart shows who reports to whom; the downward arrows indicate a supervisor-to-supervisee relationship.

Org charts represent reporting hierarchies very well, but they don’t show how coworkers interact with one another; and although they show business relationships, they do not show personal relationships. Network diagrams were first introduced by the Romanian sociologist Jacob Moreno in the 1930s. They are more useful in understanding which employees work with and know one another, and they’re often used by management consultants to diagnose problems in structural organization, productivity, or efficiency.

Below is the network diagram from a one-month survey of an Internet start-up company (the company was eventually sold to Sony). The diagram shows who interacted with whom during the month surveyed; the interactions shown are dichotomous, without attention to the number or quality of interactions. The diagram reveals that the founder (the node at the top) interacted with only one other person, his COO; the founder was on a fund-raising trip this particular month. The COO interacted with three people. One of them was in charge of product development, and he interacted with an employee who oversaw a network of seven consultants. The consultants interacted with one another a great deal.

Creating a network map allowed management to see that there was one person whom nobody ever talked to, and two people who interacted extensively with each other but no one else. Various forms of network diagrams are possible, including using “heat maps” in which colors indicate the degree of interaction (hotter colors mean more interaction along a node, colder colors mean less). Network maps can be used in conjunction with hierarchical organization charts to identify which members of an organization already know one another, which in turn can facilitate creating project teams or reorganizing certain functions and reporting structures. Standard organizational behavior practice is to split up teams that aren’t functioning efficiently and to try to replicate teams that are. But because team efficiency isn’t simply a matter of who has what skills, and is more a matter of interpersonal familiarity and who works well together, the network diagram is especially useful; it can track not just which team members work together but which, if any, socialize together outside of work (and this could be represented differentially with color or dotted lines, or any number of standard graphing techniques).

Organizations can have either flat (horizontal) or deep (vertical) hierarchies, which can have a great impact on employee and manager efficiency and effectiveness. Compare these two different org charts, for a flat company (left) with only three levels and a vertical company (right) with five levels:

The command structure in corporate and military organizations can take either form, and each system has advantages and disadvantages (conventional military structure is vertical, but terrorist and other cell-based groups typically use flat structure with decentralized control and communications).

A flat structure encourages people to work together and allows for overlap in effort, often empowering employees to do what needs to be done and apply their talents outside of formal command or task structure. A drawback of flat structure is that there may be only one person who has effective decision-making authority, and that person will have too many decisions to make. Due to the lack of a hierarchy, extra effort is needed to establish who has responsibility for which tasks. Indeed, some form of vertical structure is essential to achieve coordination among the employees and their projects, to avoid duplication of effort, and to ensure coherence across different components of a project. The additional advantage of vertical structureis that employees can more easily be held accountable for their decisions and their work product.

Tall vertical systems usually encourage specialization and the efficiencies that come from it. But tall structures can also result in employees being isolated from one another and working in silos, unaware of what others are doing that might be closely related to their own work. When a vertical system becomes too tall (too many levels), it can take too much time for instructions to filter down to the ground from higher up, or vice versa. Railroad companies led the way to more complex organization in the business world, and in the last fifty years the level of complexity has grown so that in many cases it is impossible to keep track of what everyone is doing. Fifty companies in the world have more than a quarter million employees, and seven companies in the world have more than one million.

Companies can be thought of as transactive memory systems. Part of the art of fitting into a company as a new employee, indeed part of becoming an effective employee (especially in upper management), is learning who holds what knowledge. If you want the 2014 sales figures for the southeastern region, you call Rachel, but she has the figures only for framistans; if you want to include your company’s business in selling gronespiels, you need to call Scotty; if you want to know if United Frabezoids ever got paid, you call Robin in accounts payable. The company as a whole is a large repository of information, with individual humans effectively playing the role of neural networks running specialized programs. No one person has all the knowledge, and indeed, no one person in a large company even knows whom to ask for every bit of knowledge it takes to keep the company running.

A typical story: Booz Allen Hamilton was given a big contract by the Fortune 100 company where Linda worked as the executive assistant to the CEO. Their assignment was to study the organization and make suggestions for structural improvement. While interviewing employees there, the Booz consultants discovered three highly trained data analysts with similar skill sets and similar mandates working in three entirely separate columns of the company’s org chart. Each data analyst reported to an assistant manager, who reported to a district manager, who reported to a division manager, who reported to a vice president. Each data analyst was ultimately responsible to an entirely different vice president, making it virtually impossible for them, their bosses, or even their bosses’ bosses to know about the existence of the others. (They even worked in different buildings.) Booz consultants were able to bring the analysts together for weekly meetings where they pooled their knowledge, shared certain tricks they had learned, and helped one another solve common technical problems they were facing. This led to great efficiencies and cost savings for the company.

Vertical structures are necessary when a high degree of control and direct supervision over employees are required. Nuclear power plants, for example, tend to have very tall vertical structures because supervision is extremely important—even a small error can result in a disaster. The vertical structure allows managers to constantly check and cross-check the work of lower-level managers to ensure that rules and procedures are followed accurately and consistently.

RBC Royal Bank of Canada is a $30 billion company, serving 18 million customers. Its corporate culture places a high value on mentorship, on managers developing subordinates, improving their chances of being promoted, and ensuring gender equity. Its vertical structure allows for close supervision of employees by their managers. Liz Claiborne, Inc., was the first Fortune 500 company to be founded by a woman. When Liz Claiborne was designing the structure of her company, she chose flat—four levels for four thousand employees—in order to keep the company nimble and able to respond quickly to changing fashion trends. There is no evidence that structure in and of itself affects the profitability of a company; different structures work best for different companies.

The size of an organization tends to predict how many levels it will have, but the relationship is logarithmic. That is, while an organization with 1,000 employees on average has four hierarchical levels, increasing the number of employees by a factor of 10 does not increase the number of levels by 10; rather, it increases the number of levels by a factor of 2. And after an organization reaches 10,000 employees, an asymptote is reached: Organizations with 12,000, 100,000, or 200,000 employees rarely have more than nine or ten levels in their hierarchy. The principle of minimum chain of command states that an organization should choose the fewest number of hierarchical levels possible.

These same descriptions of structure—flat and vertical—can be applied to a corporate website, or the file system on your own computer. Imagine that the flat and vertical structure drawings on page 273 are site maps for two different versions of a company’s website. Both sites might present visitors with the same data, but the visitors’ experience will be very different. With a well-designed flat organization, the visitor can obtain summary information in one click, and more detailed information in two clicks. With the vertical organization, that same visitor might also locate desired summary information in one or two clicks, but the detailed information will require four clicks. Of course sites aren’t always designed well or in a way that allows a visitor to find what she’s looking for—Web designers are not typical users, and the labels, menus, and hierarchies they use may not be obvious to anyone else. Hence the user may end up doing a great deal of searching, fishing, and backtracking. The flat organization makes it easier to backtrack; the vertical makes it easier to locate a hard-to-find file if the visitor can be sure she’s in the correct subnode. Still, there are limits to flat organizations’ ease of use: If the number of middle-level categories becomes too great, it takes too long to review them all, and because they themselves are not hierarchically organized, there can be redundancies and overlap. Visitors can easily become overwhelmed by too many choices—deep hierarchy offers fewer choices at once. The same analysis applies to the folders within folders on your hard drive.

But the organization of people is radically different from the organization of a website. Even in a deep vertical structure, people can and need to have agency from time to time. The lowliest transit worker sometimes needs to jump on the track to rescue a woman who fell; an investment bank secretary needs to be a whistle-blower; a mailroom worker needs to notice the disgruntled coworker who showed up with a rifle. All those actions fulfill a part of the company’s objectives—safety and ethical dealings.

In any hierarchically organized firm or agency, the task of carrying out the company’s objectives typically falls to the people at the lowest levels of the hierarchy. Cell phones aren’t built by the engineer who designed them or the executive who is in charge of marketing and selling them but by technicians on an assembly line. A fire isn’t put out by the fire chief but by the coordinated efforts of a team of firefighters on the street. While managers and administrators do not typically do the main work of a company, they play an essential role in accomplishing the company’s objectives. Even though it is the machine gunner and not the major who fights battles, the major is likely to have a greater influence on the outcome of a battle than any single machine gunner.

Decision-Making Throughout the Hierarchy

Anyone who has ever owned something of high value that needs repairs—a home or a car, for example—has had to contend with compromises and has seen how a management perspective is necessary to the decision-making process. Do you buy the thirty-year roof or the twenty-year roof? The top-of-the-line washing machine or the bargain brand? Suppose your mechanic tells you that you need a new water pump for your car and that, in descending order of price, he can install an original equipment manufacturer (OEM) part from the dealer, a functionally identical part from an overseas company, or a warrantied used part from a junkyard. He can’t make the decision for you because he doesn’t know your disposable income or your plans for the car. (Are you getting ready to sell it? Restoring it for entry in a car show? Planning to drive it through the Rockies next July, where the cooling system will be pushed to its limits?) In short, the mechanic doesn’t have a high-level perspective on your long-range plans for your car or your money. Any decision other than the OEM part installed by an authorized dealer is a compromise, but one that many people are willing to make in the interest of satisficing.

Standard models of decision-making assume that a decision maker—especially in economic and business contexts—is not influenced by emotional considerations. But neuroeconomics research has shown this is not true: Economic decisions produce activity in emotional regions of the brain, including the insula and amygdala. The old cartoon image of the angel on one shoulder and the devil on the other, giving competing advice to a flummoxed head in the middle, is apt here. Benefits are evaluated deep inside the brain, in a part of the striatum closest to your spine (which includes the brain’s reward center, the nucleus accumbens), while costs are simultaneously evaluated in the amygdala, another deep structure, commonly thought of as the brain’s fear center (the region responsible for the fight-or-flight response during threats to survival and other dangers). Taking in this competing information about costs and benefits, the prefrontal cortex acts as the decider. This isn’t the same thing as the experience we have of consciously trying to decide between two alternatives; decision-making is often very rapid, outside our conscious control, and involves heuristics and cognitive impulses that have evolved to serve us in a wide range of situations. The rationality we think we bring to decision-making is partly illusory.

Major decisions are usually not made by any one individual, nor by any easily defined group of individuals. They emerge through a process of vastly distributed discussion, consultation, and sharing of information. This is both a positive and a negative feature of large organizations. When they work well, great things can be accomplished that would be impossible for a small number of people to do: designing and building the Hoover Dam, the plasma TV, or Habitat for Humanity. As suggested at the beginning of this chapter, when communications or the exercise of competent and ethically based authority do not work well, or the optimal checks and balances aren’t in place, you end up with bridge collapses, Enron, or AIG.

In general, in a multilevel vertical organization, the chain of authority and direction travels downward with increasing specificity. The CEO may articulate a plan to one of his VPs; that VP adds some specificity about how best he thinks the plan can be accomplished and hands it to a division manager with experience and expertise in these sorts of operations. This continues on, down the line, until it reaches the individuals who actually do the work.

We see this in the organization of military authority. The general or commander defines a goal. The colonel assigns tasks to each battalion in his command; the major to each company in his battalion; the captain to each platoon in his company. Each officer narrows the scope and increases the specificity of the instructions he passes on. Still, the modern army gives a fair degree of situational control and discretion to the soldiers on the ground. Perhaps surprisingly, the U.S. Army has been among the organizations most adaptable to change, and has thought deeply about how to the apply findings of psychological science to organizational behavior. Its current policy strives to empower people throughout the chain of command, “allowing subordinate and adjacent units to use their common understanding of the operational environment and commander’s intent, in conjunction with their own initiative, to synchronize actions with those of other units without direct control from the higher headquarters.”

