Click here for footnotes, Ch.6. 

Research was conducted at the Motorola Corporation at their formal quality briefing, "Stalking Six Sigma", on May 16, 1991. Motorola was one of the first-year winners of the Baldrige Award, which was presented to the entire corporation in 1988, rather than to a single plant or site location.

Motorola was founded by Paul V. Galvin in 1928 as the Galvin Manufacturing Corporation in Chicago. The company began by producing a "battery eliminator" for home radios and shortly thereafter began to produce radios for the burgeoning automotive industry. They expanded over a 40-year period to producing a variety of consumer, commercial and defense electronics and telecommunications equipment. In the face of Japanese competition in the 1970s, Motorola was forced to sell their television business; in 1980, they bowed out of car radios as well, to concentrate on high-tech commercial, industrial and defense electronics. The company remains a family-owned enterprise: Robert Galvin assumed the reins of the business from his father in 1959, and it appears that his son, Christopher will assume the chairmanship when Robert Galvin retires.

Currently headquartered in Phoenix AZ, Motorola has manufacturing and distribution facilities across the U.S. and worldwide, in Canada, Europe, Mexico, Asia, and Australia. It employs approximately 104,000 people. In 1990, it was listed as 42nd in the Fortune 500; is the third ranking electronics manufacturer in the United States and 14th in total exports. It is also the only US company to supply cellular phones to Japan. Motorola produces electronic equipment, systems and component products, including telecommunications equipment (cellular phones, pagers and 2-way radios, and supporting infrastructure systems); integrated circuits and discrete semiconductors; defense and aerospace electronics, data communications and information processing equipment. The company has a strategic commitment to the field of high-tech electronics, a fairly volatile market. In 1990, the company grossed between $7-8 billion in sales worldwide.

Competition is fierce, particularly in the low-end market in communications, and there is significant competition in both microprocessors and cellular phone technology. Major competitors in the semiconductor area include NEC, Toshiba, Hitachi in Japan; Texas Instruments in the US; and both Philips and SGS-Thomson Microelectronics in Europe. Matsushita Electric Industrial Company is the primary competitor in cellular phone technology, and L.M. Ericsson poses significant competition in the European telecommunications market.

Motorola phased out of consumer electronics in the 1970s. Currently, about 80% of their customers are commercial and industrial, 5% are government, and about 10% of their customers are in the military and defense industries. Lifecycles have shortened for products across the board and particularly in the low-end market. Minimizing the time required for the development cycle, from design inception to market, is critical in both low end and middle markets, and is characterized by fierce competition, particularly from Japanese competitors.

Much of the labor force in manufacturing and assembly has only a high school education or its equivalent. Moreover, as Motorola ramped up its education and training in the early 1980s to ensure computer literacy throughout the workforce, they discovered that many of the plant employees were functionally illiterate and could not perform at the 7th grade level in mathematics competency. Many employees in U.S. plants did not speak or read English, but managed to do their jobs from verbal instruction from bi-lingual co-workers and supervisors.

Motorola has a tradition of being a "family-oriented" business. It is not uncommon for three generations of family members to be employed by the company. Much employee social life traditionally centers around the workplace and stability of employment is the rule. Although no guarantees exist of permanence of employment for new hires, after 10 years of service, an employee becomes a member of the Motorola Service Club. This means, among other things, that the employee cannot be fired for any reason without the written consent of the Chairman of the Board. Senior managers, upon retirement from active service, remain officers of the company. (There are presently 250 officers.) The company also tends to have a coherent corporate perspective, since most senior managers, including the Chief Financial Officer, are engineers. Paul Noakes, Sr. Vice President for External Quality, characterized Motorola as a strong "people" culture which "invites and requires employee participation in managing the affairs of the company."

Motorola has used a variety of pieces of quality strategy from "Quality Gurus" like Deming, Juran and Crosby, but they found that none of the strategies as available met their needs and expectations. In the early 1980s, they obtained a videotaped course by Juran to use for all manufacturing employees. They found that their employees lacked the basic knowledge of statistical process control assumed by the course, and discovered that it did not solve the problems they hoped it would solve. They did not have the infrastructure to be able to put the information to use, and it was not well-suited to Motorola's needs.

According to Robert Galvin, Chairman of Motorola Inc., the genesis of the organized quest for Total Quality occurred at an Annual Officers meeting in 1978. In spite of the fact that the company was "profitable and growing" (though suffering from Japanese competition in consumer electronics), National Sales Manager Art Sundry announced that "though our agenda is going fine, we are missing [i.e. deficient] in the area of quality." Shortly thereafter, Bill Smith, then a senior engineer, called for a meeting with Galvin to discuss problems with latent defects in product, a cause of early life failure which is attributable to errors in the manufacturing process. Galvin and other officers set about investigating customers' perceptions of quality, and how satisfied they were with Motorola's performance. At the same time, an executive task force was formed to spend several months looking at other companies to investigate what Motorola could do to improve quality of both service and manufacturing.

Over the next several years, Motorola reset their policies and processes, focusing on "Total Customer Satisfaction". They spent entire days speaking with the operational personnel in customer shops, those who installed, repaired, purchased and processed payments for Motorola equipment. They discovered that mistakes were rampant in ordering and billing. From Xerox corporation, they learned about benchmarking, which involves finding companies who perform outstandingly well in particular operations, and using their level of performance as a goal, and investigating their process to enhance one's own performance. A Quality Officer was appointed to oversee the transformation of Motorola's departments into a Total Quality Operation via education, communication of principles and practices, new management structures and practices, and quality system reviews. Company-wide goals of five, ten and one hundred-fold improvement were set in both "hard" and "soft" activities, in support and administrative departments as well as design, purchasing and manufacturing. Councils for quality improvement were initiated in most departments, including shipping, engineering and service departments. A common rating system was developed, based on the "Six Sigma" concept developed by Motorola which is described below. This measurement system extended to all departments and operations, and was intended to provide a common language and metric for improvement throughout the operation. A major focus of the quality strategy along with reducing defects was the reduction of cycle time, the time required to complete any process. Tremendous attention was focused on the analysis and restructuring of processes to eliminate defects and achieve enormous cycle time reduction.

Motorola crafted its own quality strategy based on their "Six Sigma" statistical foundation, and utilizing current technology and techniques. "Six Sigma" refers to the number of standard deviations from the mean of a manufacturing process with normal (random) variation. If there is no shift of the mean over the entire period of manufacture (or, in general, production), then four sigma will produce an acceptably low rate of defects. However, Motorola engineers discovered that in batch processes, the shift of the mean over time approaches plus or minus 1.5 sigma. In this case, four sigma acceptability produces an unacceptably high rate of defects: 6210 parts per million. However if "six sigma" (as defined by Motorola) of a product being produced falls within the acceptable tolerance limit, then the defect rate for that product will be 3.4 per million--virtually defect-free. This goal is particularly crucial in products with a large number of parts or sub-assemblies, since the defects of the whole will depend on the product of the defect rates of the parts. Conversely, the engineering design challenge is to design fault-tolerant products, minimizing the number of parts while maintaining standards of technical performance and features.

This model is applied to all areas of Motorola's business, including administration and support functions such as accounts receivable and accounts payable. To standardize a metric for all corporate operations, Motorola has devised a normalized measure applicable to processes of different complexity. They calculate PPM (parts per million) as equal to Total Defects Per Unit (or operation) divided by Opportunities for Defects per Unit (or operation), from which sigma can be calculated. This is as applicable to office processes, such as invoices, orders, or personnel processing forms as it is to manufacturing operations. Motorola has found that the common language and metrics have been extremely important to company-wide success. The corporation has been able to mandate, and largely achieve, company-wide goals of five times and ten times improvement in a wide variety of areas over a five-year period.

In the context of "Six Sigma", many other techniques and technologies have been employed to serve the quality improvement goals. Concurrent engineering of process and product, with extensive use of computer simulation has been a cornerstone of the development process. They have modified use of Just-In-Time inventory and production techniques. Sophisticated information technology is also an important component of Motorola's success, but the technology has not been implemented as CIM per se. The information systems improvements (including selected automation) have occurred as a result of systematic infrastructure and process improvements, i.e. to streamline tasks and communications after process analysis and improvement has been accomplished, rather than using information technology as a direct force to create change.

The company's success in monetary terms has been impressive. Through defect elimination between 1987 and 1989, Motorola saved over $550 million in 1990 manufacturing costs, with a cumulative savings for the four year period of over $1.5 billion. It estimates that there may be a potential for savings in non-manufacturing costs of about $1 billion additional.

Motorola has embarked on an intense and vigorous program of education for all employees, through the Galvin Center for Continuing Education and Motorola Manufacturing Institute, opened in 1986, and later through Motorola University, established in 1989. One order of priority was to bring the skills the non-literate and math-deficient personnel up to required levels for their new responsibilities. Another was to train employees in the new methods for quality improvement. In 1987, 14,000 engineers were trained in Design for Manufacturability and Design for Assembly over a 12-month period. Statistical Process Control, and Short Cycle Management "how to" courses were implemented the same year, along with the first wave of training for Motorola Suppliers. In 1988, a course on Structured Methods for Software Engineers was mandated, and 54,000 employees received their fundamentals in "Understanding Six Sigma". After the founding of Motorola University in 1989, Total Customer Satisfaction training expanded throughout the corporation, followed by an expanded curriculum for employees and suppliers in all facets of quality implementation, from effective team processes and participative management, to software development strategies for reusability. 1990 also saw the founding of the "Six Sigma Research Institute", whose mission is to "accelerate achievement of Six Sigma Quality; [develop] R & D theoretical framework and supporting tools; [and to] facilitate transfer of knowledge and tools."

