The Strategy Paradox An excerpt from the newest perspective on those surprise twins, success and failure

According to author Michael Raynor, most strategies are built on specific beliefs about the future. Despite our fixation on forecasts, however, no one can predict the future, making it risky to bet the house on even the most brilliant strategy. Yet breakthrough success requires true commitment to a strategy. How can companies resolve this paradox?

Chapter 1: What strategy paradox?

Most strategies are built on specific beliefs about the future. This is a problem because the future is deeply unpredictable. Worse, the requirements of breakthrough success demand implementing strategy in ways that make it impossible to adapt should the future turn out differently than planned. The result is the Strategy Paradox: strategies with the greatest possibility of success also have the greatest possibility of failure. Resolving this paradox requires a new way of thinking about strategy and uncertainty.

Here is a puzzling fact: the most successful firms often have more in common with failed organizations than with those that have managed merely to survive. In fact, the very traits that we have come to identify as determinants of success are also the ingredients of failure. And so it turns out that the opposite of success is not failure, it is mediocrity.


Why is this the first you have heard of the strategy paradox? After all, there is no shortage of well-designed and well-executed studies that have offered useful insights into the defining characteristics of successful firms. Similarities with failed firms and the importance of luck have not tended to feature prominently.

And so it turns out, behaviorally at least, the opposite of success is not failure, but mediocrity.

The reason most business research misses the strategy paradox is because, understandably enough, few studies ever examine failure. Instead, most investigators tend to study success; after all, who wants to learn how to fail? In some cases this is because pursuing the secrets of success seems more rewarding than picking through the wreckage of failure. In other cases it is simply a flawed method: the researchers embrace the idea that by studying winners they can discern the secrets of success, forgetting that the factors differentiating winners from losers can only be identified by analyzing both. Finally, there is the reality that failures are often harder to document, or, in the case of failed companies, the organization is no longer available for study.

So, researchers compare companies that have been very successful over ten or 15 years (focal companies) with companies that have been less successful over that same time period (comparison companies). Some studies look for firms that have done very poorly over that time, and others look for comparison companies that have actually done pretty well – just not nearly as well as their focal companies. Either way, however, comparison companies have at least survived for the period in question, and over a ten year period, mere survival is actually a pretty high bar.1

What this means is that these, and many other, studies have based their conclusions upon comparisons of success with the mediocre. Because these studies systematically seek out successful companies, each will necessarily find those that in the past made the right commitments. And because the comparison companies are always firms that have performed less well, but not failed completely, they will necessarily be firms that have avoided the high-risk, high-return strategy of commitment.

By examining primarily those companies that have guessed right and comparing them with those that have avoiding guessing, what has been largely missed is the critical importance of managing uncertainty. The gallant charge and the cowardly retreat are not the only alternatives to catastrophe, for there is a way to mitigate risk without compromising performance, and describing that solution is the promise of this book.

Accepting the strategy paradox forces us to accept mediocrity, giving up a chance at greatness as the price of our continued corporate existence. Resolving it will free us from a debilitating tradeoff between risk and return and allow us to strive to be first without giving up the hope that we will last.


The cause of the strategy paradox is as obvious as it is overlooked. A successful strategy allows an organization to create and capture value. To create value, a firm must connect with customers. For a firm to capture value, its strategy must be resistant to imitation by competitors. Satisfying customers in ways competitors cannot copy requires significant commitment to a particular strategy – strategic commitments – to unique assets or to particular capabilities.

Not just any commitment will do, however. Success stems from committing to what enough customers want, but few competitors have. Commitments are a powerful determinant of success because they make a strategy difficult to imitate, but only because competitors will only want to imitate you once it becomes clear you have made what happen to be the right commitments. Since they waited while you committed, you will enjoy a period of relatively little competition because it will take time for your rivals to replicate the capabilities you have so painstakingly created. For example, new products snapped together from off-the-shelf components are usually easily imitated by competitors, while those based upon proprietary technologies developed over years are far likelier to be the foundation of a durable franchise. The downside of commitment is that if you make what happen to be the wrong commitments, it can take a long time to undo them and make new ones.