The value of limited autonomy and the exercise of discretion by subordinates is not a recent development in organizational strategy, for companies or for the military. Nearly one hundred years ago, the 1923 U.S. Army Field Service Regulations manual expected that subordinates would have a degree of autonomy in matters of judgment, stating that “an order should not trespass upon the province of a subordinate.”

Smooth operation within the military or a company requires trust between subordinates and superiors and an expectation that subordinates will do the right thing. The current edition of the U.S. Army Training Manual puts it this way:

Our fundamental doctrine for command requires trust throughout the chain of command. Superiors trust subordinates and empower them to accomplish missions within their intent. Subordinates trust superiors to give them the freedom to execute the commander’s intent and support their decisions. The trust between all levels depends upon candor. . . .

Army doctrine stresses mission command, the conduct of military operations that allows subordinate leaders maximum initiative. It acknowledges that operations in the land domain are complex and often chaotic, and micromanagement does not work. Mission command emphasizes competent leaders applying their expertise to the situation as it exists on the ground and accomplishing the mission based on their commander’s intent. Mission command fosters a culture of trust, mutual understanding, and a willingness to learn from mistakes. . . . Commanders . . . provide subordinates as much leeway for initiative as possible while keeping operations synchronized.

Superiors often resist delegating authority or decisions. They rationalize this by saying that they are more highly skilled, trained, or experienced than the subordinate. But there are good reasons for delegating decision-making. First, the superior is more highly paid, and so the cost of the decision must be weighed against the benefit of having such a high-paid individual make it. (Remember the maxim from Chapter 5: How much is your time worth?) Along the same lines, the superior has to conserve his time so that he can use it for making more important decisions. Secondly, subordinates are often in a better position to make decisions because the facts of the case may be directly available to them and not to the superior. General Stanley McChrystal articulated this with respect to his leadership during the United States–Iraq conflict:

In my command, I would push down the ability and authority to act. It doesn’t mean the leader abrogates responsibility but that the team members are partners, not underlings. They’d wake me up in the middle of the night and ask “Can we drop this bomb?” and I’d ask “Should we?” Then they’d say, “That’s why we’re calling you!” But I don’t know anything more than they’re telling me, and I’m probably not smart enough to add any value to the knowledge they already have from the field.

Steve Wynn’s management philosophy endorses the same idea:

Like most managers, I’m at the top of a large pyramidal structure, and the people who are below me make most of the decisions. And most of the time, the decisions that they make are of the “A or B” type: Should we do A or should we do B? And for most of those, the decision is obvious—one outcome is clearly better than the other. In a few cases, the people below me have to think hard about which one to do, and this can be challenging. They might have to consult with someone else, look deeper into the problem, get more information.

Once in a while a decision comes along where both outcomes look bad. They have a choice between A and B and neither one is going to be good, and they can’t figure out which one to choose. That’s when they end up on my calendar. So when I look at my calendar, if the Director of Food Services is on there, I know it’s something bad. Either he’s going to quit, or he’s got to make a decision between two very bad outcomes. My job when that happens is not to make the decision for them as you might think. By definition, the people who are coming to me are the real experts on the problem. They know lots more about it, and they are closer to it. All I can do is try to get them to look at the problem in a different light. To use an aviation metaphor, I try to get them to see things from 5,000 feet up. I tell them to back up and find out one truth that they know is indisputable. However many steps they might have to back up, I talk it over with them until they find the deep truth underlying all of it. The truth might be something like “the most important thing at our hotel is the guest experience,” or “no matter what, we cannot serve food that is not 100% fresh.” Once they identify that core truth, we creep forward slowly through the problem and often a solution will emerge. But I don’t make the decision for them. They’re the ones who have to bring the decision to the people under them, and they’re the ones who have to live with it, so they need to come to the decision themselves and be comfortable with it.

It is just as important to recognize the value of making difficult decisions when necessary. As former New York mayor Michael Bloomberg notes:

A leader is someone willing to make decisions. Politicians can get elected if voters think they will do things, even if they don’t support all those things. W [President George W. Bush] was elected not because everyone agreed with him but because they knew he was sincere and would do what he thought needed to be done.

Ethics necessarily come into play in corporate and military decision-making. What is good for one’s own self-interests, or the company’s interests, is not always consonant with what is good for the community, the populace, or the world. Humans are social creatures, and most of us unconsciously modify our behavior to minimize conflict with those around us. Social comparison theory models this phenomenon well. If we see other cars parking in a no-parking zone, we are more likely to park there ourselves. If we see other dog owners ignoring the law to clean up after their dogs, we are more likely to ignore it, too. Part of this comes from a sense of equity and fairness that has been shown to be innately wired into our brains, a product of evolution. (Even three-year-olds react to inequality.) In effect, we think, “Why should I be the chump who picks up dog poo when everyone else just leaves theirs all over the Boston Commons?” Of course the argument is specious because good behaviors are just as contagious as bad, and if we model correct behavior, others are likely to follow.

Organizations that discuss ethics openly, and that model ethical behavior throughout the organization, create a culture of adhering to ethical norms because it is “what everyone does around here.” Organizations that allow employees to ignore ethics form a breeding ground for bad behavior that tempts even the most ethically minded and strong-willed person, a classic case of the power of the situation overpowering individual, dispositional traits. The ethical person may eventually find him- or herself thinking, “I’m fighting a losing battle; there’s no point in going the extra mile because no one notices and no one cares.” Doing the right thing when no one is looking is a mark of personal integrity, but many people find it very difficult to do.

The army is one of the most influential organizations to have addressed this, and they do so with surprising eloquence:

All warfare challenges the morals and ethics of Soldiers. An enemy may not respect international conventions and may commit atrocities with the aim of provoking retaliation in kind. . . . All leaders shoulder the responsibility that their subordinates return from a campaign not only as good Soldiers, but also as good citizens. . . . Membership in the Army profession carries with it significant responsibility—the effective and ethical application of combat power.

Ethical decision-making invokes different brain regions than economic decision-making and again, because of the metabolic costs, switching between these modes of thought can be difficult for many people. It’s difficult therefore to simultaneously weigh various outcomes that have both economic and ethical implications. Making ethical or moral decisions involves distinct structures within the frontal lobes: the orbitofrontal cortex (located just behind the eyes) and the dorsolateral prefrontal cortex just above it.These two regions are also required for understanding ourselves in relation to others (social perception), and the compliance with social norms. When damaged, they can lead to socially inappropriate behavior such as swearing, walking around naked, and saying insulting things to people right to their faces. Making and evaluating ethical decisions also involves distinct subregions of the amygdala, the hippocampus (the brain’s memory index), and the back portion of the superior temporal sulcus, a deep groove in the brain that runs from front to back behind the ears. As with economic decisions involving costs and benefits, the prefrontal cortex acts as the decider between the moral actions being contemplated.

Neuroimaging studies have shown that ethical behavior is processed in the same way regardless of whether it involves helping someone in need or thwarting an unethical action. In one experiment, participants watched videos of people being compassionate toward an injured individual, or aggressive toward a violent assailant. As long as the people in the video were behaving in an ethically appropriate and socially sanctioned way, the same brain regions were active in the participants who watched the videos. Moreover, such brain activations are universal across people—different people contemplating the same ethical acts show a high degree of synchronization of their brain activity; that is, their neurons fire in similar, synchronous patterns. The neuronal populations affected by this include those in the insula (mentioned above in the discussion of economic decision-making), our friend the prefrontal cortex, and the precuneus, a region at the top and back of the head associated with self-reflection and perspective taking, and which exists not just in humans but in monkeys.

Does this mean that even monkeys have a moral sense? A recent study by one of the leading scientists of animal behavior, Frans de Waal, asked just this question. He found that monkeys have a highly developed sense of what is and is not equitable. In one study, brown capuchin monkeys who participated in an experiment with another monkey could choose to reward only themselves (a selfish option) or both of them (an equitable, prosocial option). The monkeys consistently chose to reward their partner. And this was more than a knee-jerk response. De Waal found convincing evidence that the capuchins were performing a kind of moral calculation. When the experimenter “accidentally” overpaid the partner monkey with a better treat, the deciding monkey withheld the reward to the partner, evening out the payoffs. In another study, monkeys performed tasks in exchange for food rewards given by the experimenters. If the experimenter gave a larger reward to one monkey than another for the same task, the monkey with the smaller reward would suddenly stop performing the task and sulk. Think about this: These monkeys were willing to forgo a reward entirely (a tempting piece of food) simply because they felt the organization of the reward structure was unfair.

Those In Charge

Conceptions of leadership vary from culture to culture and across time, including figures as diverse as Julius Caesar and Thomas Jefferson, Jack Welch of GE and Herb Kelleher of Southwest Airlines. Leaders can be reviled or revered, and they gain followers through mandate, threat of punishment (economic, psychological, or physical), or a combination of personal magnetism, motivation, and inspiration. In modern companies, government, or the military, a good leader might be best defined as anyone who inspires and influences people to accomplish goals and to pursue actions for the greater good of the organization. In a free society, an effective leader motivates people to focus their thinking and efforts in ways that allow them to do their best and to produce work that pushes them to the highest levels of their abilities. In some cases, people so inspired are free to discover unseen talents and achieve great satisfaction from their work and their interactions with coworkers.

A broader definition of leadership promoted by Harvard psychologist Howard Gardner includes individuals who significantly affect the thoughts, feelings, or behaviors of a significant number of individuals indirectly, through the works they create—these can be works of art, recipes, technological artifacts and products . . . almost anything. In this conception, influential leaders would include Amantine Dupin (George Sand), Picasso, Louis Armstrong, Marie Curie, and Martha Graham. These leaders typically work outside of corporate structure, although like anyone, they have to work with big business at some contractual level. Nevertheless, they don’t fit the standard business-school profile of a leader who has significant economic impact.

Both kinds of leaders, those inside and outside the corporate world, possess certain psychological traits. They tend to be adaptable and responsive, high in empathy, and able to see problems from all sides. These qualities require two distinct forms of cognition: social intelligence and flexible, deep analytic intelligence. An effective leader can quickly understand opposing views, how people came to hold them, and how to resolve conflicts in ways that are perceived to be mutually satisfying and beneficial. Leaders are often adept at bringing people together—suppliers, potential adversaries, competitors, characters in a story—who appear to have conflicting goals. A great business leader uses her empathy to allow people or organizations to save face in negotiations so that each side in a completed negotiation can feel they got what they wanted (and a gifted negotiator can make each side feel they got a little bit more than the other party). In Gardner’s model, it is no coincidence that many great leaders are also great storytellers—they motivate others around them with a compelling narrative, one that they themselves embody. Leaders show greater integration of electrical activity in the brain across disparate regions, meaning that they use more of their brain in a better-orchestrated fashion than the rest of us. Using these measures of neural integration, we can identify leaders in athletics and music, and in the next few years, the techniques promise to be refined enough to use as screening for leadership positions.

Great leaders can turn competitors into allies. Norbert Reithofer, CEO of BMW, and Akio Toyoda, CEO of Toyota—clearly competitors—launched a collaboration in 2011 to create an environmentally friendly luxury vehicle and a midsize sports car. The on-again, off-again partnership and strategic alliance between Steve Jobs at Apple and Bill Gates at Microsoft strengthened both companies and allowed them to better serve their customers.

As is obvious from the rash of corporate scandals in the United States over the last twenty years, negative leadership can be toxic, resulting in the collapse of companies or the loss of reputation and resources. It is often the result of self-centered attitudes, a lack of empathy for others within the organization, and a lack of concern with the organization’s long-term health. The U.S. Army recognizes this in military and civic organizations as well: “Toxic leaders consistently use dysfunctional behaviors to deceive, intimidate, coerce, or unfairly punish others to get what they want for themselves.” Prolonged use of these tactics undermines and erodes subordinates’ will, initiative, and morale.

