Not long ago, continuity and stability were the watchwords of corporate operations. Now, with companies facing ever more demanding customers and tougher competition, ongoing operational performance improvement has become a strategic imperative. Executives are constantly trying to wring improved performance from their operations by following yellow-brick roads marked ERP Implementation, Balanced Scorecard, Supply-Chain Integration -- or with the names of any of a dozen other popular programs, including today's favorite, Six Sigma.
But although each of these performance-improvement initiatives can boost operating results, they need to be positioned under a process-management umbrella if they are to be successfully integrated. Otherwise, companies face the risk of program proliferation, of being burdened with a multitude of disconnected improvement efforts. Program proliferation is dangerous: It dissipates resources and creates confusion as people try to understand how each effort relates to the others. It engenders harmful competition among specialists, with each group promoting its program as the most deserving of resources. It also fosters cynicism; employees believe (correctly) that management cannot be serious about so many programs and hence think the company is serious about none.
Companies such as Bombardier, Air Products and Chemicals, Johnson & Johnson and Merck & Co. are avoiding the dangers of program proliferation by integrating their performance-improvement initiatives under the banner of process management. In so doing, they ensure that these efforts complement rather than compete with one another. And they are able to manage their portfolio of initiatives in an organized way. In particular, positioning six sigma under the process-management umbrella ensures that a company benefits from this important tool while avoiding the trap of applying it where it does not fit.
Process management is a structured approach to performance improvement that centers on the disciplined design and careful execution of a company's end-to-end business processes. Formally, a business process is an organized group of related activities that work together to create a result of value to customers (for example, order fulfillment, product development and post-sales support). The two most important words here are organized and together. All activities in a business process must work together; they must be aligned for the common purpose of serving customer needs. People must operate as a team instead of focusing narrowly on individual tasks and protecting turf. All activities in a business process also should be guided by a design that specifies which activities are to be done when and by whom. A process design ensures repeatability and consistency.
Generally, companies find it effective to identify five to 10 major business processes. Consider order fulfillment, which represents all activities from the time a customer places an order until the customer has received shipment and paid. That order-to-cash process transcends functional boundaries and integrates customer service, logistics, finance and even manufacturing to serve a common goal. Other top-level processes are similarly broadly encompassing.
Although a functionally organized company performs all the activities involved in turning an order into payment, it does not perform them as a process. The activities are handled by departments largely unaware of one another and pursuing different performance objectives. Conflict and suboptimization ensue, as well as increased overhead and non-value-adding work. Moreover, because no one is responsible from start to finish, there is no one to establish and enforce a precise and repeatable overall design; variation and improvisation are the consequences.
Process management ensures that activities are thought of, designed and performed in a process context. When employees recognize that their individual activities are part of something larger, they align around customers and common goals. When a process has an explicit end-to-end design, people can perform it consistently, and managers can improve it in a disciplined way. Process management ensures that a company's business processes are well-designed, that the designs are followed, and that they are kept upto-date. (1) (See "First Steps Toward Process Management.")
The central figure in process management is the process owner. (2) The process owner needs to ensure that the people performing the process understand it, are trained in it, have the required tools and are following the specified design. The process owner will make minor changes to the process design in order to fix a flaw or to address new issues. When there is a major gap between the current performance of the process and the desired performance, then the process owner must lead an effort to create an entirely new process design -- that is, a reengineering project.
The payoffs of process management are dramatic. Establishing a rigorous process design leads to a boost in performance because resources and time are not wasted on pointless efforts and projects do not fall between the cracks. For instance, in the mid-1990s Caterpillar introduced a formal, four-stage process for creating and introducing new models. As a result, cycle time for new-product development dropped from eight years to as little as three years. Similarly, a manufacturer of two-way private radio systems introduced a detailed 25-step selling process that concentrated resources on winnable and profitable sales oppurtunities. As a result, the manufacturer saw a quintupling of profitability.
Process management also delivers benefits by aligning everyone around a common, customer-oriented goal. A semiconductor manufacturer did that in the early 1990s. Customer orders had been taking up to 180 days to fill because the responsibility for filling orders was spread across multiple departments and locations, each of which had a different performance goal and none of which had overall responsibility for the order. By creating universal awareness of the whole process and introducing cycle time as a metric for which everyone was held accountable, the manufacturer eliminated suboptimizations, arguments and buck-passing -- and reduced cycle time by up to 75%.