According to Bill Smith, Motorola University's goal is to bring "the right training to the right people at the right time." The goal is for every employee at Motorola to spend a minimum of 40 hours per year in training. In 1990, 90% of employees did so, and the target projected is 100% by 1993. Many employees spend more time than the minimum in training, depending on the needs of their jobs. 40% of all training is in the area of quality improvement processes, principles, techniques and objectives. Another major area of training is in group process techniques for effective teamwork. This training is usually given to people in the teams they will work with, and is considered extremely important, since team processes are the backbone of the "Six Sigma" process. Everyone in the company is trained in basic statistical analysis techniques essential to the "Six Sigma" process, which they are expected to know and be able to apply to their work area. Additional training in advanced statistics and analysis is required for people in technical areas. In addition, training in general problem-solving as well as specific job-related training is provided on-site, or through local community colleges or trade schools. Job-related training and college coursework related to career advancement is sponsored and supported as well.

In addition to the common training in Six Sigma and Total Customer Satisfaction, top-level managers and officers participate in the Senior Executive Program, an annual course focused on topics to improve quality performance in all its aspects, with action for improvement following the program. Among the major results of this course are the fact that it "gave all officers a common language", facilitating rapid progress in the development of a coherent quality culture; the officers developed a clear understanding of tools and techniques used by the workforce in the pursuit of quality; and the company has undertaken continuous benchmarking of what other companies are doing to improve quality, a major source of successful quality ventures.

Managers (including directors and vice presidents from all functional areas) participate in courses offered by the Motorola Management Institute which focus on manufacturing, engineering and operations excellence. Courses include world-class excellence in manufacturing and engineering, systems approach to global management, development of proactive management styles, efficient use of resources, and the overview of the new measurement system implemented by Motorola management. This training resulted in work-in-progress inventory reduction of over 50%; reduction of over 40% in manufacturing cycle time; factory and support services space reduction of 40%; and more than 35% quality yield improvement.

At this point, Motorola claims a return on investment rate of thirty-to-one on the training they have done. Their goal is to create the "Best In Class Workforce". To do so, they believe that the key elements, on which they are currently focused, are to select the right employees for the right jobs, based on testing on job-related tasks; assess skills of incumbents with constant measurement and feedback; maintain in-house training and education to develop prerequisite skills and knowledge using curriculum based, ongoing training and education for every employee to keep pace with changing job requirements. In addition, they are looking at how to improve people's learning abilities (learning to learn) and their critical thinking skills, and have embarked on projects to see whether these approaches are successful in enhancing employees work performance.

The use of teams has been termed "a cultural standard" at Motorola since the beginning of their Total Quality journey. All employees are members of one or more teams. Managers participate in routine cross-functional management teams. Non-managerial employees are all members of standing work groups, and are routinely tapped as members of task-based or problem-related cross-functional teams.

The key to developing new products at Motorola is to identify products which will be a clear improvement over what is available on the market today (not just a clear improvement over the in-house product it replaces); engineer it with a more fault-tolerant design; improve the process of production; reduce the cycle time. All new product development is done with cross-functional teams, with early supplier involvement, and extensive use of computer simulation, under a program manager (not an engineering manager) who takes responsibility for the total success of the project. After a "Blitz" phase, where five or six product managers identify a desirable product to develop, the development process goes to the "Contract Book Phase". In this phase, a multifunctional management team determine what groups of people will be involved by identifying all functional areas which will touch the process of development through sales and shipment to customer. After "hashing out" a detailed implementation plan and determining functional roles and needs, this information is entered in a Contract Book, and signed by all participants as a formal commitment. The plan cannot be changed unless everyone, from marketing to shipping, agrees. Budgeting and staffing is provided to match the plan. Only then does the project enter the development phase. In addition to the cross-functional design team, designs go through independently staffed cross-functional reviews. Finally, the project can enter the Program Completion phase, as the production plan is carried into action. As part of the total monitoring system, there is ongoing management review and measurement of the process, though not of the individual decisions and plans.

Cross-functional problem-solving teams are crucial to both the day-to-day plant operations for existing products, and to the increased quality and reduced cycle times in staff and administrative areas. When a problem, is identified by any employee and brought to the attention of a manager, they are empowered to assemble a team of operational level employees from every function which touches the process in question. The first job of the cross-functional team (made up of people who actually perform the work) is to build a formal map of the process, identify bottlenecks and non-value-added parts of the process. They are empowered to redesign the process and implement the improvements. These teams are fairly autonomous. When they are given a problem, they are responsible for developing all aspects of the solution. This strategy has worked extremely well according to Kathryn Linder, one of the Baldrige Award team members. In fact, she notes, one of the things that causes failures in these efforts is the manager who fails to relinquish enough authority and responsibility to these teams.

Teams have had enormous impact on cycle time reduction and, concomitantly, on quality in both manufacturing and administrative support operations. One striking example is the impact that the new team-based product development strategy has had on soldering defects in integrated circuit products. Most soldering defects can be traced to problems with the manufacturability of design. Soldering defects prior to 1987 were running as high as 6000 defects per million. In 1987, the defect rate was reduced to 70 per million. By 1990, soldering defects were reduced to the Six Sigma goal, no more than 3.4 defects per million parts.

Similarly impressive strides were made in the speed with which customer orders were processed for delivery. In the pager manufacturing division, before the Total Quality initiative was mounted, it used to take 40 days to process customer orders, from the time the order was taken by a sales person until the product was received by the customer. Through a series of team processes, combining parts, improving all component processes (including sales, billing, shipping as well as manufacturing), and installing a computer system to facilitate and implement some of the processes, the cycle time has been reduced to one hour and forty minutes.

Julie Fenton, an operations level employee in the finance department identified the problem of excessive errors made in invoicing customers. She organized a cross-functional team to solve the problem and implement the solution. Within the first year after implementation, there was a 70 times improvement. In the following 15 months, NO errors were found in invoices.

The entire focus of Motorola's quality approach is centered on analyzing and correcting process. According to Motorola research, under the best conditions (of continuous, accurate timely feedback) unless the process which produces the output of interest is analyzed and corrected, the best performance which can be expected is 3-4 sigma. It was pointed out explicitly at the briefing: People don't make mistakes because they are stupid. Intelligent, educated people make just as many errors as those of average intelligence, about 4 sigma. To change the number of errors or defects, change the system (to make it easier/more foolproof to use) and provide specific training so your people can redesign and use the system. The combination of better ways of doing things, with immediate feedback, and the tools to analyze the feedback is the way to improve performance.

The byword at Motorola is that if something is considered important, it must be measured, fed back, analyzed, and improved. To achieve continual improvement and total customer satisfaction, Motorola relies on measurement systems to track progress at both macro and micro levels, with common metrics, applied uniformly throughout the company. At the operations level, this is accomplished by operational personnel tracking and analyzing with statistical techniques. At the management level, it is accomplished by Quality Audits. These audits are designed to ensure that departments and divisions understand and are pursuing key corporate goals; behaving in accord with key beliefs; according to key corporate initiatives (the approved processes for analyzing and improving operational processes); and finally that they are obtaining the desired and mandated improvements in performance.

Measurements must be reflective of the overall corporate goals. To support Motorola's goal of Total Customer Satisfaction, a re-examination of a variety of corporate measures and models has been undertaken by a committee of the Six Sigma Research Institute. Among other measurement models, they are reformulating such models as the cost accounting system. On a more mundane level, they used to track the difference between the time the factory promised delivery of product to the customer, and the time it was actually shipped. Now they measure the difference between the time the customer desires to receive it, and when it actually arrives.

Bill Smith states that you must measure what you believe will drive the business. The importance of measurement of vital objectives seems obvious. However, at the seminar, he followed this statement with a quiz for the 250 managers present. "First", he said, "how many of you believe that if you make fewer mistakes, and do it faster, you will improve your business?" Virtually all hands were raised in agreement. "Now, how many of you measure Defects per Unit and cycle times?" Only six hands went up in response.

According to Motorolan Val Feigenbaum, you may be able to make a cultural change by teaching everyone the philosophy, but even if you can, it takes too long. Motorola made dramatic cultural changes in 9-12 months at the beginning of the Six Sigma initiative. Smith explained. Motorola did not have a quality program as an overlay on top of the existing system. "Instead, we said that everyone is responsible for quality, therefore, everyone will measure quality." People measure themselves to provide feedback for the improvement process, and are empowered to act on that feedback to correct problems, in co-operation with other operational employees. The truth according to Motorola is: if you measure operations and evaluate based on quality results, people will change the way they think as well as changing their behavior. Their experience seems to bear them out.

