The three rules that provided the title for Raynor and Ahmed’s recent book on exceptional performance are based on the large-scale and detailed study of American corporations. But do these findings mean anything outside of the American context? Do they need to be adapted? Are they even relevant?
In The Three Rules: How Exceptional Companies Think we make the case that three decision-making principles are the most important, identifiable, and consistent drivers of long-term, superior corporate profitability. They are:
The rules emerged from our ongoing Exceptional Company project and are based on the large-scale and detailed study of American corporations. But do these findings mean anything outside of the American context? Can they be adopted as is? Do they need to be adapted in any material way? Or are they utterly irrelevant?
This matters for at least two equally important reasons. One, US-based companies seeking to globalize might well benefit from understanding the ways and extent to which what drives superior performance in their home market applies when venturing abroad. Two, companies based outside the United States will very likely want to know what they can learn from the very best American companies in order that they might not only inform their domestic strategies but also their own global ambitions.
When it comes to finding their way on foreign shores, companies of all sizes, regardless of their home base, are well past the clichéd tourist compensating for his lack of language skills by shouting. Rather, they tend to have well-tuned sensibilities about how to leverage the relevant strengths by adapting to local conditions. Few corporations of any moment, for example, simply translate their domestic marketing messages, fail to adapt to and comply with local regulations, or ignore the implications of all manner of cultural differences on everything from distribution to website design.
Somewhat ironically, however, there is a need to tread especially carefully when adapting lessons from one country to another precisely because we know so much. The pitfalls of trusting our intuition and experience when data are available are well-documented. Worse, it turns out that the more we (think we) know, the more trouble we can get into—especially when it comes to venturing abroad.
For example, as a college student studying in Mexico, one of us (Raynor) had ample schooling and months of immersion under his belt. One morning, recounting his misadventures with trinket vendors on the beaches of Mazatlan, he remarked to his instructors that he felt embarazado [sic] that he had not negotiated more vigorously. Hilarity ensued when it was pointed out that he had claimed to have gotten pregnant purchasing a beach towel. Raynor’s facility in Spanish fell short of fluency, which proved his undoing— confident enough to hazard a guess, he was tripped up by “false friends”: words that look and sound similar but that have different meanings. (Note to Raynor: avergonzado.)
So it is for many companies today. Seeking to apply the three rules abroad, and justifiably confident of their abilities to adapt what works at home to foreign arenas, they are at even greater risk of overreaching thanks to entirely reasonable, but potentially entirely erroneous, assumptions. The consequences could easily be far worse than embarrassment.
For example, in countries where targeted market segments are less wealthy than in the United States it can be critical to have lower prices and hence lower cost structures than in domestic markets. But does this mean that one must compete on price and cut costs in non-US markets?
Thomas & Betts (T&B), a maker of wiring systems and one of our Miracle Workers, illustrates what can go wrong as a result of this line of thinking.1 Having enjoyed almost two decades of exceptional profitability, in the 1980s the company moved into electronics, which tended to be a more nearly global market than its core operations in electrical systems. Faced with lower cost competitors based in Taiwan, Malaysia, and other Asian countries, T&B embarked on a decade-long project not merely to reduce its cost but also to compete on price. Unfortunately for T&B, but consistent with the rules, this strategic path proved fruitless: Competing on price but unable to best its competitors on cost, the company was ultimately acquired by Swiss automation giant ABB.
In contrast, Wrigley, the gum maker, expanded internationally very successfully.2 Although it had lower prices and costs in overseas markets than in the United States, it maintained its commitment to the rules. When moving into China, for instance, product launches in 1989 were preceded by significant advertising campaigns on radio, television, and outdoor media. In addition, a large sales force was deployed to ensure extensive retail distribution. By 1999 China was second only to the United States in sales for the company. Perhaps most remarkable of all, by 2005 Wrigley had a 60 percent market share and had driven the largest domestic candy company, Guangdong Fanyu Candy Co., to close.
What these anecdotes suggest is that the rules that drive exceptional performance in the US market can also drive exceptional performance outside the United States. But “for example” isn’t proof. The challenge, and the opportunity, is to determine whether the rules systematically drive the success of high-performing companies around the world, or if they are instead false friends, or worse, entirely lost in translation.
