Feature

BackgroundRecent changes in regulatory requirements for clinical trials pose considerable challenges for the clinical research community. The clinical research Quality Assurance (QA) profession as a whole has not historically utilised quality management tools, and QA professionals may have little to no formal training in continuous improvement methods, to be aware of their value in assessing and improving the impact and value of the QA department to any organisation. This article will present Deming’s 14 Principles for Management, paraphrasing them to demonstrate how easily they can apply to the clinical trial project management, Contract Research Organisation (CRO) and clinical department management.

Deming’s 14 PrinciplesDr. William Edwards Deming is known as the father of the Japanese post-war industrial revival and was regarded by many as the leading quality guru in the United States.

Deming created 14 Principles for Management that summarised his business philosophy. The principles became a basis for transformation of industry. The 14 principles apply anywhere, from small organisations to large ones, to the service industry as well as to manufacturing. They apply to any division within a company.

1. Create Constancy of Purpose Towards ImprovementReplace short-term reaction with long-term planning. Approaching a new clinical trial project, keep in mind that it must be better performed and with a lower budget than the previous ones.

Having the above in mind, analyse previous pitfalls, difficulties, problems, obstacles and identify the current and anticipated problems. Hint: findings from system and clinical trials audits may be a good source of such information. Can these recur? Most probably ‘yes’, unless you take measures to prevent or, at least, to mitigate them. Hence start with planning and allocate resources for training and education, based on what you’ve learned from your past experience, and make risk analysis and contingency plans. Acting this way you can constantly improve your service.

2. Adopt the New PhilosophyManagement should actually adopt the ‘quality philosophy’, rather than merely expect the workforce to do so.

When the new clinical trial project is thoroughly planned and the quality matters are carefully embedded into it by the study management, while creating the atmosphere of striving for the best performance, it will penetrate into the study team at all levels, both in-house and at sites.

When the management looks forward, not at the competitors, but at the customers, that will be the moment of major change.

3. Cease Dependence on InspectionQuality does not come from inspection; mass inspection is unreliable, costly and ineffective.

Of course, we can never cede the monitoring, however ‘quality embedded’ projects will allow us to reduce the level of scrutiny over clinical sites without giving up on the quality of deliverables (i.e. clinical data), genuinely improving cost-effectiveness.

The data collected by monitoring, i.e. inspection, should be used for better process control.

4. Move Towards a Single Supplier for any One ItemWork toward a single source and long term relationship.

If you’re a pharmaceutical company, invest your time in finding ‘the one and only’ CRO. It will not necessarily be a single one, but if you create a solid group of CROs delivering up to your standards, nurture the relationships, establish a mutual trust and reliance between purchaser and vendor, make them feel like partners, this will benefit you immeasurably. When outsourcing your activities, always remember that price alone has no meaning. Change focus from lowest initial cost to lowest total cost.

If you’re a CRO – strive to be a part of such a group. Working with the same customers over time will make your operation more efficient, meaning lower prices and higher profits.

Deming’s 14 Principles

10 Quasar

11Quasar

Deming’s 14 Principles

9. Break Down Barriers Between DepartmentsThe concept of the ‘internal customer’ means that each department serves not the management, but the other departments that use its outputs.

Often a company’s departments or units are competing with each other or have goals that conflict. They do not work as a team; therefore they cannot solve or foresee problems. Even worse, one department’s goal may cause trouble for another.

There is no better example of the involvement of a multi-disciplinary team than a clinical trial project: data management and clinical supplies, laboratory personnel, logistics, operations, administration, and the site personnel. So, first know your internal suppliers and customers. Integration, co-operation, open communication lines, mutual understanding and support are necessary pre-requisites for the success of the project.

We can easily depict interrelations within the team as customer-supplier relationships: operations as customer of clinical supplies, data management as customer of operations, etc.

From the very beginning of the ‘quality embedded’ clinical project, all parties involved shall have a perception of the common goal and comprehension of success as a result of teamwork; therefore promote the teamwork. This is the only environment in which continual improvement is possible in. Every department shall make its contribution, presenting its needs and expectations on the one hand as a customer, on the other hand meeting needs and delivering expectations as a service provider.

Feature

5. Improve Constantly and ForeverQuality starts with the intent of management. Management is obligated to continually look for ways to improve quality.

First of all, teamwork in clinical trial project planning is fundamental. Second, we must remember that chronic problems have chronic causes. Treating the symptoms does not solve the problem. Putting out fires is not improvement of the process. Improvement efforts must shift focus from improving the personnel to improving the processes.