All employee evaluation policies have been reviewed, and modified as necessary to support the goals of Six Sigma. Evaluation for promotion of managers is tied closely to their participation in successful co-operative quality activities with their crossfunctional management team and departments, rather than being evaluated on the individual department success which often comes at the expense of other departments. Managers are evaluated on other criteria as well. Since it is the policy of Motorola to empower employees, managers are rewarded for their pro-active stance in encouraging participative management. Since training of employees is a top priority, managers and workers arrange a schedule for the worker to participate in classes during working hours, a burden for the manager. Part of the worker's evaluation includes successful completion of the agreed upon coursework. The managers are also evaluated on whether or not their employees completed the scheduled training. Failures of employees to complete the classwork counts against managers as well.

Motorola also noticed that some existing policies were responsible for behavior which did not support the corporate goals. For example, Kathryn Linder, a former sales representative notes that there used to be a problem getting salespeople in the Communications sector to turn in orders to be processed in a timely fashion. This was a critical impediment to reducing order cycle time. Sales people were rewarded according to their quarterly sales, and since they would rather be out selling than in the office processing orders for shipment. In fact, sales representatives would often hold orders all week, process them in batches, and submit them to a supervisor at the last minute. The supervisors, unaware of the orders which had been placed, were nervous at not meeting their expected sales levels, were then enormously relieved, and sales representatives would appear heroic. The corporate focus on Total Customer Satisfaction changed the evaluation of sales representatives. Now they are evaluated on their average time-to-submit, the time it takes from the initial writing up of the sales order at the customer's request to the time it is received at the area office. With no monetary investment except the installation of FAX machines, the average time-to-submit was reduced from 8.1 days to l.7 days. Changes such as these are occurring systematically throughout Motorola. As one senior executive said, "Your people will do what managers want and reward them for doing."

Senior executives are compensated based on the achievement of corporate goals as well, not primarily on the basis of the individual success of their functional area. Everyone is evaluated based on the achievement of Total Customer Satisfaction.

Motorola also believes in sharing their financial successes with all employees. This is primarily accomplished through a profit-sharing plan voted in by the employees in 1947, which has remained virtually unchanged since. All executives are members of the Motorola Executive Incentive Plan (MEIP) whereby they share in the profits of the Corporation as a whole. All other employees participate in the Participative Management Program (PMP) and are compensated according to their division, sector, or local organization profits. All rewards are linked to Key Corporate Initiatives. The Key Initiatives drive sector and division long-range planning strategies and goals. Long range plans drive sector and division annual goals. Sector and division goals establish the executive incentive goals (MEIP), which in turn determine the balance of the employee (PMP) initiatives and goals. Thus, the goals and rewards are consistent and congruent throughout the organization.

There are two special employee recognition awards of note. SABA, the Science Advisory Board Associate award instituted in 1970 provides recognition for exceptional creative engineering talent and ability. In 1988, there were 24 inductees including 2 quality professionals. In the prior 3 years, another 6 quality professionals had been inducted. The "Crown Jewel" of awards is the Chief Executive Officer's Quality Award, instituted in 1984. This award is designed to reward individuals and groups who have made significant contributions to quality. Receipt of this award is publicized throughout the company and mentioned in the annual report. So far, 40 awards have been given to a total of 5,100 people since 1984.

In addition to formal awards Motorola has a tradition of "spreading the word" about success stories, providing both informal employee recognition, and creating grass-roots quality heros. Motorola also routinely recognizes exceptional contributions with promotions and management positions. Julie Fenton, who initiated the solution to the invoice error problem a few years ago, is now the manager of a manufacturing plant.

Motorola's entire strategy is characterized by closed, short feedback loops: the tracking of all important information by people who are empowered to act to correct deficits and improve performance. Motorola is explicitly aware of the importance of closed loop (feedback processes). Open loop processes, which have delayed or largely absent or sporadic feedback, cannot create quality, in the words of Motorola quality leader Paul Noakes. The best that can be expected in an open loop process is the achievement of four sigma. Six sigma demands closed loops with timely feedback (preferably "continuous, immediate feedback"), with correction of process. (If process is not corrected 3-4 sigma is still the best that can be expected.)

In addition to tracking internal performance, Motorola has recognized that at least two classes of people outside its organization are crucial to its survival, and therefore are important elements of the "Motorola system": customers and suppliers. Routine, personal contact between Motorola executives with operational personnel at customer organizations is one important aspect of the feedback loop. Another is the incorporation of customer input early in the new product design process. Ongoing formal tracking of customer satisfaction is a third.

Motorola's relationship with suppliers is even more extensively looped in. In addition to working with suppliers early in the development program, and developing ongoing, guaranteed relationships with suppliers who meet quality and performance criteria, Motorola places additional demands and offers additional support. It requires all of its major suppliers to apply for the Baldrige award within the next five years, to ensure that they have analyzed their organization, and received professional feedback, on their organizational quality infrastructure. In return, Motorola offers advice, assistance, and access to a variety of courses at Motorola University to enable suppliers and Motorola to work more closely toward mutual quality goals.

The coherence of the fundamental infrastructure at Motorola is obvious and impressive. Goals are not only clear and well-communicated, they are supported functionally with ongoing audits, measurements, evaluations, incentives, and reward systems, all of which are congruent with corporate goals. Feedback loops are closed in virtually every operation and behavior of interest, at each level of organization as well as vertically throughout the organization. Decision-making (including error correction) is as proximal as possible to the choice or error. Audits ensure that meta-level process criteria are met and that results are achieved, while the process design and corrections are handled by crossfunctional teams of operational personnel involved in the process. Personnel at each level are systematically trained in the specific skills required for them to take maximum responsibility for their work, individually and collectively, and are granted the authority to do so. The version of the concurrent engineering paradigm adopted for the design of new products is isomorphic with the description of the ideal heterarchical control system described in chapter III.

Communication and co-ordination of planning and evaluation is streamlined not only by common language and goals, but by common metrics as well. Process correction is the sine qua non of the organization. Motorola provides a most striking example of an explicit cybernetic infrastructure along the lines this work has described, and in fact exceeded all expectations of what might be found in industry.

Research at IBM's Rochester, Minnesota plant was conducted at their Quality Showcase on July 17, 1991. Unlike Xerox and Motorola, whose corporations as a whole were recipients of the Baldrige Award, IBM's Rochester plant itself received the award. Nevertheless, the history, traditions, and quality initiatives from the IBM parent organization were instrumental in the Rochester plant's achievements.

Founded in the 1880s, the company which was to become IBM was originally known as the Computer-Tabulating-Recording Company of Elmira, Ohio. It had developed an electrical punchcard system which was used for the 1890 national census, and branched out to produce commercial scales, meat slicers and adding machines. In 1914, the faltering C-T-R Company was taken over by Thomas Watson, Sr., a veteran of the aggressively marketing-driven NCR Corporation, who turned C-T-R around and gave the company its reputation for entrepreneurship, concern for employees welfare, and employment security. Renamed International Business Machines in 1924, the company focused on producing timeclocks, punchcard tabulators and other data processing machines for business and industry. IBM pioneered the electric typewriter in the 1930s and 1940s. After Remington-Rand developed the UNIVAC computer in 1951, which replaced IBM equipment at the Census Bureau, IBM aggressively developed business-oriented mainframe computers. During the 1960s and 70s, IBM became the premier producer of business computers and computer-based systems, holding 80% of the market, and began to produce related products including high-end paper copiers and personal computer systems.

IBM International now has manufacturing, distribution and research facilities worldwide, and dominates most of the markets in which it participates. From the company's inception, Thomas Watson Sr.'s three key personal beliefs were incorporated into its business doctrines: respect for the individual; best possible service for the customer; and the pursuit of excellence. These beliefs have stood IBM in good stead, and provide a firm cultural basis for the pursuit of Total Quality. However, the principles alone were not enough to keep up with the competition in the modern complex environment.

The IBM Rochester was opened in 1961, and from its inception has been one of the most productive plants with the highest morale and success rates of any plant in the IBM corporation. Located in Rochester, MN and currently employing about 8100 people, it has the traditions of the rural Midwest work ethic, an outstanding local public education system, and a homogeneous workforce.

The first 5-1/4" floppy disk drive system was developed at the plant, as were the original System 3X (34, 36, 38) series of midsized business computers. IBM Rochester currently manufactures DASDs (Direct Access Storage Devices, or hard disks) used in their AS400, PS/2, RS/6000, and 9370 systems. It has most recently developed and manufactures the AS400 system itself. These systems are used primarily in meeting the data processing needs of small to medium-sized companies with 20-500 employees. They also supply DASDs to other OEM manufacturers and, competitively, to other IBM operations.

In 1988, the Rochester plant received specific encouragement from Corporate headquarters to apply for the Baldrige Award, since there were several factors both culturally and historically in their favor. Under the leadership of Larry Osterwise, IBM Rochester applied for the Baldrige Award in 1989 but failed to qualify. With the insights gained from the Baldrige evaluation, they made a number of internal changes and reapplied the following year. In 1990, IBM Rochester (MN) received the Baldrige Award. Based in part on insights gained from the Rochester experience, IBM Corporation as a whole initiated an aggressive Market-Driven Quality Initiative (MDQI) that year to radically improve the company's operations and customer service.

The market in mid-sized computer systems and peripherals is volatile. There is intense competition for IBM Rochester's DASDs, largely from Hewlett Packard, Seagate and Connors. Major competitors in the area of mid-sized computer systems include Digital Equipment Corporation, Hewlett Packard, and UNISYS in the US; Siemens, Bull, and ICL in Europe; and Fujitsu, NEC, Toshiba and Hitachi in Japan. Many US and European competitors rely on offshore operations to cut their labor costs.