Call it an “emotional paradox”... loving and hating—can have far more in common with each other than a seemingly intermediate state.

The strategy paradox, then, arises from the collision of commitment and uncertainty. That is, the most successful strategies are those based on commitments made today that are best aligned with tomorrow’s circumstances. But no one knows what those circumstances will be, because the future is unpredictable. Should one have guessed wrong and committed to the wrong capabilities, it will be impossible to adapt – after all, a commitment that can be changed was not much of a commitment. As a result, success is very often a result of having made what turned out to be the right commitments (good luck), while failed strategies, which can be similar in many ways to the successful ones, are based on what turned out to be the wrong commitments (bad luck). In other words, the strategy paradox is a consequence of the need to commit to a strategy despite the deep uncertainty surrounding which strategy to commit to; call this strategic uncertainty.

New research detailed in Chapter 3 suggests that often the main factor separating success and failure is indeed luck. Mediocre firms – those that survive but do not prosper – avoid commitments that expose them to the vagaries of luck. The price of avoiding that risk is the lost opportunity for greatness; the reward is an increased chance of survival. For now, these seem to be the only alternatives to failure, and firms are forced to choose. There is no intrinsic merit in opting for greater returns over survival or vice versa; the problem is that firms must choose at all.

Sony’s case, referred to above, is not unique. The strategy paradox is more than a theoretical possibility or a curiosity; it is a general condition. As recounted in Chapter 3, an analysis of the competitive strategies of several thousand operating companies reveals that organizations pursuing the most commitment-intensive strategies generate the highest returns, but they also suffer the highest mortality rates. Seen in this light, Sony’s failures were not a consequence of avoidable mistakes but were instead an inevitable result of making commitments – the defining element of successful strategy – despite inescapable uncertainty. And when those uncertainties were resolved to Sony’s detriment, it paid the price.

The strategy paradox rests on two premises: commitments cannot be adapted should predictions prove incorrect; and predictions are never reliably or verifiably correct. Are these premises true?


For all we might think we know about how to make organizations agile, flexible and adaptive, the data suggest strongly that, if anything, competitive advantage is eroding faster than ever. This acceleration is interpreted by some to mean that there is a greater need than ever for adaptable enterprises. Such an observation is entirely correct. Unfortunately, the acuteness of the need does not mean it can be satisfied.

Some measure of adaptability is visible in most organizations. However, as explained in Chapter 4, it is far less useful than we might like. Specifically, adaptation is only viable when the pace of organizational change can match the pace of environmental change. When the environment changes either faster or slower than the organization, adaptation is no longer effective; every organization will at some point face either “fast change” or “slow change,” and each can prove debilitating.

Fast change leaves an organization handcuffed because it leaves an organization’s capabilities optimized for an environment that suddenly no longer exists. For example, when the price of oil rose 400% in a matter of weeks, North American auto makers found that the mainstay of their product lines – full-sized cars – were singularly inappropriate to the new competitive conditions. Unfortunately, it took those same automakers years to design, manufacture and market more fuel efficient models, and they lost valuable market share to better-positioned competitors. (Ironically, to have the tide turn in their favor in the mid-1990s when cheap gas made SUVs all the rage, and then to see their good fortune evaporate once again in the face of high-priced petrol.)

Slow change prompts an organization to adapt to incremental changes in the environment around it, and because of these incremental adaptations, the company often fails to see the need for a more fundamental transformation. The auto sector’s response to the current oil crisis may be subject to this slow change pathology. As oil prices have crept up, auto makers have responded by extending the life of the internal combustion engine by dramatically increasing fuel economy and creating hybrid electric engines, among other technological advances. But the day may well come when, due either to the need to limit carbon emissions for environmental reasons or the inability of the general economy to absorb still further increases in oil prices, the internal combustion engine must be abandoned. Should that day arrive, companies that have been exploiting their adaptive capacity by extending the life of this century old technology will be largely unable to respond.