Leaders are found in all levels of the company—one doesn’t have to be the CEO to exert influence and affect corporate culture (or to be a storyteller with the power to motivate others). Again, some of the best thinking on the subject comes from the U.S. Army. The latest version of their Mission Command manual outlines five principles that are shared by commanders and top executives in the most successful multinational businesses:

Build cohesive teams through mutual trust.
Create shared understanding.
Provide a clear and concise set of expectations and goals.
Allow workers at all levels to exercise disciplined initiative.
Accept prudent risks.

Trust is gained or lost through everyday actions, not through grand or occasional gestures. It takes time to build—coming from successful shared experiences and training—a history of two-way communication, the successful completion of projects, and achievement of goals.

Creating shared understanding refers to company management communicating with subordinates at all levels the corporate vision, goals, and the purpose and significance of any specific initiatives or projects that must be undertaken by employees. This helps to empower employees to use their discretion because they share in a situational understanding of the overriding purpose of their actions. Managers who hide this purpose from underlings, out of a misguided sense of preserving power, end up with unhappy employees who perform their jobs with tunnel vision and who lack the information to exercise initiative.

At McGill University, the dean of science undertook an initiative several years ago called STARS (Science Talks About Research for Staff). These were lunchtime talks by professors in the science department who described their research to the general staff: secretaries, bookkeepers, technicians, and the custodial staff. These jobs tend to be very far removed from the actual science. The initiative was successful by any measure—the staff gained an understanding of the larger context of what they were doing. A bookkeeper realized she wasn’t just balancing the books for any old research lab but for one that was on the cusp of curing a major disorder. A secretary discovered that she was supporting work that uncovered the cause of the 2011 tsunami and that could help save lives with better tsunami predictions. The effect of Soup and Science was that everyone felt a renewed sense of purpose for their jobs. One custodian commented later that he was proud to be part of a team doing such important work. His work improved and he began to take personal initiative that improved the research environment in very real and tangible ways.

The third of the army’s five command principles concerns providing a clear and concise expression of expectations and goals, the purpose of particular tasks, and the intended end state. This furnishes focus to staff and helps subordinates and their superiors to achieve desired results without extensive further instructions. The senior manager’s intent provides the basis for unity of effort throughout the larger workforce.

Successful managers understand that they cannot provide guidance or direction for all conceivable contingencies. Having communicated a clear and concise expression of their intent, they then convey the boundaries within which subordinates may exercise disciplined initiative while maintaining unity of effort. Disciplined initiative is defined as taking action in the absence of specific instructions when existing instructions no longer fit the situation, or unforeseen opportunities arise.

Prudent risk is the deliberate exposure to a negative outcome when the employee judges that the potential positive outcome is worth the cost. It involves making careful, calculated assessments of the upsides and downsides of different actions. As productivity expert Marvin Weisbord notes, “There are no technical alternatives to personal responsibility and cooperation in the workplace. What’s needed are more people who will stick their necks out.”

Some employees are more productive than others. While some of this variation is attributable to differences in personality, work ethic, and other individual differences (which have a genetic and neurocognitive basis), the nature of the job itself can play a significant role. There are things that managers can do to improve productivity, based on recent findings in neuroscience and social psychology. Some of these are obvious and well known, such as setting clear goals and providing high-quality, immediate feedback. Expectations need to be reasonable or employees feel overwhelmed, and if they fall behind, they feel they can never catch up. Employee productivity is directly related to job satisfaction, and job satisfaction in turn is related to whether employees experience that they are doing a good job in terms of both quality and quantity of output.

There’s a part of the brain called Area 47 in the lateral prefrontal cortex that my colleague Vinod Menon and I have been closely studying for the last fifteen years. Although no larger than your pinky finger, it’s a fascinating area just behind your temples that has kept us busy. Area 47 contains prediction circuits that it uses in conjunction with memory to form projections about future states of events. If we can predict some (but not all) aspects of how a job will go, we find it rewarding. If we can predict all aspects of the job, down to the tiniest minutiae, it tends to be boring because there is nothing new and no opportunity to apply the discretion and judgment that management consultants and the U.S. Army have justly identified as components to finding one’s work meaningful and satisfying. If some but not too many aspects of the job are surprising in interesting ways, this can lead to a sense of discovery and self-growth.

Finding the right balance to keep Area 47 happy is tricky, but the most job satisfaction comes from a combination of these two: We function best when we are under some constraints and are allowed to exercise individual creativity within those constraints. In fact, this is posited to be the driving force in many forms of creativity, including literary and musical. Musicians work under the very tight constraints of a tonal system—Western music uses only twelve different notes—and yet within that system, there is great flexibility. The composers widely regarded as among the most creative in musical history fit this description of balancing creativity within constraints. Mozart didn’t invent the symphony (Torelli and Scarlatti are credited with that) and The Beatles didn’t invent rock ’n’ roll (Chuck Berry and Little Richard get the credit, but its roots go back clearly to Ike Turner and Jackie Brenston in 1951, Louis Jordan and Lionel Hampton in the 1940s). It’s what Mozart and The Beatles did within the tight constraints of those forms, the enormous creativity and ingenuity they brought to their work, that pushed at the boundaries of those forms, leading to them being redefined.

But there is a critical point about differences between individuals that exerts arguably more influence on worker productivity than any other. The factor is locus of control, a fancy name for how people view their autonomy and agency in the world. People with an internal locus of control believe that they are responsible for (or at least can influence) their own fates and life outcomes. They may or may not feel they are leaders, but they feel that they are essentially in charge of their lives. Those with an external locus of control see themselves as relatively powerless pawns in some game played by others; they believe that other people, environmental forces, the weather, malevolent gods, the alignment of celestial bodies—basically any and all external events—exert the most influence on their lives. (This latter view is artistically conveyed in existential novels by Kafka and Camus, not to mention Greek and Roman mythology.) Of course these are just extremes, and most people fall somewhere along a continuum between them. But locus of control turns out to be a significant moderating variable in a trifecta of life expectancy, life satisfaction, and work productivity. This is what the modern U.S. Army has done in allowing subordinates to use their own initiative: They’ve shifted a great deal of the locus of control in situations to the people actually doing the work.

Individuals with an internal locus of control will attribute success to their own efforts (“I tried really hard”) and likewise with failure (“I didn’t try hard enough”). Individuals with an external locus of control will praise or blame the external world (“It was pure luck” or “The competition was rigged”). In school settings, students with a high internal locus of control believe that hard work and focus will result in positive outcomes, and indeed, as a group they perform better academically. Locus of control also affects purchasing decisions. For example, women who believe they can control their weight respond most favorably to slender advertising models, and women who believe they can’t respond better to larger-size models.

Locus of control also shows up in gambling behaviors: Because people with a high external locus of control believe that things happen to them capriciously (rather than being the agents of their own fortunes), they are more likely to believe that events are governed by hidden and unseen outside forces such as luck. Accordingly, they are likely to take more chances, try riskier bets, and bet on a card or roulette number that hasn’t come up in a long time, under the mistaken notion that this outcome is now due; this is the so-called gambler’s fallacy. They are also more likely to believe that if they need money, gambling can provide it.

Locus of control appears to be a stable internal trait that is not significantly affected by experiences. That is, you might expect that people who experience a great deal of hardship would give up any notions of their own agency in the face of overwhelming evidence to the contrary and become externals. And you might expect that those who experience a great deal of success would become internals, self-confident believers that they were the agents of that success all along. But the research doesn’t bear this out. For example, researchers studied small independent business owners whose shops were destroyed by Hurricane Agnes in 1972, at the time, the costliest hurricane to hit the United States. Over one hundred business owners were assessed for whether they tended toward internal or external locus of control. Then, three and a half years after the hurricane, they were reassessed. Many realized big improvements in their businesses during the recovery years, but many did not, seeing once thriving businesses deteriorate dramatically; many were thrown into ruin.

The interesting finding is that on the whole, none of these individuals shifted their views about internal versus external locus of control as a function of how their fortunes changed. Those who were internals to begin with remained internals regardless of whether their business performance improved or not during the intervening time. Same with the externals. Interestingly, however, those internals whose performance improved showed a shift toward greater internality, meaning they attributed the improvement to their hard work. Those who were externals and who experienced setbacks and losses showed a shift toward greater externality, meaning they attributed their failures to a deepening of the situational factors and bad luck that they felt they had experienced throughout their lives. In other words, a change of fortune following the hurricane that confirmed their beliefs only caused them to increase the strength of those beliefs; a change in fortune that went counter to their beliefs (an internal losing everything, an external whose business recovered) did nothing to change their beliefs.

The locus-of-control construct is measurable with standard psychological tests and turns out to be predictive of job performance. It also influences the managerial style that will be effective. Employees who have an external locus of control believe their own actions will not lead to the attainment of rewards or the avoidance of punishment, and therefore, they don’t respond to rewards and punishments the way others do. Higher managers tend to have a high internal locus of control.

Internals tend to be higher achievers, and externals tend to experience more stress and are prone to depression. Internals, as you might expect, exert greater effort to influence their environment (because, unlike externals, they believe their efforts will amount to something). Internals tend to learn better, seek new information more actively, and use that information more effectively, and they are better at problem solving. Such findings may lead managers to think they should screen for and hire only people with an internal locus of control, but it depends on the particular job. Internals tend to exhibit less conformity than externals, and less attitude change after being exposed to a persuasive message. Because internals are more likely to initiate changes in their environment, they can be more troublesome to supervise. Moreover, they’re sensitive to reinforcement, so if effort in a particular job doesn’t lead to rewards, they may lose motivation more than an external, who has no expectation that his or her effort really matters anyway.

Industrial organization scientist Paul Spector of the University of South Florida says that internals may attempt to control work flow, task accomplishment, operating procedures, work assignments, relationships with supervisors and subordinates, working conditions, goal setting, work scheduling, and organizational policy. Spector summarizes: “Externals make more compliant followers or subordinates than do internals, who are likely to be independent and resist control by superiors and other individuals. . . . Externals, because of their greater compliance, would probably be easier to supervise as they would be more likely to follow directions.” So the kind of employee who will perform best depends on the kind of work that needs to be done. If the job requires adaptability and complex learning, independence and initiative, or high motivation, internals would be expected to perform better. When the job requires compliance and strict adherence to protocols, the external would perform better.

The combination of high autonomy and an internal locus of control is associated with the highest levels of productivity. Internals typically “make things happen,” and this, combined with the opportunity to do so (through high autonomy), delivers results. Obviously, some jobs that involve repetitive, highly constrained tasks such as some assembly-line work, toll taking, stockroom, cashier, and manual labor are better suited to people who don’t desire autonomy. Many people prefer jobs that are predictable and where they don’t have to take personal responsibility for how they organize their time or their tasks. These workers will perform better if they can simply follow instructions and are not asked to make any decisions. Even within these kinds of jobs, however, the history of business is full of cases in which a worker exercised autonomy in a job where it was not typically found and came up with a better way of doing things, and a manager had the foresight to accept the worker’s suggestions. (The sandpaper salesman Richard G. Drew, who invented masking tape and turned 3M into one of the largest companies, is one famous case.)

On the other hand, workers who are self-motivated, proactive, and creative may find jobs with a lack of autonomy to be stifling, frustrating, and boring, and this may dramatically reduce their motivation to perform at a high level. This means that managers should be alert to the differences in motivational styles, and take care to provide individuals who have an internal locus of control with autonomous jobs, and individuals who have an external locus of control with more constrained jobs.

Who was the most productive person of all time? This is a difficult question to answer, largely because productivity itself is not well defined, and conceptions of it change through the ages and over different parts of the world. But one could argue that William Shakespeare was immensely productive. Before dying at the age of fifty-two, he composed thirty-eight plays, 154 sonnets, and two long narrative poems. Most of his works were produced in a twenty-four-year period of intense productivity. And these weren’t just any works—they are some of the most highly respected works of literature ever produced in the history of the world.