The best way to assess whether the rules translate well, we felt, was to determine whether exceptional performers in those countries also owed their success to applying them.3 To that end, we accessed Compustat’s database of publicly traded companies for every country and for all years for which data are available.
There is considerable variance in the extent and magnitude of the coverage. At one extreme, there are data on 23,223 US-domiciled companies from 1966 to 2011 for a total of 333,743 company-year observations. Toward the other end, we find Zambia with 12 listed companies—but each with its own, often inspiring story.4
Not surprisingly, the bulk of the data is from developed markets and covers significant time periods: Over 90 percent of the non-US data come from countries for which we have at least 23 years of coverage (see figure 1).
We then applied our analytical model to this global database, revealing the global population of exceptional companies.5 Figure 2 provides summary statistics on the 15 countries identified in figure 1.
Note the differences in the relative frequency of exceptional performers as a percentage of the population of companies in each country. The range seems fairly wide, stretching smoothly from a low of 4.8 percent in Italy to a high of 8.2 percent in South Korea and Malaysia (8.2 percent is the frequency of exceptional performers in the United States as well). Japan is the outlier, with a frequency of exceptional performance of over 14 percent.
There are two primary and related statistical determinants of how long a company must deliver top-level performance in order to be exceptional: the period of observation and the volatility of performance for the system within which a company functions. For the purposes of this analysis, system-level volatility is measured at the country level and captured in the notion of “stickiness,” or the likelihood that a company’s performance next year will be the same as it was this year.
When stickiness is high, companies need to deliver longer streaks of strong performance to be statistically exceptional. After all, if there is, say, a 50 percent chance that 9th decile performance will be repeated simply by virtue of system-level characteristics, one shouldn’t be too impressed at a run of three 9th decile years in a row: The first one might have been a lucky break, and there was then a 25 percent chance of delivering two more 9th decile years thanks to nothing more than the echo of that initial good fortune.
The non-US markets that we can study in detail have only 25 years of observation compared to 47 for the United States. The longer period of observation in the United States allows companies more opportunities to compile the sorts of performance streaks that add up to exceptional performance, so it is perhaps not too surprising that we see the United States with among the highest relative incidence of exceptionalism.
It would be a mistake, however, to chalk up the lower relative incidence of
exceptional performers in other countries to nothing more than the shorter observation period. Other countries show dramatically lower levels of stickiness, especially in high-end performance. In the charts below (figure 3) the y-axis shows the relative stickiness of performance at each decile of performance (the x-axis) for 14 countries, when compared to the United States.
Consider figure 3, panel A. The stickiness of 9th decile performance in the United States is 49 percent. This means a company domiciled in the United States has a very nearly even chance of repeating top-decile performance thanks entirely to the residual impact of this year’s results. In other words, it can expect, 49 percent of the time, to repeat a 9th decile year without having to change in a way that is distinctive to that company.
In Italy, however, 9th decile performance has a stickiness that is 11 percentage points lower than in the United States: The highest performing Italian companies have only a 38 percent chance of repeating top-level performance in the following year. So it is for pretty much all other countries: comparable stickiness at the lower and middle ranges of performance, and lower stickiness toward the higher end of the performance spectrum.
This lower level of stickiness offsets the shorter time periods: Since it is harder to stay in the 9th decile in these countries, companies need stay in the 9th decile for relatively shorter periods of time before they qualify as exceptional performers. Yet their incidence of exceptional performance is typically lower than is observed for the United States.
Except in Japan, where we see both a higher incidence of exceptional performance and dramatically higher high-end stickiness: At 54 percent for the 9th decile, it is the highest of all the countries for which we have sufficient data to estimate this particular parameter.
Our modeling of company performance within these systems compensates for these differences, and so higher stickiness with a shorter observation period should, all else equal, lead to a lower incidence of exceptional performance. As is frequently the case, however, all else would appear not to be equal: It simply turns out that when there is structurally lower stickiness to high-end performance we tend to observe lower frequencies of exceptional companies that are able to get beyond the variance of the system.
