Using Scenario Analysis to Manage the Strategic Risks of Reengineering

Major business reengineering efforts represent an organization’s commitment of millions of dollars for redesigning internal organizational processes, changing fundamental product delivery and customer service procedures, and often reexamining and repositioning corporate strategy. These efforts are inevitably accompanied by millions of dollars for replacing the information infrastructure and developing new application code to support the new processes, procedures, and strategies.

Just as inevitably, after completing a reengineering project, the organization lacks both resources and will to undertake a second reengineering effort to resolve the first project’s major deficiencies; thus reengineering generally constitutes a lasting legacy, and whatever is decided about the organization’s future design and built into the information systems will constrain the corporation for years.

Reengineering projects are inherently risky and uncertain. While the individual risk components associated with the projects are the same as those of any other large systems undertaking, the specific risk profile of reengineering projects is fundamentally different. In particular, the risks either of building the wrong systems or of terminating prematurely and thus completing no systems are both greater.

Successful business reengineering must begin by examining an organization’s future and its operating environment; based on this, reengineering must determine which of the organization’s fundamental assumptions about the future to reexamine and which strengths based on these assumptions must change. Techniques like scenario planning can greatly reduce risk and help executives properly focus their reengineering efforts.

Risks of Reengineering

Anecdotal evidence, supported by the high priests of reengineering, suggests that reengineering is the greatest advance to hit U.S. business since Henry Ford and the assembly line. (For an introduction to business reengineering, please see the sidebar.) Michael Hammer and his colleagues would have us believe that, although reengineering is difficult and risky, those companies that persevere will inevitably be rewarded with productivity and competitiveness gains comparable to those from introducing the assembly line.1 The partner who directs the global reengineering practice area for a major accounting firm believes that reengineering efforts can even restore U.S. competitiveness, given advances in this area beyond Europe and Japan.

However, a larger body of evidence suggests that many, even most, reengineering efforts ultimately fail. But it is not enough to warn executives that reengineering is difficult and that they must persevere to have their shot at competitive payoffs. If we cannot tell executives why reengineering is difficult, why so many efforts are disastrously expensive failures, and how to manage their risks, we are advising them to “buy reengineering lottery tickets.” We are recommending that they bet their careers, and sometimes their companies, on winning productivity and competitiveness in a reengineering “lottery.”

During the past several years, I have participated in numerous reengineering efforts, as scholar and consultant, in an attempt to understand the risks of large-scale reengineering efforts and help companies manage their risks. Based on my experiences in diverse industries, ranging from securities trading and stock exchanges to manufacturers of consumer goods like soap and detergent, I have concluded that most companies fail at reengineering either because they do not attempt the radical internal organizational change required by changes in their business and operating environment, or because they do! That is, reengineering can fail because companies do not attempt necessary change, reflecting a corporate failure of vision, or because they actually attempt to make the changes, which causes organizational resistance to change and reflects a corporate failure of will.

This apparent “damned if we do and damned if we don’t” paradox requires some explanation. The two greatest risks associated with business reengineering — or with any strategic change program — are:

• Functionality Risk. The risk of making the wrong changes to systems and processes, or making inadequate changes that do not accommodate strategic changes in business, strategy, technology, customer needs, or other rapidly changing competitive factors.
• Political Risk. The risk that the organization will not complete the project, either because of serious internal resistance to the proposed changes or because of a more gradual loss of will to continue the project.

Not surprisingly, efforts to undertake necessary radical change, thereby addressing functionality risk, are almost always associated with a significant increase in political risk. That is, a corporation’s willingness to address functionality risk and make the difficult, disruptive, even radical decisions to address critical issues successfully in redesigning itself will almost inevitably lead to major political risk. Established power centers will be threatened, and some senior executives will be vehemently opposed, although often behind the scenes and invisible. Subversive efforts, euphemistically termed organizational “defensive routines,” come into play.2 Firms either attempt too little, creating functionality risk, or attempt whatever change is necessary, creating political risk.