We must identify the systemic policies, practices, belief systems etc. that are dysfunctional and change them. As long as these systemic causes remain, the resulting systemic problems will remain. One of the most efficient tools to reveal the causes is quality audit followed by corrective actions that include investigation, a.k.a. root cause analysis. Clinical trial audits and system audits may open our eyes and indicate the process deficiencies and flaws. Appropriate handling of the latest will inevitably lead to improvement.

6. Institute TrainingIf people are inadequately trained, they will not all work the same way, and this will introduce variation.

In our business, every new project calls for study specific training, even when very experienced personnel are involved, both in-house and at site. The more we invest in the development of an extensive and comprehensive training programme, the more field, office and site staff are trained, training effectiveness evaluated and re-training carried out when the need is identified, the more uniformity in results, meaning statistically we will achieve valid clinical data.

Management must remove the inhibitors to good work and provide the setting where workers can be successful.

7. Institute LeadershipDeming makes a distinction between leadership and mere supervision.

First of all managers, or should we say leaders, must know the work they supervise. That will assist them to know the difference between special and common cause of variation, or in other words to distinguish between a mere mistake and a weakness of a process.

CRA Managers or Study Managers are not policemen, neither are QA personnel. We all have a common goal – to complete our clinical trial in time, at the lowest possible cost, with scientifically valid results, while strengthening the relationships with investigators and subcontractors/clients. It’s everyone’s job and everyone’s responsibility. When the management message, not verbal but though action devoted to quality, it will make its way to the field personnel, also reaching to the site’s staff.

8. Drive Out FearDeming sees management by fear as counter- productive in the long term, because it prevents workers from acting in the organisation’s best interests.

There’s no better way to make personnel feel that every single one of them is responsible for improvement than to allow a free discussion of any subject at any point of time. Attentiveness and responsiveness of the Study Management to the field personnel requests, open door policy, willing to assist and support, fast troubleshooting, backup in case of conflicts and trust, will not only encourage the personnel to report the problems, or to admit their own mistakes, but also will create an atmosphere of partnership which is the most favourable for the project success.

Moreover, remember that the more information, both positive and negative we have, the better we can manage not only the current project, but also the future ones.

10. Eliminate SlogansIt’s not people who make the most mistakes - it’s the process they are working within. Harassing the workforce without improving the processes they use, is counter-productive.

Eliminate slogans, exhortations and numerical targets for the workforce. These never help anybody do a good job.

Facing problems and obstacles raised by audit non-conformities, study management must not think of penalties, replacement, rebuke, or any other person-orientated actions as these have proved ineffective in terms of contributing to improvement measures, but rather look carefully at the process, analyse it, understand its strengths and weaknesses, consider the possible resolutions, and only then act. Following this approach, study management will most probably find out that the measures, i.e. corrective actions rather than mere corrections, are process-oriented, and therefore, target-oriented. Moreover, once implemented, corrective actions will improve the entire process as a result, preventing the study team from making additional mistakes.

11. Eliminate Management by Objectives Deming saw production targets as encouraging the delivery of poor-quality goods.

Have you participated in ‘competitive recruitment’ to a clinical study? I’m sure you have. Have you ever analysed the number of protocol violations, damage to relationships with site personnel, CRAs’ frustration, illegible data for statistical analysis data as a result of such a process?

Quotas only take into account numbers, not quality or methods. A person, in order to hold a job, will try to meet a quota at any cost, including doing damage to their company.

Initiate a ‘seasonal’ project when fast recruitment is crucial, encourage your teams by creating the best foundations for the most efficient implementation, i.e. all administrative and regulatory issues resolved, study teams trained appropriately, logistics and clinical supplies standing by. Then set measurable and realistic (!) objectives of performance and make sure they are understood. This way the importance of timely and prompt execution will be perceived as the natural course of things and the employees and the site teams will be proud in their workmanship.

12. Remove Barriers to Pride of WorkmanshipPeople are eager to do a good job and distressed when they cannot. Too often, misguided supervisors, faulty equipment and defective materials stand in the way of good performance. These barriers must be removed.

When the CRA encourages the recruitment, but the site is running out of clinical supplies that for some reason couldn’t be delivered in time, or hasty allocation of resources makes the field personnel struggle through more work than they are able to handle appropriately, when the site personnel are inadequately trained or communication in an electronically managed trial constantly fails, what pride can people possibly take of their workmanship?

Feature

12 Quasar

‘It’s not people who make most mistakes - it’s the process they are working within. Harassing the workforce without improving the processes they use is counter-productive.’