One could also make a case for Thomas Edison, who held nearly eleven hundred patents, including many that changed history: electric light and power utilities, sound recordings, and motion pictures. He also introduced pay-per-view in 1894. One thing these two have in common—and share with other greats like Mozart and Leonardo da Vinci—is that they were their own bosses. That means to a large degree the locus of control for their activities was internal. Sure, Mozart had commissions, but within a system of constraints, he was free to do what he wanted in the way he wanted to do it. Being one’s own boss requires a lot of discipline, but for those who can manage it, greater productivity appears to be the reward.

Other factors contribute to productivity, such as being an early riser: Studies have shown that early birds tend to be happier, more conscientious and productive, than night owls. Sticking to a schedule helps, as does making time for exercise. Mark Cuban, the owner of Landmark Theatres and the Dallas Mavericks, echoes what many CEOs and their employees say about meetings: They’re usually a waste of time. An exception is if you’re negotiating a deal or soliciting advice from a large number of people. But even then, meetings should be short, drawn up with a strict agenda, and given a time limit. Warren Buffett’s datebook is nearly completely empty and has been for twenty-five years—he rarely schedules anything of any kind, finding that an open schedule is a key to his productivity.

The Paperwork

Organizing people is a good start to increasing value in any business. But how can the people—and that’s each of us—begin to organize the constant flood of documents that seem to take over every aspect of our work and our private lives? Managing the flow of paper and electronic documents is increasingly important to being effective in business. By now, weren’t we supposed to have the paperless office? That seems to have gone the way of jet packs and Rosie the Robot. Paper consumption has increased 50% since 1980, and today the United States uses 70 million tons of paper in a year. That’s 467 pounds, or 12,000 sheets, of paper for every man, woman, and child. It would take six trees forty feet tall to replenish it. How did we get here and what can we do about it?

After the mid 1800s, as companies grew in size, and their employees spread out geographically, businesses found it useful to keep copies of outgoing correspondence either by hand-copying each document or through the use of a protocopier called the letter press. Incoming correspondence tended to be placed in pigeonhole desks and cabinets, sometimes sorted but often not. Cogent information, such as the sender, date, and subject, might be written on the outside of the letter or fold to help in locating it later. With a small amount of incoming correspondence, the system was manageable—one might have to search through several letters before finding the right one, but this didn’t take too much time and could have been similar to the children’s card game Concentration.

Concentration is a game based on a 1960s television game show hosted by Hugh Downs. In the home version, players set up a matrix of cards facedown—perhaps six across and five down for a total of thirty cards. (You start with two decks of cards and select matched pairs, so that every card in your matrix has an identical mate.) The first player turns over two cards. If they match, the player keeps them. If they don’t, the player turns them back over, facedown, and it is the next player’s turn. Players who can remember where previously turned-over cards were located are at an advantage. The ability to do this resides in the hippocampus—remember, it’s the place-memory system that increases in size in London taxicab drivers.

All of us use this hippocampal spatial memory every day, whether trying to find a document or a household item. We often have a clear idea of where the item is, relative to others. The cognitive psychologist Roger Shepard’s entire filing system was simply stacks and stacks of paper all through his office. He knew which pile a given document was in, and roughly how far down into the pile it was, so he could minimize his search time using this spatial memory. Similarly, the early system of finding unsorted letters filed in cubbyholes relied on the office worker’s spatial memory of where that letter was. Spatial memory can be very, very good. Squirrels can locate hundreds of nuts they buried—and they’re not just using smell. Experiments show that they preferentially look for nuts that they buried in the places they buried them, not for nuts buried by other squirrels. Nevertheless, with any large amount of paperwork or correspondence, finding the right piece in the nineteenth century could easily become time-consuming and frustrating.

The cubbyhole filing system was among the first modern attempts to externalize human memory and extend our brains’ own processing capacity. Important information was written down and could then be consulted later for verification. The limitation was that human memory had to be used to remember where the document was filed.

The next development in the cubbyhole filing system was . . . more cubbyholes! The Wooton Desk (patent 1874) featured over one hundred storage places, and advertising promised the businessman he would become “master of the situation.” If one had the prescience to label the cubbyholes in an organized fashion—by client last name, by due date for order, or through some other logical scheme—the system could work very well.

But still the big problem was that each individual document needed to be folded to fit in the cubbyholes, meaning that it had to be unfolded to be identified and used. The first big improvement on this was the flat file, introduced in the late 1800s. Flat files could be kept in drawers, in bound book volumes, or in cabinets, and they increased search efficiency as well as capacity. Flat files were either bound or unbound. When bound, documents tended to be stored chronologically, which meant that one needed to know roughly when a document arrived in order to locate it. More flexible were flat files that were filed loosely in boxes and drawers; this allowed them to be arranged, rearranged, and removed as needed, just like the 3 x 5 index cards favored by Phaedrus (and many HSPs) in Chapter 2.

The state of the art for flat file storage by the late nineteenth century was a system of letter-size file boxes, similar to the kind still available today at most stationery stores. Correspondence could be sewn in, glued in, or otherwise inserted into alphabetical or chronological order. By 1868, flat file cabinets had been introduced—these were cabinets containing several dozen drawers of the dimensions of a flat letter, something like oversize library card catalogues. These drawers could be organized in any of the ways already mentioned, typically chronologically, alphabetically, or topically, and the contents of the drawers could be further organized. Often, the drawer contents were left unsorted, requiring the user to have to look through the contents to find the right document. JoAnne Yates, professor of management at MIT and a world expert in business communication, articulates the problems:

To locate correspondence in an opened box file or a horizontal cabinet file, all the correspondence on top of the item sought had to be lifted up. Since the alphabetically or numerically designated drawers in horizontal cabinet files filled up at different rates, correspondence was transferred out of active files into back-up storage at different rates as well. And the drawers could not be allowed to get too full, since then papers would catch and tear as the drawers were opened. Letter boxes had to be taken down from a shelf and opened up, a time-consuming operation when large amounts of filing were done.

As Yates notes, keeping track of whether a given document or pile of documents was deemed active or archival was not always made explicit. Moreover, if the user wanted to expand, this might require transferring the contents of one box to another in an iterative process that might require dozens of boxes being moved down in the cabinet, to make room for the new box.

To help prevent document loss, and to keep documents in the order they were filed, a ring system was introduced around 1881, similar to the three-ring binders we now use. The advantages of ringed flat files were substantial, providing random access (like Phaedrus’s 3 x 5 index card system) and minimizing the risk of document loss. With all their advantages, binders did not become the dominant form of storage. For the next fifty years, horizontal files and file books (both bound and glued) were the standard in office organization. Vertical files that resemble the ones we use today were first introduced in 1898. A confluence of circumstances made them useful. Copying technology improved, increasing the number of documents to be filed; the “systematic management movement” required increasing documentation and correspondence; the Dewey Decimal System, introduced in 1876 and used in libraries for organizing books, relied on index cards that were kept in drawers, so the furniture for holding vertical files was already familiar. The invention of the modern typewriter increased the speed at which documents could be prepared, and hence the number of them needing to be filed. The Library Bureau, founded by Melvil Dewey, created a system for filing and organizing documents that consisted of vertical files, guides, labels, folders, and cabinetry and won a gold medal at the 1893 World’s Fair in Chicago.

Vertical files function best when alphabetized. One factor that prevented their earlier invention was that, up through the eighteenth century, the alphabet was not universally known. The historian James Gleick notes, “A literate, book-buying Englishman at the turn of the seventeenth century could live a lifetime without ever encountering a set of data ordered alphabetically.” So alphabetizing files was not the first organizational scheme that came to mind, simply because the average reader could not be expected to know that H came after C in the alphabet. We take it for granted now because all schoolchildren are taught to memorize the alphabet. Moreover, spelling was not regarded as something that could be right or wrong until the eighteenth and nineteenth centuries, so alphabetizing was not practical. The first dictionaries were faced with the puzzling problem of how to arrange the words.

When vertical files became the standard around 1900—followed by their offspring, the hanging file folders invented by Frank D. Jonas in 1941—they offered a number of organizational advantages that probably seem obvious to us now, but they were an innovation hundreds of years in the making:

Papers could be left open and not folded, so their contents could be easily inspected.

Handling and ease of access: Papers filed on edge were easier to handle; papers before them in the sequence didn’t have to be removed first.

Papers that were related to one another could be kept in the same folder, and then subcategorized within the folder (e.g., by date, or alphabetically by topic or author).

Whole folders could be removed from the cabinet for ease of use.

Unlike the bound systems previously in use, documents could be taken out of the system individually, and could be re-sorted or refiled at will (the Phaedrus principle).

When folders became full, their contents could be easily redistributed.

The system was easily expandable.

Transparency: If properly labeled and implemented, the system could be used by anyone encountering it for the first time.

Vertical files don’t solve every problem, of course. There is still the decision to make about how to organize the files and folders, not to mention how to organize drawers within a filing cabinet, and if you have multiple filing cabinets, how to organize them. Running a strictly alphabetical system across a dozen different cabinets is efficient if every folder is sorted by name (as in a doctor’s office), but suppose you’re filing different kinds of things? You may have files for customers and for suppliers, and it would be more effective to separate them into different cabinets.

HSPs typically organize their files by adopting a hierarchical or nested system, in which topic, person, company, or chronology is embedded in another organization scheme. For example, some companies organize their files first geographically by region of the world or country, and then by topic, person, company, or chronology.

How would a nested system look today in a medium-size business? Suppose you run an automotive parts company and you ship to the forty-eight states of the continental United States. For various reasons, you treat the Northeast, Southeast, West Coast, and “Middle” of the country differently. This could be because of differential shipping costs, or different product lines specific to those territories. You might start out with a four-drawer filing cabinet, with each of the drawers labeled for one of the four territories. Within a drawer, you’d have folders for your customers arranged alphabetically by customer surname or company name. As you expand your business, you may eventually need an entire filing cabinet for each territory, with drawer 1 for your alphabetical entries A–F, drawer 2 for G–K, and so on. The nesting hierarchy doesn’t need to stop there. How will you arrange the documents within a customer’s file folder? Perhaps reverse chronologically, with the newest items first.

If you have many pending orders that take some time to fill, you may keep a folder of pending orders in front of each territory’s drawer, filing those pending orders chronologically so that you can quickly see how long the customer who has been waiting the longest has been without their order. There are of course infinite variations to filing systems. Rather than file drawers for territory, with customer folders inside, you could make your top-level file drawers strictly alphabetical and then subdivide within each drawer by region. For example, you’d open up the A file drawer (for customers whose surnames or company names begin with A) and you would have drawer dividers inside, for the territorial regions of Northeast, Southeast, West Coast, and Middle. There is no single rule for determining what the most efficient system will be for a given business. A successful system is one that requires the minimum amount of searching time, and that is transparent to anyone who walks in the room. It will be a system that can be easily described. Again, an efficient system is one in which you’ve exploited affordances by off-loading as many memory functions as possible from your brain into a well-labeled and logically organized collection of external objects.

This can take many forms, limited only by your imagination and ingenuity. If you find you’re often confusing one file folder for another, make the folders different colors to easily distinguish them. A business that depends heavily on telephone or Skype calls, and has clients, colleagues, or suppliers in different time zones, organizes all the materials related to these calls in time-zone order so that it’s easy to see whom to call at which times of day. Lawyers file case material in numbered binders or folders that correspond to statute numbers. Sometimes simple and whimsical ordering is more memorable—a clothing retailer keeps files related to shoes in the bottom drawer, pants one drawer up, shirts and jackets above that, and hats in the top drawer.

