Note: While reading a book whenever I come across something interesting, I highlight it on my Kindle. Later I turn those highlights into a blogpost. It is not a complete summary of the book. These are my notes which I intend to go back to later. Let’s start!

  • The social psychologist Eliot Aronson observed that people are not rational beings so much as rationalizing beings. We want explanations. We want the world around us to make sense. We may not know exactly why Lego ran into a brick wall, or why WH Smith fell on hard times, or why Wal-Mart has done so well, but we want to feel that we know what happened. We want the comfort of a plausible explanation, so we say that a company strayed or drifted. Or take the stock market, whose daily fluctuations, edging higher one day and a bit lower the next, resemble Brownian motion, the jittery movement of pollen particles in water or of gas molecules bouncing off one another. It’s not very satisfying to say that today’s stock market movement is explained by random forces. Tune in to CNBC and listen to the pundits as they watch the ticker, and you’ll hear them explain, “The Dow is up slightly as investors gain confidence from rising factory orders,” or, “The Dow is off by a percentage point as investors take profits,” or, “The Dow is a bit higher as investors shrug off worries about the Fed’s next move on interest rates.” They have to say something. Maria Bartiromo can’t exactly look into the camera and say that the Dow is down half a percent today because of random Brownian motion.

  • During World War I, an American psychologist named Edward Thorndike was conducting research into the ways that superiors rate their subordinates. In one study, he asked army officers to rate their soldiers on a variety of features: intelligence, physique, leadership, character, and so on. He was struck by the results. Some men were thought to be “superior soldiers” and were rated highly at just about everything, while others were thought to be subpar across the board. It was as if officers figured that a soldier who was handsome and had good posture should also be able to shoot straight, polish his shoes well, and play the harmonica, too. Thorndike called it the Halo Effect.

  • The Halo Effect is a way for the mind to create and maintain a coherent and consistent picture, to reduce cognitive dissonance. The Halo Effect is not just a way to reduce cognitive dissonance. It’s also a heuristic, a sort of rule of thumb that people use to make guesses about things that are hard to assess directly. We tend to grasp information that is relevant, tangible, and appears to be objective, and then make attributions about other features that are more vague or ambiguous. For example, we may not know if a new product is good, but if it comes from a well-known company with an excellent reputation, we might reasonably infer it should be of good quality. That’s what brand building is about: creating Halos so that consumers are more likely to think favorably of a product or service. Or take a well-documented setting for the Halo Effect — the job interview. What’s the most relevant and tangible information we first have about job candidates? Probably the school where they earned their degree, their grade-point average, and what honors they received. With this information clearly in mind — relevant, tangible, and seemingly objective — interviewers tend to shade their evaluations about other things that are less tangible, such as a candidate’s personal manner or the quality of answers to general questions. A strong record from an excellent school? The job candidate often appears to be a little brighter, with smarter answers and greater potential for success. A modest record from an unheralded local school? The very same answers may sound a little less intelligent, the same appearance a bit less impressive. Which is exactly what Thorndike found in his study about army officers and their soldiers all those years ago. Now consider companies. What’s the most relevant and tangible information we often have about a company? Financial performance, of course. Whether the company is profitable. Whether sales are growing. Whether the price of its stock is on the rise. Financial performance looks to be accurate and objective. Numbers don’t lie, we like to say — which is why Enron, Tyco, and a handful of other recent scandals shake our confidence so deeply. We routinely trust financial performance figures. And it’s natural that on the basis of this performance data, people make attributions about other things that are less tangible and objective. All of which helps explain what we saw at Cisco and ABB. As long as Cisco was growing and profitable and setting records for its share price, managers and journalists and professors inferred that it had a wonderful ability to listen to its customers, a cohesive corporate culture, and a brilliant strategy. And when the bubble burst, observers were quick to make the opposite attribution. It all made sense. It told a coherent story. Same for ABB, where rising sales and profits led to favorable evaluations of its organization structure, its risk-taking culture, and most clearly the man at the top — and then to unfavorable evaluations when performance fell. Journalistic hyperbole? To some extent, sure. But more importantly, a natural human tendency to make attributions based on cues that we think are reliable. Halos in the Business World Financial information is far from the only data on which people make attributions.

  • Picture a group where people express their views vigorously and passionately, even arguing with one another. If the group performs well, participants might reasonably look back and say that open and forthright expressions of opinion were a key reason for success. They’ll say: We were honest, we didn’t hold back — and that’s why we did so well! We had a good process! But what if the group’s performance turned out to be poor? Now people might recall things differently. We argued and fought. We were dysfunctional. Next time we should follow a respectful and disciplined process. But now imagine a group where people are calm, polite, and respectful of one another. They speak quietly and in turn. If the group does well, participants might look back and credit their courteous and cooperative nature. We respected one another. We didn’t fight. We had a good process! But if the same group’s performance was poor, people might say: We were too polite. We censored ourselves. Next time, we should be more direct and open, not so concerned about one another’s feelings. The fact is, a wide variety of behaviors can lead to good decisions. There’s no precise way to engineer an “optimal” discussion process. We may try to avoid extremes, sure, but between those extremes is a wide range of behavior that might be conducive to success. And because we really don’t know what makes an optimal decision process, we tend to make attributions based on other things that are relevant and seemingly objective — namely, what we’re told about performance outcomes.  