The Rochester MN plant employs approximately 8100 people. About 1300 of these are part time or temporary employees, receiving no benefits or job security. Wages are generally the same as those paid to a first year employee in the same job category. The maximum period of employment for a temporary employee is generally one year. The balance of the workers are regular IBM employees. They come under the company's no-layoff policy, which includes provisions for transfers and retraining of employees who are no longer needed in their present location or position.

Located in a small city in rural Minnesota, IBM Rochester has been a major employer in an area beset with a fluctuating local economy. The Rochester workforce has demonstrated a powerful middle America work ethic. The plant has traditionally made good, successful products, and the workforce has consistently had the highest morale of any workforce in the corporation, demonstrated by results of an annual corporate-wide employee survey. The employee turnover rate is extremely low. The safety record at the plant is 60% higher than the industry average.

About 66% of the employees at Rochester are engineers, programmers and other professionals, who generally hold Bachelor or Master of Science degrees or the equivalent. Most manufacturing employees have a high school education, and some have two year vocational college degrees, or equivalent training.

According to IBM veteran Bob Mesarchik, IBM Corporation as a whole initiated quality efforts in 1981, when Fujitsu began moving into European and US markets, and the Japanese threatened to take over the computer industry as they had the auto industry. Over the last ten years, IBM has taken a variety of focused approaches to improve the quality of products as well as customer service and operational effectiveness beginning with a Zero Defects program in 1981. However, CEO John Akers observed, that wasn't enough, because the competition kept getting better.

After visiting a number of other firms involved in the quality effort in both the US and Japan in the late 1980s, they began in earnest to develop their own quality initiative, drawing in part on Motorola's Six Sigma concept.

IBM Rochester began with a Zero Defects program and PRIDE (people responsibly involved in developing excellence) initiative in 1981, focusing on cost of quality and product performance and reliability improvement. In 1984, they concentrated on manufacturing cycle time improvement and focused on business and manufacturing processes themselves. 1986 marked the beginning of active customer and supplier involvement, along with competitive benchmarking, and improving the development cycle time. During 1989, the year they first applied for the Baldrige Award, IBM Rochester shifted significantly toward an systemically integrated approach of total customer satisfaction, focus on total business process (rather than the tight product focus of previous years, a focus criticized by the Baldrige Award evaluation), and "closed loop quality/management system." This phase focused on reducing total cycle time to market, process development for the prevention of defects, and functional integration around the market-driven strategy. The 1990-94 plan is organized around six "Critical Success Factors" and embodies a model now embraced by IBM Corporation as a whole. The six factors are: product strategy and planning; improvement of customer requirements definition; Six Sigma defect-level quality strategy; creation and deployment of an Excellence in Education plan; enhance employee involvement; reduction of total cycle time (inception to delivery). A system of closed loop measurements are designed and continue to be enhanced to ensure continuous improvement toward Rochester's customer satisfaction goal. Each senior manager at the Rochester plant "owns" one of these factors, meaning that they are responsible for the process planning, implementation and, ultimately, its success.

Several generic industry practices are employed and tailored to achieve these goals. Concurrent engineering is used endemically in the development process, with extensive use of simulation of the entire system being developed, rather than simply the simulation of functional component performance, resulting in major improvements in reducing development time, errors, and subsequent problems. Statistical process control, based largely on the Six Sigma principles is used throughout the factory, as is a tailored Just-In-Time production system (Continuous Flow Manufacturing). Work groups are used throughout the manufacturing operations, where all members are cross-trained to do each others' jobs. Early and continuous customer involvement and supplier quality certification and early involvement are of primary importance.

Overall these strategies have yielded impressive results. IBM Rochester productivity improved 30% between 1986 and 1989; time for development of the AS400 was cut in half, compared to projected estimates for development time based on its System 3X predecessors; the manufacturing cycle time has diminished 60% since 1983; and product reliability has increased threefold during that time.

IBM engages in ongoing classroom and on-the-job education. All employees are required to pursue at least 40 hours of job or career related training every year, some of which is available in onsite classes and seminars, and in on-line courses. All employees receive one week of orientation training within 30 days of being hired.

For management employees, this training is given at company headquarters in New York and consists primarily of IBM philosophy, people management, and policies and procedures. This is followed by one week of local management training. An additional week of standardized management training is required at the beginning of the manager's second year to be sure the manager is "on-track" with IBM corporate practices, values and attitudes. Each subsequent year, 40 hour of additional training is required, part standardized and part elective.

Non-management employees receive one week of training in IBM practices and procedures, in addition to being familiarized with IBM product development process. Manufacturing employees receive additional training in safety and chemical awareness, as well as specific job task training in both classroom and on the job settings. A large number of in-house courses exist for further training of manufacturing employees as needed, in statistical technique, group and team processes, and additional job cross-training. All non-management employees are similarly required to take at least forty hours of additional job or career related training each year.

With the advent of the total quality effort, all managers and key technical people are required to take leadership courses on how to empower employees. Empowerment, as defined by IBM, is the practice of managers giving their subordinates responsibilities and the authority to act on them without losing management accountability. This requires managers to shift from being doers (or detail-specifiers) to being leaders, i.e. goal setters and coaches. The General Manager of Rochester acknowledges that "it's the right concept, we're committed to it, but we still have a long way to go." As part of the IBM corporate Market Driven Quality Initiative, IBM has instituted corporate-wide Transformational Leadership training, delivered on a cascade model: managers train their immediate subordinates, from the top down. This training includes the concepts, vision, philosophy, terminology and a general overview of process management. In addition, Six Sigma concepts and general statistical information will be provided to all employees by their management.

Teams of every variety are used in all areas at IBM. Teamwork is a traditional part of IBM, but so deeply embedded in the culture as a way of doing business that they are not uniformly mandated, formally tracked, or formally empowered as new decision-making entities. Tom Dutke, a human resources specialist from the Rochester Baldrige Award Application team reports that it was awkward to answer that portion of the Malcolm Baldrige application because they had no formal numbers or statistics. Most teams at IBM naturally form and disband and change composition as needed. This observation was corroborated by the comments of both management and manufacturing employees, as well. He estimates that now almost 100% of IBM employees at the plant are on teams of one sort or another. The most common of these in manufacturing is the standing work group. Official management estimates at Rochester place team membership at about 82% of the employees at any given time.

Speakers from the manufacturing operations described their participation on Process Improvement Teams and as part of standing working groups. Standing work groups are comprised of employees who work co-operatively on various manufacturing tasks. Members are typically cross-trained in each others' jobs, and have a considerable degree of autonomy over the performance and co-ordination of their tasks.

Process Improvement Teams are ad hoc and typically cross-functional organizations with informal structure. One member may start the meeting and/or keep the minutes, but these roles are informal and may rotate among members of the group. Manufacturing PITs are empowered to develop solutions in manufacturing areas and work with Process Owners (persons officially designated to be responsible and accountable for process results) on their implementation. Membership is voluntary and informal PIT, and PITs often contain a mix of line employees, technical people and managers.

In their own words, IBM Rochester quality managers describe the nature and importance of these improvement teams:

As a result of process improvement teamwork, many processes have been analyzed, streamlined and incorporated into an integrated on-line system. Operations like the Engineering Change process has been successfully moved to an on-line environment, with significant gains in reduction of product cycle time, efficiency and communication effectiveness.(QJC) The new Continuous Flow Line was developed with the co-operation among quality, manufacturing and test engineering departments, reducing AS400 manufacturing time from 24 to 14 hours.

Product development teams were originally composed of design engineers, with some informational input or participation from the Marketing Department. As successive quality initiatives were pursued, development teams evolved to include not only marketing, design, and manufacturing personnel, but also representatives from purchasing, site support, plant facilities and any other departments which might impact or be impacted by the development project. As a result of this evolutionary activity the Product Development Manual, the IBM "bible" for the development process, was updated recently to reflect this experience. Almost all team activity consists of face-to-face meetings augmented and assisted by electronic communications among participants. All IBM Rochester personnel have terminals and E-mail facilities for communications.

Development teams using simulation, early manufacturing and continuous customer involvement, and parallel/concurrent development techniques have reduced the development cycle time for the AS400 system at least 40%. Customer satisfaction with the product is evidenced by the fact that IBM's share of the world market for mid-sized computers increased in both 1988 and 1989. In addition, engineering changes to designs have dropped 45% since 1983.

All teams have autonomy within their mandates. The rule is that process solutions be developed and implemented with consent and co-operation of those affected by the change. Ultimate authority and responsibility for process changes rests with the respective process owner(s), who are often members of the team themselves.

As the corporation adopted Market Driven Quality Initiative, fifteen key processes were identified at the Corporate level, and a number of processes at sub-system levels. Each key process required an executive process owner, who is responsible for the achievement of goals associated with the process. For instance, the Chief Financial Officer became responsible for a 10 times reduction of billing errors in the company over a two year period. Failure of the company to achieve those goals will cost the executive a portion of his salary. As previously mentioned, Rochester executives "own" critical success factors, and key manufacturing and development processes are similarly "owned" by lower level managers throughout the plant.