Linda describes the particularly robust system that she and her colleagues used at an $8 billion company with 250,000 employees. Documents of different kinds were separated into designated cabinets in the executive offices. One or more cabinets were dedicated to personnel files, others for shareholder information (including annual reports), budgets and expenses for the various units, and correspondence. The correspondence filing was an essential part of the system.

The system for correspondence was that I would keep hard copies of everything in triplicate. One copy of a letter would go in a chronological file, one in a topic file, and one alphabetically by the name of the correspondent. We kept these in three-ring binders, and there would be alphabetical tabs inside, or for a particularly large or often-used section, a custom tab with the name of that section. The outside of the binder was clearly labeled with the contents.

In addition to the hard copies, Linda kept a list of all correspondence, with keywords, in a database program (she used FileMaker, but Excel would work as well). When she needed to locate a particular document, she’d look it up in her computer database by searching for a keyword. That would tell her which three binders the document was in (e.g., chron file for February 1987, topic binder for the Larch project, volume 3, or alpha binder by letter writer’s last name). If the computers were down, or she couldn’t find it in the database, it was nearly always found by browsing through the binders.

The system is remarkably effective, and the time spent maintaining it is more than compensated for by the efficiencies of retrieval. It cleverly exploits the principle of associative memory (the fire truck example from the Introduction, Robert Shapiro’s and Craig Kallman’s annotated contacts lists in Chapter 4), that memory can be accessed through a variety of converging nodes. We don’t always remember everything about an event, but if we can remember one thing (such as the approximate date, or where a given document fell roughly in sequence with respect to other documents, or which person was involved in it), we can find what we’re looking for by using the associative networks in our brains.

Linda’s decision to move correspondence to three-ring binders reflects a fundamental principle of file folder management: Don’t put into a file folder more than will fit, and generally not more than fifty pages. If your file folders contain more than that, experts advise splitting up the contents into subfolders. If you truly need to keep more pages than that in one place, consider moving to a three-ring binder system. The advantage of the binder system is that pages are retained in order—they don’t fall or spill out—and via the Phaedrus principle, they provide random access and can be reordered if necessary.

In addition to these systems, HSPs create systems to automatically divide up paperwork and projects temporally, based on how urgent they are. A small category of “now” items, things that they need to deal with right away, is close by. A second category of “near-term” items is a little farther away, perhaps on the other side of the office or down the hall. A third category of reference or archival papers can be even farther away, maybe on another floor or off-site. Linda adds that anything that needs to be accessed regularly should be put in a special RECURRENCE folder so that it is easy to get to. This might include a delivery log, a spreadsheet updated weekly with sales figures, or staff phone numbers.

—

An essential component of setting up any organizational system in a business environment is to allow for things that fall through the cracks, things that don’t fit neatly into any of your categories—the miscellaneous file or junk drawer, just as you might have at home in the kitchen. If you can’t come up with a logical place for something, it does not represent a failure of cognition or imagination; it reflects the complex, intercorrelated structure of the many objects and artifacts in our lives, the fuzzy boundaries, and the overlapping uses of things. As Linda says, “The miscellaneous folder is progress, not a step backward.” That list of frequent flyer numbers you constantly refer to? Put it in the RECURRENCE folder or a MISCELLANEOUS folder in the front of the drawer. Say you take a tour of a vacant office building across town. You’re not really looking to move, but you want to save the information sheet you received just in case. If your filing system doesn’t have a section for relocating, office lease, or physical plant, it would be wasteful to create one, and if you create a single file folder for this, with a single piece of paper in it, where will you file the folder?

Ed Littlefield (my old boss from Utah International) was a big proponent of creating a STUFF I DON’T KNOW WHERE TO FILE file. He’d check it once a month or so to refresh his memory of what’s in it, and occasionally he’d have accumulated a critical mass of materials with a theme to create a new, separate file for them. One successful scientist (and member of the Royal Society) keeps a series of junk drawer–like files called THINGS I WANT TO READ, PROJECTS I’D LIKE TO START, and MISCELLANEOUS IMPORTANT PAPERS. At home, that little bottle of auto body paint the shop gives you after a collision repair? If you keep a drawer or shelf for automotive supplies, that’s the obvious place to put it, but if you have zero automotive supplies except this little bottle, it doesn’t make sense to create a categorical spot just for one item. Better to put it in a junk drawer with other hard-to-categorize things.

Of course CEOs, Supreme Court judges, and other HSPs don’t have to do all this themselves. They simply ask their executive assistants for the Morrow file or, more often, their EAs bring them a file with instructions about what needs to be done on it and by when. But their EAs need to follow logical systems, and there is often room for improvement in these. Key is that the system be transparent so that if an assistant falls ill, someone else, even without specific training, can find what the CEO needs.

Linda says that in training new EAs, the biggest point is to “remember that you’re organizing people, not files and documents. You need to get to know your boss’s routines—they might tend to put things in piles and you might have to go through their things, or you might have to keep copies for them if they tend to lose things. If you’re working for more than one person—or if your boss interacts with others on a regular basis—it’s a good idea to have a separate folder on your desk for each one of them so that if they drop by unexpectedly, you’ve got the essential information right in front of you.”

Linda’s advice for time management from Chapter 5 is worth revisiting here. “For deadlines, you might need to keep a tickler file. For example, as soon as you learn about any kind of deadline, you need to talk to the boss about it and see how long they think they’ll need. Then you put a tickler on the calendar on the day they’re supposed to start working on it.” Other EAs even put a tickler a few days before that so their boss can start thinking in advance about the project he’ll be working on.

“The things that typically require organization in an office are correspondence, business documents, presentations, things you need for meetings (including information in advance that needs to be reviewed), to do lists, the calendar, contacts, and books and journals,” Linda adds. The first four are usually best organized in files, folders, boxes, or binders. To Do lists, calendars, and contacts are important enough that she recommends the redundancy of keeping them on paper and on the computer. This only works if the number of contacts is small enough to fit on paper—Craig Kallman, the CEO of Atlantic Records, who has 14,000 contacts, has to rely on his computer for the entire list, but he keeps frequently used ones stored in his cell phone. If he kept more than the frequently used ones, it would be too cumbersome and time-consuming to search them.

E-mail filing and sorting is increasingly time-consuming as many HSPs report receiving hundreds of e-mails a day after the spam filter has gotten rid of nuisance mail. Craig Kallman gets about six hundred e-mails every day. If he spent only one minute on each of them, it would take him ten hours a day to get through them. He uses weekends as catch-up time and, when possible, forwards e-mails to others to attend to. But, as with many HSPs, he got into his line of work because he actually loves the work. Delegating a project lessens his involvement with it and the joy he derives from it, not to mention that his expertise and experience are not easily matched by others—the work product benefits from his involvement, but the e-mail stream alone, apart from telephone calls, snail mail, and meetings, is a big time commitment.

How does the White House organize communications? To begin with, the president and vice president do not have piles of documents on their desks, and they do not use e-mail, both for national security reasons. All communications go through the executive secretary, who decides what has priority and what needs to be worked on now. The president and vice president do receive briefing books about particular topics. For example, if the president wants to know everything about a pipeline project in Minnesota, staffers pull together information that has been obtained from phone calls, meetings, e-mails, faxes, letters, and so on and put it into the binder.

Each individual staffer has the autonomy to decide how he will sort or file the papers and communications that they need to do their work—there is no “White House Standard” or anything like that. As long as they can put their hands on a needed communication, they can organize it as they please. This distributed system of organization is a powerful reminder that a top-down approach (like that used in the Fremont GM plant) is not always the most effective.

Mike Kelleher, the director of the Office of Correspondence in the White House during the first Obama administration, says that every week the office received 65,000 paper letters, 500,000 e-mails, 5,000 faxes, and 15,000 phone calls. Spending even just one minute on each of these would require 9,750 person hours, or the equivalent of 244 employees working full-time. Numbers like this require the kind of quick sort and prioritizing system that Ed Littlefield used for his mail when he served on the boards of Wells Fargo, Chrysler, and Utah International. Kelleher’s office employs 49 full-time staff, 25 interns, and a small army of volunteers. Paper letters get sorted into one of more than a hundred bins or cubbyholes, as in the back of a post office, depending on recipient (first lady, first dog, children, VP, cabinet offices, such as HUD, DOT, DOD). With these kinds of numbers, delegation is essential. The White House can’t declare e-mail bankruptcy, as Lawrence Lessig suggested in Chapter 3. As you might imagine, although hundreds of thousands of letters and e-mails are addressed to the president, many of the questions are about policies that come under the jurisdiction or mandate of specific departments within the administration. Questions about health care, economic policy, or veterans’ benefits are referred to their respective departments. Much of the correspondence includes requests for the president to write letters of congratulations for various life events, such as making Eagle Scout, turning one hundred, being married fifty years, and so on, which the White House does try to honor. These end up in the Office of Presidential Correspondence. And again, there are no centralized guidelines for sorting and filing e-mails; staffers use whatever method they see fit—as long as they can produce an e-mail when requested.

Increasingly, people who use e-mail have separate accounts. HSPs might have two business accounts, one for the people they deal with regularly and another that their EAs monitor and sort, in addition to one or two personal accounts. Having separate accounts helps to organize and compartmentalize things, and to restrict interruptions: You may want to turn off all your e-mail accounts during a productivity hour except for the one your assistant and your boss use to reach you right away. An efficient way to deal with multiple accounts is to use a single computer program to collect them all. Most e-mail programs, including Outlook, Apple Mail, Gmail, and Yahoo!, allow you to download into their interface any mail from any provider. The advantage is that if all your different accounts show up in one interface, it’s easier to locate what you’re looking for if you don’t have to log into several accounts to find a specific e-mail or document. Also, categories have fuzzy boundaries. That dinner invitation from a coworker might have shown up in your business account, but you need to coordinate with her husband, who sent his schedule to your personal account.

To reiterate a point from Chapter 3, some people, particularly those with attention deficit disorder, panic when they can’t see all of their files out in the open in front of them. The idea of filing e-mails on a computer is stressful, and so for them, adopting Linda’s system of printing them out is often necessary. Open filing carts and racks exist so that physical files don’t need to be hidden behind a drawer. Other people simply cannot set up or maintain filing systems. The idea of putting things in little compartments is incompatible with their cognitive style, or disengages their creative mode. This relates to the two attentional systems introduced in Chapter 2. Creative people are at their most innovative while engaging the daydreaming mode. Putting things into little compartments requires not only engaging but staying in the central executive mode. Recall that these modes operate in a see-saw relationship—if you’re in one mode, you are not in the other. Consequently, many creative people resist the kinds of geeky, compartmentalized systems outlined here. Compartment-resistant people are found across all walks of life, in a variety of professions, from the law to medicine, and from science to art. In these cases, they either hire assistants to do all their filing for them, or they declare filing bankruptcy and just let piles accumulate.

Jeff Mogil is a very creative and productive behavioral geneticist. His desk is preternaturally clear and the only things ever on top of it are things he’s working on at the moment, arranged in neat piles. His filing system is impeccable. At the other extreme is Roger Shepard, whose office always looked like the aftermath of a natural disaster. Piles of papers had covered his desk for so long that even he didn’t remember what color the surface was. The piles extended to every available space in the office, including a coffee table, the floor, and windowsills. He had barely enough room to walk a path from the door to his desk. But he knew where everything was, thanks to an exquisite temporal and spatial memory. “These piles over here are from five years ago,” he said, “and these are from this month.” When I was a student, walking down the hall from Roger Shepard’s office to Amos Tversky’s was a sobering study in contrasts. Amos had the kind of clean and tidy office that visitors found incredibly intimidating; there was never anything on his desk. Years later, a colleague confided, “Yes, the desk was clean. But you wouldn’t want to look into his drawers and cabinets!” Neat and organized are not necessarily the same thing.