One possible explanation for this phenomenon lies in the need for companies to amass the resources and capabilities necessary for extended periods of exceptional performance through actually having delivered exceptional performance. That is, a company that earns high performance—whether by good luck or good planning—and then is able to “catch its breath” while still enjoying the afterglow has a greater probability of achieving subsequent superior performance deliberately. But if superior performance is too fleeting, companies perhaps prove less able to prolong their runs beyond the parameters of the system.
By way of analogy, imagine learning to ride a two-wheeled bicycle with only a five- or ten-foot driveway to practice on before you have to stop to avoid rolling out into traffic. Imagine now that you have a long, gentle downward slope. In this second case, there’s a much better chance that you’ll stay on two wheels by chance alone just long enough to actually learn how to ride a bike with intent. In other words, good luck coupled with structural endowments gives you a better chance of being great on purpose.
It is worth noting that this relationship is relatively weak, and fails to rise to the level of a statistically significant correlation. However, the scatterplot in figure 4 suggests strongly that Japan is an outlier and that the United States lies toward the extreme end of the range. That Japan and the United States should be potentially of a kind in this regard is perhaps somewhat surprising. After all, the US economy is by many measures one of the world’s most dynamic, and one would expect exceptional performance to be especially difficult to come by. Yet the United States has the second-highest relative frequency of Miracle Workers—second only to Japan, the only country with higher high-end stickiness.
As an aside, these findings speak to the nature of how one appraises the performance of managers of national subsidiaries or divisions. It is not uncommon to expect, and perhaps even demand, that performance—and especially relative performance—improve year after year. Yet how often are these deep structural differences across countries in how persistently a company can deliver sustained high performance taken into account?
It is one thing to allocate capital to its most profitable opportunities—that can and should be done independently of these considerations. It is another to evaluate management based on relatively prolonged runs of high performance when there is a very good chance that the manager of the US division is in large part the beneficiary of structural advantages over their French or German counterparts. Where the systems within which companies or operating divisions must compete are different in this regard, we must be careful to separate how we evaluate financial performance from how we evaluate managerial acumen.
Whatever the relative incidence of exceptional performance, our primary inquiry is into the drivers of the exceptional performance that exists. We looked at all the years in which Miracle Workers and Long Runners in a given industry (as defined by its four-digit SIC code) enjoyed a performance advantage compared with the relevant Long Runners and Average Joes.6 For all possible pair-wise comparisons in each country, we decomposed the ROA difference into three constituent elements: gross margin, other costs, and asset turnover.7
This sort of analysis is profligate with data, and in many cases we find that there are too few observations, even in countries with otherwise well-developed public markets, to make statistically valid claims. Figure 5 shows the observations and industry concentrations for each pair-wise comparison by country.
These tables reveal definitively what one would likely intuit: that the distribution of companies by industry in the US market is far more balanced than in other countries in our analysis. No industry accounts for more than 13 percent of the total observations in any one pair-wise comparison (Chemical and Allied Products in the Miracle Worker vs. Long Runner comparisons), and every pair-wise comparison has at least more than twice the number of total observations as the next closest market. Consequently, claims about the United States are much more likely to capture true central tendencies independently of industry effects.
In contrast, some of the markets in our database have as few as four unique comparisons (Italy: Miracle Worker vs. Long Runner), and even when there are hundreds of observations, single industries can account for almost 80 percent, and often more, of that total (for example, Canada; Miracle Workers vs. Average Joes; Oil and Gas Extraction: 78 percent of 428 comparisons). Consequently, a naïve comparison of the decompositions of ROA advantage by country would confound country-level and industry-level effects.8
With this limitation of our data in mind, observe the differences in the structure of performance advantages across countries. It is striking that, regardless of the pair-wise comparison, there are three distinct clusters of countries, each defined by the relative importance of a gross margin advantage to superior performance. The panels in figure 6 identify them. In Group 1 countries, exceptional companies rely almost entirely (and in some cases, like the United States, Japan, and Great Britain, more than entirely) on gross margin for their performance advantages. Group 2 countries rely predominantly on gross margin. In Group 3 countries, lower costs account for between 37 percent and 126 percent of exceptional companies’ advantages over their relevant comparisons.
