To pull away from the pack one needs to break performance tradeoffs by getting better in several ways at once. The struggle for greatness, however, is far more complex and subtle. Prevailing over capable adversaries requires accepting and exploiting tradeoffs and very often seeking an advantage in only a very small number of very carefully identified ways, while frequently accepting a performance disadvantage along other dimensions.
Perhaps the most fundamental concept in strategy is the “trade-off.” In automobiles, for example, no one car is all of the biggest, fastest, most fuel efficient, safest, lowest carbon-emitting, most stylish and least expensive. Every vehicle manifests different trade-offs among these various dimensions of performance, and the choices customers make reveal which set of trade-offs they are most comfortable with. Competition in the automobile industry turns on the degree to which different automakers are able to discern these preferences and satisfy them. Since many of these trade-offs are rooted in laws of physics (e.g., it takes more energy to move faster or to move a larger car), it seems foolhardy to do other than to work within them in the pursuit of valuable differentiation.
At the same time, trade-offs are much maligned as a foundation for goal-setting and decision-making. The “tyranny of the OR” we are told, is an unnecessary limitation on the heights to which we can rise. Not lower cost or higher quality, but both; not customization or volume, but both; not higher performance or lower price, but, both. Focusing on trade-offs leads all too often to their unquestioned and frequently unnecessary acceptance.
There is merit in this view, for although there is a trade-off between, say, acceleration and fuel economy at any given point in time, cars have improved dramatically on both dimensions over time. We have expanded the limits of the possible rather than simply accepting historical constraints as inviolable constants.
When compared to the very good companies, the very best companies do not deliver highly differentiated solutions at low prices; they pick one or the other.
In other words, where strategy turns on exploiting trade-offs, progress depends on breaking them. How are we to know when and how to do which?
Our investigation into the drivers of superior performance has shed some light on this critical question. Perhaps somewhat surprisingly, it turns out that for all the rhetorical fireworks, the tyranny of the OR is very often your best friend when seeking to break through from very good to truly exceptional.When compared to the very good companies, the very best companies do not deliver highly differentiated solutions at low prices; they pick one or the other.
On the other hand, the difference between mediocre and noteworthy very often does turn on the ability to break trade-offs. That is, average companies are almost always bested by good and great companies on both differentiation and price. Consequently, those stuck in the middle of the pack typically face the daunting task of improving on multiple dimensions simultaneously.
In short, whether to accept or break trade-offs is a function of over whom one seeks a performance advantage. To pull away from the pack one needs to break performance trade-offs by getting better in several ways at once. The struggle for greatness, however, is far more complex and subtle: Prevailing over capable adversaries requires accepting and exploiting trade-offs and very often seeking an advantage in only a very small number of very carefully identified ways, while frequently accepting disadvantages along other dimensions.
Founded in 1949 and headquartered in Minneapolis, Medtronic has long been synonymous with cardiac pacemakers and has been described as “the best medical device company in the world.”1 A leading innovator in pacemaker technology since the 1960s, the company nevertheless struggled on occasion with product recalls: Among the most significant were the 1975 recall of the Xytron and the 1984 recall of the Model 6972.
When attempting to learn from the behaviors of putatively high-performing companies, perhaps the most important step is the first one: How do you know which companies have delivered performance worthy of our admiration?
In our monograph “A Random Search for Excellence” (see www.deloitte.com/us/persistence) we describe in detail how we have chosen to approach this problem. For now, it is perhaps sufficient to point out that for us the key is to separate “luck” from “skill.” Doing this objectively requires employing specific statistical tests that allow us to say with quantifiable confidence which firms have delivered sufficiently high performance for sufficiently long periods of time that luck is very unlikely to be a central part of any explanation.
In our research, we have identified two categories of exceptional performers: Miracle Workers (MW) and Long Runners (LR). Miracle Workers are our gold standard; Long Runners are remarkable—but not as remarkable. Keep in mind, also, that whether or not a company qualifies for one of these exalted categories is a function of its lifetime performance: A relatively short run of five or even ten years of standout results is simply not enough, statistically speaking, to be heard above the white noise of competitive markets. Our method accepts the possibility of ignoring companies that are exceptional (i.e., a potentially high rate of false negatives) in exchange for a very high degree of confidence that the companies that are identified as standouts truly are noteworthy (i.e., a low rate of false positives).
