Maybe it’s time to have a look behind the generally accepted rules for successful, sustained performance. As readers will learn, and as these authors write, the rules that really drive companies to superior performance are quite different than what many of us have been led to believe.
The evergreen objective of many – and perhaps most – inquiries into how corporations function is to determine how to improve performance. Some investigators dig deeply into specific functions, such as finance or operations, while others look at more nebulous but equally important objectives such as talent development or innovation. For many, however – and we place ourselves in this category – the most seductive siren song is the promise of understanding the determinants of superior corporate behavior writ large.
Many others have responded to this same call. In fact, the search for the drivers of exceptional, long-term corporate performance has become a sub-genre in the field of popular management science over the last thirty years. In Search of Excellence launched and defined the “success study” business book in 1982, and since then there have been, by our count, almost two dozen high-profile attempts to tease out the secrets of corporate success using the genre’s defining method:
1) Identify high performing companies;
2) Figure out what they did that made them high-performing companies;
3) Conclude that any company that adopts those behaviors will also be high performing, if not better.
Yet, despite a field crowded with supernova management gurus – from Tom Peters to Jim Collins and beyond – we felt compelled to try our hand. Such conceit seems to demand that, before telling you our version of the truth, we take a step back and explain why our approach just might be different enough to warrant imposing two more consultants with a book on the world.
Advice on advice
Our frustration with most success studies is not that any particular directive is wrong. Rather, somewhat ironically, it is that all too often many researchers offer prescriptions that could never be wrong; that is, their claims cannot be proven to be false.
For example, one researcher tells us that a key to long-run performance is the willingness to confront the brutal facts. One could perhaps imagine circumstances in which hiding from the truth and self-delusion might be advisable, though it does seem a stretch. (Note that the advice is not to tell the truth to others, but to face the truth yourself). In other words, this advice falls short because it cannot be proven to be false.
Worse, this advice is not actionable. Of all the facts worth confronting, we are told to pay special attention to the “brutal facts.” But which facts are those? The “brutal” qualifier connotes “unpleasant,” and perhaps also “important.” But these adjectives do not help much, since we lack an objective, reliable way to assess whether a given fact is indeed unpleasant or important.
This is not to say that confronting the brutal facts is incorrect advice anymore than, say, “Buy only stocks that go up” is incorrect advice. But it is neither falsifiable nor actionable, and so it is fundamentally bad as advice.
Unfortunately, the success study genre seems, at times, to have little to offer but recommendations that fail these two tests. Take, for example, the admonition that a successful strategy must be “clear and focused.” Does anyone ever craft or seek to implement a strategy that they felt violated these principles? Or what about the suggestion that sustained success depends upon finding a “big enough market insight?” “Get the right people on the bus.” “Be agile and disciplined.” “Be specific, methodical and consistent.” “Do not abandon your core prematurely.” As opposed to what? Building a business on a market insight you know is too small, getting the wrong people on the bus, being sclerotic and unruly, being vague, feckless and fickle? And should you ever do anything prematurely? (These examples are all taken from popular or credible success studies.)
Frustrated by this proclivity to offer Spike Lee-like advice (“Do the right thing”), we undertook our own success-study effort. But we were also keenly aware that the very works we found so unsatisfying were the fruits of accomplished and careful researchers; we have no reason to believe that we were any smarter than they. If we were to avoid facile conclusions we would need to design and execute our research in a fundamentally different way.
Moving from complaining about the weather to doing something about it ran us headlong into a particularly vexing feature of the success-study design. The bedrock assumption of every success study is that the companies chosen for examination have in fact achieved noteworthy success. Typically, the researcher selects a time period and sets a standard that is intuitively exigent. Popular choices are periods of 10 or 15 years, with performance that exceeds an industry or overall market benchmark by, say, ten-fold.
It is tempting to believe that such performance merits our close attention. But real science can’t be based on what seems right, and no success study we are aware of tests this assumption for its high performers. In particular, as far as we know, there has been no accounting for the variation of the system within which all companies compete. It turns out that the rough-and-tumble of competitive markets creates tremendous noise in company performance: a series of lucky breaks can catapult a company to the top of its peer group, and perhaps even the top of the market, for far longer than you might think. Consequently, the level and duration of performance required to stand out in a meaningful way goes beyond the somewhat arbitrary standards set in most success studies.