Evaluation for all employees below the highest level of management has always been an individually tailored process between the employee and his or her immediate manager, known as Performance Planning, Counseling and Evaluation. Specific goals, tasks, and areas for improvement are identified and negotiated between employees and supervisors, with few corporate standards explicitly specified. No formal changes have been implemented in this process, but process ownership and concomitant success may be assumed to be incorporated in the individual evaluations of managers as a reflection of IBM Rochesters' commitment to quality initiatives, and each managers interest in their own success, which is tied to the performance of their employees. According to Tom Dutke from Rochester's Human Resources Department, as Market Driven Quality Initiative values are incorporated into IBM culture, PPC&Es are converging to embody and reflect these values across the board. In addition, he indicated there may be plans to incorporate important aspects of these values, such as effective teamwork and quality focus, into the PPC&E process more formally.

IBM does not have any form of profit-sharing, though most IBM employees are stockholders in the company. IBM has, somewhat paradoxically, been an individually oriented company, where pay, bonuses and promotions were based solely on individual merit. One of the most renowned bonuses in industry is the IBM special bonus pay for suggestions. If the suggestion is adopted, the employee is awarded a percentage of the money saved or profits gained as a result of the suggestion, which can often run into tens of thousands of dollars, and occasionally much higher.

Even when success above and beyond the call of duty was produced by an informal team effort, however, individual tasks could usually be distinguished and special rewards would go to outstanding individuals who participated. More effort lately has been committed to granting team awards, split equitably among team members.

Basic structural categories of awards are given by Corporate Headquarters. Each IBM site has a slightly more detailed structure, and titles for each award. Management teams in various functional areas have a wide dollar range for each award, and enormous latitude in both the actual dollar amount for each award, and the criteria by which it is bestowed. Some management department teams have specific criteria stated for each award, for consistency's sake, but generally managers rely on their own judgement and informal consultation with other managers and teams as a "sanity check" on their judgements. Awards may be given to individuals or to teams as a whole. Managers at Rochester felt that little official recognition is given to teams or individuals outside of the monetary awards, though Rochester's Baldrige Award was given tremendous notice throughout IBM and recognition for a variety of activities of IBM employees is provided by the company magazine. No official company policy seems to exist for the publication of contributions to quality which are not rewarded monetarily.

It is recognized that having appropriate measures is of the utmost importance to success at every level of the organization. According to Jack Scheetz, Program manager for MDQI, "The organization you have is a direct result of the measurements and evaluations you use." (from videotape) Hundreds of separate measurements exist at all levels, in interlocking and cascading feedback/action loops. Measures themselves are not uniform, following the principle of IBM culture: maximum autonomy. Six Sigma analysis and statistical process control per se are applied only in manufacturing. Development processes use minimal and specific statistical tools, primarily the X-bar chart, standard deviation calculations and other tools developed for specific needs. Administrative services, similarly, does not use standard Six Sigma or SPC terminology or processes. They do, however, use a variety of data-based analytical techniques, and are held to the same kind of order of magnitude improvements as the manufacturing and development side of the house.

After unsuccessfully applying for the Baldrige Award in 1989, IBM Rochester was chided by examiners to focus on process rather than on product, and to focus on the success of the enterprise as a whole. As a result, IBM Rochester shifted away from evaluating individual functional subsystem performance and focused instead on evaluating the total process integration and outcome. This approach has been applied to both business activities (evaluating and encouraging co-operation among functional departments) and to the computer systems development process.

According to development team member Roy Bauer, the AS 400 system was developed using extensive simulation modeling for both the hardware and microcode, and a historically based bug detection model for associated software. However, instead of simulating each component or chip individually, a more extensive model was developed to simulate the entire hardware and microcode system to the I/O level. Customers were brought in at milestones in this development process to provide feedback on the simulated performance. The results were that no prototype was ever built by R&D. Manufacturing built the first system with no false starts, due both to the extensive simulation modeling and the fact that manufacturing had been involved in the development process from the beginning. No hardware changes were required for the first two years that the system was in use. And the entire system was designed and built in twenty-eight months. Simulation modeling and the cross-functional team process cut an estimated ten months from the hardware development time and sixteen months from the software development time.

IBM has not changed incentive systems for their employees explicitly except for the process ownership and its consequences. However, IBM has had an extremely strong cultural bias in favor of internal co-operation in its plants, particularly in Rochester. Co-operation at the Rochester plant increased as required to meet company goals; most incentives were exceedingly flexible and tailorable to current needs as a matter of IBM standing policies and practice. The culture also supports feedback from employees to managers on their management styles and effectiveness, via an annual (on-line) anonymous survey of employee attitudes and observations about their workplace and working conditions. The information is analyzed, and results fed back to the appropriate managers at all levels. Though no formal and official ramifications exist for managers who do not measure up in the eyes of their employees, there is an extremely strong cultural bias toward management excellence, and toward the legitimacy and importance of this feedback. Managers at Rochester were quite emphatic that this evaluation was enormously effective in getting managers to change their behavior, even though there were no top-down sanctions associated with the feedback.

The Human Resources department at Rochester reports that though the empowerment of employees is hard to measure, responsibility has definitely shifted downward. Layers of management have decreased from eight levels to five in the past three years and spans of control have increased from 8 or 9 people per supervisor to 12 to 15. Given the equal or greater number of decisions which must be made, managers must count on team leaders and technical people to act on their own to implement changes. All line employees in manufacturing have the authority to shut down not only the line they are working on but the entire plant, if required to correct quality problems. Both of these things have happened within the last year. Finally, the Human Resources department analyzes the annual anonymous IBM employee survey, which is intended to feed back information on employee perceptions and satisfaction. The results of three or four key questions indicates that employees feel significantly more empowered on their jobs over the last three years.

IBM Rochester, as mentioned above, has begun to elicit customer input early in the development process and on an ongoing basis throughout the process to be sure the product will fit customer needs. Early customer input on designs is provided to Rochester by a Customer Advisory Council who meet several times a year. In a radical departure from previous tradition, the company discloses its development plans two or three years in advance to elicit customer opinions and suggestions. To enable ongoing customer participation during development, Rochester has created test labs where customers can try out the simulated system at various stages of development to give feedback on ease of use and adequacy of the system based on their expectations and desires. This way, potential problems can be addressed in the planning stages, before products or even prototypes are constructed.

IBM Rochester has no direct partnership ties with the IBM marketing and distribution organization. To improve customer service, and elicit first-hand customer feedback on the AS400 system, Rochester initiates a call to every customer who orders an AS 400 within 30 days of its installation for an open-ended query on customer satisfaction and needs. If the customer is having any problems with the system or the installation, Rochester initiates a call to the customer's local distribution center which must contact the customer and within 10 days develop a comprehensive plan to fix the problems and inform Rochester in writing.

IBM Corporation has developed a new information system with a graphical interface is on-line providing electronic access to key information about the business as part of the move toward data analysis for quality improvement based on facts. Information is available and integrated worldwide, to IBM locations, marketing and service as well as manufacturing, and to customers, business partners and suppliers. IBM locations, at the customer's request, can access customers' computers in remote locations to diagnose and analyze problems and can often repair them on-line.

Similarly, as a matter of overall IBM policy, Rochester has a much closer working relationship with a select group of both suppliers and third party development companies who produce peripheral or companion products to IBM systems. In contravention of traditional IBM proprietary confidentiality, IBM works with both of these groups early on in the development process to share new product plans and to utilize their technical expertise and advice regarding new product development. In addition to auditing suppliers' quality assurance processes and results, IBM also requires their suppliers to go through the Baldrige evaluation process, whether or not they apply for the award. In support of this process, they supplies assistance in both the documentation and evaluation process using specially trained IBM employees.

IBM Rochester exemplifies many of the formal aspects of the ECCO system model: cross-functional teams, extensive and specific training to enhance employees ability to make decisions about their tasks and the authority to do so, closed feedback loops with suppliers, customers, business partners and employees themselves. There is less in the way of alteration of formal incentive and reward structures, and while appropriate and extensive measurement systems are the rule, their uniformity of approach is not. It will be useful to examine these differences and how the company's success in their apparent absence might be explained.

In some ways, IBM culture resembles that of companies in Japan, paralleling the Japanese permanent employment practices, bolstered by a significant temporary workforce, embodying a strong personal commitment of employees to the corporation and its values. IBM cross-trains and cross-utilizes its employees, and routinely transfers its managers through different departments and locations. Employees' perspectives and loyalties lie with the company as opposed to a functional specialty or profession. Among IBM Rochester employees, cultural homogeneity, loyalty, morale and work ethic is if anything stronger than that found in the corporation as a whole.

It can be argued that the culture, which has already well-established values and practices supportive of ECCO system effectiveness, implicitly provides much of the intrinsic alignment which is provided explicitly in other companies. Certainly the effect is the same. When the vision is promulgated from corporate headquarters, the cultural alignment of values, expectations and behaviors adjusts and rectifies the measures, evaluation systems and rewards without explicit intervention or direction from the top. Evaluation systems and rewards are highly flexible anyway: There is little hard core policy which requires restructuring. So long as there are upper level manager process owners, and managers negotiate their subordinates' evaluation criteria, the mechanism for achieving alignment is potentially in place.