Lew Goldberg, a personality psychologist who is known as the father of the Big Five personality dimensions, devised a system for filing correspondence and for reprints of articles by other scientists. The reprint collection had seventy-two topic categories, and each article was represented on—you guessed it—a 3 x 5 index card. The cards lived in wooden library card catalogues and were cross-referenced by author, title, and topic. He’d look up an item in his catalogue, which would direct him to one of several hundred three-ring binders that took up a wall in his office from floor to ceiling. While his system has worked for him for fifty years, he acknowledges that it isn’t for everyone. His University of Oregon colleague Steve Keele, a pioneer in the study of timing mechanisms in the brain, was a Roger Shepard–like piler and stacker. “Steve was reputed to have had the world’s messiest office, and yet he could always find everything. Piles and piles and piles all over. You could walk in and say, ‘Steve, I know this isn’t your area, but I’ve developed an interest in how humans fixate on a moving visual target.’ And he would say, ‘Oh, I happen to have a student who wrote a paper about this in 1975, and I haven’t graded it yet, but it’s right . . . here.’”

But there are piles and then there are piles. Oftentimes, pile and stack makers are procrastinating on a decision about whether to keep something or throw it out, whether it is relevant or not. It’s important to go through piles on a regular basis to whittle them down, trim them, or re-sort them—not everything in them remains relevant forever.

Recall the system that Microsoft senior research fellow Malcolm Slaney advocated in Chapter 3, to keep everything on your computer. Jason Rentfrow, a scientist at Cambridge University, agrees, adding that “although Gmail doesn’t organize your files, it does allow very easy access and searchability to them. In some ways that—and ‘spotlight’ or ‘find’ on your computer—is like applying a Google Internet search strategy to your own computer. Maybe it’s not even worth bothering having folders anymore—you could have just one folder with everything in it and then use search functions to find anything you want. You can limit to date, content, name, etc.”

From Multitasking to Planning for Failure

In Chapter 5, I wrote about some of the evidence against multitasking as a strategy for getting more work done. But is it realistic to give it up—isn’t it what we have to do in the business world? Stanford professor Clifford Nass assumed, as most people do, that multitaskers were superhumans, capable of doing many things at once with high success: juggling phone calls, e-mails, live conversations, texting. He further assumed that they had an unusually high ability to switch attention from one task to another and that their memories could differentiate the multiple tasks in an orderly way.

We all bet high multitaskers were going to be stars at something. We were absolutely shocked. We lost all our bets. It turns out multitaskers are terrible at every aspect of multitasking. They’re terrible at ignoring irrelevant information; they’re terrible at keeping information in their head nicely and neatly organized; and they’re terrible at switching from one task to another.

We all want to believe that we can do many things at once and that our attention is infinite, but this is a persistent myth. What we really do is shift our attention rapidly from task to task. Two bad things happen as a result: We don’t devote enough attention to any one thing, and we decrease the quality of attention applied to any task. When we do one thing—uni-task—there are beneficial changes in the brain’s daydreaming network and increased connectivity. Among other things, this is believed to be protective against Alzheimer’s disease. Older adults who engaged in five one-hour training sessions on attentional control began to show brain activity patterns that more closely resembled those of younger adults.

You’d think people would realize they’re bad at multitasking and would quit. But a cognitive illusion sets in, fueled in part by a dopamine-adrenaline feedback loop, in which multitaskers think they are doing great. Part of the problem is that workplaces are misguidedly encouraging workers to multitask. Nass notes a number of societal forces that encourage multitasking. Many managers impose rules such as “You must answer e-mail within fifteen minutes” or “You must keep a chat window open,” but this means you’re stopping what you’re doing, fragmenting concentration, Balkanizing the vast resources of your prefrontal cortex, which has been honed over tens of thousands of years of evolution to stay on task. This stay-on-task mode is what gave us the pyramids, mathematics, great cities, literature, art, music, penicillin, and rockets to the moon (and hopefully—soon—jet packs). Those kinds of discoveries cannot be made in fragmented two-minute increments.

It is a testament to our cognitive flexibility and neural plasticity that we are able to go against all this evolution, but at least until the next evolutionary leap in our prefrontal cortex, multitasking leads to not more work but less, not better work but sloppier work. Adding to this, every day we are confronted with new Facebook and Instagram updates, new YouTube videos, Twitter streams, and whatever new technology will replace them in the next year or two. As of this writing, there were thirteen hundred apps for mobile devices being released every day. “Cultural forces, and the expectation that people will respond instantly, and chat and talk and do all these things all at once, means all the pressure is going that way,” Nash says.

The companies that are winning the productivity battle are those that allow their employees productivity hours, naps, a chance for exercise, and a calm, tranquil, orderly environment in which to do their work. If you’re in a stressful environment where you’re asked to produce and produce, you’re unlikely to have any deep insights. There’s a reason Google puts Ping-Pong tables in their headquarters.Safeway, a $4 billion grocery chain in the United States and Canada, has doubled sales in the last fifteen years under the leadership of Steven Burd, who, among other things, encouraged employees to exercise at work, through salary incentives, and installed a full gym at corporate headquarters. Studies have found that productivity goes up when the number of hours per week of work goes down, strongly suggesting that adequate leisure and refueling time pays off for employers and for workers. Overwork—and its companion, sleep deprivation—have been shown to lead to mistakes and errors that take longer to fix than the overtime hours worked. A sixty-hour work week, although 50% longer than a forty-hour work week, reduces productivity by 25%, so it takes two hours of overtime to accomplish one hour of work. A ten-minute nap can be equivalent to an extra hour and a half of sleep at night. And vacations? Ernst & Young found that for each additional ten hours of vacation their employees took, their year-end performance ratings from their supervisors improved by 8%.

It is now well known that some of the most productive companies—Google, Twitter, Lucasfilm, Huffington Post—provide perks such as in-house gyms, gourmet dining rooms, nap rooms, and flexible hours. Google paid for 100,000 free employee massages, and its campus boasts wellness centers and a seven-acre sports complex with basketball, bowling, bocce ball, and roller hockey. The statistical software giant SAS and Toyota distributor JM Family Enterprises feature in-house health care; Atlantic Health System offers on-site acupressure massage; Microsoft’s campus has a spa; SalesForce.com provides free yoga classes; Intuit lets employees spend 10% of their time on any project they’re passionate about; Deloitte encourages employees to donate time to nonprofits for up to six months by offering full benefits and 40% of pay. Giving employees environments like these seems to pay, and it makes sense from a neurobiological standpoint. Sustained concentration and effort is most effective not when fragmented into little pieces by multitasking, but when apportioned into big focused chunks separated by leisure, exercise, or other mentally restorative activities.

Multitasking results from information overload, trying to attend to too many things at once. When the many things we’re attending to require a decision, how much information do we need to make optimal decisions? Optimal complexity theory states that there is an inverted U function for how much information or complexity is optimal.

Too little is no good, but so is too much. In one study, experimenters simulated a military exercise. Players in the simulated game were college students in teams who were either invading or defending a small island country. Players were allowed to control the amounts of information with which to make their decisions—they received a document that read:

The information you are receiving is prepared for you in the same way it would be prepared for real commanders by a staff of intelligence officers. These persons have been instructed to inform you only of important occurrences. You may feel that these men do not give you sufficient information or do not give you adequate detail. On the other hand, you may feel that the information you are receiving is too detailed and you are presented with some unimportant information. You may instruct these intelligence officers to increase or decrease the amount of information they present to you. We would like you to decide this matter for yourself. Please do not consult the other commanders on this issue at any time. We will adjust the information flow according to the majority opinion in your group. Please check your preference in comparison to the immediately preceding game period:

I would prefer to:

receive much more information

receive a little more information

receive about the same amount of information

receive a little less information

receive much less information

In actuality, the players did not have control over the information, but their response was used to study optimal levels of information. They received either two, five, eight, ten, twelve, fifteen, or twenty-five pieces of information within the thirty-minute period of the game.

According to the theory of optimal information, players would perform better with about ten to twelve pieces of information during the course of the game, and the experiment confirmed this. The amount of additional information players requested decreased for those who were already receiving fifteen or twenty-five pieces of information. This leads to the upside-down U-shaped curve.

But although optimum performance came with the ten to twelve pieces of information, players at every level asked for more information, even though this caused them to exceed the optimal amount of information and enter into a condition of information overload, and—when that additional information put them over the ten to twelve pieces of optimal information—caused their performance to decline. What motivated them to ask may have been the belief that a key piece of information lay just around the corner in the next bulletin. But as we now know, additional information carries a cost.

These findings suggest that consumers will have finite limits for how much information they can absorb and process within a given period of time. Let’s call it the load effect. In fact, this has been shown empirically—consumers make poorer choices with more information. This mechanism is similar to the load effect we saw in Chapter 4 that leads to incorrect social judgments. A separate study examined the effects of additional information on the decision to purchase a home.It found the maximum number of parameters that can be processed is around ten. The interesting thing is that the parameters can be either attributes of choice or alternatives. In other words, if you are trying to decide between two houses, you don’t want to be keeping track of more than ten pieces of information about them combined. Or, if you can trim your list to two pieces of information you’re interested in—perhaps square footage and quality of the school district—you can compare ten houses. In the home-buying studies, consumers were given up to twenty-five attributes to keep track of on up to twenty-five different homes. Their decision-making ability began to suffer when either parameter was greater than ten. Above ten, however, it didn’t matter if there were fifteen, twenty, or twenty-five parameters—once the consumer hits information overload, still more information doesn’t significantly affect the already saturated system. This limit of ten is the maximum. The optimal number is closer to five and is consistent with processing limits of the brain’s central executive. This may remind you of the problem with online dating sites mentioned in Chapter 4—that more information is not always better and, in that context, has been found to lead to poorer selectivity and poorer choices as online daters become overwhelmed by irrelevant information and suffer both cognitive overload and decision fatigue.

Another important factor, shown by Duke economist and author Dan Ariely, is that consumers perform better when they have a particular type of internal locus of control, that is, when they can actually control the type of information they receive. In a series of experiments, he demonstrated that if the consumer can choose which parameters to receive information about, as well as how much, they make better decisions. This is primarily because the consumer can choose information that is relevant to them or that they are best able to understand. For example, in shopping for cameras, Consumer X may care mainly about size and price, while Consumer Y may care mainly about resolution (number of pixels) and type of lens. Information that would be distracting or impossible to interpret for one type of consumer causes information overload and interferes with optimal decision processing. Separate research by Kahneman and Tversky shows that people are unable to ignore information that is not relevant to them, so there is a real neural cost of being presented with information you don’t care about and can’t use.

Then the question becomes not one of how many things you can do at once, but how orderly you can make the information environment. There is considerable research into the difference in utility between simple and complex information. Claude Shannon, an electrical engineer who worked at Bell Laboratories, developed information theory in the 1940s. Shannon information theory is among the most important mathematical ideas of the twentieth century; it has profoundly affected computing and telecommunications, and is the basis for the compression of sound, image, and movie files (e.g., MP3, JPEG, and MP4 respectively).

A fundamental problem in telecommunications, signaling, and security is how to transmit a message as briefly as possible, to pack the maximum amount of data into the minimum amount of time or space; this packing is called data compression. Back when telephone service was carried along a single pair of copper wires (what telecommunications nerds call POTS, for “plain old telephone service”), there was a limited amount of call volume that could be carried across the main telephone wires (trunks), and the cost of running new lines was prohibitively expensive. This led to perceptual experiments and the finding that the telephone company didn’t need to transmit the entire frequency range of the human voice for speech to remain intelligible. The so-called telephone band transmitted only 300–3300 hertz, a subset of the full range of human hearing, which spans 20–20,000 hertz, and gave telephone transmissions their characteristic “tinny” sound. It wasn’t hi-fi, but it was intelligible enough for most purposes—it satisficed. But if you’ve ever tried to explain on POTS that you’re talking about the letter f and not the letter s, you’ve experienced the bandwidth limitation, because the acoustical difference of those two letters is entirely within the range that Bell cut out. But in doing so, the telephone company could squeeze several conversations into the space of one, maximizing the efficiency of their network and minimizing hardware costs. Cell phones continue to be band limited for the same reason, to maximize the ability of cell towers to carry multiple conversations. This bandwidth limitation is most apparent if you try to listen to music over the telephone—the low frequencies of the bass and the high frequencies of cymbals are almost completely absent.