Process management also provides a framework for reengineering -- the deliberate and holistic redesign of work. Enterprise success depends on the effective performance of well-designed processes. No matter how hard people work, they cannot exceed the capability of a process as it has been designed. Continuous improvement requires an improved design.
Consider how Progressive Casualty Insurance Co. made a quantum leap in its claims-handling performance. Originally, a claimant would report an accident to an agent, who passed the information to one of Progressive's customer-service representatives, who in turn passed it to a claims manager, who would batch the claim with others in the same territory and assign it to an adjuster, who would schedule a time to see the vehicle. It typically took seven to 10 days from the time of the accident until an adjuster showed up. Because rapid initiation of claims is a major shaper of customer satisfaction, Progressive resolved to improve. Recognizing that it could not slash waiting time merely by pushing employees harder (they were already working hard) or eliminating flaws in execution (there were none), the company decided in 1992 to redesign the whole process. Today, Progressive's adjusters are organized into teams, with one member handling claimant calls while the others are in claims vehicles in the field. When a ca ll comes in, an adjuster takes it and does whatever possible over the phone; if an inspection is needed, the adjuster contacts a team member in the field and schedules the visit. Progressive now measures the elapsed time from call to inspection in hours rather than days.
Companies as diverse as IBM Corp., Duke Power Co., United Parcel Service and Cadbury Schweppes have made process management the central theme of their programs of operational performance improvement. However, process management is more than a way to improve the performance of individual processes; it is also a way to operate and manage a business. Traditional management systems -- whether focused on measurement and compensation or on organizational structure and managerial roles -- are inimical to processes, having been designed for functionally centered organizations. But as process management becomes ingrained, all the organization's management systems refocus to support processes. People work in teams, not departments; their compensation is linked to results, not activities or seniority; managers are coaches, not supervisors. Computer systems are integrated to support end-to-end processes, not individual departments, and the culture encourages both individual accountability and collective responsibility. A n enterprise that has institutionalized process management and aligned management systems to support it is a process enterprise; it is centered on its customers and managed around its processes. (See "The Traditional Versus the Process Enterprise.")
The Centrality of Process
Besides directly leading to better performance, process management also provides a framework and context for integrating performance-improvement initiatives. Some of the most popular contemporary approaches to improving performance are based on technology deployment; others center on installing new performance-measurement systems; still others are more strategic in nature. Yet a careful look at virtually any of them reveals a common underlying theme: business process.
Enterprise-Resource-Planning Systems Technically, an ERP system is a family of software modules that share a common database and readily interface with one another. In practice, because its modules are so tightly integrated, an ERP system is a tool for supporting end-to-end business processes. Implementing an ERP system surfaces and transforms a company's processes. When companies pursue ERP without understanding their own processes, they are courting disaster. An ERP-implementation effort needs to be positioned and managed primarily in terms of creating new business-process designs and secondarily in terms of installing a software system to support those designs.
Supply-Chain integration A supply chain is just an intercompany business process that crosses corporate as well as functional boundaries. It includes all the activities needed to create and deliver results to a company's customers -- activities performed by the company, its suppliers, their suppliers and so on. Companies find the techniques of process redesign useful for supply-chain and other intercompany processes.
Merger Integration Turning previously independent enterprises into a single company is, in effect, an exercise in business-process standardization. The leaders of the merged company must decide which processes are to be common across the enterprise and which may vary. They also must decide if a standardized process should be based on an existing process or be entirely new.
Globalization Creating a global business means installing uniform business processes for all geographies. Instead of country managers deciding how their businesses will operate, global process owners create uniform processes and country managers implement them. In that way, global customers interface with the company as a single entity, not as a loose confederation of autonomous geographical units.
E-Business Now that Internet hype has abated, it is possible to recognize the real meaning of e-business: providing outside parties (particularly customers and suppliers) with electronic access to a company's systems and the business processes they embody. However, putting a Web site in front of lousy processes merely advertises how lousy they are. Success in e-business requires processes that permit easy external access and operation.
Other Initiatives Balanced-scorecard and economic-value-added (EVA) techniques develop measurement systems that translate corporate objectives into process-performance goals. Customer-relationship management (CRM) allows all of a customer's interactions with the company to be handled as a single process. The list goes on.