So we took a very different approach to identifying high performers, employing cutting-edge statistical techniques and the most comprehensive database we could find – one that covered more than 25,000 publicly traded, U.S.-based companies and 45 years of performance. This allowed us to be more confident that the companies we chose to study were more than just lucky… they had gone beyond the right tail of the performance distribution and had delivered the kind of results that we could, with confidence, lay at the feet of superior management. (See “Where Have You Gone Joe DiMaggio,” Ivey Business Journal, May/June 2009.)
We defined two types of exceptional companies: Miracle Workers and Long Runners. In their lifetimes, Miracle Workers land in the 90th percentile of performance often enough to have at least a 90 percent chance of being more than just lucky. Long Runners land in the 60th to 80th percentile band, often enough to clear the same probability benchmarks.
Based on these criteria, it turns out that our concerns were well-founded. When we applied our method for assessing company performance to the samples selected by high-profile success studies, we found that only rarely did putatively high-performing companies perform in a way that warranted examination. The table below shows how the samples of a selection of success studies fared.
Of course, there are many ways of measuring corporate performance. We focused on return on assets (ROA), a measure of profitability. Shareholder return is also a popular measure, as is revenue growth; most success studies use one of these two measures. It is therefore perhaps not surprising that previous efforts fare rather poorly in terms of exceptional performance as measured by ROA. What surprised us is that – even when looking at these other measures – such a small percentage was in the right tail of the distributions of those measures. Half of the companies in the samples of the success studies we mentioned were below the 70th percentile of performance for both growth and shareholder returns.
Comparing success studies
Our take-away from this table is that few success studies are doing what they think they are doing. With small sample sizes, and with less than 15 percent of their high-performing companies clearing the bar as measured by ROA, less than 18 percent in the ninth decile as measured by growth rate, and less than 30 percent in the ninth decile as measured by shareholder returns, the success-study genre itself might well be thought of as “in search of excellence.” But it has far more typically found only mediocrity.
In contrast, our sample, by design, consists only of companies with exceptional ROA. As it turns out, these same companies have a lifetime growth rate that is comparable to the allegedly high-performing companies in other studies, and a lifetime shareholder return that is every bit as good. In short, companies with great ROA seem to have only marginally lower growth and equally strong shareholder returns as those companies singled out in previous work.
With a sample in hand that we felt we could trust, we then faced another oft-ignored bugbear of management research: establishing causal connections between behavioral differences and differences in performance. Although, as 18th century philosopher David Hume taught, causality can never be observed, only inferred, some claims of causality are stronger than others.
For example, leadership is frequently singled out as a driver of superior performance, based on seemingly strong correlations between differences in leadership and differences in performance. Yet rarely can researchers quantify these relationships: how much of the difference in enterprise value over a decade can be tied to differences in leadership? The inability to establish these links does not undermine the indisputable fact that leadership matters; but it does mean that we cannot know, at least not yet, how much or in what way leadership matters. We must build our causal explanations on more solid foundations than what should be true, lest we fall into the trap of simply proving what we already believe.
Our performance measure (ROA) provided the opportunity to establish a more direct link between behavior and performance than much prior work. A company’s ROA is the product of return on sales (ROS) and total asset turnover (TAT). We exploited this arithmetic identity to identify the drivers of ROA differences among the trios in each industry. For example, in trucking we looked at Heartland Express (Miracle Worker), Werner Enterprises (Long Runner) and P.A.M. Enterprises (Average Joe). Heartland endured a TAT disadvantage versus Werner that was more than offset by a strong ROS lead. This pointed us to Heartland’s high degree of customer focus, which translated into a 10 percent price premium over its closest competitors – which, based on our financial models, accounted for most of Heartland’s ROS lead.
The constraints imposed on our behavioral explanations by the underlying financial structure of the performance advantages gave us a whole new level of confidence in our findings. What we found is that Miracle Workers tended, by far, to derive their performance edge from higher return on sales, driven, in turn, by higher prices born of a differentiated competitive position.
This observation held in even the most unlikely of circumstances. For example, Family Dollar, a U.S. discounter, has bested the legends in discount retailing since the mid-1970s. When compared to its much larger, but less profitable, competitors, Family Dollar (like Heartland) tolerates a lower asset turnover, but realizes a much higher return on sales. What might surprise you (it certainly surprised us) is that – despite the fact that many of the company’s customers are poor – Family Dollar achieves higher ROS through higher prices (rather than lower costs), which it can charge because it offers superior convenience and selection: its smaller stores are easier for customers to get to, and many shoppers buy small amounts of a wide variety of goods.