Hints of a potentially major impediment for IBM Corporation as a whole to receiving the Baldrige Award surfaced during the discussion at the end of the briefing. Marketing, sales and service are a separate corporate function, and currently have a reward system which might be incongruent with overall company effectiveness, or so it was felt in Rochester. Sales compensation is awarded largely on a commission basis. This was thought to lead to sales representatives preferring to be more knowledgeable about and to sell larger systems to maximize commissions, regardless of real company needs. This choice makes sense for salespeople in the short term, though not necessarily for customers, plants which manufacture small and medium-sized systems, or service people. Some customers, according to a user from the Mayo Clinic, must bypass the sales organization to get the information they need from the factory. Though a method has been described above for the factory to obtain feedback directly from customers and require action from the distribution organization, the method gives no direct authority to the factory to ensure compliance on the part of the distributor. It can be predicted on the basis of the thesis advanced by this work that until the incentive structures for distributor organizations (and other internal structures with similar incongruities) are formally aligned with the desired ends of the organization as a whole, IBM Corporation as a whole will not be a recipient of the Baldrige Award.

Research on the Xerox Corporation was conducted at their briefing, The Xerox Quality Forum, on June 4, 1991 in Long Beach, California. Xerox Business Products and Systems was one of two large manufacturing businesses awarded the Baldrige Award in 1989, competing in a field of 40 applicants.

Today, Xerox headquarters is located in Stamford, Connecticut, and it has locations throughout the US, Japan, and other parts of the world, employing a total of over 100,000 people. Xerox Business Products and Services (winner of the Baldrige Award) has approximately 50,200 of those employees at 83 locations in the United States, about 4,500 of whom belong to the Amalgamated Clothing and Textile Workers Union. Xerox Corporation manufactures more than 250 types of document processing equipment copiers, laser printers, computer workstations for business and engineering, LAN systems, integrated desktop publishing systems, document scanning and imaging systems, FAX terminals, and electronic typewriters.

Xerox Corporation was originally known as the Haloid Company, a small photographic paper and supply business. In 1945, they decided to invest in a promising but rudimentary technology, electrophotography, invented by Chester Carlton. The name of the process was changed to "Xerography" (from the Greek xeros, meaning "dry", and graphein meaning "to write"). It took fifteen years to develop the technology to the point where it was truly marketable, at which time, in 1959, the company introduced the 914 model office copier, which succeeded beyond the company's wildest expectations. Profits soared from $2.5 million just prior to the introduction of the 914 to $138 million in 1969.

Xerox held a virtual monopoly on the market for almost fifteen years after the introduction of the 914 copier. During that time, it developed a number of phenomenal technologies for document and data handling at Xerox Palo Alto Research Center (PARC), not the least of which was the Alto, the first personal computer. A model of managerial and interfunctional myopia allowed that emerging technology to be consigned to relative oblivion, to be eclipsed in the personal computer market by Apple and IBM.

The same structures which squelched the timely implementation of personal computer technology also presented a major barrier to Xerox addressing both process and quality improvements in the copier market. When IBM entered the copier market in the early 1970s, it had trouble producing a reliable machine, due in part to its lack of familiarity with dry copier technology. When Eastman Kodak entered the market in 1975, after ten years of development, the copy quality produced by their machines was significantly superior to that produced by Xerox machines. "But like too many other activities at Xerox, the yardstick that measured progress was calibrated entirely by Xerox's past, present, and future. Xerox did not look beyond Xerox for its standard of quality because Xerox copies by definition--by acclamation--were the best. Until Kodak."

At Xerox, increasing layers of management were placed over engineering design, with their own financial and marketing objectives. Many of the managers came from the conservative auto industry with classical bureaucratic and financial-based models of company organization and performance measures, in contrast to the Haloid spirit of cutting edge technology. The final judgement over design was given to non-engineer managers. Thus designs were developed in the absence of customer input, and evaluated in terms of internal company expectations and requirements, also having little to do directly with customer satisfaction. As described by Smith and Alexander, Xerox developed a devastatingly outdated management style.

In the words of Phil Fishbach at the Xerox Quality Forum, "Xerox lost its vision, as a monopoly, and developed practices and procedures which would not support world class performance and competition." In short, Xerox was being managed as a classical Weber-Taylor bureaucracy.

In the early 1970s, Japan entered the copier market, targeting it for dominance as it had with the automotive industry, and would soon do with the computer industry. At the same time, a Federal Trade Commission decision ordered Xerox to allow international access to several of its key patents. In 1979, Xerox still held 18.5% of the US copier market. By 1984, its share had shrunk to 10%. In 1983, CEO David Kearns initiated a major, long-term, comprehensive quality program which would affect all phases and aspects of the business. It was a bold bid for survival.

Competition in the plain paper copier market is fierce. As soon as key patents were made available in the 1970s, the Japanese attacked the low-end and mid-priced copier market, and other US companies set their sights on the high-end market. Major US competition comes from Eastman Kodak Corporation and from IBM; Japanese competitors include Canon, Ricoh and Minolta.

By the late 1970s there were over 100 companies selling copiers, with over 300 models to choose from. By the early 1980s, Japanese competitors were bringing new products to market in less than half the time Xerox could. When CEO David Kearns visited Japan in the early 80s, he discovered that the Japanese were manufacturing, shipping, taking a profit on, and giving distributors and dealers profits on comparable copiers for less than the unit manufacturing cost of those produced by Xerox in the United States. Domestic operations used twice the number of people to produce copiers as the Japanese, and the quality wasn't as good. In the words of CEO David Kearns, in 1983 there was serious doubt as to whether the company would be in existence by 1990.

The average Xerox entry level employee has a two year technical degree. Some manufacturing, assembly and unskilled support personnel have at least a high school education. No formal permanence of employment practices exist at Xerox. In 1980, Xerox signed a landmark contract with the ACTWU which represents about 4,500 of Xerox Corporation's manufacturing and production workers. The contract broke ground in exploring a joint union-management quality process, calling for cross-functional problem solving teams to incorporate assembly workers, engineers, managers and other employees to improve quality. This Quality of Work Life process improved job satisfaction for workers, as well as shop-floor problem-solving.

Xerox crafted its own strategy, using pieces of almost every quality formula on the marketplace. However, Xerox contributed at least one major original quality strategy now used by almost every serious quality effort in industry: Benchmarking. When CEO David Kearns went to Japan to visit Fuji-Xerox and other Japanese copier manufacturers, his single most striking observation was that the Japanese had much higher expectations of themselves and their business than did companies in the US. In order to change the Xerox model of using its own past performance and reputation as the standard against performance was measured, Kearns insisted that internal goals be set based on performance of the best of the competition. Benchmarking comparisons first look at the industry average for the performance in a particular area; then at the performance of the best companies in the industry, or competitive benchmark; and finally at the companies with the best performance in a particular area from any industry--the best of the best--known as the world class benchmark. Rather than looking at the selected benchmark simply as a goal to be achieved by any means, the most important thing is to look at the processes the company uses to achieve the goal, and consider how to implement and enhance the process in the Xerox setting.

In the early 1980s, after Kearn's visits to Japan, a year-long Senior Executives Project was implemented to develop a Strategic Quality Plan. The top 25 senior operating executives from Xerox worldwide operations took 15 months to develop the Strategic Quality Plan, known as Leadership Through Quality. Leadership Through Quality is described as "a long-term process aimed at fundamentally changing the way Xerox people work and manage so they can continuously improve the way they meet the requirements of their customers."

One major example of quality practices came from an internal source: Fuji Xerox.

The new Strategic Quality Policy written by senior management reads as follows:

To implement this strategy, which the senior executive team realized would require massive change in Xerox culture, six mechanism were identified. These principles are:

Paul Fishbach reports that the result of implementing the strategic plan was that it indeed changed the way people managed and worked, recognizing and rewarding these efforts and using them as criteria for promotion. It also ensured that everyone was measured and evaluated in the same way, and that they operated according to the same principles. "For the first time, all our people were working with objectives based on top management objectives. Every employee can [now] say how what they are working on supports which objectives at the department or local organization up to the presidents-level goals."

Using teams at every level, including consultative team relationships with the labor union, Xerox has improved phenomenally in the last eight years. It has also implemented the essentials of statistical process control throughout the company and extensively in manufacturing. In addition to eliminating 90% of suppliers and cultivating long-term relationship with the remaining 10%, Xerox has streamlined the manufacturing process with some standard and tailored Just-In-Time production techniques, and has benchmarked and improved the phased Product Delivery Process (new product development and implementation) to reduce cycle time and improve quality. It has standardized its analytical processes, tools, and measurements and instituted standard problem solving processes to give the company a common language for communication and decision-making. One of the most important original tools is the Cost of Quality calculation, a way of evaluating the true cost of doing it right. Cost of quality evaluates not only the additional costs of changing an operation, and the expected benefits, but also the costs in customer dissatisfaction and losses due to not changing.

Since the inception of Leadership Through Quality, there has been a major strategic shift to comprehensive, cross-functional team planning processes. Over 500 teams have been involved in creating new business plans using the new tools and techniques of quality. Planning is now a system of gathering of customer requirements, its translation into Xerox terms, the development of processes to make it happen (using benchmarking) and continuous tracking to be sure it is producing the right results.

