![]()
PART ONE
How Bad Practices Prevail
![]()
CHAPTER 1
Weâre Suckers for Success
At the end of World War II, in 1945, the Japanese economy was devastated. Industrial areas were destroyed in almost all major cities, transportation networks were severely damaged, a quarter of Japanâs national wealth was annihilated, and more than two million people in its workforce were killed. The countryâs population suffered for a decade.
Yet, by the 1960s, Japanâs economy started growing rapidly and unabatedly. In what became known as Japanâs âeconomic miracle,â it quickly rose to become the second-largest economy in the world. Where, in 1965, Japanâs gross domestic product stood at about $90 billion, by 1980, it had increased to over $1 trillion. Much of this was achieved through exports. Japanese car manufacturers, for example, were outcompeting the traditional Western brands in their home markets. Japanese carsâsuch as Toyota, Honda, and Mitsubishiâwere more affordable but also of significantly higher quality. American car manufacturers, among others, were eager to learn how the Japanese companies managed to achieve this.
A large part of their success could be ascribed to a set of new management practices referred to as total quality management (TQM).
In the 1950s, Americans W. Edwards Deming and Joseph Juran had gone to Japan to lecture on quality management, and their views concurred with those of local management professor Kaoru Ishikawa. TQMâas the practices they advocated became widely knownâprescribed how all aspects of an organization should be committed to delivering quality: not only top to bottom within the organization, but also start to finish in the entire product life cycle (see the sidebar âThe Core Components of TQMâ).
The Core Components of TQM
Worker involvement
Whereas, in most companies, management and engineers organized and controlled the production process, while line employees were merely expected to execute their prescribed tasks, TQM advocated much greater involvement of ordinary workers in the organization of the process.
Top managementâs responsibility
Senior managementâs foremost task was to create an environment in which workers would feel safe and able to contribute and improve the functioning of the organization. Managers should remove all organizational systems that create fear, such as punishment for poor performance or appraisal systems that explicitly ranked employees.
Intrinsic motivation
TQM assumed that employees have an intrinsic motivation to perform their tasks well, that they enjoy the feeling of accomplishment, and that they are motivated to contribute to the prosperity of the organization. The assumption resulted, for instance, in relying on self-managed work teams as a form of empowerment.
Cross-functional solutions
Since organizations are systems of interdependent parts, TQM advocated for ad hoc cross-functional teams that could identify and solve quality problems. Other permanent cross-functional teams were then responsible for the improvement of processes over the long term.
Evidence-driven decisions
Front-line employees are the ones who should analyze and control processes. TQM also advocated for the use of systematic data collection, statistics, and testing solutions by experiment to solve problems and generate process improvements, rather than relying on impressions and conjectures.
Continuous improvement
On top of that, organizations should constantly collect and analyze data in order to improve the accuracy of conclusions, aid consensus and decision making, and allow for predictions based on past data. In the TQM view, quality improvement should be treated as a never-ending quest, involving all employees in the organization.
Customer and supplier relationships
Finally, creating quality was thought to transcend the boundaries of the organization, and should be determined by the requirements of customers. Moreover, companies should establish long-term supplier partnerships rather than engage in ad hoc transactions. Thus, TQM was a comprehensive management approach that involved all stakeholders in an organization.
Overall, TQM required a long-term, cooperative, and comprehensive approach to doing business. In order to be successful, organizations needed to continuously improve their processes and deliver high-quality products and services that were useful to customers. By doing so, they would grow and gain market share while uniting employees around a common goal and providing them with job satisfaction and security. From all this, profits would follow.
And, indeed, Japanâs multinationals kept growing and flourishing with TQM.
So itâs no surprise that American companies, many of which were watching their market share dwindle, were eager to replicate Japanese companies that adopted TQM and, with it, their successes. Consulting companies also jumped onto the bandwagon and began offering TQM workshops and implementation programs. If it worked in Japan, why not in the United States?
TQM became all the rage in the 1980s and early 1990s, but there was one problem: the American companies that adopted it didnât find nearly the same successes as their Japanese counterparts.
Imperfect Copies
Western companies ran into trouble because they created imperfect copies of TQM. They often only adopted its most visible, tangible elements or they set up cross-functional teams and gave people access to data about quality levels, but failed to create the organizational climate needed for employees to feel secure enough to suggest and implement improvements. They abolished end controls without having nurtured a sense of responsibility for errors throughout the organization. Top managers also thought they could outsource TQMâs implementation to consultants, without realizing their close involvement was necessary to signal quality management as the key focus of the entire organization. Thus, since the resulting TQM copy was imperfect, lacking the nuances and complexities of its original, it became useless at best and harmful at worst.
Replicating a successful management practice is a very difficult task. As professors Sidney Winter from the Wharton School and Gabriel Szulanski from INSEAD argued, organizational practicesâespecially those that lead to competitive advantageâare often highly complex.1 They consist of multiple components, some of them tangible (such as technologies and procedures), but many of them tacit and intangible (such as organizational culture and informal networks), that are subtlety interwoven. These practices, which developed and evolved over many years, are often so complex that the firms that implemented them arenât completely sure how they work. So itâs no surprise that Western firms erred. They took a highly complex system and reduced it to something much simpler.