Since then, however, the company has been on a nearly three-decade tear. Its gradual but steady improvement in return on assets (ROA) has put the company consistently in the top decile of performance, even when compared to the full population of publicly traded U.S.-based companies. It is what we call a “Miracle Worker” (MW): a company that has been good enough for long enough that we can be confident that there is more than just luck behind these results.
Of course exceptional performance is a relative concept: Medtronic’s performance is only remarkable compared to the performance of other companies. In the medical devices industry, a good comparison is Stryker, founded in 1941 and headquartered in Kalamazoo, Michigan. For much of its early existence, the company focused on relatively low-tech segments of the medical devices market and was a pioneer with innovations such as the mobile hospital bed. By the time it went public in 1978, Stryker had moved into more technologically advanced gear including joint implants and emergency equipment. Solid but gradually declining performance resulted in a string of 6th–8th decile performances, but since 1998 the company has been on a strong upward trajectory. The net result has been performance that goes beyond what one would expect by chance alone. Although less remarkable than Medtronic, Stryker is still what we call a “Long Runner” (LR).2
Figure 1. The anatomy of profitability in three medical device companies
Comparing two different types of exceptional performance—a “gold medal” winner with a “silver medal” winner—can be revealing, but a more complete picture emerges when we include a company with “average” performance: average duration, average profitability and average variation. We call such companies “Average Joes” (AJ)—and it is important to emphasize that “average” performance is by no means poor, since by definition many companies do worse than average. The AJ we identified in the medical devices sector is Invacare, public since 1983 and headquartered in Elyria, Ohio. Focusing on mobility aids (e.g., wheelchairs) and home care equipment (e.g., beds), the company’s ROA has been generally improving, although it has ranged in decile rank from 0 (in 2006 when it was in the red) to 7 (in the early 1990s).
To learn something useful, we must decompose each company’s ROA... the product of two very different elements of a company’s operations: return on sales (ROS) and total asset turnover (TAT).
Figure 1 identifies specific periods for each company that warrant careful attention. Note in Panel A that Medtronic’s MW status is driven by a streak of 9th decile ranks from 1986–2010. To understand what made Medtronic exceptional, we should focus our comparisons with Stryker and Invacare on this period. In contrast, Stryker is a consistent LR, and so we can compare the full period of overlap between Stryker and Invacare (1983–2010).
A simple comparison of the ROA values among these three firms reveals little, however: Knowing that Medtronic ran two percentage points per year of ROA ahead of Stryker and 9.7 ahead of Invacare, while Stryker bested Invacare by 8.4 percentage points per year tells us only that Medtronic did better than Stryker, who did better than Invacare—and we knew that already.
To learn something useful, we must decompose each company’s ROA, which exposes the underlying structure of the relevant profitability advantages. Specifically, ROA is the product of two very different elements of a company’s operations—return on sales (ROS) and total asset turnover (TAT).
Equation 1. The components of ROA
One company’s ROA advantage over another need not be a function of advantages in each of ROS and TAT. Rather, an ROA advantage can be driven by superior ROS or superior TAT. Through the magic of ROA decomposition, differences in ROA can be attributed to differences in each of ROS and TAT (see figure 2).
Figure 2. The components of ROA advantage in the medical devices trio
During Medtronic’s streak of 9th decile performance, its profitability advantage over Stryker is driven entirely by an ROS advantage; Stryker enjoys a material TAT advantage over Medtronic. Medtronic’s much larger ROA lead over Invacare is similarly dominated by its ROS component. But when we compare Stryker with Invacare there is less evidence of this seeming trade-off, as Stryker enjoys superior ROS and a TAT disadvantage that is proportionally half that of Medtronic’s, even though the performance lead Stryker enjoys over Invacare is comparable to Medtronic’s.