To the extent that a company sees reengineering as an update of organizational infrastructure or aims at reducing costs of underlying business operations, it will see reengineering as an outgrowth of the systems function. This often leads to personnel who lack the requisite skills and authority managing reengineering efforts. Reengineering leaders frequently lack the authority to undertake sweeping organizational change, so they limit the scope of reengineering efforts by focusing on minor changes and process improvements. They often lack the strategic vision to redesign the corporation and its strategy and assess its new set of critical resources and core competencies. And often they do not have the requisite political skill to manage the problems associated with change, even after identifying appropriate changes.

After decades of systems development, why aren’t organizations better at managing risk? After briefly describing the risks in large-scale implementation efforts and explaining the inability of many previously successful organizations to manage them, I will describe concrete, inexpensive steps, involving scenario planning, that executives can take to protect their investments in reengineering, manage the inherent risks, and increase the likelihood of winning the reengineering lottery.

Changing Risk Profiles

There are five components in the risk profiles associated with large-scale implementation efforts. I have ranked them according to how they were perceived by systems development professionals in the 1970s and early 1980s.

1. Financial Risk. When mainframe computers were extraordinarily expensive and software engineering was not even an emerging discipline, the costs of hardware and software could not be estimated accurately. Projects were plagued by endless delays and cost overruns. Financial risk — the risk that the project could not be brought in on time and on budget — might jeopardize the justification for a systems investment.
2. Technical Risk. When mainframes were small and slow, what the organization wanted to do was simply beyond the existing technology’s capability. Early mainframes, like the IBM 1401, had processors that were unimaginably slow by current standards, memory of 16K, no disk drives, and, of course, no database management software. The limited capabilities were easily exceeded by overly aggressive systems design.
3. Project Risk. MIS projects are complicated, with hardware and software selection, systems design and implementation, data conversion and entry, and staffing and training. All phases must be properly sequenced, and all hardware and software carefully matched. Data processing personnel may not understand new technology, compounding problems. During conversion lasting months or even years, the hybrid collection of old and new hardware, data, and applications must work together seamlessly to process millions of transactions nightly.
4. Functionality Risk. Completed systems do not have the right capabilities, either because systems designers misunderstood the organization’s needs, or because the needs have changed.
5. Political Risk. Systems are not completed because of organizational resistance to change or the gradual loss of commitment to the project.

Naturally, when hardware’s cost and performance was hundreds of times less attractive than today, and when hardware and software were not available for many now-routine tasks, financial and technical risk would have ranked at the top of the problems faced by every systems effort in every systems group. Functionality risk —the danger that we could not figure out what we wanted our new accounts payable system or statement processing system to do — was less significant. Political risk —the danger that clerks would successfully resist changes to the organization — was seldom a factor in any analysis. The data processing staff learned to manage the risks that really mattered; many became quite competent at managing the standard risk profile.

The first computers I used, despite their multimillion-dollar prices and glass-house installations, did not have the processing power or memory now common in a toaster oven or digital wristwatch. So, naturally, the risk profile of current systems projects is different. In particular, the ranking has probably reversed; financial risk is reduced, given the inexpensive hardware and packaged software applications. Thus, while the risks of the 1970s and 1980s have diminished, political and functionality risk have increased and are now more important.

Designing de novo for an uncertain future is certainly difficult, especially if we don’t know what customers will want, what competitors will attempt, what capabilities we will have, or what our regulators will permit, require, or prohibit. Design problems caused by strategic uncertainty are especially severe if we have to operate for a decade within the constraints imposed by the reengineering systems design.

In a clean-slate redesign, some divisions are no longer necessary, and many senior employees lack the skills for their new positions. Not surprisingly, they may be determined not to lose their jobs, and, despite the apparent inevitability of change, may try to protect their positions and defend the value of their (obsolete) expertise.

While the individual risk components in reengineering are the same as those associated with any large systems undertaking, the specific risk profile of reengineering projects is fundamentally different. Thus there is an enormous difference between the MIS risk profile that senior personnel have learned to manage and the MIS risks that these same people are now called on to manage; the most critical difference is the shift of functionality and political risk from least important to most significant. This single difference — the difference between risks that personnel have learned to manage and risks that these personnel now need to manage — explains why reengineering efforts frequently fail. Fortunately, once this is understood, it is possible to take defensive measures.