Deming’s Fourteen Points of Management

1Create constancy of purpose for improvement of product and service 8

Drive out fear

2Adopt the new philosophy

9Break down barriers between staff areas

3Cease dependence on mass inspections 10

Eliminate slogans, exhortations, and targets for the work force

4End the practice of awarding business on the basis of price tags alone 11

Eliminate numerical quotas

5Improve constantly and forever the system of production and service 12

Remove barriers that rob people of pride of workmanship

6Institute training

13Encourage education and self-improvement for everyone

7Adopt and institute leadership

14Take action to accomplish the transformation

13. Institute Education and Self-improvementInvest in the education of your work forces (including sub-contractors), encourage self-education and introduce training programmes, worship knowledge and professionalism. This is the only recipe for commitment to lifelong employment and business success.

As the CRA (that does not necessarily have a clinical background) aquires more knowledge, of the therapeutic area of the study protocol, the better the CRO team is trained on the internal sponsor’s processes (SOPs, internal communication lines, etc.), as the site staff have more understanding in regulations, the more study management is educated in the topics like teamwork, statistical techniques, risk management, project management and, of course, quality tools, the more chances that our mission - to complete our clinical trial in time, at the lowest possible cost, with scientifically valid results while strengthening the relationships with investigators and sub-contractors/clients – will be successfully accomplished.

14. The Transformation is Everyone’s JobIt will require a special top management team with a plan of action to carry out the quality mission. The critical mass of people should be included in the change. Workers cannot do it on their own, nor can managers.

The leadership model must replace the bureaucratic model of the boss deciding on what and how everyone else must change. Once consensus is reached, it is feasible to focus collective energy, intellect, experience, etc. on priorities for improvement.

SummaryDeming preached that to achieve the highest level of performance requires more than a good philosophy – the organisation must change its behaviour and adopt new ways of doing business. Indeed, his 14 Principles pose a challenge for many companies to figure out how to apply them in a meaningful way; however CRO or clinical departments will benefit immeasurably by finding this way and implementation of the principles in the management practice.

Author Biog

Rita Gabay

Rita is currently owner and manager of QA INSIGHT - a company providing independent QA, GCP and ISO 13845 consultancy and audit services.

With a clinical background, and a MSc in Quality Assurance and Reliability, Rita has over 20 years of experience in marketing, Clinical Quality Assurance, and management of Quality Systems compliant with ISO 9001:2008 and 13485:2003.

She served as QA Manager and Auditor in pharmaceutical, medical devices and in-vitro diagnostic companies.

She lectures at the professional and educational forums on GCP, ISO and Medical Devices regulations

Rita is a certified auditor. She holds the Certificate of International Register of Certified Auditors, and the Health Canada Certificate of CMDCAS Auditor Qualification Programme.

13Quasar

Quality Management and Juran’s LegacySoren Bisgaard

Eugene M. Isenberg School of

Management, University of

Massachusetts Amherst,

Massachusetts, and Institute for

Business and Industrial Statistics,

University of Amsterdam,

The Netherlands

ABSTRACT Quality management provides the framework for the industrial

application of statistical quality control, design of experiments, quality

improvement, and reliability methods. It is therefore helpful for quality

engineers and statisticians to be familiar with basic quality management

principles. In this article we discuss Dr. Joseph M. Juran’s important contri-

butions to modern quality management concepts, principles, and models.

Many people have contributed to modern quality management. However,

through his extensive writings covering more than six decades, Juran has

managed to distill and synthesize the subject. He has provided us with a

coherent framework and terminology and anticipated many of the princi-

ples that subsequently became incorporated under the Six Sigma umbrella.

We briefly outline Juran’s framework and discuss a number of key contribu-

tions he has made to the subject of managing for quality.

KEYWORDS Design for Six Sigma, quality engineering, Six Sigma

INTRODUCTION

Quality management provides the framework within which modern

statistical quality control, quality improvement, and reliability operate. Only

when imbedded within a managerial framework do statistical tools and tech-

niques become fully operational and effective. Quality management provides

the structure, policies, and organizational environment in which statisticians

and quality professionals can get work done. However, quality engineers and

statisticians are often primarily focused on tools and techniques. They are

typically not well versed in the managerial aspects related to how quality

activities are effectively organized and managed within an organization.

Indeed, Six Sigma advocates seem at times preoccupied with statistical tools

and road maps and pay less attention to the larger managerial framework.