As part of our case-study analysis, we also built rudimentary financial models to test whether the behaviors we had identified could plausibly account for the timing and magnitude of the performance differences to be explained. So, when comparing Merck (Miracle Worker) with Eli Lilly (Long Runner) in the pharmaceutical industry we found that asset turnover was a key driver of Merck’s ROA advantage in the mid-1970s and again in the late 1980s. It turned out that Merck’s sales were growing much faster than Eli Lilly’s during these periods, thanks to more rapid international expansion, and that non-U.S. sales had a higher ROA than the company’s domestically-generated revenues. Couple this with the observation that the pattern and magnitude of differences in international exposure of the two companies coincided with Merck’s superior overall ROA, and we have what seems to us to be a strong case for a causal link between specific behaviors and performance differences.
Doing versus thinking
Perhaps not surprisingly, each exceptional company had dramatically different drivers of superior performance. A Miracle Worker in Grocery (Weis Markets) faces very different competitive forces and reaches for very different levers than one in medical devices (Medtronic). Yet our project would be meaningless unless we could find some set of general principles that spoke to the deep structure of what put these companies on top.
Extracting the signal from the noise proved challenging. One hypothesis after another was toppled by contradictory observations. At first, merger activity seemed relevant, but as we worked through the cases, sometimes high-performing companies did big deals, sometime small ones. Yet the same could also be said of companies with dramatically different performance profiles.
Product proliferation, international expansion, business-line diversification, innovation…you name it. Every general strategy or discipline that we could measure and connect to performance was in evidence with equal frequency in all three performance categories. We were worried that our answer to the question “What drives superior performance?” would amount to little more than a two-word sentence of surrender: It depends.
We began to make some headway when we shifted our focus from what exceptional companies did to how they chose what to do. This reframing allowed us to look past differences in behavior to underlying similarities in apparent motivations. We ultimately distilled the order that emerged from the chaos into three rules for exceptional profitability. These were rules not for action but for deciding what action to take.
Rule #1. Better before cheaper
Perhaps the most fundamental choice a company must make is how it is going to create value for customers that will differentiate it from its competition. It is a choice that can usefully be framed in rather stark terms: provide superior price value (in the form of lower prices) or superior non-price value, in the form of better reliability, durability, convenience, brand, fashion – in short, every sort of benefit other than price.
Miracle Workers don’t compete on price. In our sample, eight of the nine top-performing companies were focused on non-price dimensions of value. They systematically chose to find ways to distance themselves from the competition with product or service performance rather than seek to draw customers closer with lower price.
Take, for example, Linear Technology, a semiconductor company that began life as a second-source supplier to the U.S. Department of Defense. Contracts were characterized by narrowly-specified performance criteria, limiting the scope for non-price differentiation and emphasizing price leadership.
Micropac Industries, a Long Runner in this industry, had mastered this dynamic and met the DoD’s exacting standards. But it had done so in a way that kept its costs down, enabling it to be price competitive, yet still enjoy worthwhile levels of profitability. Long Runner status, after all, is nothing to sneeze at.
Linear was profitable enough, but it lived in Micropac’s performance shadow. It wasn’t until the early 1990s that Linear was able to find its recipe for top-tier profitability, by shifting to consumer markets. Rather than continue to navigate between the Scylla of demanding performance and the Charybdis of price constraints, the company chose to tackle a different challenge: to provide nearly-bespoke technology for a large number of relatively small accounts. By the mid-1990s, no single customer accounted for more than 10 percent of sales and Linear was focused on high-performance, mission-critical integrated circuits that typically did not account for a high proportion of its customers’ total cost. This combination of attributes allowed Linear to charge relatively higher prices and so capture much of the value it created.
For example, a representative Linear customer sold high-performance, mobile data scanners for thousands of dollars. Linear’s chips improved battery life, a key differentiator for this product, yet one that accounted for less than five percent of the total materials cost for each scanner. Consequently, this customer typically looked elsewhere to ensure the cost competitiveness of its products. This gave Linear significant pricing power, mainly because of its highly diversified customer base. If a Linear customer became especially price sensitive, the company was less compelled to accede to pricing pressures than, say, Micropac, which had a much more concentrated customer base. With a portfolio of more than 15,000 such customers, Linear was constantly in a position to capture a greater proportion of the value it created through higher prices than most other competitors.