Business “Revisioning” and Functionality Risk

Successful business reengineering is often about developing a new vision of the company and its mission, customers, and potential competitors; it is about business “revisioning” rather than merely business process redesign or continuous improvement. This inevitably produces unanticipated consequences, some unacceptable.

The experience of member firms at the London Stock Exchange (LSE) after “Big Bang” deregulation in 1986 is suggestive. The associated information technology gave traders, regardless of their location, easier market access, causing London to be by far the most international stock exchange in the world.3 This appeared to create a significant competitive advantage for the LSE. Unfortunately, the same technology eliminated the information advantage that had enabled member firms to “buy low and sell high” when trading for their own accounts and effectively destroyed the member firms’ profitability. All of them would have earned higher rates of return in risk-free government bonds than they did in their extremely difficult and risky trading operations. It is probably fair to say that no London firm fully anticipated the functionality risk associated with these strategic changes; in many firms, the senior officers, experienced traders who managed their firms through Big Bang, are being eased into early retirement as neither they, nor their firms, appear suited for the current competitive environment.

Consumer credit card organizations and interbank credit card franchises like MasterCard and Visa offer another example. Both MasterCard and Visa evolved when thousands of banks (issuers) issued credit cards, and thousands of banks (acquirers) literally “bought the paper charge receipts” from merchants and acquired the merchants’ receivables. MasterCard and Visa facilitated inter-bank settlement and clearing, allowing issuers and acquirers to “settle up” in a timely, cost-effective manner, and were responsible for publishing lists of lost and stolen cards.

What function do these franchises serve now? There has been marked consolidation in both issuing and acquiring; for any charge transaction in New York, for example, it is highly probable that the issuer and acquirer will be Citibank, Chase, Chemical, AT&T Universal, or an MBNA affinity card. Authorization is handled by the merchant’s acquirer or by a third-party provider. Why should the transaction be routed through St. Louis (MasterCard) or northern California (Visa)? Why not just allow the major East Coast banks to enter into bilateral settlement agreements, as in the United Kingdom? Of course, as this begins to occur, either with MasterCard’s and Visa’s cooperation or despite their opposition, it will become harder to justify their data processing budgets. In fact, it will be difficult to think of them as other than professional associations that maintain standards, design logos, and orchestrate advertising campaigns. What mission can they serve? Again, no amount of cost cutting or systems performance improvements may be adequate to assure their continued operation in current forms.

Both organizations appear uncertain about their future direction; indeed, AT&T and GM credit card launches, Microsoft’s acquisition of Intuit (Quicken), and the joint Microsoft/Visa transaction exchange network suggest that the future direction and innovation will come from outside the association and its traditional banking members. Judging from Schoemaker’s work, this may not be very unusual: one of the greatest limitations to strategic change is the considerable number of things successful incumbents know about their industry that unfortunately are no longer true.4

Difficulties in Changing Strategic Vision

Cost cutting, downsizing, and continuous improvement of current processes and procedures require commitment and follow-through; the difficulty of continuous improvement should not be underestimated. But these are relatively easy forms of change; an outside observer or a competent, disinterested internal team can determine what is required, even if the follow-through and implementation require sustained effort. However, revising a firm’s fundamental business mission and competitive strategy to reflect massive, discontinuous change in its operating environment and then revising all processes, operating procedures, job descriptions, work flow, and information systems to support these changes is a much more difficult undertaking. Just determining what is needed and revising the vision or “revisioning” is hard enough, without making the detailed supporting changes.

In all organizations, there are significant obstacles to “revisioning”; paradoxically, they are frequently most severe in the most successful organizations.5 Two obstacles are:

· Overconfidence and Intellectual Arrogance.

We all believe that we are quite good at what we do, and we all greatly overestimate our knowledge and understanding. This may lead us to drive faster than we should, confident in our ability to handle future obstacles, roadway obstructions, and the unanticipated actions of those with whom we share the road. Overconfidence may also make us believe that we know how to lead companies into the future, confident in our abilities to handle strategic challenges, regulatory obstructions, obstacles created by new entrants and substitute products, and competitors’ unanticipated actions.

· Anchoring and Availability Bias.