Information theory came up in Chapter 1 in discussing the number of simultaneous conversations that a person can follow, and the information processing limits of human attention being estimated at around 120 bits per second. It is a way to quantify the amount of information contained in any transmission, instruction, or sensory stimulus. It can apply to music, speech, paintings, and military orders. The application of information theory generates a number that allows us to compare the amount of information contained in one transmission with that contained in another.

Suppose that you want to convey instructions to someone on how to construct a chessboard. You could say

Make a square and color it white. Now make another square adjacent to that and color it black. Make a square adjacent to that and color it white. Make a square adjacent to that and color it black. Make a square adjacent to that . . .

You could continue this type of instruction until you get to eight squares (completing one row) and then you’d have to instruct your friend to go back to the first square and put a black square just above it, and then proceed square by square to fill the second row, and so on. This is a cumbersome way to convey the instructions, and not very streamlined. Compare that to

Make an 8 x 8 matrix of squares, alternately coloring them black and white.

The first instruction refers to each of the 64 squares individually. In binary arithmetic, 64 pieces of information requires 6 bits of information (the number of bits is the exponent of the equation 2ⁿ = 64. In this example, n = 6 because 2⁶ = 64). But implementing a rule such as “alternately color the squares” requires only 1 bit: A given square is either black or white and so there are two choices. Because 2¹ = 2, we need only 1 bit (1 is the exponent, which determines the amount of information). The two additional facts that the grid is eight squares wide and eight squares long makes for a total of three pieces of information, which take 2 bits. If you want to specify what pieces are on which squares, you’re back up to 6 bits because each bit needs to be specified individually. So an empty chess board can be fully specified in 2 bits, a chessboard with its 32 pieces on it takes 6 bits. There is more information on a loaded chess board than on an empty one, and now we have a way to quantify how much more. Even though Shannon and his colleagues at Bell Labs were working in an analog, precomputer world, they were thinking ahead to when computers would be used for telecommunications. Because computers are based on binary arithmetic, Shannon opted to use the measurement units of digital computers, the bit. But it doesn’t have to be that way—we could talk about all this in regular numbers and leave bits out of it if we wanted to: The instructions to make an empty chess board require a minimum of 3 pieces of information, and the instructions to re-create a loaded chess board require a minimum of 64 pieces of information.

The same logic applies to re-creating photographs and pictures on your computer. When you are looking at a JPEG or other image file on your screen, you’re looking at a re-creation of that file—the image was created right there, on the spot, as soon as you double-clicked on the filename. If you were to look into the actual file, the computer file that your operating system uses to construct the picture, you’d see a string of zeros and ones. No picture, just zeros and ones, the vocabulary of binary arithmetic. In a black-and-white picture, every little dot on the screen—the pixels—can be either black or white, and the zeros and ones are telling your computer whether to make a pixel black or white. Color pictures take more instructions because they are represented by five different possibilities: black, white, red, yellow, and cyan. That’s why color picture files are larger than black-and-white picture files—they contain more information.

Information theory doesn’t tell us how much information we could use to describe things, it tells us the minimum amount of information we need—remember, Shannon was trying to figure out how to cram as much telephone conversation onto a single pair of copper wires as he could, to maximize Ma Bell’s capacity and to minimize the investment in new infrastructure (telephone poles, wires, network switches).

Computer scientists spend a lot of time trying to condense information in this way so that their programs can run more efficiently. Another way of looking at Shannon information theory is to consider two strings of letters 64 characters long:

ababababababababababababababababababababababababababababbababab
qicnlnwmpzoimbpimiqznvposmsoetycqvnzrxnobseicndhrigaldjguuwknhid

Number 1 can be represented with a 2-bit instruction:

64 items, ab, alternate

Number 2, being a random sequence, requires 64 individual instructions (6 bits) because the instruction itself must be exactly the same as the string:

qicnlnwmpzoimbpimiqznvposmsoetycqvnzrxnobseicndhrigaldjguuwknhid

How do we determine whether or not a sequence of numbers or letters is random? The Russian mathematician Andrey Kolmogorov introduced an influential idea about this. He said that a string is random if there is no way to describe it or represent it in an abbreviated form. By his definition, number 1 above is not random because we can come up with a scheme (computer scientists call this an algorithm) to represent it in brief. Number 2 is random because there is no scheme we can come up with apart from simply listing every element, one at a time, as it is in the actual sequence.

Kolmogorov complexity theory encapsulates it this way: Something is random when you cannot explain how to derive a sequence using any fewer than the number of elements in the sequence itself. This definition of complexity meshes with our everyday, lay use of the term. We say that a car is more complex than a bicycle, and surely it takes a far larger set of instructions to build a car than a bicycle.

Information theory can be applied to organizational systems like the file and folder hierarchy on your computer, or to org charts in a company. And, according to Kolmogorov complexity theory, if the org chart can be described by a small number of simple rules, the company is said to be highly structured. Compare these two descriptions. For company one, starting at the top with the CEO, everyone supervises three people, and this extends all the way down through four levels, after which everyone supervises fifty to one hundred people. This model might apply to a telephone, water, electric, or gas company that has four layers of management and then a large number of workers out in the field repairing or installing lines, or reading meters. This could also be a technology company with customer service and technical assistance agents at the bottom level. This org chart could be completely and accurately specified in 2 bits.

A company with a less systematic and regular structure requires as many bits as there are elements because there is no discernible pattern, similar to the random letters in example number 2 above:

The more structured a system is, the less information required to describe it. Conversely, more information is required to describe a disorganized or unstructured system. At the extreme, the most disorganized system possible is a random arrangement of everything—because there is no pattern whatsoever in a random system, each element needs to be described individually. This requires an enormous amount of communication or, as Shannon called it, information. This is a counterintuitive formulation of things that can be hard to get your head around. We’re taught that more information is better. When you’ve got a tough medical decision to make, the more information you obtain from your doctor and from research studies, the better situated you are to make a sound decision. But this all coheres. If a medical condition is well understood and its literature well organized, it doesn’t take much information to convey the treatment. “If you have pneumococcus, take an antibiotic.” That’s easy. But cancer and multiple sclerosis and lupus are much less well understood; there are a lot of ifs, ands, and buts, a lot of exceptions and different factors to balance; hence, they require more information to convey.

The power of information theory is that it can be applied to anything—website structure, legal and ethical domains, even directions you give to someone trying to find your house. Recall the discussion of flat versus vertical organization as applied to websites or computer file hierarchies. Shannon information theory can be applied to quantify the level of structure or information they contain (here, we’re talking about the information in the hierarchical structure itself, as distinct from the information content contained on the website).

Or take legal systems. They contain a great number of redundancies, exceptions, and specifics because they are attempting to cover all possible cases. Nearly all civilized societies have laws against rape, murder, robbery, extortion, maiming, assault, battery, and slander, for example. The codes take up a great deal of space as they are encoded in books and on computers. From an information theoretic standpoint, these could all be minimized with a short algorithm: Don’t do anything to someone who would not want it done to them (this is essentially the Golden Rule).

Similarly, compare two ways of a friend giving you directions to his house:

Take Highway 40 East to Highway 158 East, then turn left on Main Street, right on Basil Avenue, left on South Lake Road, continue straight on North Lake Road, a slight right on Main Street (not the same Main Street you were on before) to a quick left on Big Falls Road, a right on 8th Trail until you get to number 66 on your right, just before the park.
Take Highway 40 East and follow the signs to Big Falls Regional Park—my house is just before the entrance.

Algorithm 2 has less Kolmogorov complexity. Notice that it accomplishes this in part by following a dictum of Chapter 2: Off-load as much information as possible to the external world—here, the road signs that already exist.

Given an org chart, then, one can compute the amount of information contained in it and use that as a measure of the organization’s complexity, or by using the reciprocal of the measure, one can calculate the degree of structure (or organization) within a business, military unit, or any other work or social unit. Here, structure is high when complexity is low—this is equivalent to saying that the Shannon information content is low. Again, this may seem counterintuitive, but a business has a greater degree of structural organization if its org chart can be described in a simple rule containing few words, and there are no exceptions to the rule.

Whether the degree of structure of a company predicts efficiency, profitability, or job satisfaction remains an empirical question, one that has not been investigated. On the one hand, individuals clearly differ in their ability to supervise others, and so, naturally, some bosses will have more employees simply because they are adept at handling more. Individuals also differ widely in their skills, and a nimble and efficient organization should allow employees to use their strengths for the good of the company. This can lead to ad hoc reporting structures and special arrangements. Even the best-planned hierarchies become circumvented for the larger good of the company.

One such case occurred in Linda’s company. A data analyst had a set of skills unavailable elsewhere in the company, and his supervisor arranged for him to take on a special project, for which he reported to a manager two levels above him and in a different vertical column on the company org chart. This ad hoc arrangement would require 2 extra bits to represent in the company structure, but the arrangement was profitable for the corporation, ultimately allowing them to introduce a new product that greatly increased revenues. Incentives figure into such arrangements. Those increased revenues were accrued to the account of the divisional manager where the work was completed, not to the account of the divisional manager who lent out the skilled data analyst. As in many large corporations, the organizational structure and incentive schemes put too much emphasis on profits accorded to a district or division, and not enough emphasis on the shared goals of the entire company. As I wrote earlier, Booz Allen Hamilton spent several months interviewing employees at Linda’s company to better understand their skills, the kinds of problems they were working on, and their job descriptions. After they reported their recommendations, the incentive structure and corporate vision statement were reworked to include cross-division teamwork. It seems obvious to us, but in a company with 250,000 employees, big ideas can get lost.

Ad hoc reporting arrangements can facilitate collaboration, but there are downsides. Too many exceptions to the straightforward org chart become difficult to follow; employees with too many bosses can be difficult to manage, and their hours can be difficult to track. In general, a business that is highly structured is more resilient under stress. If a manager leaves her job, smooth, continuous operation of the company is achieved if the replacement can step into a well-defined job with clear reporting structure and fewer ad hoc arrangements. Clearly defined roles promote continuity and efficiency, and give upper management more flexibility in reassigning managers and workers. It is also easier to keep track of and remember who is who in a highly systematic, well-structured organization because, by definition, it can be communicated in very few words, such as “Every division manager has four districts.”

In setting up any kind of structured system—the way file folders are arranged within drawers, or files on a computer, or employees within a company, a successful system is one that requires the minimum amount of searching time and is transparent to anyone who walks in the room. It will be easily described. This reduces the Shannon information content and reduces Kolmogorov complexity. Work flow charts can be similarly analyzed using the same approach:

—

We can make our business worlds more orderly by paying close attention to information flows and escaping the illusions of multitasking. But is that enough? Chapter 3 introduced the idea of planning for failure, a strategy session in which you try to figure out anything that could possibly go wrong and how it would go wrong, and then put systems in place to either prevent it or recover from it. At home, the kinds of things that go wrong typically cause inconvenience for us and our families. At work, things that go wrong can affect thousands of people and cost serious money. The planning-for-failure procedure is to think about everything that could go wrong. Then, come up with a way to minimize the likelihood of those things happening, and a backup or fail-safe plan in the event that they do happen. Leaving your keys by the front door minimizes the likelihood of your forgetting to take them with you. Hiding a key in the garden allows you to gracefully recover (without breaking a window or calling a locksmith) if you do forget the key. What does planning for failure mean in the business world?