Compounding the difficulties of responding to fast and slow changes is the fact that most competitive environments are characterized by multiple rates of change, creating the impossible organizational task of changing at different rates at the same time.

As a result, firms cannot expect to resolve or even mitigate the strategy paradox through adaptation.


The first half of the paradox is commitment: companies cannot adapt their commitments should they turn out to be the wrong ones. Might they instead predict the future accurately enough to consistently make the right commitments in the first place? Futurists make such predictions, believing that by analyzing the past and present they can identify and interpret the trends that will ultimately define the future. Since there always seems to be someone with a track record of successful predictions, we might be tempted to write off the paradox as an illusion.

We would be wrong for three reasons, which are explored in Chapter 5. First, no one can legitimately claim to have a meaningful ability to foresee the future in anything like the level of detail required to make consistently successful strategic commitments. Any such claims can always be explained by the law of large numbers: with so many people predicting so many things, it is inevitable that someone is going to get something right occasionally. Since we cannot know who that someone is going to be, or what they are going to get right, the fact that some predictions turn out to be accurate is useless.

Second, predictions in the form of point estimates betray a fundamental misunderstanding of what the future actually is. The future is a range of possible outcomes, not a specific set of circumstances that will inevitably come to pass. A prediction of the future as seen from the perspective of today would have to describe each of the events that could happen and their associated probabilities. Unfortunately, there is no way to compare our probability-based description of the future with the true probabilities as they are today. For instance, if I say that there is a 10% chance of rain tomorrow, whether it rains tomorrow or not tells you nothing about the accuracy of my prediction, for either outcome is consistent with a 10% chance of rain.

One could, of course, note all those occasions when I predicted a 10% chance of rain and then see if it rained on 10% of those days for which I predicted a 10% chance of rain. When it comes to weather forecasting, this is reasonable, but when it comes to strategic forecasting, the outcomes of interest are rarely repeated events. As a result, this kind of track record is impossible to establish. Consequently, we have no way of determining if someone can provide accurate, probability-based descriptions of the range of possible future events.

The third reason accurate prediction is impossible is the ubiquity of randomness. Randomness generally can be thought of as the absence of the kind of order that allows us to predict what comes next in a series. That is, we might be able to identify a pattern, but unless that pattern repeats in ways that allow us to foresee what follows, the series is ultimately random. It turns out that the systems we hope to understand and predict for the purposes of making strategic commitments are subject to two main sources of randomness.

First, competitive systems are subject to exogenous shocks such as new technologies or regulatory regimes that create new competitors and upset long-standing equilibria. It is tempting to believe that we can overcome this problem by simply expanding the boundaries of our analysis, but it is the nature of this source of randomness that as soon as we begin expanding our scope we don’t know when to stop. Before long we find ourselves compelled to build a “theory of everything” in order to predict anything at all.

Second, even if the dynamics of a particular system are predictable, competitive dynamics are highly sensitive to past commitments – what systems dynamics theorists call “initial conditions.” What constitutes the initial conditions of a system is a judgement call, and getting it wrong makes any subsequent predictions highly suspect.

We can see these first two sources of randomness at work in the evolution of Toyota from its humble beginnings to a world-class auto manufacturer. The oil crisis of the mid-1970s was an exogenous shock that created a surge in demand in North America for smaller, fuel efficient automobiles. Toyota had a model line-up of precisely these kinds of cars, and many customers who would not otherwise have purchased a Toyota were motivated to purchase one for the first time. Many were pleasantly surprised at the value Toyota offered, and stuck with the brand as the pressure on gas prices receded. The oil shock was, in a sense, the lucky break Toyota needed to gain access to the mainstream of the U.S. auto market.

For General Motors to have predicted this shift in competitive fortunes, it would have had to focus not on its would-be competitor, but on the geopolitics of the Middle East. GM would also have had to predict accurately the impact of that shock on consumer buying behavior, and that the “temporary” interest in Toyota would prove to have long-term implications.

