Many redesign efforts do not deliver the step-function improvements in performance that, by rights, they should. The reasons for these shortfalls display a fairly consistent pattern: lack of "hands-on" top management involvement, lack of staying power, and lack of focus. Of these, the third—focus—is both the most important and the most difficult to get right. Striking the right balance is hard. Managers of successful redesign programs have learned how to be moderate in selecting which activities and processes to redesign, but immoderate in their ambitions to improve those that they do select.

The jury is partly in, and the news looks good. Many of the process-focused change programs recently unleashed in corporations around the Triad are delivering notable improvements in cost, time, and quality. In a wide variety of industries, efforts to redesign or re-engineer such core processes as customer service, order fulfillment, or new product development have begun to pay handsome dividends.

But the news is also bad. These benefits have accrued mostly at the individual project level. All too often the sum of these benefits does not add up to anything like the quantum leap in overall corporate performance that managers seek and competition demands.

Redesign programs often display a marked lack of focus

Symptoms vary, of course, but this kind of shortfall in performance follows a disturbingly consistent pattern. First, redesign programs often display a marked lack of focus. At one financial institution, for example, an ambitious, multi-year, and multi-project effort involving over fifty initiatives turned out to be so complex and unwieldy that management spent most of its time cutting back the scope of individual projects and severing the linkages between them. As a result, the program's cost reduction goal fell from 30 percent to just 5 percent. Worse, the resulting fragmentation of effort still placed a huge demand on management time.

Second, after redesign programs have been launched, senior levels of management often prove reluctant to roll up their sleeves and get personally involved with the nuts and bolts of infrastructural change—especially with the details of information technology. When, for example, European insurance companies began to experiment with direct marketing by mail and telephone, their efforts varied widely in profitability. Those that delegated IT issues to the technologists and gave little top-level attention to prototyping and piloting new screen-based systems did poorly. By contrast, companies that got their management deeply involved in what the screens should display, and in specifying the data flows needed to construct those displays, did well.

Substantial, lasting improvement comes only from a series of initiatives spaced out over a period of years

And third, because redesign efforts are so disruptive, managers often want to treat them as "one-off " initiatives of finite duration, after which things can be allowed to return to a stable, if different, routine. Longitudinal studies of truly successful redesigns show, however, that substantial, lasting improvement comes only from a series of initiatives spaced out over a period of years. The success of American Hospital Supply, which has developed a radically new process for establishing linkages with customers, took not one but nine separate large-scale efforts spaced over nearly thirty years.

How, then, should managers scope and structure redesign efforts so that they will be most likely to generate and capture the full measure of potential performance improvement? What follows is a set of diagnostic questions that my colleagues and I have found useful in helping managers work toward the best answer for their organizations.

When to redesign?

Any serious process redesign is an exceptionally complex and painful task. Determining the needed changes in information technology is usually the easy part. Far more difficult, for most organizations, are building new front-line skills and refocusing entrenched middle-management attitudes. This disparity in difficulty still comes as a surprise to many managers. And their surprise, in turn, helps explain why so many redesign efforts either miscarry or underperform.

One service-oriented company tried to base its redesign effort on the notion of using more effective teamwork to provide each of its customers with a single point of access. What it discovered, however, was that it would take front-line clerical employees up to five years to build the broader range of skills needed to operate in such an environment. With hindsight, it was obvious that the skills and skill mix built gradually over many years could not be changed overnight—no matter how attractive the new process might be for customer service.

The rule of thumb should be not to maximize change for its own sake but, instead, to focus on the minimum amount of change necessary to reach the desired level of performance improvement

At bottom, remember, redesign implies far-reaching behavioral change, which is almost always traumatic, expensive, and disruptive. As a practical matter, therefore, redesign efforts should be as tightly focused, as carefully bounded, as possible. The rule of thumb should be not to maximize change for its own sake but, instead, to focus on the minimum amount of change necessary to reach the desired level of performance improvement. Focus, not volume or intensity of effort, is the key. Product designers follow this implicit rule when they plan new models based on shared or outsourced components. Software programmers follow it when they write programs based on joining existing blocks of code together with standard bought-in modules. The watchword is economy of effort.