Thinking of Initiatives in Process Terms Managing such efforts in process terms has two salutary effects. First, it enhances the initiatives' effectiveness by ensuring they are appropriately carried out. When ERP implementation, for instance, is positioned not as a technical, CIO-led project but as the installation of software to support process designs authorized by process owners, it is more successful. Second, it enables companies to position initiatives in a larger context. Employees need not worry about previous efforts being abandoned for the fancy of the moment; each program merely adds new support for the existing framework of process management. Rather than continually deploying specialized groups that will inevitably compete with other groups implementing other initiatives, companies can give process owners responsibility for deciding which technique is right for each situation.
The process-management framework is particularly important when it comes to the latest performance program to capture managers' imagination, six-sigma quality.
Making Sense of Six Sigma
Technically, six sigma is a specific measure of quality -- namely, 3.4 defects per million opportunities. Most companies operate at a much lower level of quality (typically in the two-to-three-sigma range, from 66,000 to 300,000 defects per million opportunities). But the technical definition is irrelevant. Some companies don't need to reach that level, whereas others need to go even higher.
In practice, six sigma has become a code name for a set of methodologies and techniques used to improve quality and reduce costs. The six-sigma methodology that is most widely used -- the one promoted in best-selling books -- is known as DMAIC (define, measure, analyze, improve and control). DMAIC offers a structured and disciplined methodology for solving business problems. (3)
The premise of six sigma is that companies need consistently higher levels of quality and lower levels of cost and that a disciplined, organized approach will root out the variance, waste and errors that plague operations. Six sigma does not address business problems at a superficial, phenomenological level but by attacking root causes. If defects are too high, one does not reduce them by inspecting for defects and then discarding defective items; rather, one carefully measures and analyzes operations to determine exactly how and why defects occur and then takes steps to address those causes.
With DMAIC, a problem is first defined and quantified then measurement data are collected to bound and clarify the problem; analytic tools are deployed to trace the problem to a root cause; a solution for the root cause is identified and implemented; and finally, the improved operations are subjected to ongoing control to prevent recurrence. The six-sigma tool kit includes a variety of techniques, primarily from statistical data analysis and quality improvement. Many tools are familiar from the era of total quality management (TQM); others are more recent and more sophisticated. (4)
Six sigma is a project-based methodology. That is, the unit of activity is a project that applies the DMAIC methodology to solve a specific performance problem recognized by the organization. Projects are performed by project teams led by six-sigma black belts (expert practitioners). A rule of thumb is that a project should take three to six months to complete and should yield savings of $150,000 to $500,000; a black belt should be able to lead four to six projects per year.
The power of six sigma lies in the discipline it provides for coping with the complexity of business operations. Many different factors could be the cause of a quality problem: miscalibrated machine, raw material that is not up to specification, an operator who performs a task incorrectly. Rather than trying random solutions, a company using six sigma pinpoints the cause of a problem and applies only appropriate solutions. Six sigma has been aptly compared to detective work: filtering through clues in a logical way to solve a problem.
The Wrongly Accused Supplier A pharmaceuticals company was experiencing rapid growth in demand for a compound, but the compound's production line had low yields. The company thought the fault lay with poor quality in a raw material from a supplier. However, six-sigma analysis uncovered the real problem: variation in temperature in different parts of the production facility. Adjusting the HVAC system to standardize temperature improved yield by 60%, leading to direct savings of more than $17 million. Moreover, the improved yields enabled the manufacturer to postpone a $500 million capital investment for constructing an additional production facility.
The Case of the Inexperienced Forklift Operator A polymers manufacturer fielding customer complaints about product contamination hypothesized numerous causes. Six-sigma analysis narrowed the problem to products shipped from a particular warehouse in boxes that for some reason had holes in the bottom. It transpired that, when picking up a box for shipping, inexperienced forklift drivers were inadvertently puncturing the box stacked behind it. The solution was to shorten the forks on the forklifts.