When we plan for the future, we may think it will be packaged a little differently, but it will fundamentally be a continuation of the present. We anchor on the present because it is a readily available basis for our predictions. There is surprisingly little difference between AT&T’s exhibit at Disney World’s Epcot Center and the comparable exhibit at the 1964 New York World’s Fair, despite developments like personal computing, the Internet, structural unemployment, and nearly ubiquitous automatic weapons. As the chairman of a small, extremely profitable upstart carrier currently competing with American Airlines has commented, “Companies that believe that they are successful probably no longer are!”

There are many examples of failure in corporate and military history engendered by success and continuous improvement of previously successful strategies. France prepared for World War II by continuous improvement of its World War I strategy and the associated technology of trench warfare, culminating in construction of the Maginôt Line. In contrast, Germany prepared by mastering air power and blitzkrieg. As the French general staff intended, Germany went around the Maginôt Line and through the supposedly impenetrable Ardennes Forest.6 Without artillery, which could not get through the forest, German forces attacking through the Ardennes were supposed to be cut to pieces at the first defended river crossing. German forces under von Manstein substituted air power (Stukas) for artillery support and crossed the river despite French defenses; France fell in a matter of weeks.

The difficulty of business revisioning is exacerbated by various factors. These range from Cartesian traditions and the search in U.S. business schools and boardrooms for a single right answer to the prevalence of spreadsheets and ease of plugging projections for next year’s numbers into last year’s demonstrably successful forecasting model. However, what if the business is encountering or about to encounter truly disruptive change, events that we might consider an environmental discontinuity? In this case, last year’s spreadsheet will produce results that have enormous precision (lots of digits before the 000s) but very little accuracy (even the first digits are wrong!). We need not look for events as dramatic and unpredictable as the asteroid impacts that apparently led to the rapid extinction of dinosaurs and the beginning of the age of mammals; the combination of factors that hit the Detroit auto industry — high interest rates on auto loans and unprecedented increases in gasoline prices — were nearly as disastrous. Chrysler almost followed the Deinonychus into oblivion, and Ford and GM had some very bad years indeed. The Age of Detroit was about to be replaced with the Age of Honda, Nissan, and Toyota.

Clearly Chrysler, Ford, and GM should have been able to predict that if environmental changes made their cars seem unacceptably expensive to buy and other related changes made them unacceptably expensive to operate, new competitors would be attracted and the Big Three’s market share and margins would erode. Their corporate planners, armed with spreadsheets, should not have been lulled into planning exercises that were about as accurate as predictions 65 million years ago that dinosaurs would enjoy another successful 200-million-year run.

Here we see the effects of overconfidence and intellectual arrogance and anchoring and availability bias. Moreover, Tushman and others have explained that these problems are most extreme when organizations have been most successful:7

• Behaviors that have previously been successful will be seen as having produced success and will be believed to lead to future success.
• Behaviors that have previously been successful will be rewarded and reinforced. (Actions that have led to promotions, bonuses, and interviews with The Wall Street Journal are likely to be repeated.)

While problems are associated with making change in any previously successful organization, and while Tushman does not explicitly attempt to relate them to business reengineering, the connection with functionality and political risk associated with reengineering should be made clear.8

People internalize behaviors that have been rewarded and reinforced. These behaviors will be difficult to change. Thus they will — explicitly or implicitly, consciously or unconsciously — provide much of the basis for future planning. Personnel will specify designs consistent with these behaviors and resist designs leading to actions that are inconsistent with them. Most important, they will see actions that diminish the value and importance of previously rewarded skills as destructive to the value of their competence and will resist those most strongly.

Dangers of Attempting Too Little

The difficulties in planning frequently lead companies to do too little when undertaking reengineering, and they confuse cost cutting, downsizing, and continuous improvement with courageous strategic change. The causes of this overly conservative approach may be clear, but we have not yet examined the costs and risks.

While reengineering is not about systems and automation, it is inextricably linked with them. Any change in a company’s vision and strategy and the processes and procedures for implementation requires comparable changes in systems applications and infrastructure. Increasingly, it is simply not cost effective to implement a strategy without appropriate systems support. The dependence on systems for strategy execution means that systems efforts occasionally determine organizational strategy, even if only by default. In organizations where the planning process has become detached from the systems group, the head of systems implementation sometimes becomes the de facto head of strategic planning, whether qualified for this position or not.