Be moderate in defining which processes need to be prioritized or performed in-house, but immoderate in the ambition to improve those that do

At the same time, however, there must be no foolish economy—that is, no short-sighted limitation on the scope of activities to be included within an integrated, multi-project redesign effort. The challenge is to re-engineer only those activities that together create distinctive, cross-functional, value-adding processes—and then to make that focused re-engineering a success by devoting adequate managerial time and attention to it. In other words, be moderate in defining which processes need to be prioritized or performed in-house, but immoderate in the ambition to improve those that do.

How to measure?

Most comprehensive redesign programs set themselves the legitimate objective of developing world-class performance within a medium-term time frame of, say, two to four years. To date, many have found that their most useful tool for measuring the performance gap that needs to be closed is the international benchmarking of best practice. But this has its limits. Benchmarking what leading competitors do, no matter how good they are, may reflect only the old levels of best performance, not those achievable through a zero-based redesign. There is, of course, great value in making performance targets concrete. The danger, however, is that benchmarking can often be little more than a beauty parade among dinosaurs.

In the international insurance industry, many traditional multi-line companies have become less and less profitable during the past few years, outmaneuvered by specialist companies with leaner and more focused business systems. Long-established modes of operation have often proved wanting when measured against direct marketing by telephone or by dedicated salesforces. It is against the performance achieved by these new distribution channels, not by traditional competitors, that benchmarking should take place, for it is their levels of efficiency that will define the future shape of industry competition. Similarly, even the tiny handful of branch-based retail banks that have driven their costs as a percentage of their income down into the low 50s or high 40s are vulnerable to the more efficient direct operations being set up by successful niche players, including those in the insurance industry.

Thus, successful redesign begins not with standard industry classifications or comparisons, but with greenfield thinking about the best levels of performance in any relevant business processes, wherever they occur.

How to focus?

If the goal of redesign is to carry out the minimum amount of change needed to achieve the desired performance improvement, priorities must be set—and set early. But when, as is often the case, such efforts begin with managers splitting up a company into twenty or more core processes, how can meaningful priorities be established? Even deciding to focus only on those processes most obviously broken—and hence offering the greatest opportunities for improvement—will not necessarily lead to zero-based thinking or make possible world-class performance. Instead, it may well end with the fixing of an old business system, rather than the design of a new one.

Focus initially on corporate capabilities, not just on processes

One way to avoid this trap is to focus initially on corporate capabilities, not just on processes. "Capabilities" here refers to individual skills, as well as efficient processes, combined with an appropriate set of decision-making mechanisms. Together these three elements comprise the foundation on which any new business model or approach must rest. No greenfield vision can be successfully implemented unless it can call on the necessary organizational capabilities. Outsourcing aside, these internal capabilities can be built (or upgraded) only by improving the skills of the organization's people, by developing decision-making support systems that boost their effectiveness, and by improving the executional processes that flow from each decision.

In retail banking, the key capabilities are credit management, transaction processing, customer service, and salesforce effectiveness. The leading banks in the world have been able to ratchet up their performance by improving these capabili-

ties and rethinking the way each is applied—that is, by changing their established business models, for example by centralizing credit service and transaction processing or focusing branch operations on new selling techniques.

Some of the major US retail banks, like Banc One, have moved quickly to acquire smaller independents and build up a multi-state presence. This strategy has led to a redesign focus: first on developing transaction processing systems able to handle the business volume of the acquired banks, and second on establishing shared customer service processes. By contrast, some of the West Coast regional banks, such as Wells Fargo and Security Pacific, focused initially on removing credit, transaction, and service processing from their back offices in order both to break down their entrenched "branch credit" culture and to reorient their front-line people toward sales. Both these strategies—one centered on acquisition, the other on profitability—have proved successful. But their different emphases have, quite naturally, led to different sets of priorities for capability building and business model redesign.

How to achieve critical mass?

Any redesign effort large enough to produce a significant bottom-line impact has to stir up and sustain a critical mass of change in an organization's culture. Xerox reversed the decline in its market share and international standing by building a new service culture. British Airways made itself one of the world's most profitable airlines by adopting a similar focus on service. Such shifts in culture are enormously difficult to achieve. They are virtually impossible without major redesign activity.

At one European bank, a major change program overlooked the need to reorient the credit culture of branch managers. It left local credit processes and decision-making mechanisms in place, and asked the same people who already ran them to be responsible for building new sales and service skills. Being far more comfortable sticking with their traditional skills than developing new ones, they preferred not to, and so resisted the change process. As a result, some three years into the program, nothing had really gelled at the front-line level, new skills had not been developed, and corporate performance had not improved. This failure so weakened the already shaky credibility of top management that they could not resist a hostile takeover.

