The Triumph of Process
Revolutions often begin with the intention of only improving the systems they eventually bring down. The American, French, and Russian revolutions all started as efforts to ameliorate the rule of a monarch, not to end it. Reform turns into revolt when the old system proves too rigid to adapt. So, too, the revolution that has destroyed the traditional corporation began with efforts to improve it.
For some twenty years managers of large American corporations have been engaged in a relentless effort to improve the performance of their businesses. Pressured by suddenly powerful international (especially Japanese) competition and ever more demanding customers, companies embarked on crusades to lower costs, improve productivity, increase flexibility, shrink cycle times, and enhance quality and service. Companies rigorously analyzed their operations, dutifully installed the newest technological advances, applied the latest management and motivational techniques, and sent their people through all the fashionable training programs--but to little avail. No matter how hard they tried, how assiduously they applied the techniques and tools in the management kit bag, performance barely budged.
The problems motivating managers to make these efforts were not minor. The operating performance of established corporations was grossly unsatisfactory, especially when compared with that of aggressive international competitors or hungry start-ups. Some cases in point:
- Aetna Life & Casualty typically took twenty-eight days to process applications for homeowner's insurance, only twenty-six minutes of which represented real productive work.
- When buying anything through their purchasing organization, even small stationery items costing less than $10, Chrysler incurred internal expenses of $300 in reviews, sign-offs, and approvals.
- It took Texas Instruments' Semiconductor Group 180 days to fill an order for an integrated circuit while a competitor could often do it in thirty days.
- GTE's customer service unit was able to resolve customer problems on the first call less than 2 percent of the time.
- Pepsi discovered that 44 percent of the invoices that it sent retailers contained errors, leading to enormous reconciliation costs and endless squabbles with customers.
This list could be extended indefinitely. The inefficiencies, inaccuracies, and inflexibilities of corporate performance were prodigious. This was not a new phenomenon; it was just that by 1980 these problems were starting to matter. When customers had little choice and all competitors were equally bad, there was little incentive for a company to try to do better. But when sophisticated customers began deserting major companies in droves, these problems rocketed to the top of the business agenda. The persistence of performance problems in the face of intense efforts to resolve them drove corporate leaders to distraction.
After a while, understanding gradually dawned on American managers: They were getting nowhere because they were applying task solutions to process problems.
The difference between task and process is the difference between part and whole. A task is a unit of work, a business activity normally performed by one person. A process, in contrast, is a related group of tasks that together create a result of value to a customer. Order fulfillment, for instance, is a process that produces value in the form of delivered goods for customers. It is comprised of a great many tasks: receiving the order from the customer, entering it into a computer, checking the customer's credit, scheduling production, allocating inventory, selecting a shipping method, picking and packing the goods, loading and sending them on their way. None of these tasks by itself creates value for the customer. You can't ship until it's been loaded, you can't pack until it's been picked. A credit check by itself is simply an exercise in financial analysis. Only when they are all put together do the individual work activities create value.
The problems that afflict modern organizations are not task problems. They are process problems. The reason we are slow to deliver results is not that our people are performing their individual tasks slowly and inefficiently; fifty years of time-and-motion studies and automation have seen to that. We are slow because some of our people are performing tasks that need not be done at all to achieve the desired result and because we encounter agonizing delays in getting the work from the person who does one task to the person who does the next one. Our results are not full of errors because people perform their tasks inaccurately, but because people misunderstand their supervisor's instructions and so do the wrong things, or because they misinterpret information coming from co-workers. We are inflexible not because individuals are locked into fixed ways of operating, but because no one has an understanding of how individual tasks combine to create a result, an understanding absolutely necessary for changing how the results are created. We do not provide unsatisfactory service because our employees are hostile to customers, but because no employee has the information and the perspective needed to explain to customers the status of the process whose results they await. We suffer from high costs not because our individual tasks are expensive, but because we employ many people to ensure that the results of individual tasks are combined into a form that can be delivered to customers. In short, our problems lie not in the performance of individual tasks and activities, the units of work, but in the processes, how the units fit together into a whole. For decades, organizations had been beating the hell out of task problems but hadn't laid a glove on the processes.
It wasn't surprising that it took managers a long time to recognize their mistake. Processes, after all, were not even on the business radar screen. Though processes were central to their businesses, most managers were unaware of them, never thought about them, never measured them, and never considered improving them. The reason for this is that our organizational structures for the last two hundred years have been based on tasks. The fundamental building block of the corporation was the functional department, essentially a group of people all performing a common task. Tasks were measured and improved, the people performing them were trained and developed, managers were assigned to oversee departments or groups of departments, and all the while the processes were spinning out of control.
Slowly and even reluctantly, American corporations began in the 1980s to adopt new methods of business improvement that focused on processes. The two best known and most successful were total quality management (TQM) and reengineering. Through a long period of intensive application of these techniques, American businesses made enormous headway in overcoming their process problems. Unnecessary tasks were eliminated, tasks were combined or reordered, information was shared among all the people involved in a process, and so on. As a result, order of magnitude improvements were realized in speed, accuracy, flexibility, quality, service, and cost, all by at last attending to processes. The application of process-oriented business improvement programs played a major role in the competitive resurgence of American companies and the revitalization of the American economy in the 1990s.
So far, so good. But to paraphrase an infamous statement from the Vietnam era, process-centered improvement techniques saved companies by destroying them. By bringing processes to the fore, the very foundations of the traditional organization were undermined. A disregard for processes had been built into the structure and culture of industrial era corporations. The premise on which modern organizations were founded, Adam Smith's idea of the specialization of labor, was in fact a rejection of process. It argued that success was based on fragmenting processes into simple tasks and then resolutely focusing on these tasks. By attending to processes instead, the new improvement efforts created stresses that could not be papered over.
Who would have control over the newly recognized and appreciated processes? Consisting as they did of diverse tasks, processes crossed existing organizational boundaries and thereby imperiled the protected domains of functional managers. The new ways of working did not fit into the classical organization. They often entailed the use of teams, groups of individuals with various skills drawn from different functional areas. But such teams had no place in the old organizational chart. Whose responsibility would they be? The new processes often called for empowered frontline individuals who would be provided with information and expected to make their own decisions. This was heresy in organizations where workers were considered too simple to make decisions and where the need for supervisory control was considered a law of nature. In short, it quickly became clear that the new ways of working that marvelously improved performance were incompatible with existing organizations: their structure, personnel, management styles, cultures, reward and measurement systems, and the like.