The differences in the composition of each company’s ROA advantage begin to lay bare fundamental differences in strategy and execution between the comparisons. Medtronic pursues a highly differentiated position in the market for medical devices. Establishing this position requires significant and sustained investments in R&D (to develop the products) and SG&A (to administer and sell them effectively). Its products, which rival or surpass the most sophisticated consumer electronics in their complexity of design, engineering and manufacturing, command a significant price premium—one that more than recoups these expenses, resulting in the strong ROS advantage.
Medtronic has exploited its technical expertise by broadening its product portfolio through both internal development and acquisitions. Stryker, on the other hand, has grown by exploiting the popularity of a relatively narrower range of products. Consequently, Medtronic’s growth has resulted in a disproportionately larger asset base, which leaves it with a TAT disadvantage compared to its leaner, more focused competitor.
The reality of trade-offs must be accepted; but the fact that trade-offs are often broken must be acknowledged. When should we pursue each?
None of this is to say that Medtronic has “poor” asset turnover or that Stryker has “poor” ROS. Rather, it implies that the two companies have very different strategies that imply very different trade-offs between ROS, a measure of how effectively sales are turned into income, and TAT, a measure of how efficiently assets generate sales. Medtronic lives with the lower asset turns its more highly differentiated position and diversified product line imposes, more than making up for the resulting drag on profitability with superior ROS. In other words, what seems to separate Medtronic from Stryker and from Invacare—an MW from an LR and an AJ, respectively—is the willingness to accept a core trade-off and exploit it effectively.
In contrast, the structure of Stryker’s (LR) advantage over Invacare (AJ) is much different. Although Stryker captures its performance advantage through superior ROS, it has essentially no TAT disadvantage. It does not appear to accept a different set of trade-offs compared to Invacare; it seems instead to break the very ROS/TAT tension that defines Medtronic’s performance edge. In other words, Stryker’s superior differentiation compared to Invacare is achieved with a similarly lean balance sheet.
In this one case, at least, Miracle Worker status flows from accepting a trade-off. In contrast, if Invacare were to try and catch up with Stryker, it would have to improve its ROS without sacrificing its asset efficiency—essentially breaking the very trade-off that Medtronic accepts.
Drawing on the full population of U.S.-based companies listed on U.S. exchanges between 1966 and 2010, we identified every company that meets the definition of exceptional performance. This census of the exceptional yielded 87 Miracle Worker/Long Runner pairs, 94 Long Runner/Average Joe pairs, and 147 Miracle Worker/Average Joe pairs. By quantifying and decomposing the structure of the ROA advantage in each of these pairings we were able to determine the frequency with which ROA advantage in each comparison was driven by either breaking or accepting a trade-off between ROS and TAT.
Note first, however, that there is no theoretical reason to believe that our results will turn out one way or another. For example, a company could have lower prices, which leads to higher volumes and lower cost of goods sold, yielding an ROS advantage over its competitors. Higher volume can also drive asset efficiency, resulting in an asset turnover lead. But if that high volume is driven by lower prices without concomitantly lower costs, a profitability leader could well be accepting an ROS disadvantage and driving profitability through asset efficiency. And of course, an ROS lead driven by higher prices might depress volumes resulting in a TAT disadvantage. In short, we cannot reason our way to the answer: We must analyze the data and see.
What the data reveal is captured perhaps most intuitively with a simple count (see figure 3).
Figure 3. Frequency of breaking or accepting an ROS/TAT trade-off by performance category comparisons: Counts and percentages
A simple statistical analysis on these frequencies reveals that the distribution of “accept” and “break” the ROS/TAT trade-off is anything but random: All of the differences prove significant. Further, while MW’s and LR’s are much likelier to break the ROS/TAT trade-off in besting AJ’s, MW’s are far likelier to accept an ROS/TAT trade-off in establishing their ROA lead over LR’s.