Most corporations are like giant jellies. Unless you can fundamentally reform the culture that holds them together, they will swiftly wobble back into their old form

If you put your people in a new business model, your hope is that they will behave differently. But if you merely exhort them to change their ways, an early shift in attitude may swiftly relapse into old behaviors once initial enthusiasm runs down. Most corporations are like giant jellies. You can force them briefly into a new shape. But unless you can fundamentally reform the culture that holds them together—the personal skills focus of middle management, for example—they will swiftly wobble back into their old form.

How to be distinctive?

Most organizations spend resources and energy supporting capabilities that are not essential to their competitive value propositions. In such cases outsourcing, not redesign, may be the best option. After a leading European retailer had defined its core capabilities as product-range planning, operational merchandising, and integrated logistics, it decided to outsource all non-core capabilities, such as IT and accounting. Another went further and outsourced parts of its all-important logistics capability. Too often, however, these critical make versus buy decisions fail to receive the top-level attention they deserve.

While trying to improve its new product development capabilities, one European insurance company got stuck in the middle of an over-ambitious program to redesign its internal IT-based product-support system from the ground up. The redesign team feared that using a standard software package would deny them any basis for competitive advantage. In fact, exactly the opposite was true. The package was so flexible that new products could rapidly be tailored to the needs of different customers and markets. Once the decision had been taken to use such a package and to redesign the company's processes around it, improvement came so swiftly that the company was able to break into new markets in less than a year.

How extensively to redesign?

The basic question to be answered is how extensive—that is, how radical—the redesign should be

If the central challenge for managers who face redesign is to strike the right balance between tight focus and aggressive ambition for performance improvement, much depends on the ability to know when tinkering with individual processes is enough, and when circumstances call for more radical change. Exhibit 1 indicates several different points at which an effective balance can be struck. Determining where to peg any particular effort requires careful thought about its scope (single- or multiple-project focus) and the organization's expectations for performance yield (one-off quick hits or integrated, business-system-wide improvements). The basic question to be answered, of course, is how extensive—that is, how radical—the redesign should be.

Process or program?

Without a clear, shared view about the proper level of radicalism early on, management expectations will vary all over the map, making disappointment likely and confusion inevitable. The work of reaching such a view, however, is usually treated as an informal art: Let's get all the good guys in a room and brainstorm. It should be more of a science: Let's quantify the impact of varying levels of redesign through decision engineering, comparing international best practice, and piloting alternative business models.

For one retail bank, we found that changing the branch-based credit process, as well as the clerical procedures associated with it, could reduce costs by 5 to 10 percent. Had this been all that was needed, it could have been accomplished by redesigning that one process. Benchmarking against world-class competitors, however, showed that the best performers were achieving both levels of cost reduction that were much higher and improvements in decision-making effectiveness that were reflected in significant loan loss reductions.

What the bank's competitors had discovered was that generalist branch managers usually spread their attention so thinly over so many areas—customer service, transaction processing, general administration, and the like—that they were unable to build and retain professional credit skills. In fact, a simple test showed that less than one manager in five still possessed basic credit skills. This discovery inexorably led the banks to a more radical change in their business model: centralizing the credit function completely and supporting it with sophisticated decision software based on expert systems. Further study showed that this new approach offered the potential for a 50 to 60 percent cut in costs, as well as major improvements in loss ratios.

Not surprisingly, the branch managers of these banks had strongly opposed such changes. They feared their jobs would be lost or downgraded. Top management had to think long and hard about whether the change was worth the pain and disruption it would cause. The answer—given the huge bottom-line impact of a fundamental shift in credit processes, as well as the knock-on effects of eliminating the old "credit culture" from the branches—was clearly yes. Redesign had to aim higher than cosmetic adjustments to a single process. A radical program was justified.

Is IT essential?

All redesign efforts, whatever their scope, involve the building of new skills. Even so, managers usually have a hard time quantifying or calibrating that task. One insurance company relied on developing multi-skilling among its clerical staff as the basis for redesigning its claims process. Along the way, however, it found that the redesign weakened individuals' specialist skills to such an extent that the gain in productivity was more than offset by much poorer performance resulting in the overpayment of claims. How, then, to estimate the likely effects of skill-building requirements on redesign efforts?