A detailed and extensive Product Delivery Process, detailing the phased and integrated development cycle for new products, from "Pre-Concept" through maintenance and ultimately product withdrawal and service discontinuance, has been developed by Xerox over the past 10 years. The result of engineering, operational, and management teams analysis and process development as well as support from benchmarking and training activities, the PDP is a comprehensive guide to the new product development process. It integrates the management decision process and related checkpoints for feasibility evaluation with the process elements, or operational requirements and participation, at every phase, in order to produce a seamless, orderly transition through phases of product development. Keystones of the process are the incorporation of cross-functional operational level employee participation, early supplier involvement, and customer participation.

As a result of the new quality practices, development cycle times have been shortened by a full year. Manufacturing costs have been cut in half, and Xerox has simultaneously achieved a ten times overall reduction in manufacturing defects.

Applying the new problem solving and measurement techniques to non-manufacturing areas has resulted in a number of significant improvements. Sales returns (the result of sending the wrong machine, or sending the right machine to the wrong address) have been reduced 47% in the last two years. Errors in billing have been reduced from 4.5% to 1.6% during the same period. Problem solving methods, which now address root causes rather than superficial symptoms, have had collateral benefits. Sales personnel turnover was noted to be extremely high, reducing the effectiveness of the sales force due to turnover and the dearth of experienced sales representatives. In the old culture, according to Fishbach, the response was to increase their compensation. Under the new problem-solving processes, it was discovered that the real problem was that sales managers weren't trained to be able to provide adequate support for the sales representatives. The staff addressed this problem, resulting in a 25% decrease in turnover the following year.

Customer satisfaction ratings have been extensively measured and tracked. In the last two years these survey results have gone from 80% to 89% of customers who rate themselves as "satisfied or very satisfied". In the new culture, that isn't viewed as satisfactory improvement. The deficit of 11%, the customers who are not fully satisfied, is viewed as a "disaster". The goal is 100% customer satisfaction over the next two years.

David Kearns and his top executives were trained in the selected quantitative evaluation tools and group problem-solving methods adopted by Xerox in 1983. The training was then cascaded down through the company, with each of the newly trained functional managers training their immediate reports in "family groups", so that people who worked together on a day-to-day basis trained together as well. At each level, a model of Learn-Use-Teach-Evaluate was used. Each manager would first learn the methods and then, with the training group, apply them to solve a real problem. After being evaluated by the trainer, the managers would each go out and teach their subordinates in the same way, and then evaluate the results of their efforts. Sam Malone, Xerox's Project Manager for Corporate Communications, noted that since managers are notoriously the biggest impediment to quality efforts, the four-step cascade training method gives you "four shots at each manager: as student, user, teacher, and inspector" which seems to expedite their buying into the quality process.

Xerox management decided consciously not to track the costs of training, since they determined it was pre-requisite for quality improvement, and not to be evaluated in terms of "cost-effectiveness". They estimate, however, that approximately four million man-hours have been spent in training at an approximate cost of $125 million. It took about four and one-half years for the cascade training to reach the lowest level employees in the company.

The cascade training consists of an introduction to the Leadership Through Quality initiative; basic statistics and statistical process control techniques to be used by employees in all areas of the company; the 9-step quality improvement process (QIP) including process flowcharting; the 6-step problem-solving process (PSP); benchmarking; and Cost of Quality calculation. These form the core of the common tools and language used throughout the company in quality efforts.

Additional training in advanced statistical process control and related techniques are provided to technical and manufacturing employees as needed. Specialized courses are available as needed and tailored to support specific teams involved in the integrated Product Delivery Process, including technical topics from a PDP curriculum, and teambuilding and team facilitation sessions.

Employees in other functional areas may also receive specialized training related directly to their job responsibilities. In Marketing and Customer Operations, for instance, each new hire receives training in interactive skills and communications within 45 days of employment.

A variety of teams with specialized cross-functional composition are used throughout the phases of the Product Delivery Process. Its main purpose is to link the development and manufacturing process with the marketing and customer operations end of the business. It integrates the process as a whole, to eliminate the phenomenon of "throwing it over the wall" both between design and manufacturing and between manufacturing and sales/marketing. It is designed to "front-load" the process to bring both operational personnel from all related functions and the resources to support them into the earliest phases of the development process in order to "plan early and in parallel" their parts of the product strategy.

The product development teams are a far cry from early Xerox development methods. According to John Lawrence, former Manager of Engineering for International Trade, the process used to be schedule driven, resulting in a premature movement from phase to phase of the development process. As a result, as the product was going on the market, engineering was still doing final development. In his view, the new process heralded a complete shift in responsibility and organization. The walls between functional departments were broken down, and the product development team included operational personnel from procurement, design, manufacturing, marketing, and other departments as well as suppliers, including people from every department or organization involved in the process. These teams were headed by a chief engineer, and focused on internal quality process, not primarily on financial requirements. The teams create the design of product and process along with plans and requirements for collateral support. Consultation on, and final approval of financial requirements is incorporated in, but no longer drives, the process.

Teams are routinely called together to analyze identified Business Processes. Business processes, such as the billing process, virtually always cross functional barriers. Billing requires information from (at least) sales, manufacturing and shipping as well as accounts receivable. Identified "process owners" bring managers and operational personnel from the departments involved together to improve processes and solve problems.

Teams are routinely used for a variety of ad hoc problems encountered throughout the organization, and are composed of operational personnel with knowledge of and/or involvement in the process under study. Recently, an "alloy fractionalization study team" was assembled to solve a quality of copy problem. It is difficult to deposit the toner chemical on the drum or belt for transference to a paper copy. A particular product was producing a 60-80% success with the processes they experimented with. The team investigated the more than 600 variables associated with the process, identified 12 of the variables which were critical, and using SPC techniques, increased the success rate to 98%.

Another area where teams are ubiquitous is in the Marketing and Customer Operations division. The division consists of three main functional areas: Sales, service, and administration. Though the three managers at each level retain their functional responsibilities, they co-operate as a team and are rewarded and evaluated as such. Operational level employees for each service area also function as a team, with considerable autonomy. Service teams are assigned a particular service area and are empowered among other things to internally assign job duties, coverage schedule, and backup plans within the group.

The evaluation and improvement of process is central to the quality effort at Xerox. Currently, about 75% of all benchmarking is done on best practices (processes), the remaining 25% on products and services. The focus in benchmarking practices involves analyzing the processes of the benchmarked organization as well as the current processes used by the organization trying to improve. One of the major "lessons learned" about benchmarking as a whole is that it should not be used for budget cutting, comparison, or "beating people over the head": It should only be used for improving process.

John Lawrence, former Manager of Engineering for International Trade and currently Manager for Quality Assurance, USMG at Xerox, states unequivocally: "The results you get are yielded from a particular process. To improve results, you must improve the process." Xerox training and standardized techniques are all designed to do exactly that.

Xerox has made extensive and successful changes in their incentive and employee evaluation systems since undertaking the Leadership Through Quality initiative. It was recognized particularly by top level management that changes were necessary to develop the personnel behaviors which would support the achievement of quality goals.

Prior to 1987, senior executives were compensated based on their own area of functional expertise: Finance V.P. was rewarded based on the achievement of the year's cash profit plan, the Marketing V.P. based on Xerox meeting the year's market share goal, and so on. According to Paul Fishbach, Vice President of Finance for the United States Marketing Group (USMG), this caused major problems in the boardroom. Arguments over strategy were characterized by executives taking positions on issues such that their own goal achievement would be met, which typically conflicted with the positions of others. Now, all Xerox's top executives are compensated based on the same figures: those which reflect the company's overall success and customer satisfaction.

Line managers in all areas of the company are responsible for implementation of quality processes in their organization. Success in this implementation constitutes a substantial part of their evaluation criteria for raises and promotion. Managers in the USMG specifically are evaluated for promotion a the basis of customer satisfaction, successful implementation of quality processes, and on their effectiveness in dealing with human resources (human interactions with subordinates and co-workers).

The USMG has developed its own version of joint goals and rewards by conjoining the Sales, Service and Administrative managers at each level in their "Sales Partnership". At every level, V.P. to local front line managerial positions, these three managers retain their functional responsibilities. However, instead of being evaluated solely on their own unit's performance, 35% of their evaluation for raises, bonuses and promotions is based on meeting a single goal: customer satisfaction, as represented by a survey of Xerox customers in their geographical service area.

According to both Vice Presidents and front line managers, it has changed the way they work. One front line service manager described the prior conflicts at the local level. Sales representatives and sales managers used to be compensated based on making a sale--market share. They would therefore often sell less expensive machines than would be adequate for the customer's requirements, just to make the sale. The machine would be used beyond its design capacity, leading to frequent breakdowns which resulted in customer dissatisfaction with the equipment and many service calls. The service department would be overworked, and unable to provide prompt service, and the figures for service expenditures would look bad to the district. Service had no means for correcting the essential problem: providing an adequate machine for the customer's needs. On the other hand, Administration was rewarded on the basis of the soundness and accountability of their paperwork, rather than providing it in a timely fashion, ultimately obstructing sales and annoying customers. With the advent of the new partnership system which based evaluation largely on customer satisfaction, the managers now operate co-operatively and in support of one another, since each others' goals are also their own.

As noted before, there was a major shift in the measurement of day-to-day operations, in terms of what aspects were to be measured, as a result of the new Business Plan for customer satisfaction. According to Burt Tornstrom, Manager for Quality Promotion, "All of the key measurements, and most of the internal measurements are congruent with the requirements of the customers" as a result of the Plan. This was a radical departure from the standardized quality and productivity measures of the early 1980s.