A subtler but more revealing approach is to compute the probability of breaking a trade-off conditional on the comparison, the duration of the advantage and the magnitude of the advantage. That is, holding constant by how much and for how long an MW is superior to an LR (and so on for all the other comparisons), do we see a difference in the probability of accepting or breaking an ROS/TAT trade-off (see figure 4)?3
Figure 4. Estimated probability of breaking the ROS/TAT trade-off by performance category comparisons
The size of the performance difference to be explained is shown on the horizontal axis. The likelihood that this performance difference is built on both an ROS advantage and a TAT advantage—that is, by breaking an ROS/TAT trade-off—is shown on the vertical axis. This chart shows that smaller performance differences (the left end of the horizontal axis) are less likely to be built on breaking trade-offs than are larger differences for all category comparisons (MW vs. LR, MW vs. AJ, LR vs. AJ). In addition, Long Runners are far likelier to break trade-offs when compared to Average Joes than are Miracle Workers when compared to either Long Runners or Average Joes for any given magnitude of performance advantage.
In addition, when the trade-off is accepted it is much more frequently the case that a company will seek higher ROS at the cost of lower TAT. Once again, a simple frequency count reveals this most starkly (see figure 5).
Figure 5. Frequency of accepting an ROS/TAT trade-off in favor of ROS
Having motivated our exploration of the ROS/TAT trade-off with a case study, perhaps it is fitting that we explore the significance of the findings using the same vehicle.
Abercrombie & Fitch (A&F) is a Miracle Worker in the clothing retail industry. Originally founded in 1892 as an adventure outfitter, the brand has been reincarnated several times, most recently purchased in 1988 by The Limited and then spun out as a public company in 1996, when our observation period begins. With its edgy branding, sometimes controversial advertising and focus on a differentiated in-store experience, the company grew quickly and profitably thanks to its premium pricing across a range of casual apparel targeted largely at high school and college-age customers.
The retail industry, however, is characterized by a well-understood and seemingly inviolable archetypal pattern of growth and decline. A new retail format identifies or creates a new niche in the market based on an ineffable, inimitable, unimaginable, but in retrospect inevitable combination of ingredients—everything from style to branding to store fixtures. With every mall in America lying before it, the format grows at double digits as it increases the number of stores. Then, when the segment or the retail real estate available to it is saturated, growth slows dramatically. The market eventually tires of what has become yesterday’s news, and it’s on to the next. What was the hottest new thing stagnates and, not infrequently, withers and dies.
Figure 6. Abercrombie & Fitch's continued relative excellence despite decreasing absolute ROA values
The first part of this equation describes A&F’s rise through the late 1990s. A seemingly dramatic reversal in the pattern of A&F’s ROA growth in 2000 might lead one to conclude that the second act was also going to follow the script. However, remember that what we care about is performance relative to other companies, not compared to A&F’s past. The genius of A&F during this second era is that through the astute launching, growing and pruning of brands and formats, the company preserved its relative standing even as its absolute ROA fell (see figure 6).
Our comparison for A&F is Finish Line, a retailer of manufacturer-branded athletic footwear and apparel. Founded in 1976 and going public in 1991, Finish Line is the Long Runner of the trio, while Syms, a discount fashion retailer is the Average Joe of the three.
For present purposes, the most salient fact of A&F’s performance profile is that the company’s pattern of absolute ROA results can meaningfully be divided into two eras, indicated by the two regression lines in figure 6.
Figure 7. The components of ROA advantage in the clothing retail trio
During A&F’s first era, from 1995–1999, the structure of its performance advantage over Finish Line broke the ROS/TAT trade-off. Commanding a price premium even over other store-branded retailers such as American Eagle Outfitters, the company enjoyed a significant ROS advantage over Finish Line, as one might expect from a brand-dependent strategy. Yet, A&F was also vertically integrated from design through manufacturing and retail, enabling the company to respond quickly to and shape effectively the rapidly changing fashion tastes of its target markets. As a result, its asset turns were also high, contributing to almost a third of its total ROA advantage. In short, during this relatively brief period of rapid ascent, A&F broke the ROS/TAT trade-off against both Finish Line and Syms, in keeping with our observation that significant leads result from superiority on both dimensions (see figure 7).