The Air-Bubble Mystery As part of quality control, a pharmaceuticals manufacturer routinely took samples of tablets, dissolved them in a liquid and then conducted tests on the solution. Unfortunately, the test results often conflicted, necessitating high rates of rework. Six sigma traced the problem to a step in which air was added to the liquid to facilitate tablet dissolution. Exactly how much air to add was not specified in the step, and consequently different testers added different amounts. The answer was to designate precise parameters for aeration and to reassign testers who could not perform the task as required. Rework was essentially eliminated, quality improved, and costs were reduced. (5)
The Trajectory of Six Sigma
In the 1980s, building on its work in TQM, Motorola developed and organized the six-sigma methodology and achieved both financial benefit and widespread recognition. However, relatively few other companies followed suit -- until General Electric Co. did so in 1996. GE maintains it has saved billions of dollars through the use of six sigma; former CEO Jack Welch once described six sigma as the most important initiative GE had ever undertaken.
With GE's highly publicized success, interest in six sigma went from a trickle to a tidal wave. At least 25% of the Fortune 200 claim to have a serious six-sigma program, including Ford Motor Co., Bank of America Corp., Eastman Kodak Co., DuPont and American Express Co. In June 2001, attendees of a conference on process-based performance improvement were surveyed regarding their companies' use of six sigma. Of the 65 responding companies, 40 were already using it, and most of the others expected to begin soon. Companies' commitments are not minor. Ford, for instance, has trained 2,500 black belts and has nearly 2,000 projects under way.
Despite the extravagant claims of its advocates (one widely used book begins with the assertion that six sigma is "the most important breakthrough management tool ever devised"), six-sigma success is not synonymous with business success. (6) Some of its early adopters -- Kodak, Xerox Corp. and Polaroid Corp., among others -- have experienced significant business reversals recently. Even Motorola has seen its performance fall and rise and fall again, despite its ongoing practice of six sigma. GE is virtually alone among six-sigma practitioners in showing consistently superior performance. Moreover, many companies have successfully improved operations without six sigma. For instance, IBM's revival in the mid-1990s began after it had experimented with and then abandoned six sigma.
Even companies that have been successful with six sigma have learned that it cannot do everything. In that regard, the experiences of Bombardier are instructive. Bombardier is a Canadian corporation active in aerospace, rail transportation, recreational products and financial services. In 2000, it had more than 80,000 employees and revenues of $16 billion (Canadian). Having had a positive experience with TQM in the early 1990s and having learned about six sigma, the company's leaders decided in 1996 to commit to it: 200 black belts were trained in the DMAIC methodology; more than 500 projects were launched. Bombardier achieved five-year net cash-flow savings of $137 million (Canadian) on an investment in six sigma of $21 million.
In 1999, Bombardier undertook a formal assessment of its six-sigma efforts and concluded that despite the success achieved, those efforts suffered from serious limitations:
* Most six-sigma projects were narrowly focused, concentrating on low-level and small-scale activities, typically within one functional unit of the organization. When managers attempted to apply six sigma to larger-scale projects, the results were unsatisfactory until the scope was narrowed.
* Six sigma was not well aligned with the strategy of the organization as a whole. Although each individual project was worthwhile, in the aggregate the projects did not contribute to larger corporate goals.
* The six-sigma efforts had not gotten at the company's basic assumptions or its functional organizational structure. Because breakthrough improvements in performance require just such fundamental change, six sigma's impact was limited.
In other words, although DMAIC-based six sigma can lead to higher quality and lower costs, it is not effective at generating dramatic improvements in business performance. In the last decade, companies such as IBM and Allmerica Financial Corp. have reduced costs by hundreds of millions and even billions of dollars while simultaneously achieving huge increases in customer satisfaction; they have done so through concerted programs of transformation, not through DMAIC. The managers at Bombardier also concluded that DMAIC was not a vehicle for business transformation. Other early implementers of six sigma, such as American Express, Merck, Motorola, and Air Products and Chemicals, have reached similar conclusions.
Six sigma's limitations are inherent in its nature as a project-oriented, problem-solving regimen. It deploys statistical analytic tools to uncover flaws in the execution of an existing process. Those tools do not raise questions about whether there is an entirely different way of performing the process. Six sigma assumes that the existing process design is fundamentally sound and just needs minor adjustments to be more efficient. That assumption is not the road to dramatic improvement.