And why would GM ever think that Toyota would be a threat to its business rather than historical rivals Ford and Chrysler? Was Toyota blessed with a prescience or adaptability that would allow it to exploit unforeseen events that others were blind to? Hardly: in Toyota’s home market of Japan gasoline had long been more expensive than in the United States, while economizing on space had long been a priority. These “initial conditions” are features of the Japanese market that can be attributed to geographic, political, and cultural traits that are centuries old. It is not as though Toyota had a “copy GM” strategy in the 1950s, but changed course when it foresaw the oil crunch; neither did Toyota adapt to the embargo when it occurred. Rather, as a consequence of market pressures in its home market the company committed a particular strategy (small, fuel-efficient, inexpensive cars) that was eventually appropriate for the North American market for reasons few had foreseen. This takes nothing away from Toyota’s success; as Louis Pasteur put it, “fortune favors the prepared mind.” But the corollary of this insight is that preparation without fortune – or worse, coupled with bad fortune – amounts to the wrong commitments.

In sum, prediction cannot resolve the paradox any more than adaptability can.

...the same behaviors and characteristics that maximize a firm’s probability of notable success also maximize its probability of total failure.

1.5 Managing strategic uncertainty

The strategy paradox is a consequence of the conflict between commitment and uncertainty, i.e., strategic uncertainty. Commitments are what allow an organization to create and capture value. Uncertainty creates risk and opportunity. The answer to the paradox lies in separating the management of these two elements of strategy, charging some with the responsibility of delivering on the commitments the organization has already made, and others with the task of mitigating risk and providing exposure to promising opportunities.

The foundation for this division of labor is the traditional organizational hierarchy. As explained in Chapter 6, well-functioning hierarchies are defined by clear separation between levels according to the time it takes for those at that level to know whether or not they have made the right decisions. Most who work in large organizations have an intuition that this basis of organization is right. The CEO should be thinking about the long term, while divisional management (typically an operating division) is worried about the medium term, and in-the-trenches line management has to deliver the goods.

In the short run, there is very little strategic uncertainty. We typically know (even if we cannot implement) the best way to create and capture value in the present. We cannot know how best to create and capture value ten or 20 years from now. The range of strategies that might be optimal in the future only expands as we lengthen the time horizon under consideration. Consequently, there is a great deal of strategic uncertainty when considering the long run.

Senior management, because it is responsible for longer time horizons, should therefore focus its efforts on managing strategic uncertainty. Those lower down in the hierarchy, because they are responsible for shorter time horizons, should focus on delivering on the commitments already in place. This new organizing principle is called Requisite Uncertainty because each level of the hierarchy is defined by its relationship to managing strategic uncertainty.

The implications of separating the management of uncertainty from the management of commitments are more far-reaching that it might seem. In the first place, Requisite Uncertainty provides a foundation for the widely-held yet often violated belief that senior management should not be concerned with short-term results. There is very little that pulling on the strategy lever can do to improve this quarter’s cash flow, and any CEO that is compelled to intervene frequently on issues that can affect current financial results is likely not able to pay enough attention to strategy.

Explicitly identifying strategic uncertainties and requiring that senior management attend to them might well be quite different from the more bottom-up risk management processes many organizations have in place. Focused on important but shorter-term uncertainties such as the supply chain, the company’s reputation, and so on, much of established risk-management practice overlooks the risk that the company has committed to the wrong strategy, and what to do about that.

Perhaps more controversially, applying the principles of Requisite Uncertainty implies that CEOs should not see their role in terms of making strategic choices, i.e., commitments. Rather, they should focus on building “strategic options,” that is, creating the ability to pursue alternative strategies that could be useful, depending on how key uncertainties are resolved. It implies also that the board should not concern itself as much with engaging the substance of a firm’s strategy as with determining the most appropriate exposure to strategic risk and opportunity. Only by shifting the emphasis at the top of the hierarchy from making and executing strategy to managing strategic uncertainty can corporations hope to mitigate strategic risk while simultaneously creating strategic opportunities.

In Chapter 7 a case study of Vivendi Universal, a French media conglomerate, illustrates what happens when a CEO succumbs to the temptation of top-level, hands-on strategy-making. Examinations of the diversified Canadian telecommunications company BCE Inc. and U.S. software company Microsoft highlight the benefits of an option-creating top management. In particular, the BCE and Microsoft cases show how corporations can manage strategic uncertainty in ways that investors cannot replicate.