Where Linear’s performance improved with a strong movement toward non-price differentiation, Maytag’s decline illustrates the perils of moving away from non-price differentiation. A Miracle Worker thanks to a two-decade run of 9th decile performances beginning in 1966, Maytag is known even today for its iconic “Ol’ Lonely” repairman, endlessly idled by the legendary reliability of Maytag washers. From the mid-1960s until the early 1980s, Maytag’s distinctive products were supported by a highly motivated network of over 10,000 independent dealers. Its marketing and strong network was a powerful combination that supported the sort of price premium that typically drives superior profitability.
In the mid-1980s, however, the rise of the big-box retail format began to undermine the viability of Maytag’s traditional channels. These newly ascendant retailers wanted to deal with a relatively limited number of suppliers that could provide a more complete range of products and price points. Maytag responded by broadening its portfolio beyond high-end washers and dryers through sizeable acquisitions.
However reasonable this response might have been, it did not work. Unable to compete effectively on price due to stubbornly high costs, and having compromised its non-price value position, Maytag’s profitability began an almost uninterrupted decline. The company was acquired by Whirlpool in 2006.
Every Miracle Worker creates its non-price value position in its own way, but the recurring theme is unmistakable: Differentiation based on non-price dimensions is systematically a primary driver of superior long-term profitability.
Rule #2. Revenue before cost
The logic of competition reduces the choice of how a company can create value to one of price- or non-price-based extremes. The arithmetic identities that define return on assets similarly distill the choice of how to capture value in the form of profits into a choice between revenue and cost. Companies with superior profitability must have disproportionately higher revenue – driven either by higher unit prices or higher unit volumes – or disproportionately lower cost.
Miracle Workers rely on revenue. In fact, not only do Miracle Workers rely on superior revenue to drive their ROA advantage, they are as likely as not to have higher relative cost than Long Runners. In other words, Miracle Workers create superior non-price value and accept the higher cost they must incur to do so. But they more than recoup that investment in relatively higher revenue.
The relationship between higher price and better profitability is relatively straightforward. Several of the examples discussed so far illustrate this: Heartland in trucking, Medtronic in medical devices, Linear in semiconductors, Maytag in appliances (in its glory days). Driving revenue through superior volume is perhaps somewhat less intuitive, and is illustrated in high relief by Wm. Wrigley Jr. Company, the confectioner best-known for its chewing gum.
When compared to the Long Runner in the category, Tootsie Roll Industries, Wrigley has a strong non-price position based on brand equity. Through significantly higher relative advertising expenditures – which consist of not only mass market media but also extensive co-op arrangements to secure preferred placement in retail outlets – Wrigley’s portfolio of products are among the best known in the industry. In contrast, Tootsie Roll has a long history of cost-consciousness, and has traded on the durability of the brands in its stable. Where Wrigley products are advertised on TV and occupy the top shelf in most convenience stores, Tootsie Roll depends more on customers’ long memories and transient candy preferences across the generations. It is content to see its product on the middle or bottom shelves.
Despite its price premium, and during much of its run of superior performance, Wrigley has actually had lower return on sales than Tootsie Roll. As well, Wrigley’s gross margin advantage was more than eroded though consistently higher SG&A expenses, made up largely of marketing and sales.
Consequently, Wrigley’s higher ROA has perennially been a function of higher asset turnover, which has in turn been a function of higher revenue: From 1986 through to 2007 (when Wrigley was acquired), Wrigley grew revenues by over ten percent per year while Tootsie Roll grew at just over seven percent per year. Certainly, Tootsie Roll’s growth is impressive (it bears repeating: the company is a Long Runner), but in the competition for relative profitability, Wrigley’s higher growth rates carried the day. Perhaps most important, Wrigley’s superior volume was not, contrary to the prescriptions of your first-year university economics course, a result of lower price. Rather, Wrigley drove volume through non-price differentiation.
The story of Thomas & Betts, a manufacturer and distributor of electrical wiring components, offers a salutary example of the pitfalls resulting from competing on volume-driven cost reduction when seeking volume through lower price. A one-time Miracle Worker, T&B enjoyed top-decile performance through the mid-1980s, thanks to its consistent investment in R&D and non-price differentiation in products and services.
In a manner strongly reminiscent of Maytag’s woes, however, T&B found itself subject to industry-level shifts that seemed to mandate a thoroughgoing strategic transformation. The rise of electronic (vs. electrical) components suggested that strong growth lay outside of T&B’s traditional markets. As a result, the company made several large acquisitions and sought to expand both its production and its markets overseas.