  • For all the books written about leadership, most people don’t recognize good leadership when they see it unless they also have clues about company performance from other things that can be assessed more clearly — namely, financial performance. And once they have evidence that a company is performing well, they confidently make attributions about a company’s leadership, as well as its culture, its customer focus, and the quality of its people.

  • Companies that listen to their customers, that design products and services to meet customer needs, and that work hard to satisfy their customers should, all else equal, usually outperform companies that don’t. But you don’t discover these companies by asking: Are you customer oriented? All you’ll get is the self-reporting Halo, cued by company performance. If you want to measure customer orientation, you have to rely on measures that are independent of performance. The same holds for corporate culture. It stands to reason that when employees share common values and don’t need to be told what to do, decisions are made more quickly and people collaborate more easily. But you don’t measure the strength or the fit or the adaptability of a corporate culture just by asking people who already have an opinion about company performance. Instead, you have to look for specific actions or policies or behaviors that are not shaped by perceptions of performance.

  • Inferring causality from correlation trips up many studies about business. Take something as basic as, say, the relationship between employee satisfaction and company performance. It’s logical to think that having satisfied employees ought to lead to high performance. After all, satisfied employees might be willing to work harder and longer, and might care more about keeping their customers happy. It sounds right. We know not to measure employee satisfaction simply by asking employees, “Are you satisfied?” since the answers will likely be colored by the Halo Effect. But suppose we look at a measure that is not tainted by Halos — say, the rate of employee turnover — and we find a high correlation with performance. Now the challenge is to untangle the direction of causality. Does lower employee turnover lead to higher company performance? Perhaps, since a company with a stable workforce might be able to provide more dependable customer service, spend less on hiring and training, and so forth. Or does higher company performance lead to lower employee turnover? That could be true as well, since a profitable and growing company might offer a more stimulating and rewarding environment as well as greater opportunities for advancement. Knowing which leads to which is critical if managers want to know what to do — how much they should invest in greater levels of satisfaction versus other objectives.  
  • Or suppose we want to capture the impact of executive education on company performance. As a first step, we have to avoid the Halo Effect by measuring executive education in ways that are not shaped by perceptions of performance, like total spending on education, number of days of education per employee, range of educational opportunities available, and so forth. Suppose we find that companies that spend more on executive education also tend to be high performers. How do we interpret the results? Can we say that investing in executive education leads to high performance? No, because it could be that profitable companies have the funds to afford a greater investment in education. As long as we gather data at one point in time — cross-sectionally — we won’t know. Psychologist Edwin Locke made the point emphatically: “While the method of correlation may be useful for the purposes of suggesting causal hypotheses, it is not a method of scientific proof. A correlation, by itself, explains nothing.”  
  • One way to improve our ability to explain causality is to gather data at different points of time, so that the impact of one variable on some subsequent outcome can be more clearly isolated. This approach, called a longitudinal design, is more time-consuming and expensive to carry out, but it stands a better chance of avoiding mistaken inferences from simple correlation. That way we could, for example, tell whether the advice given by a consulting company in one time period led to better performance in subsequent periods.  
  • One recent study, by Benjamin Schneider and colleagues at the University of Maryland, used a longitudinal design to examine the question of employee satisfaction and company performance to try to find out which one causes which. They gathered data over several years so they could watch both changes in satisfaction and changes in company performance. Their conclusion? Financial performance, measured by return on assets and earnings per share, has a more powerful effect on employee satisfaction than the reverse. It seems that being on a winning team is a stronger cause of employee satisfaction; satisfied employees don’t have as much of an effect on company performance. How were Schneider and his colleagues able to break the logjam and answer the question of which leads to which? By gathering data over time. It is far easier, of course, to rely on data from a single point in time and make an assumption about the direction of causality. But that way delusions lie.  

  • So many things contribute to company performance that it’s awfully hard to know exactly how much is due to one particular factor versus another. Even if we try to control for many things outside the company, like environmental turbulence and competitive intensity and industry and firm size, we can’t control for all the many different things that go on inside the company.