This suggests strongly that the results that emerged from our analysis of US companies—namely, that gross margin is the primary driver of exceptional performance—does not necessarily hold up in all other countries. We cannot be sure for the moment whether this difference is driven by country or industry. Either way, however, it implies that we must be circumspect when applying the three rules beyond the populations we have studied carefully.
For example, it appears that exceptional performance can be a function of not just different, but polar opposite, drivers depending on circumstances. It is tempting to suggest that the Italian exceptional companies’ reliance on other costs is an industry effect, with 50 percent of observations coming from heavy construction. However, German exceptional performers are almost as dependent on cost advantages, and business services are about a quarter of total observations. At the same time, business services are about a quarter of total observations in Great Britain, a country where exceptional performers are among the most dependent on gross margin. This suggests that the structure of exceptional performance is strongly influenced by country-level effects.
Consequently, successfully leveraging one’s domestic business model in new geographic regions is likely to require far more than mere adaptation. Even though much might look the same—as in the case of professional services in Germany and the United Kingdom—the underlying drivers of exceptional performance can be radically different. The path to exceptional performance therefore appears to admit of no shortcuts: One must understand customers, competitors, industries, and country-level factors.
At the same time, there is one remarkable and strong similarity in the structure of exceptional performance across the countries we have examined, and it speaks to how best to extend a performance advantage.
Beyond the simple structure of performance advantages, we can examine how exceptional companies’ performance advantages change with an increase in their edge in each of the three drivers. That is, for a given increase in gross margin, other costs, or asset turnover advantages over industry median performance levels, how much of an increase in ROA advantage do we see? We call this the efficiency ratio associated with superior performance, for it measures how efficiently superior performance in a driver of ROA is translated into ROA advantage.
Our analysis reveals that there are no statistically significant differences among countries for any pair-wise comparison; the figures for the United States and Rest of World are presented solely for the sake of interest. Miracle Workers tend to see their ROA advantage increase when their gross margin advantage increases, and in expectation will see a 0.50 percentage point increase in their ROA advantage for every percentage point increase in their gross margin advantage.
Advantages in other costs also serve to increase Miracle Workers’ advantages over industry median performance. Revealingly, however, the efficiency ratio for other costs is statistically significantly less than the efficiency ratio for gross margin: Extend your other costs’ advantage by 1 percentage point and expect to see your performance lead go up by only 0.42 percentage points. This means that Miracle Workers have profitability formulas that favor revenue before cost.
Long Runners and Average Joes have, respectively, decreasing efficiency ratios with a smaller advantage accruing to gross margin. The implication is that companies with business models that do not respond strongly and preferentially to increasing gross margin are less likely to deliver exceptional profitability.
The larger message is that, while attaining exceptional performance could well require attention to different drivers in different countries or industries, extending a performance advantage is everywhere best accomplished by focusing on gross margin: The returns to an increase in gross margin advantage are greater than those of other costs no matter where you are.
We all, in our own way, strive to be happy. Travel the world and you will experience many and profound differences among people as they pursue their own versions of this universal aspiration. Culture, language, styles of food and dress, modes of thinking, and models of humanity’s place in the cosmos can differ dramatically as a function of place. If we are to be wise, temperate, and open-minded, we must understand and embrace these differences so that they can enrich our lives.
Yet the similarities that bind us together, whatever these other differences might be, warrant our respect as well. As John F. Kennedy put it, “Our most basic common link is that we all inhabit this small planet. We all breathe the same air. We all cherish our children's future. And we are all mortal.”9
The findings of this research do not speak to such profound differences and similarities. Even so, all companies seek their own form of greatness in superior profitability. And, as we look at how this has been achieved around the world, there are important differences that warrant careful study. Yet the similarities, too, appear to matter enormously, for as much as markets and customers and companies and managers adapt to and shape their environments, there is much that remains immutable, and so much we can learn from each other. And so, in the world of exceptional performance, as in so many other domains, there is much that is lost, but perhaps more that is found, in the translation of the rules that guide us in our pursuit of greatness.