T&B proved unable to compete effectively in these new markets, yet each setback spurred only a deeper commitment and additional acquisitions in a quest for scale-driven cost advantages that never materialized. It was only through a return to its historical strengths – non-price differentiation in electrical components and services – that the company was able to right itself. It was ultimately acquired by ABB.
The ways in which a company might find valuable non-price differentiation are almost infinite. In contrast, there are only two ways companies can drive superior revenue: price and volume. Six of our eight revenue-driven Miracle Workers relied on price-driven, gross-margin leads. But for the other two (Wrigley and Merck), volume-driven revenue was a viable alternative. Either way, however, the underlying rule is clear: revenue before cost.
Rule #3. There are no other rules
We couldn’t find anything else that mattered in a systematic way. Whatever superior-performing companies did that contributed to their superior performance was consistent with rules #1 and #2. Whenever their actions undermined that performance, they were behaving in ways that were inconsistent with the rules.
It is for that reason that we see M&A on both sides of the high-performance ledger. When T&B went in search of scale to cut costs so it could compete on price, it lost its Miracle Worker-level performance. In contrast, when Heartland almost doubled in size through a series of acquisitions, it was easy to infer that the company was zealous – even religiously so – about pruning back acquired revenue that was not sufficiently profitable and even those customers that the company could not serve in its traditionally distinctive way.
Similarly, when Maytag globalized in order to reduce its costs while seeking new revenue through lower price, it foundered. In contrast, Merck’s international efforts were born of an ability to generate new revenue on the back of its underlying differentiation.
The absence of other rules doesn’t give you permission to shut down your thinking. You still have the responsibility for searching actively – and with an open mind – to figure out how to follow the rules in the face of what may be wrenching competitive change. It takes enormous creativity to remain true to the first two rules.
For example, Abercrombie & Fitch has stayed on top of a constantly changing retail-clothing market by being willing to invent new images for itself and new product lines. For example, its Abercrombie brand is aimed at grade-schoolers, while Hollister is for the 14-to-18s and Gilly Hicks is for women, a feat that the company accomplishes without wavering in its dedication to a position driven by a brand-intensive, non-price value, and a higher-price-driven profitability formula. A&F has avoided markdowns and promotions, and has typically sold its clothing at about 70 percent of full price, which is higher than the comparable figure for many apparel retailers. When the recession hit in 2008, A&F resisted following other clothing companies down the discount path, a choice that earned the company no small amount of criticism from analysts, mainly because its same-store sales dropped more than its competitors’. The company’s persistence, has, however, preserved its brand cachet, and with the recent economic recovery, A&F is now returning to a level of profitability that its competitors are having a tough time matching, especially after it let its customers know that T-shirts don’t have to cost $30 after all.
Using the Three Rules
These rules for exceptional performance are by no means a comprehensive blueprint for success. They leave a great many important questions unanswered. For example, knowing that “better before cheaper” is the way to go does nothing to tell you precisely how your company needs to be better today in order to differentiate yourself on those non-price dimensions of performance, dimensions that you know the customer will pay for and value. Neither can “revenue before cost” reveal whether you should focus on price or volume to drive your revenue advantage.
Instead, think of the rules as a compass: collectively, they point you in the direction you want to go. But you cannot, of course, simply put your head down and walk north. To head north without incident, you will sometimes have to head east or even double back and head south for a while. Watching only the compass could well lead you into quicksand or off a cliff. Perhaps just as important, the rules help you avoid following red herrings or giving in to dangerous temptations. Travelers can be led astray by mirages or the sirens of Greek myth, misled into believing that illusory or temporary pleasures are preferable to the hard-won satisfaction of true virtue.
Similarly, following the rules might occasionally require you to deviate somewhat, for a time, from the rules themselves. The emergence of low-cost competitors, rising commodity prices, or a recession might well have you paying much closer attention to price competitiveness and your relative cost position than the rules might typically imply. Likewise, pressures for quick wins or the desire to “make the numbers” can delude managers into pursuing the false certainty of cost-cutting at the expense of maintaining a commitment to the seeming uncertainties of investments in non-price value. That’s why the rules are “better before cheaper” and “revenue before cost.” Sometimes price competitiveness matters, and sometimes cost control is critical, but in the long run, don’t forget what matters most.
Our research shows that what looks like the easy, sure path of price and cost is actually more likely a dead end. If it is superior, long-term performance that you seek, the three rules show not just the road less traveled, but also the road best traveled. And that can make all the difference.