Firms that accept strategic risk reap either great reward or utter ruin.

However, it will be apparent from these examples that the success of an options-based corporate strategy has depended largely on the charisma, influence, and power of the CEO – it has required people such as Bill Gates and Rupert Murdoch to do this successfully. Not every company is led by such a titan. Consequently, we need an explicit, process-based description of how managers at all levels can contribute to managing strategic uncertainty in ways that mitigate risk and position a firm to capture emerging opportunities. This is the subject of Chapter 8, a case study of how Johnson & Johnson (J&J), the diversified U.S. pharmaceutical, medical devices, and consumer products company, has tackled this challenge.

J&J is breaking new ground in the management of strategic uncertainty. The company’s operating divisions are not exclusively responsible for managing long term strategic uncertainty. Rather, divisional leadership’s primary role is to commit to a specific strategy, because making a commitment is the only way to create and capture value. But the range of strategies available to the operating divisions is a function of the opportunities created by the corporate office. That is, operating divisions enjoy a level of strategic flexibility – a deliberate oxymoron that qualifies the irreversibility of strategic commitments without undermining their competitive power.

Operating divisions that manage their own long-term strategic uncertainty will most likely end up mediocre performers, avoiding high risk bets to increase their odds of survival. In addition, since great performance demands relentless focus on a particular strategy, devoting resources – especially management time and attention – to creating options is typically beyond the capacity of an operating division. Consequently, Strategic Flexibility is not something a successful operating division can typically create for itself. Only by focusing the corporate office on the management of uncertainty can the overall corporation achieve high results (thanks to commitment-focused divisions) at lower risk (thanks to the uncertainty-focused corporate office).

Highlighted here is the profound difference “growth” options and true strategic options. Often, companies will make a small investment in a new venture and see it as an “option” on future growth: if the venture takes off, then invest more in it; if it falters, let it wither, or perhaps even expedite its demise. Strategic options, however, enable established divisions to pursue fundamentally different strategies. A strategic option is an option on an element of an alternative strategy that might or might not be implemented, not simply an option on further investment in a new business that might or might not succeed.

Making the corporate office responsible for managing uncertainty is a significant break from much current thinking about the role of the corporate office. Most prescriptions on what a corporate office should do start with the premise that all competition takes place at the product market level – that is, in the domain of operating divisions.2 This is true, but the conclusion that all corporate office activity must, therefore, be directed at improving the current competitiveness of operating divisions – by, for example, facilitating the capture of synergies between operating divisions – does not necessarily follow.

Drawing too direct a line between the actions of the corporate office and the performance of the operating divisions is an unhealthy side effect of our collective obsession with generating returns. The frameworks for developing competitive strategy that have emerged over the last 30 years have given us unparalleled insight into how companies can succeed. And competitive strategy remains enormously important, but it should be the preserve of divisional management. But just as there can be no left without a right, there is no return unaccompanied by uncertainty: operating divisions worry about competitive strategy to create and capture value; corporate strategy should be focused on the management of strategic uncertainty. Requisite Uncertainty delegates to corporate management responsibility for managing strategic risk and opportunity, and Strategic Flexibility is the framework for fulfilling that responsibility through the creation of strategic options.

It would be a mistake to think that the distinction between corporate and competitive strategy, and hence Requisite Uncertainty, applies only to Fortune 500 corporations. Any organization with greater size and complexity than a sole-proprietor corner store – or simply the ambition to achieve said size and complexity – can benefit from thinking carefully about separating generating returns from managing uncertainty. If your organization has operating managers who report to still more senior managers, there is not simply the chance but the likelihood that there is an unhealthy overlap between the jobs each level thinks it is doing. In fact, the smaller the organization, the greater the temptation of senior management to involve itself in operating decisions, with the unfortunate consequence of leaving the management of uncertainty largely to…chance.

Reprinted with Permission.