Prior to the quality initiative, Xerox had no consistency in problem solving approaches. With the advent of Leadership Through Quality, that began to change. Now, basic Statistical Process Control techniques are used to analyze data throughout the company, in R&D, manufacturing, marketing and customer service, as well as business areas. Additional advanced tools are applied in manufacturing and related areas as needed. The same problem analysis approach (Quality Improvement Process) and problem solving process (PSP) are used throughout the company, as well as benchmarking and Cost of Quality analyses. "These tools are how we all look at data in the company," according to Barry Latham. They form the basis for presentations, analyses, evaluations that everyone, in all functional areas, can understand; and that means they can understand each other better.

Now, common measures and tools provide a common language for the entire company, which streamlines communications and standardizes evaluations. Cost of Quality analysis is one of Xerox's homegrown (and now widely imitated) measures. Though it is not cost accounting, nor does it replace the cost accounting system, it is a management tool for understanding and communicating priorities. It accounts for many previously invisible costs, whose existence radically changes strategic and new product/services planning.

Cost of Quality is calculated from three figures: Cost of Conformance, Cost of Non-Conformance, and Costs of Lost Opportunities. Cost of Conformance includes investment, training, communications and other costs associated with doing the job right, fulfilling all customer requirements. Cost of Non-Conformance include costs associated with fixing failures and inadequacy of product during the development and manufacture period as well as those occurring after the product has been delivered to the customer, and those costs associated with "unnecessary extras" which exceed or are irrelevant to customer requirements. Cost of Lost Opportunities includes profits not earned due to loss of existing or potential customers as a result of poor quality.

Fishbach claims that it allows Xerox to identify opportunities for quality improvement in a cost effective way," and a way of communicating how funds can be most profitably invested. It also provides a dollar figure estimate of potential improvement given a specified amount of resource investment, and ensures the congruence of corporate priorities and project level decisions. Cost of Quality calculations are routinely performed by work groups contemplating changes in process which are likely to require additional resources.

Benchmarking, which has been described above, has been described by Xerox executives as their answer to Juran's question: How do you integrate quality into your business plan? Benchmarking establishes goals, and at the same time gives insight into the process by which those goals might be achieved. It is a far cry from the days when Xerox looked only to its past for the standards to which it aspired.

Rewards and recognition

In addition to regular compensation and bonuses, Xerox provides special team and individual cash bonuses for service in support of the quality process above and beyond the call of duty. It also provides informal and formal recognition for these efforts. Two of these are associated specifically with Leadership Through Quality. The first is the Team Excellence Award. Cash awards are given, on a competitive basis, to teams for excellence in "using Quality tools to achieve outstanding results." The teams receiving cash awards are selected by the Senior Staff, after competing at Regional and District levels. All teams who achieve the rank of Excellent, however, are awarded a certificate.

The second special recognition event is a series of "Teamwork Days", which honor and share (through team booths, exhibits and speakers) the achievement of PSP and QIP teams. These events are held at District, Regional and functional levels, as well as two annual Corporate sponsored Teamwork Days on the East and West Coasts. Previously open to Xerox employees, Corporate Teamwork Days are now open to the public, beginning with the West Coast Teamwork day scheduled for October 22, 1991 at the Los Angeles Convention Center.

Shortened, closed feedback loops

Xerox has explicitly and systematically shifted responsibilities downward to the operational personnel closest to the problem or process to be controlled. Manufacturing plant employees are organized into "family groups", work groups of eight to twelve people who work co-operatively with little or no direct supervision, and are empowered to solve problems, including shutting down the line if necessary, having to do with their aspects of quality production. Service technical employees in the USMG are divided into groups, which collectively have the responsibility for arranging training needs and schedules, backup coverage, and other job responsibility distribution, formerly determined by management. Service people are authorized to waive up to $200 worth of service charges to customers on their own initiative, and have the authority not to bill calls at all if they feel a problem was the result of previous errors on the part of the service department. Similarly, sales people have a range of prices they are independently able to negotiate with customers, with no higher approval.

The extent to which this local autonomy seems to be true of Xerox is characterized by the story of a consulting visit which Xerox executives paid to a manufacturing corporation in Chicago. The CEO of the Chicago company was skeptical about how much autonomy Xerox employees really had, so he demanded, on extremely short notice, to talk with one of Xerox's service teams. The executives contacted the local service organization who pulled a team of five employees out of the field and sent them to the CEO's office within two hours of the request. The CEO had just one question: When was the last time you asked your boss's permission for anything? The team huddled for a few moments, and the employee who had been in the field the longest reported: No one could really remember for sure, but they determined it had been at least 18 months, and could have been as long as two years since they had asked permission for anything. This indicates that the level of authority granted to service employees matches closely with the possible disturbances or problems they are likely to encounter in their work.

The organization is also mandated to consult the union before farming out work which might cost union jobs. Xerox has opened the corporate books to union workers and formed union/management teams to investigate and review these situations and to recommend solutions. As a result of these team efforts, 250 union jobs remained in-house and Xerox saved about $7 million due to the adoption of the report's recommendations regarding work process improvement and equipment investments. Most recently, union workers helped to design the newest Xerox plant.

According to CEO David Kearns, "If you share information with your workforce, you don't need layers of middle management analyzing it and military style command and control." As a result, the entire organization is flatter. In USMG, two layers of middle management were eliminated out of a total of five managerial levels, bringing the highest levels of management closer to operations and the customer. There is less supervision, both qualitatively and quantitatively. The ratio of managers in the service line to service technicians has gone from 1:12 to 1:32.

Other decisions which used to be made by middle and upper level managers is being delegated to lower levels. It used to be mandatory to run a credit check on every customer buying from Xerox, and credit approval required many signatures and formal written approval from headquarters before a sale could be closed. In reviewing this as part of the "business simplification" process, it was discovered that credit approval took a long time to obtain, and no one was ever denied credit. As a result, the authority to extend credit rests at the district level office, with the caveat that bad debts show up on the district level profit and loss statement. So far, there have been no credit problems, only happier customers and more satisfied operational personnel.

Early in the quality initiative period, local service offices were authorized to grant "customer accommodations" for up to $300. However, higher level management approval was required to grant larger accommodations. One service tech wrote to CEO David Kearns in frustration that a $600 accommodation he had requested on behalf of a customer had been in the works for six weeks. Kearns called for a QIP team investigation, which discovered that six management signatures were required for accommodations over $300; that 90% of all accommodations were for less than $1000; and that no request for accommodations had ever been turned down. Now, District offices are authorized to approve accommodations of up to $10,000, cutting out many layers of management approval. Subsequent requests for accommodations did not rise, as feared. They diminished 25%. Similar delegations of authority have occurred in all parts of the company as a result of specific complaints as well as the "business simplification" process, with similar positive results.

Closing feedback loops is an essential part of Xerox quality strategy. It has been reiterated by all companies visited that just because an action seems like "common sense", doesn't mean it is undertaken. In the same light, Barry Lathan (V.P. Western Operations) emphasizes you must ask whether the process in question is delivering the desired result. If not, the process must be changed; if so, how can we pursue continual improvement. He notes, "It's surprising how many times you get involved in changing the process without paying attention to whether it's giving you the results you want." As a result of the Baldrige Award effort, Xerox is even more focusing on results, analysis of process and actively tracking whether the changes in process get the results anticipated and needed. This closed loop process is intimately and formally integrated into the everyday efforts of people at Xerox.

Xerox has cut the number of suppliers it uses to 10% of its pre-quality initiative years, and has actively involved them in the early stages of the Product Delivery Process. The importance of suppliers involvement cannot be overemphasized, not only for the value of planning and their technical expertise, but also because almost 80% of the Xerox product cost is in purchased parts from suppliers. Suppliers serve as team members in the development process. Xerox has also required suppliers to participate in quality audits, and a Baldrige style analysis, performed by qualified Xerox personnel, with Xerox assistance in improving supplier operations, if they so desire.

Customer input and involvement is also a primary focus at Xerox. New employees spend six weeks in the field with service technicians, so that they can see first-hand what matters to customers. Customers are formally included in early and later phases of the Product Development Process, to provide early feedback and evaluate evolving new product designs. Customer input about existing products and suggestions for new products and features is solicited from product user groups, as well as through standard market research studies. A major tool for customer input and feedback about products and services is the Customer Satisfaction Measurement System (CSMS). Each month more than 45,000 surveys are mailed to customers, soliciting their opinions on Xerox products, service, sales and administrative performance. This survey is also the basis of the customer satisfaction measures used in manager evaluation throughout the company.

Xerox exemplifies all of the control characteristics of an ECCO system organization. Probably more than either of the other two companies studied, Xerox required and achieved tremendous cultural change in order to survive, let alone succeed. It has taken the course of a highly explicit strategy, and an extremely coherent set of common measures, processes and language. It seems to have succeeded in a very short time, in no small part due to the embedded and ubiquitous interlocking feedback mechanisms and cultural alignment of measures and incentives.

It is fascinating, though predictable after reading the Japanese literature, that Xerox and other companies have not emphasized the high tech robotics and information systems as critical to their success. It implies that the control infrastructure along with soft quality tools and the education to support them, may be prerequisite (and largely sufficient) for major changes in organizational success as well as cultural change.

Back to Table of Contents On to last chapter Back to Home Page