Era 2, however, is a very different story. Even though the core brand remained strong, A&F accepted that this level of performance was very unlikely to continue. The company launched new brands and store formats, built around A&F’s historical differentiation strategy but targeting different segments with bespoke solutions. The abercrombie format targeted children; Hollister focused on high school-age customers at a lower price point than A&F; while Ruehl, launched in 2000, was an attempt to stay connected with post-collegiate consumers but was discontinued in 2010. Gilly Hicks, the most recent format and brand, launched in 2008, offers an assortment of lingerie, loungewear and accessories.
These efforts, conceived and implemented in a manner fully consistent with the strategic and operational details of A&F’s initial success, have allowed the company to defy the gravitational forces that pull so many retailers back to earth. However, sustained relative superior performance in Era 2 has had a very different structure than in Era 1. Consistent with general trends, its smaller lead, although still exceptional, no longer breaks the ROS/TAT trade-off. Now A&F is accepting the trade-off, sacrificing its historical asset efficiency in order to preserve its differentiated position and associated price premium in each of the segments its serves.
Average Joes (AJ), having achieved the praiseworthy feat of a measure of survival and profitability, prove unable to distinguish themselves.
The insight here is the recognition of the need to accept as a binding constraint what had once contributed mightily to its performance advantage. The ability to preserve what one must, and change as circumstances require, is a defining feature of exceptional companies.4
In contrast, to an even greater degree than was the case with Stryker (LR) and Invacare (AJ) in medical devices, both A&F’s and Finish Line’s performance advantages over Syms are consistently a function of breaking the ROS/TAT trade-off. Finish Line has done it very differently than has A&F. Instead of pricing premiums and highly differentiated in-store experiences, Finish Line has relied on solid pricing discipline, cost control, and rigorous inventory and other asset management, but with similar results: Finish Line’s exceptional performance is a function of both superior ROS and superior TAT when compared to Syms. And Syms, which filed for bankruptcy in November 2011, proved consistently unable to improve on both fronts simultaneously—a feat that would have been its most reasonable hope for catching up to the superior performance that has long characterized Finish Line.
The “serenity prayer” is a well-known appeal for guidance in accepting what we must, the courage to change what we can, and the wisdom to know the difference.
When it comes to the pursuit of superior profitability, a similar sort of insight has been in short supply. The reality of trade-offs must be accepted; but the fact that trade-offs are often broken must be acknowledged. When should we pursue each?
Our analysis of the trade-off between return on sales and total asset turnover in the pursuit of superior return on assets reveals a distinct pattern of when and how a trade-off between the two is best respected and when best rejected. Miracle Worker status—the pinnacle of superior, long-term profitability—brings companies up against the limits of the possible: ROA cannot be infinite, and once a company reaches the 9th decile, there is, relatively speaking, nowhere left to go. It is perhaps understandable, then, that the most profitable companies rely for their advantage on a keen ability to accept and exploit this trade-off.
When doing so, they overwhelmingly drive profitability with superior ROS while accepting inferior TAT. As the Medtronic and A&F cases illustrate, this takes the form of investing in the assets required to carve out a differentiated position in the market and earn the higher prices required to compensate for relatively inefficient asset utilization.
Both Miracle Workers and Long Runners have separated themselves from the mass of average companies by breaking these same trade-offs. Average Joes, having achieved the praiseworthy feat of a measure of survival and profitability, prove unable to distinguish themselves. Average companies are, by definition, performing below the limits of the possible. To pull abreast of the companies with exceptional performance they must improve ROS and TAT at the same time.
The implication is that if you wish to perform better than an already exceptional company, you are unlikely to do so by breaking the ROS/TAT trade-off: You can’t be better than they are at everything. Far more likely is the necessity of identifying, accepting and exploiting key strategic and operational choices that will allow you to take your advantage in ROS or TAT, but almost never both.
On the other hand, if you are an average company seeking to separate yourself from the crowd of also-rans, doing so in the way it is most often done requires you to improve on multiple fronts—if not simultaneously, at least eventually.
It is in being clear about your performance improvement objective—pulling ahead or catching up—that we find the wisdom to know the difference between the things we must accept and the things that we should change.
Read the book: The Three Rules: How Exceptional Companies Beat the Odds, by Raynor and Ahmed, by Portfolio, the business imprint of Penguin Group (USA).