One simple thing that can go wrong at the office, just as at home, is to miss a deadline or important appointment. Electronic calendar reminders scheduled to appear on your computer and on your cell phone are effective, and a phone call or in-person interruption from a colleague or EA is an effective backup for the most important appointments.

To ensure you’ll be able to locate important documents, former Google VP and CIO Douglas Merrill recommends making search a forethought rather than an afterthought. That is, file things, either electronic or physical, in a way that will allow you to quickly retrieve them. Ask yourself, “Where will I look for this when I need it?” or “How can I tag or label this item so that I’ll be able to find it?” Related is preparing for the possibility that you’ll show up at a meeting with no idea what it’s about or why you’re there. “I make sure all my calendar entries contain some context,” Merrill says. “When my assistant adds a new meeting to my calendar, she types notes directly into the appointment entry, telling me things like the topic and goals of the meeting and who the other participants will be. If I don’t know one of the attendees, my assistant may add a few notes about that person, such as job title, what he or she is contributing to the project at hand, and so on.”

Why do psychiatrists work a fifty-minute hour? They use that extra ten minutes to write down what happened. Rather than scheduling meetings back-to-back, experts advise giving yourself ten minutes to write down what happened, to make notes about what needs to be done, and other comments that will orient you to this project when you next start to work on it. And to give yourself ten minutes before a meeting to review what is going to happen there. Because attention switching is metabolically costly, it’s good neural hygiene for your brain to give it time to switch into the mind-set of your next meeting gradually and in a relaxed way before the meeting starts. When interrupted during a project, experts recommend making notes about where you left off so you can get back into the project more quickly later.

This is good advice, but there is an underlying point. Thinking ahead about what could go wrong, looking at the future and foreseeing threats—this is what an organized business mind can do, should do, must do.

Some threats are bigger than others. The smooth functioning of businesses is threatened by computer failure to a degree that many people have never taken seriously. Hard drives crash, and Internet servers go down (either your own or ones that you rely on from third parties). Many customers at restaurants, in taxicabs, or at clothing stores have experienced an unanticipated “failure to connect” of the electronic credit card machine. Taxi drivers in big cities, who can’t afford to lose a payment due to a faulty connection, often carry around an old-style plastic credit card press, a machine that makes an impression of the credit card numbers on a specially designed form from the credit card company. This illustrates planning for failure par excellence, the result of effective planning for failure. The optimists who think nothing will go wrong experience the lost sales and financial setbacks, and the realists who are prepared for things to go wrong retain revenue in spite of technological hiccups.

More serious is the loss of important records and data or, just as bad, an inability to open files that have become corrupted or outdated. There are two hazards here to worry about—disk failure and file format obsolescence. A rigorous planning-for-failure approach to your data requires thinking about the ways you can lose access to that data, and how you can set systems in place to prevent or at least minimize that loss.

As of this writing, 90% of the world’s data is stored on magnetic disks. These are vulnerable to magnetic field variations just like recording tape—prolonged exposure to magnets (such as found in loudspeakers) or radiation can corrupt the data, and temperature changes as small as 15 degrees Centigrade can double failure rates. Copying and backing up files can also introduce transcription errors—for many types of files, one wrong bit in the header can render the file completely unreadable. Beyond that, hard disks, thumb drives, CDs, and other storage media eventually all fail. (An old disk drive that has been sitting on a shelf, even one in a dust-proof, magnetically shielded case, can stop working after a few years if its bearings freeze.) And just having multiple copies of your files on the same hard drive doesn’t protect you if the hard drive fails. Probabilities that a hard drive will fail within five years reach 50% or more. A study by Microsoft engineers found that 25% of all servers suffer a disk failure within two years. These are reasons to back up your data—many IT experts have an aphorism, “It’s not if your hard drive will fail—it’s when.” For any business, having unimpeded access to all its files, both current and historical, is important. For a publicly traded company or government agency, it’s essential for legal and regulatory reasons. USB flash drives and solid state drives are more expensive than magnetic drives, but are more robustly sensitive to changes in the environment.

The recommended solution is to back up your files to at least two different hard disks, and check those hard disks regularly—once every three months is a good rule of thumb—to be sure they’re still functioning. Many companies use rolling backups and maintain an archive of files one day old, one week old, one month old, two months old, and so on, and they keep these in duplicate, triplicate, or more. If one fails, you’ve got a backup of the backup. It’s unlikely that all would fail at the same time. The only way that is likely to happen is if there’s a fire, flood, nuclear explosion, or other event that wipes out everything in a particular location. For this reason, governmental organizations and large companies spread the risk by keeping their hard disk backups in different locations. For a small business without vast resources, this is still within reach. If you’ve got a customer or close colleague (or even relative) in another city, you can hook up a remotely accessible backup disk at their home or place of business and schedule automatic backups and restores from your home base.

Backing up to the cloud, that is, to remote servers accessible via the Internet, is another way to maintain copies of files. It’s also efficient for primary access when you use several different devices and want to keep them synchronized. Suppose you have a laptop, home computer, office computer, smartphone, and tablet. How do you keep track of where certain files are, or which computer has the most recent version of the Pensky file? Tech writer Paul Boutin sums up the dispersion that is indicative of a modern digital life: “Some photos are on your smartphone. Others sit on your home computer. Your digital work documents, favorite Web clippings and notes from meetings? Scattered like confetti after New Year’s Eve.” The solution is to synchronize all your devices, but few of us actually take the time to do that. After a long day at work, it’s hard to motivate yourself to plug your phone into your computer, let alone to set up the synchronization program to function correctly, in automatic mode, in the first place. Cloud storage mitigates this problem to a large degree—you simply set all your devices to automatically upload and synchronize their files to a digital storage locker maintained by a third-party company, and when you’re looking for that photo of your dog wearing sunglasses, or the shopping list you made on the subway on your way to work, you have to search only one place and it comes to you instantly (provided you have an Internet connection).

Perry R. Cook, professor of computer science at Princeton University, points out that there are pluses and minuses to backing up your files to the cloud. The advantage is that someone else is taking responsibility for maintaining the hardware, backing up those big servers (they’re not just keeping one copy of your tax documents and family photos but multiple copies), and keeping everything running smoothly. On the other hand, Perry notes, “one of the issues of cloud storage is accessibility. MegaUpload, in addition to allowing people to store backup files, became a huge piracy site. When it was shut down by the U.S. Department of Justice in 2012, no one could get their files. All their clients, including professional photographers and moviemakers, lost everything. It’s like making a deal with your neighbor to store your lawn mower, and he gets raided by the Feds for growing pot and everything is seized. You lose your lawn mower. With MegaUpload, courts wouldn’t open it up even long enough to let legitimate users get their stuff. With the cloud, the companies could go under or be subject to regulatory or judicial restrictions and you’re out of luck. The moral: keep your own data.”

Back to planning for failure: You’ve backed up your files, but what if you upgrade your system and they won’t open?! Cook advises having a plan for migrating files.

File migration refers to the process of making readable files that are no longer readable due to system, software, and hardware updates—basically, many computer file formats become obsolete. This follows from the rapid developments in the tech sector. Both hardware and software manufacturers have an incentive to create faster and more powerful products. These cause incompatibilities with old systems. It is likely that you, or someone you know, has experienced this. Your old computer stops working and when you go to get it repaired, the technician tells you that he can’t fix it because the parts are no longer available—motherboards, logic boards, whatever. He suggests you buy a new computer. You do, and when you get it home, you realize that it comes with a wholly new and unfamiliar operating system. The new operating system won’t open files from your old computer, and you can’t simply reinstall your old operating system because the computer’s hardware won’t run it. Now you have a hard disk full of files that you can’t open—your tax returns, family photos, correspondence, projects from work—all of them unreadable.

To be proactive about file migration, you keep track of all the different file types you have on your computer. When a new operating system comes out, or a new version of a software application you use, don’t just blindly hit the upgrade-now button on your computer screen. It’s necessary to test your old files to see if they’ll open before you commit to the new system. You don’t need to test all of them, just a sample of each kind by trying to open them. (Do this on a different machine or external hard drive than the one you’re currently using.) There are three typical possibilities about what could happen.

The files will open without any issue and your new software uses the same file format as the old software.
The files open slowly because they require a conversion to the new software. The new software has a different format now (as when Microsoft moved from .doc to .docx format). If this happens, you need to begin migrating, that is, translating your files to the new format.
The files don’t open at all. You need to wait until a translation program becomes available (this happens some of the time) or figure out a way to save your old files in a different format that will render them readable on the new computer (as when you save a Word document as .rtf, a more highly readable file type, although some page formatting can be lost).

Who needs to worry about migration? It’s potentially a legal issue for corporations, publicly traded companies, research labs, and journalists to be able to put their hands on archival materials. For the rest of us who use our computers as digital archives of our lives, migration is just ordinary everyday planning-for-failure thinking.

Perry Cook counsels businesses and conscientious individuals to keep legacy (old) machines around, or to be sure that you have access to them, as well as any printers they worked with in their day (old printers don’t typically work with modern computers). If there is no way to translate an old file to a currently readable format, you can always print it out. “It’s a very retro, caveman approach to modern technology,” Perry says. “But it works. So if you want to keep that e-mail from Aunt Bertha, print it out.” Perry advises against throwing away an old computer when you upgrade but instead, making a bootable disk image and checking on your old machine every three to six months. There are still companies who kept vital information on 9 millimeter tape back in the era of big mainframes, or on 5-1/2-inch floppy disks in the first era of PCs, who never migrated. There are services for file migration in many major cities, but they’re expensive. These media are so old now that few machines exist that can read them, and the process requires several steps of translating the file up through several different formats. Librarians and IT departments in large corporations recommend having one or more full-time people just for file migration (independent of anyone who handles backups, which are a different matter).

Finally, Perry says, “It helps if your file formats are open source. Why? Microsoft and Adobe files are very fragile—if one bit is off, the computer can’t open the file at all. If it’s plain text (.txt files), almost any program can open and inspect it, and if there’s an error, it’s just one character. If it’s Open Source, somewhere there is a computer geek who can figure out how to open the file for you.”

Another aspect of planning for failure, especially for business travelers, is that we often find ourselves stuck on a plane, in an airport, or in a hotel room for longer than we thought we’d be. There’s not much we can do to prevent these un-anticipatable events, but as part of a planning-for-failure approach, we can control how we cope with it. HSPs might assemble a pouch with everything they need to make a mobile office:

extra chargers for phone and computer
USB key
pen
pencil, eraser
mini stapler
notepad
Post-its
extra computer connector cables that you often use

The key to this working is to keep it stocked up. Don’t raid it if you’re home—it is sacrosanct! Along the same lines, seasoned business travelers assemble a little emergency food pack: nuts, dried fruit, PowerBar. And they assemble a toilet kit with duplicates so that they’re not having to pack stuff from the bathroom in the rush and haze right before a trip—that’s how things get forgotten.

Planning for failure is a necessary way of thinking in the age of information overload. It is what CEOs, COOs, and their attorneys do, as well as military officers and strategists, and public officials. Performing artists do it, too. Musicians carry extra guitar strings, reeds, electronic connectors—anything that might break in the middle of a performance and bring the show to a screeching halt. All these people spend a great deal of time trying to think about all the ways that something might go wrong, how they might prevent it, and how they would recover if it does. Humans are the only species that possess this capacity. As described in Chapter 5, no other animal plans for the future or strategizes about how to act in situations that haven’t yet occurred. This kind of planning is not important just for being personally organized but is essential to successful business. It comes down to locus of control: An effective organization is one that takes steps to manage its own future rather than allowing external forces—human, environmental, or otherwise—to dictate its course.