  • The problem of untangling rival alternatives is rarely given much attention. The article by Huselid and Becker is exceptional; most articles either touch on the subject briefly in a small section toward the end, when they discuss limitations of their research, or else they ignore it altogether. Why? Drawing attention to the limitations of the findings detracts from the power of the desired story — which is to demonstrate the importance of a given variable on company performance. Many academic researchers want to show strong conclusions about cause and effect. They want to demonstrate that leadership is hugely important, or that human resource management has a major impact on company success, or that strong customer orientation significantly raises performance. Readers, too, prefer clear stories. We don’t really want to hear about partial causation or incremental effects or threats to validity. And there’s a further problem compounding all of this. As Harvard psychologist Stephen Pinker observed, university departments don’t always represent meaningful divisions of knowledge. Some of the most important questions come at the intersections among fields, such as the study of decision making, which rests at the convergence of cognitive psychology, sociology, and economics. The same holds for business performance, which is shaped by many different factors. Yet researchers often belong to one department or another. If you’re a professor of marketing, you care a lot about market orientation and customer focus, and there’s a natural tendency to want to demonstrate the importance of your specialty. Same for professors of human resource management or business ethics. There’s no real incentive to explore correlations with other factors — better to leave them safely out of view. As for the journals that publish these articles, many use a “double-blind” review process where the reviewers don’t know the name of the authors, and the authors never learn the names of the reviewers, in order to preserve impartiality. But almost everyone who reviews an article for the Journal of Human Resource Management believes in the importance of HRM — it’s their field, it’s their department, and it’s their specialty. Of course they look favorably upon articles that show the importance of human resource management. Ditto for the Journal of Business Ethics — research that shows how investments in CSR boost firm performance is welcome news, a wonderful validation of their field. And who can blame the Journal of Marketing for publishing a study that demonstrates the importance of market orientation on firm performance? It would take an unusual amount of self-discipline to point out that market orientation is correlated with so many other things that its impact is small. Of course, the tendency for exaggeration isn’t found just in academic articles — it’s also found in business press articles like the one that told us how much a Great Place to Work contributed to performance. The bigger the claim, the larger the headline — and the greater the temptation to overlook rival explanations.

  • Competitive advantage is hard to sustain. Sure, if you want to, you can look back over seventy years of business history and pick out a handful of companies that have endured, but that’s selection based on outcomes. On the whole, if we look at the full population of companies over time, there’s a strong tendency for extreme performance in one time period to be followed by less extreme performance in the next. To revise a well-known phrase, Nothing recedes like success. Suggesting that companies can follow a blueprint to lasting success may be appealing, but it’s not supported by the evidence.

  • Good to Great documented what was written and said about companies that had made the leap versus those that had not — which is completely different. At the start of his book, Collins urges his readers to be honest, to “confront the brutal facts.” Well, here’s a brutal fact we may wish to consider: If you start by selecting companies based on outcome, and then gather data by conducting retrospective interviews and collecting articles from the business press, you’re not likely to discover what led some companies to become Great. You’ll mainly catch the glow from the Halo Effect.

  • Stories will always be with us. They’re an important part of life, providing coherent explanations of complex events. They help people act by conferring a moral dimension to events. By offering what Stephen Jay Gould called “a rare beacon of hope,” stories can inspire people to action. Some of the people I’ve enjoyed quoting, like Gould and Richard Feynman, were scientists, professors at universities. They could take their time and refine their research, running additional experiments or gathering more data until they were satisfied with their answers. Managers, on the other hand, have to act. Endless debate about alternative courses of action can’t be conducive to success when, as we know, performance is relative and companies that stand still are rarely successful. Another chief executive, Harry S. Truman, famously complained that he wanted one-armed advisers — he was tired of advisers who continually said, “on one hand…and on the other hand…” Executives have to act, which may be one reason the image of the Hedgehog, focusing on one thing rather than having to know many things, is so appealing.  

  • According to Michael Porter of Harvard Business School, company performance is driven by two things: strategy and execution. Strategy is about performing different activities from those of rival companies, or performing similar activities in different ways. A strategy is not a goal or an objective or a target. It’s not a vision or mission or a statement of purpose. It’s about being different from rivals in some important way. In turn, execution is all about carrying out those choices. It refers to the way that people, working together in an organizational setting, mobilize resources to deliver on the strategy. Building high-quality products, providing customer service, managing working capital, developing and deploying talent — these usually aren’t matters of strategy because almost every company wants to do these things well. Rather, these things are the stuff of day-to-day management. They’re all about effective operations. Explaining high performance in terms of just two things — strategy and execution — may at first raise our hopes. Just two items rather than some lengthy list! Surely managers ought to be able to get two things right!

  • But a closer look shows that both are fraught with uncertainty, and makes plain why all the talk about blueprints and guarantees and immutable laws is a delusion. Together these three factors — uncertain customer demand, unpredictable competitors, and changing technology — and it becomes clear why strategic choice is inherently risky. And nowhere have the risks been higher than in high-technology industries.

  • James March of Stanford and Zur Shapira of New York University explained, “Post hoc reconstruction permits history to be told in such a way that ‘chance,’ either in the sense of genuinely probabilistic phenomena or in the sense of unexplained variation, is minimized as an explanation.” But chance does play a role, and the difference between a brilliant visionary and a foolish gambler is usually inferred after the fact, an attribution based on outcomes. The fact is, strategic choices always involve risk. The task of strategic leadership is to gather appropriate information and evaluate it thoughtfully, then make choices that, while risky, provide the best chances for success in a competitive industry setting.