The project orientation that makes six sigma manageable simultaneously limits its power. Companies undertaking it convene many largely independent project teams that address self-contained problems. It is common to have dozens, even hundreds, of such teams in operation. Organizing around a large number of small-scale projects does offer a high likelihood of success and attractive returns on investment, but it cannot lead to a systemic attack on a company's major problems.
How does six sigma relate to a company's other performance-improvement initiatives? Is it sui generis and stand-alone, or does it fit into a larger picture? Once again, process management comes to the fore. The best way for a company to manage six sigma is as one more weapon in the armamentarium of process management.
Fitting Six Sigma Into Process Management
Business processes provide the context for DMAIC. Before beginning a six-sigma project, a company should create a SIPOC (supplier, input, process, output, customer) model of its processes, relate the problem to be solved to a specific business process, and map that process (document its steps). The process map provides an analytic framework for identifying the specific actions causing the observed problem (say, aerating a liquid or picking up boxes with a forklift). Once the problem's source has been identified, the remedy (establishing specifications for aeration or installing shorter forks on the lift) is typically straightforward and localized to the problematic activity.
Although DMAIC ensures the correct execution of the steps of a process and enables minor modifications to the process design, when there is a large gap between current and desired performance, a process will need an entirely new design. Creating one is something for which six sigma, with its analytic rather than creative orientation, is unequipped. When a process is operating consistently but at a performance level lower than required, and when the reason is that the overall design is flawed, six sigma will spin its wheels. There was no single narrow root cause why Progressive's process for handling claims was taking 10 days. Indeed, there was no "problem" with the execution of that process; it was operating correctly and consistently. When market circumstances demanded that Progressive come up with a breakthrough approach that would allow claims to be handled in hours rather than days, an entirely new design was needed.
The distinction between improving process performance and creating a new process design is related to the distinction between waste and non-value-adding work. Waste comes from individual activities erroneously or inconsistently performed; DMAIC is effective at tracking down the sources of waste so they may be eliminated. Non-value-adding work, on the other hand, is work that holds the process together and so cannot be eliminated readily. Process redesign aims to reduce the amount of non-value-adding work. Identifying non-value-adding work is the easy part; the hard part is reorganizing the process so that less of it is needed.
Moreover, six sigma does not get the performers of a process aligned around common goals. With six sigma, business processes are just the framework for a problem-solving regimen. The notion that they are the central organizing theme of the enterprise is missing. As Bombardier's managers concluded in 1999, the company was doing six sigma but not being six sigma. The ideas underlying six sigma were apparent in the context of specific performance-improvement interventions, but not in the day-today operation of the business.
Six sigma is not a system for operating and managing a business. It does not transform a company into a process enterprise.
Nor does it address the changes to culture and mind-set, measurement and rewards, organizational structures or other management systems needed to make processes the central axis of an organization. In six sigma, even process ownership is only a weak afterthought and overlay on an existing structure. (7)
To overcome six sigma's limitations and get the most out of it, companies should position it in the larger context of process management. In a process-managed organization, the process owner is responsible for ensuring peak performance of business processes. When a problem appears, the relevant process owner decides whether, when and how to address it. If it appears amenable to a six-sigma solution, then the process owner convenes a project team; if broader change is needed, then a process-redesign team, employing different tools, is organized. Thus, six-sigma teams are formed only when appropriate; moreover, the process owner ensures that all six-sigma projects are aligned and integrated to achieve the strategic goals of both the process and the enterprise. (See "The Do's and Don'ts of Six Sigma.")
Bombardier took just such an approach after its assessment of its six-sigma efforts. The company instituted an organized program of process management: It identified performance goals for its major processes, supplemented the DMAIC methodology with a methodology for holistic process redesign, and instituted a mechanism for determining whether the situation called for a process-improvement team or a process-redesign team. American Express, Merck, Motorola, and Air Products and Chemicals have taken similar steps to bring DMAIC under the process-management umbrella.
Some companies want the six-sigma label even though they have recognized the limitations of DMAIC; they are stretching six sigma to include process management and process redesign. A few are applying to process design a GE-developed variant of DMAIC originally developed to guide the development of high-quality products: DMADV (define, measure, analyze, design and verify).