The costs of doing too little in a reengineering effort thus include:

• The need to write off the investment in reengineering processes, retraining personnel, and rewriting systems —often hundreds of millions of dollars or more. A major U.K. bank spent close to $1 billion replacing its branch automation hardware and software. However, it did not examine its commitment to branch banking or the branches’ future role. In the first eighteen months after system roll-out, it reduced the branches from 3,300 to 2,500; future reductions and comparable changes in the role of branch banking are likely. Clearly, the bank does not need to invest in hardware for the closed branches. More significant, some fundamental changes will occur in the way customers use branch banks; the bank may discover that any changes it wishes to make are outside the capabilities of their new branch automation system. The new systems will themselves need to be written off and rapidly replaced.
• The even larger opportunity costs of business lost, perhaps forever, to competitors. After deregulation of the airline industry, Rosenbluth Travel, a small Philadelphia firm, anticipated the changes that would occur in corporate travel and carved out a $6 billion global market share before American Express understood the changes and developed a strategic response.

Scenario Analysis in Planning for Strategic Uncertainty

The desire for a single right answer, anchoring in the present, and overconfidence in knowledge and current models all combine to yield surprisingly conservative estimates of the future. More significant, they produce dangerously conservative strategies and long-term investments for dealing with different future needs.

Scenario analysis is a technique that has proved consistently effective for dealing with strategic uncertainty in numerous companies in diverse industries.9 Rather than determining a single correct view of the future and the implicit strategic response, scenario planning embraces uncertainty and devises a range of views of an uncertain future. These are not the traditional high/low/average scenarios of more traditional strategic planning; rather, they provide competing views of the future, such as “Credit card associations like MasterCard and Visa are supplanted by networks provided by Microsoft or the Internet; associations remain, but credit card issuers are dominated by nonfinancial institutions like GM or Exxon; consumer cards are for facilitating interaction between merchants and consumers generally and no longer focus on short-term credit.” While these scenarios have much in common, they differ in critical, fundamental ways. Preparing for each scenario in a timely, cost-effective manner and developing the appropriate strategy, procedures, and infrastructure provide the necessary link between scenario planning and business reengineering.

Scenario analysis does:

• Acknowledge uncertainty and highlight the key, critical sources of uncertainty and ambiguity.
• Develop a range of possible future scenarios for exploration, acknowledging that not all are equally likely, and that the future may indeed have aspects from more than one scenario.
• Develop a range of strategies and future indicators of which strategies may become most critical.
• Acknowledge that future uncertainties may create discontinuities, after which current data become meaningless as predictors of future sales, competitor actions, and customer demand, and thus become meaningless as determinants of strategy. (Rapid and widespread adoption of private fax equipment did not just alter Federal Express’s plans for Zap mail, it caused their cancellation. Inexpensive air travel did not just alter the strategy of Greyhound Bus, it nearly killed the company.)

Scenario analysis does not:

• Hide uncertainty or ambiguity.
• Develop a single most likely answer or develop a single average prediction.
• Develop a single strategy to which the firm can commit and that the firm can pursue.
• Obtain unavailable data or make decisions on available data that may not be relevant to the future planning process.

The deceptively simple scenario analysis process uses an organization’s expertise to consider the different shapes of its future environment. First, a team of executives discusses its perceptions of key uncertainties facing the business.10 Some are environmental uncertainties that will affect the firm and define its environment, but are outside the firm’s direct control. Others are operational uncertainties, the key business decisions the firm needs to make. Environmental uncertainties define the business environment and constrain the firm’s strategic options; operational uncertainties define the firm’s strategic response.

Next, the executives rank the environmental uncertainties. Those with the greatest potential future impact and the most poorly understood likelihood of emerging are at or near the top. Executives then select two or three critical environmental uncertainties as the driving uncertainties and combine them for future scenarios. Next, they explore each selected scenario. An internal logic or story line for each emerges. Each has advantages and disadvantages. Each requires different strategic responses and investments in preparation. The broad outline of the firm’s preferred strategic responses will determine appropriate business decisions and, thus, operational uncertainties.