They also added new features to the system, some of which ran counter to the spirit of TQM. As professors J. Richard Hackman from Harvard University and Ruth Wageman from Columbia University discovered, a majority of American organizations added performance measurement and financial reward systems that rewarded employees for achieving quality goals.2 Although rewards and incentives are commonplace in many organizations, they were antithetical to the philosophy of TQM, and Deming himself explicitly argued that they were counterproductive.
This is a very common pattern. In an attempt to replicate a best practice, firms end up transforming a complex practice into a much simpler one, and this simplified version, which is much more alluring and easier to copy, is transferred from one firm to the next, becoming less and less usefulâand eventually harmful.
Yet, the American managers werenât deliberately adopting a deleterious practice; they were genuinely trying to improve their organizations. They did what managers often do: they looked around them to see what seemed to work for others. And TQM was a great success in Japan, so it seemed to make sense to try to emulate it.
But, in the end, they, like all of us, became suckers for success. They blindly adopted a practice based on nothing but its prior success, implemented it poorly, and then overlooked its nefarious effects once they committed to it.
Unfortunately, poor replication is only one of many ways a practiceâs association with success can lead organizations astray. Letâs look at a few others.
Benchmarking Is BS
Managers often unknowingly adopt bad practices when they try to benchmark their organizations against the other companies in their industries.
I see this all the time in my own home organization, London Business School. Whenever, for instance, an MBA curriculum review is being discussed, some dean or senior administrator will say, âLetâs do some benchmarking.â Which means: âLetâs see what others are doing (and then do it too).â The outcome of such a benchmarking exercise is usually a list of ten or so of our main competitor business schools (i.e., Harvard, INSEAD, Wharton, and so on), of which, say, seven have adopted a particular course structure. Then, some top dog says, âSo, we should really do it too,â and everybody nods. This is a wonderful way for a bad practice to spread.
Not everyone needs to be doing something for us to believe itâs a good thing. Sometimes just the top performers need to be doing it. This is because we are all inclined to pay the most attention to the best-performing companies in our industry and only to those. For instance, some years ago, whenever GE did something new (such as Six Sigma), many firms were inclined to immediately imitate it. Like a lot of people, the managers at these organizations assumed that GEâs leaders knew it all: âSurely, when they do it, it must be a good thing, because theyâre such a successful firm.â
Indeed, research has confirmed that organizations tend to imitate the actions of other companies that stand out as successful, even when it is clear that the newly developed practice is not the cause of the companyâs success.3
The press does it, too. Journalists habitually write about top-performing companies and interview their CEOs, rather than the average Joes. We, admittedly, do it in business schools: we teach cases about the best, blue-chip companies, ignoring the less-sexy average types.
Blame It on Perception Bias
In an experiment I performed at the London Business School with my former PhD student Xu Li (now an assistant professor at the European School of Management and Technology in Berlin), we noticed that focusing on the âbestâ is a very general and human trait.
Li and I asked people to look at a file that contained information about a thousand firms, including ten years of performance data on each and how they ranked over the course of the ten years. Firms could follow one of three strategies, which we generically labeled A, B, or C (in order not to bias people in any way toward one or the other). We displayed this information on a computer screen so that we could unobtrusively use eye-tracker technology to monitor what they were paying attention to. We then asked participants a simple question: which strategy do you think leads to the highest performance, on average?
Monitoring their eye movements, we noticed that people usually first started scanning the entire content of the file; they looked at individual firms in detail, their financial performance, and the strategies they followed. However, after a short whileâusually no more than a couple of minutesâthey gave up trying to assess every individual company: there were just too many. Then they turned to a simple course of action: they glanced over the file, looking for one particular thing among the different firms: their performance rank among the thousand companies. As soon as they spotted a high-ranking firm, their eyes stopped scanning, and they assessed the firm in more detail, specifically focusing on what strategy (A, B, or C) the particular company was following. Toward the end of the file, they had seen enough and had made up their minds; a large majority of people concluded that strategy B was the superior strategy.
However, they were wrong: strategy B, on average, led to significantly lower financial performance.
How come the majority of people got it so wrong in this experiment? Because of one simple manipulation that we did, but one that occurs all the time in reality: the different strategies (A, B, and C) also led to different levels of variance in firm performance.
The performance of the firms following strategy A or C was quite evenly distributed: most firms performed pretty well, although there were also some that slightly underperformed and a handful that performed really well.
The performance of firms following strategy B, by contrast, was much more varied. Most of them underperformed, with some doing really quite poorly, but a good number performed really, really well: they were among the top performers in the industry.
How come then that people got it wrong, thinking that strategy B on average led to the best performance (while it led to the worst)? Thatâs because they were drawn to the top performers. Since most of the very top-performing companies had adopted B, the participants concluded that B must have been the best strategy for everyone. But they were wrong. These firms were merely the exceptions to the rule. (See figure 1-1.)
We experienced real-world confirmation of this bias when we interviewed members of the Chinese pharmaceutical industry who told us that companies that engaged in new product development had outperformed those who didnât from 1991 to 2000 (the first decade of privatization in the ind...