Quite often, formal classroom training and "sitting with Nelly" apprenticeships achieve reasonably rapid skills development. But the gains are often lost when employees return to their old work environment and experience a lack of reinforcement. Far more effective are techniques that rely on screen-based training, supported by "scripts" and expert systems that both build and reinforce new skills. When such IT-enhanced approaches are used, the training time for unskilled operators in computer-

supported telephone service centers has been cut to six weeks or less, even for complex tasks like insurance underwriting or credit authorization. Equally important, with ongoing IT support the new skills do not decay.

The role of IT in building individuals' skills may well prove to be the most valuable application of information technology so far discovered

The role of IT in building individuals' skills may well prove to be the most valuable application of information technology so far discovered. Effective redesign will increasingly depend on how well technology is harnessed to the different needs—and different skill requirements—of different core processes. The parts of banks handling commercial credit, for example, will need high levels of individual skill, sophisticated on-line decision support and information access systems, and streamlined administrative processes that make the best use of skilled people's time. The parts handling personal credit will need expert systems for credit scoring, automated work scheduling and administrative processes, and screen-based scripts to guide low-skilled operators through standard procedures.

Are steep learning curves possible?

Redesign efforts depend on a well-managed organizational learning curve—one that is flexible enough to adapt to the knowledge gained along the way

Where redesign calls for major changes in the interactions between people skills, new processes, and decision-support mechanisms, it is not usually possible to design all the necessary changes in advance. Instead, redesign efforts must depend on a well-managed organizational learning curve—one that is flexible enough to adapt to the knowledge gained along the way.

Some process designers have become expert in sketching and managing such curves. A retail bank kick-started an effective salesforce improvement program by building the core of the force around a cadre of sales people who were already skilled and trained in insurance. Next, it put in place a sales training program and a separate organizational structure for the established staff. Only then did it tailor state-of-the-art decision- support technology to the needs of the new salesforce.

Managers attempting to organize technology support for such gradual learning processes may find themselves up against the doctrinal purity preached by the guardians of their company's IT architecture. Traditional architectural theory stresses the importance of developing an integrated, consistent approach to making any change in an established system—a process that may take years to complete. Planning such change can itself be extremely inflexible: a long period of study leads to a massive "systems specification" document, which is then translated into a build program that spans several more years.

Some signs of change in this rigid scheme are now beginning to emerge. Several innovative start-up companies with their own approaches to highly flexible IT architectures are being supported by a growing body of niche technology suppliers eager to take advantage of the new opportunities. With their help, one bank has developed the entire systems support for its move to a centralized, telephone-based service model with an eight-person team in only six months. It has also built in the facility to support a continuous stream of upgrades with reaction times that can sometimes be measured in minutes.

How much organizational change is necessary?

Just where do the more traditional activities of organizational change fit with the newer behavioral- and technology-based disciplines of redesign? This is not an academic question. Too heavy-handed a usage of major organizational change up front can stop a redesign program dead in its tracks while internal politics and turf battles rage.

We have found that it is usually best to use these older techniques to reinforce the values and people skills within a major redesign—either after it has had time to settle down or as part of a disciplined process surrounding its implementation. When, for example, a large utility developed a new organizational model for its field operations, it refined it in detail in the context of a major redesign program, piloted it extensively, then rolled it out location by location over a two-year period. This controlled process produced none of the dangerous side effects of more traditional changes in organization structure.

Nor did it, as sometimes happens, upset the organizational consistency of its 7S framework. Redesigned processes that require multiple skills or cross-functional teams can, if used incorrectly, lead to culture conflicts and inefficient implementation. For similar reasons, many retail banks that have used multi-skilled cross-functional teams in their branches for decades are now returning to single-skilled functional structures. In fact, it is precisely to counter such problems that some companies are starting to experiment with what might be called "logical" structures—organizational arrangements that leave traditional functional groups in place and seek to forge process linkages through well-designed information flows alone.

How to plan for redesign

Major redesign efforts often require senior management to spend at least 50 percent of their time on managing the change program

Major redesign efforts often require senior management to spend at least 50 percent of their time on managing the change program. The need for such a degree of top-level involvement is thus a key factor in deciding how radical the effort should be. If, as seems increasingly likely, such involvement is needed to ensure the success of any redesign beyond a limited tinkering with a single neatly-bounded process, it may well be sensible, other factors permitting, to make the program as radical as possible. After all, with redesign as potentially the major vehicle for a corporation to attain world-class performance, it is reasonable to ask what other top management role could be more critical.