The executives next explore different views of possible futures, using them as “flight simulators” to “test fly” alternative strategies. One important approach is to examine how a strategy and set of operational decisions developed for one strategy will fare in another. This allows the team members to examine their decisions and classify them, based on a determination of their applicability under different scenarios:

• What are the “no brainers,” the actions common to all scenarios, that will be required in all foreseeable futures? The company should undertake them.
• What are the “no regrets,” the actions that may be valuable in some scenarios, less valuable in others, but not damaging in any? The company may undertake these, but should stop if later events demonstrate that the future is evolving in ways that make them unnecessary or superfluous.
• What are the “contingent possibilities,” the actions that may be valuable only in selected scenarios?
• What are the “no ways!,” the scenarios the company deems unacceptable, and how does the company avoid their occurrence or lessen their damaging impacts?

Thus scenario analysis helps firms to determine the actions that:

• They should take now.
• They should stop if future events indicate that they are headed toward scenarios that make the actions less attractive.
• They should take, perhaps right away, if future events indicate that the business environment is headed toward scenarios that make these actions essential.

This last category includes the “contingent possibilities.” A major contribution of scenario planning to reengineering is to identify the contingent possibilities so that the company can react quickly. But, most critically, it ensures that the actions a firm takes early on do not preclude implementation of contingent possibilities later, if necessary. In addition, scenario analysis helps firms determine signposts and early warning signals so they know where they are headed, and provides sufficient time to respond.

The firm makes no attempt to determine which scenario is correct or, more important, to rank the scenarios by probability, take their weighted average, and plan for the average case occurrence. Rather, the firm determines a set of future possibilities, the appropriate strategic responses, and how to know where it is headed with sufficient lead time and speed to respond to emerging conditions.

Unlike the more traditional strategic planning exercises of strategic management consulting firms, which are based on detailed interviews outside the firm and extensive data acquisition, scenario planning encourages senior personnel to reveal their concerns and examine the implications. Thus scenario exercises are fast and surprisingly inexpensive. Schoemaker provides some excellent accounts of recent corporate experience with scenario planning.11

Despite its power, scenario planning has not been generally applied to business reengineering until recently. Most firms employed a traditional strategic planning firm to assist in their reengineering efforts, and most major strategic consulting firms were unwilling to accept the ambiguity and uncertainty of scenario planning. This meant that firms got precise, apparently certain, reengineering specifications, rather than accurate but apparently imprecise specifications that enabled or supported a range of probable future strategies. My experience suggests that reengineering must embrace uncertainty. It must provide the infrastructure to support operations that are radically different from the firm’s current operations. Most important, it must provide an infrastructure so the firm can respond rapidly to future changes that, although uncertain, are highly plausible; scenario planning, while leaving future options uncertain, at least renders them predictable and foreseeable.

Scenario analysis differs critically from sensitivity analysis. While scenario analysis acknowledges the possibility of very different future outcomes, it makes no attempt to determine the average case expected. Consequently, while with sensitivity analysis, the firm may be well prepared for the average case but poorly prepared for any actual outcome, scenario analysis ideally gives the firm an understanding of the very different scenarios it may encounter so it can respond rapidly and successfully to each.

Scenario Planning Workshops for Reengineering

Scenario workshops have been used effectively in a wide range of strategic planning exercises, in problem domains as diverse as planning for electronic distribution of soup and soap, military war games, or devising a strategy for South Africa’s developing multiethnic government. The numerous examples discussed by Schwartz and Schoemaker all have a common need for planners to embrace and plan for a high degree of strategic uncertainty.12 Reengineering efforts have the same need, as we shall see:

• The competitive environment determines strategy.
• Strategy in turn determines the firm’s tactics and operations.
• Operations determine the necessary software applications.
• In aggregate, application programs determine infrastructure, including database design, development languages, and hardware platform.

But the legacy of infrastructure decisions will frequently endure for a decade or more, constraining the applications programs that can readily be developed. Therefore, infrastructure constrains the applications that can be developed, which in turn constrain operations that can be readily implemented, which in turn constrain strategies that can be pursued. Thus, if reengineering efforts do not consider strategic environmental uncertainty in precisely the same ways that strategic planning efforts must, then the results of the reengineering will be overly constraining and may not support the firm’s essential future strategies. Therefore, the same requirements that drive the use of scenario analysis in strategic planning — understanding the full range of strategic alternatives and comprehending them early — motivates its use in reengineering.