But it's better to call things what they really are. Process management and process redesign antedate DMADV and any extension of six sigma beyond DMAIC. Stretching the definition of six sigma to encompass process redesign and process management is like stretching the definition of basketball to include baseball. It is also likely to create confusion, because to most of the business world, six sigma is synonymous with DMAIC. Moreover, process design is not the same as product design, and DMADV's effectiveness as a tool for process design is still open to debate. Six sigma should be part of process management, not the other way around.
The Big Picture
Process management is the culmination of the movement to transform business operations. It provides a unifying theme for initiatives directed at improving organizational performance. As companies move along the six-sigma learning curve, they will encounter the same challenges they confronted with previous initiatives: project proliferation, limited payback, competition with other efforts and inapplicability to larger problems. After picking the low-hanging fruit, they will find themselves in need of a broader and more robust approach. Subsuming six sigma beneath the process-management umbrella addresses those challenges and allows companies to reap its substantial benefits while keeping it away from areas where it won't work.
Because process management entails major changes to virtually all management systems, it demands absolute commitment from executives. Companies that rise to the challenge will garner extraordinary rewards: not just cost savings, but accelerated new-product introduction, major improvements in customer satisfaction and sharp increases in profitability.
Michael Hammer is the president of Hammer and Company a management-education and research company based in Cambridge, Massachusetts. His latest book is "The Agenda: What Every Business Must Do To Dominate the Decade."
(1.) For a discussion of process management and process enterprises, see M. Hammer, "Beyond Reengineering: How the Process-Centered Organization Is Changing Our Work and Our Lives" (New York: Harper Business, 1996); and M. Hammer and S. Stanton, "How Process Enterprises Really Work," Harvard Business Review 77 (November-December 1999): 108-118.
(2.) The challenges of process ownership are explored in M. Hammer, "The Agenda: What Every Business Must Do To Dominate the Decade" (New York: Crown Business, 2001), 137-140.
(3.) See, for instance, F. Breyfogle III, "Implementing Six Sigma: Smarter Solutions Using Statistical Methods" (New York: Wiley-Interscience, 1999); P. Pande, R. Neuman and R. Cavanagh, "The Six Sigma Way" (New York: McGraw-Hill, 2000); and M. Harry and R. Schroeder, "Six Sigma: The Breakthrough Management Strategy Revolutionizing the World's Top Corporations" (New York: Currency Doubleday, 2000).
(4.) Breyfogle, "Implementing Six Sigma," provides a good summary of the six-sigma tool kit.
(5.) The first and third examples were obtained through personal communication with the leaders of six sigma at the companies in question; the second example is found in Harry, "Six Sigma," 250-254.
(6.) Harry, "Six Sigma," vii.
(7.) For instance, see Harry, "Six Sigma," 175; and Pande, "The Six Sigma Way," 346.
The Traditional Versus the Process Enterprise Traditional Enterprise Process Enterprise Central Axis Function Process Work Unit Department Team Job Descriptions Limited Broad Measures Narrow End-to-end Focus Boss Customer Compensation Activity-based Results-based Manager's Role Supervisor Coach Key Figure Functional executive Process owner Culture Conflict-oriented Collaborative The Do's and Don'ts of Sigma Don't ... Do ... ... oversell its capability. ... recognize both its power and its limitations. ... position it as the only ... leverage it using process performance-improvement tool. redesign. ... apply it to all business ... use it to solve narrow problems. problems in fundamentally sound processes. ... fall prey to a flood of ... organize projects to uncoordinated projects. achieve strategic goals. ... establish it as the province ... bring it under the control of an autonomous group. of process owners. ... stretch the term six sigma ... recognize that six sigma equals beyond recognition. DMAIC (define, measure, analyze, improve and control). ... make it the umbrella for ... bring it under the process- performance improvement. management umbrella.
RELATED ARTICLE: First Steps Toward Process Management
* Identify company processes, typically five to 10.
* Make people throughout the company aware of the processes and how their own work fits in.
* Create and deploy measures of end-to-end process performance, derived from customer and share-holder needs. Assess current process performance and set targets.
* Designate process owners: senior managers with end-to-end authority for a process, responsible for ensuring consistently high performance. The process owner establishes the process design, ensures that the design is followed, obtains resources that the process requires and intervenes as needed to improve the process.
* Select two or three processes for redesign an improvement. Implement those new designs in a staggered fashion.
* Over time, align the company's management systems with the new prominence of its processes.