Is time important?

In his description of the logistics effort behind Desert Storm, General Gus Pagonis noted, "Several times in the combat zone we just stopped the 'train,' even during the ground war, brought all the commanders in a room, and put their subordinates in charge. If their subordinates could not handle the responsibilities, the commanders were to blame: they had not trained their subordinates enough."¹ In the same spirit, one of the fundamental preconditions of radical redesign is the need to free top management time by planned delegation of control-oriented responsibilities.

The senior managers who have transformed their schedule from the usual round of budget meetings and control activities to detailed involvement in redesign planning and execution know at first hand the dramatic increase in personal effectiveness that such a role switch brings. According to one chairman, who now regularly attends meetings of his company's redesign board, it was only after he saw the effectiveness of a disciplined process of translating opportunities into a sequenced program of deliverables that he realized how ineffective most of the traditional mechanisms of top-level management activity have been.

Does teamwork matter?

No major redesign programs succeed without an effective top management team in place. Given the radical thinking implied by the development of new business models, no existing policy or organizational issue can be held sacred. And given the need for speed, there can be no room for fundamental conflict at the top. The team must be in place, and all its members must be on board.

Some may not want to participate—or to stay. Listening to recommendations for a major redesign effort, one senior executive at a bank objected, "If we implement these proposals, we won't be bankers any longer." The idea clearly did not appeal; nor did being part of the team leading the change. So he left. Maybe he was right, but four years later the operating performance of the bank has become truly world class.

Believing as he did, however, he could not have been part of an effective team. Teams, after all, do not come into existence when a group of executives spend a weekend mountain climbing with a professional facilitator. As Jon Katzenbach and Doug Smith have shown, they crystallize only when "a small number of people with complementary skills [ ... ] are committed to a common purpose, performance goals, and approach for which they hold themselves mutually accountable."² Fortunately, we have found that the careful specification of a redesign effort and its organization into a sequence of initiatives often provide an excellent basis for top-level team formation.

Does sequence matter?

Traditional planning methods are not very good at integrating the technology and the behavioral aspects of a redesign effort—or, for that matter, integrating either with required changes in skill levels. This is a real problem. More than once, we have seen redesign efforts split into separate behaviorally- and technologically-driven programs that are pursued by different parts of a single organization. The result: a waste of huge investments, and minimal progress against overall redesign objectives.

One promising alternative is to sequence change through a "release program," which breaks all the separate initiatives into a series of deliverable chunks, integrated and packaged for easy implementation, and then incorporates them into a unified change program with tight performance targets. Such an approach is, in effect, an analogue of traditional strategic planning processes, where a five-year look ahead is split down into specific one-year consequences, which are then summarized into annual operating plans and budgets. In a release program, a five-year objective is carved into a three- to four-year sequence of regular "releases" of redesign—a sequence in which, to allow for technology build time, only the earlier releases are relatively fixed. Such flexibility avoids designing a change program that becomes a greater white elephant than the organization it sets out to improve.

Top management can add most value by rolling up their sleeves and involving themselves closely with the redesign activity itself

Release program management is still a new and unfamiliar discipline. There is a need, certainly, for a high-quality staff unit, both to help design the program and to monitor its progress, yet too much central bureaucracy can be dangerous: complex technology-based project management disciplines are usually inappropriate at the highest levels of program management. Ultimately, the responsibility for making a redesign stick will lie with an organization's cadre of line managers and top management team, who can add most value by rolling up their sleeves and involving themselves closely with the redesign activity itself. Using separate taskforces or delegating active involvement can easily cause programs to go astray.

When responsibility for one major redesign effort got delegated to a separate change board, the individuals on the board rapidly began to develop their own view of the corporation's future—which increasingly diverged from that of the CEO and his team. Eventually the CEO had to step in, disband the change board, and assume personal responsibility for the program—but not before large investments had been misdirected and much time lost.

The role of top management has changed forever. The line activity of the future will be managing the process of continuous change

True redesign efforts are never really completed. Top management responsibility is never really at an end. The perfectly understandable impulse, after a successful program, is to disband the program management function, allocate responsibilities back to the line, and return the organization to "normal." This is probably a mistake. By their very nature, the performance improvements achievable only through redesign may well mean that the role of top management has changed forever. The line activity of the future will be managing the process of continuous change.