A major insurance company had a well-advanced reengineering effort, with a budgeted cost of $350 million. The officers initially thought their project quite bold, with no sacred cows or unchallenged assumptions. But, at a scenario planning workshop, I coaxed from them the assumptions they thought too obvious to mention, which had informed and directed their reengineering efforts:

• “We will always sell products through a traditional agent sales force.”
• “We will always sell traditional products — property and casualty, and life insurance.”
• “Regulators will either continue the current regulatory regime or tighten restrictions on all players. There is little need to consider new entrants or competition from outside the industries that currently compete with us.”

After the first scenario workshop, when senior personnel raised their assumptions, the very applicability and incomplete scope of their prior reengineering efforts became clear. At the workshops, the executives had generated a wide range of alternative, equally plausible operating environments and had considered the infrastructure and information needed to compete effectively in each. They realized that their previously proposed strategy was designed for a world that, while possible, was not the only future environment in which they might operate, and that some of their cost-cutting changes would have precluded successful operation in other environments. They ultimately renamed their existing plan “Madame Tussaud’s Waxworks.”

Reengineering is proceeding, but based on a sounder and more complete understanding of the company’s future requirements. Each scenario that the management team developed has an internal logic and a compelling story; each has been accepted by the full senior management team and has provided the shared vocabulary for discussing future plans, threats, opportunities, and requirements. (For more detail on the driving uncertainties and scenarios generated, see the sidebar, “Scenario Planning at an Insurance Company.”)

At the workshop for a large retailer, there were too many “outside experts.” So, while management personnel raised their uncertainties and developed profound alternative visions of the future operating environment, it was easy for them to view the scenarios as products of the outside facilitators. Management never internalized or took ownership of the scenarios or recognized the threats they represented. When one senior officer commented that he could now safely say he could not imagine a future scenario in which his organization’s current skills were not sufficient to ensure its future success, I knew the workshop had failed.

When successful, scenario workshops are truly effective at reducing critical functionality and political risk. Functionality risk is reduced because future needs and strategies are surfaced, addressed, and fully explored. Organizations escape the Epcot trap of assuming that the future will be much like the present, and that current skills, strategies, products, and services will be the basis of future success. Since the alternative strategies come from the management team, from its perceptions of uncertainties and threats, it sees the need to change as very real. This too helps motivate significant strategic change, reducing functionality risk. Perhaps more significant, since the need to change is internally generated and accepted, the political risk associated with such change is likewise greatly reduced.

Conclusion

Business reengineering is difficult and risky, and success is not ensured. As we have seen, the principal risks come either from not attempting necessary change, or from actually attempting to make the changes.

The dangers of not attempting to make necessary change and the associated costs include:

• The need to write off the investment in reengineering processes, retraining personnel, and rewriting systems — often hundreds of millions of dollars.
• The even larger opportunity costs of business lost, perhaps forever, to competitors.

The dangers of actually attempting to make the changes in a reengineering effort include:

• Getting yourself fired.
• Getting your whole staff fired.
• Miring the project in controversy so that nothing is attempted or nothing significant is completed.

Addressing the balance between these dangers is particularly challenging for the executives in charge of business reengineering. At one extreme, doing too little may jeopardize the long-term viability of their corporations. At the other, actions necessary over the long term may create unacceptable resistance and rapidly destroy the executives’ own careers.

Scenario analysis has been developed as a strategic planning tool to deal explicitly with overconfidence, reliance on one certain estimate for an uncertain future, and anchoring in the present. While it cannot eliminate all the risks of strategic uncertainty, it has been used successfully in various firms to address a wide range of strategic contexts.13 It is useful in addressing the functionality and political risks of business reengineering. Properly applied, scenario analysis helps a firm avoid the trap of attempting too little — of being satisfied with attempting minor change and performance tuning — when a major reexamination of its most fundamental strategic options may indeed be required. As important, it reduces the risks in making the adjustments essential to dealing with a rapidly changing business environment.