Besides possessing the necessary technical skills, it is useful for quality

professionals to supplement their knowledge base with a basic understand-

ing of quality management principles. One of the most prolific contributors

to modern quality management was Dr. Joseph M. Juran. In this article we

review Juran’s legacy. Many have obviously contributed to quality manage-

ment. However, Juran, more than anyone else, unified the concepts and

brought them together as a system. In our estimation, he was the

quintessential synthesizer of the field. His concept of managing for quality,

comprehensively outlined in his book Juran on Leadership for Quality,

Address correspondence to SorenBisgaard, Eugene M. Isenberg Schoolof Management, 121 PresidentsDrive, Amherst, MA 01003. E-mail:Bisgaard@som.umass.edu

Quality Engineering, 20:390–401, 2008Copyright # Taylor & Francis Group, LLCISSN: 0898-2112 print=1532-4222 onlineDOI: 10.1080/08982110802317398

390

( Juran, 1989), is a generic system that we expect will

have a long lasting universality. It has provided the

blueprint for the subsequent Six Sigma movement.

Indeed, some of Juran’s ideas have yet to be adopted

by Six Sigma for that approach to be a comprehen-

sive quality management system.

The purpose of this article is twofold: (a) to describe

Juran’s comprehensive quality management frame-

work and (b) to highlight some of Juran’s most impor-

tant contributions to modern quality management. In

doing so, we hope to assist quality professionals pri-

marily familiar with statistical tools and techniques, to

gain a better understanding of Juran’s management

framework as well as developing an interest in studying

Juran’s many books and publications.

A BRIEF HISTORICAL BACKGROUND

Quality management as we know it today has

evolved over the past century from an early embry-

onic set of ideas to a comprehensive framework for

managing all aspects of quality in an organization,

private or public, for profit or not, manufacturing

or service. Early contributions began before World

War I. For example, the textile business that played

a pivotal role in the industrial revolution was an early

proving ground for many quality related ideas. How-

ever, if we seek a formal birth date, it seems fair to

assert that the publication of Dr. Walter Shewhart’s

seminal 1931 book, The Economic Control of Quality

of Manufactured Product, heralded the beginning of

modern statistically based quality management.

During the 1920s it was recognized that statistical

principles were needed to guide the development

and evaluation of sample inspection procedures.

Further, it was recognized that random variation

needed to be accounted for in the monitoring and

control of production processes. Much of the early

statistical work took place within the Bell Telephone

Company. For example, Shewhart is credited for hav-

ing invented the control chart in 1924 and H. F.

Dodge and H. G. Romig developed early acceptance

sampling theory. They all worked for Bell Labs; see

Millman (1984). Further, after graduating as an elec-

trical engineer, in 1924 Joseph M. Juran assumed a

position in the inspection department of Western

Electric Company, the production division of the Bell

System, responsible for producing communications

hardware; see Juran (2004).

Although Shewhart primarily focused on statistical

issues relevant to quality control, he also discussed

managerial issues. However, most of his managerial

emphasis was on organizing and managing the

inspection function. Juran’s first book, Management,

of Inspection and Quality Control (Juran, 1945),

published in 1945, right after he left Western Electric

to venture into consulting and teaching, had a similar

focus.

Quality management in the modern sense was not

discussed in the literature until after World War II.

Even then, the field only evolved slowly over the

following decades. In the early 1950s both Dr. W. E.

Deming and Dr. J. M. Juran were invited to consult

for the Japanese industry about the implementa-

tion of quality control. Both based their advice to

Japanese executives on their experience before and

during WWII in United States implementing indus-

trial quality control. Both had been frustrated about

the ineffectiveness of quality control when (a) upper

management only exhibited a passing and superficial

interest in managing the quality function and (b) sys-

tems were based primarily on inspection. Both came

to the conclusion that to be effective, efforts should

have strong top management leadership and should

include activities aimed at finding and eliminating

root causes for poor quality, not just passive inspec-

tion. Indeed, both came to the conviction during the

1950s that a comprehensive system for managing

quality was critical. Deming’s focus was primarily

philosophical and strategic. His main contribution

has been to educate upper management in the criti-

cal strategic role played by quality in a competitive

economy. His writing is mostly focused on quality

concepts, paradigms and philosophies. Deming’s

work is essentially setting the stage for initiating

quality efforts in an organization and as such essen-

tial. However, his writing is not particularly specific

in his advice about how quality should be implemen-

ted and organized in a company. Juran, on the other

hand, while not silent on strategic issues was more

hands-on, prescriptive, and developed compre-

hensive and detailed practical guidelines for the

development and implementation of a quality man-

agement system. His ideas clearly evolved over the

years, but the fundamental concepts were already

evident in the first edition of his quality hand-

book ( Juran, 1951). A side-by-side comparison of

Deming’s (1986) book Out of the Crisis with Juran’s

391 Quality Management and Juran’s Legacy

(1989) Juran on Leadership for Quality brings out

these differences. However, it should also be noted

that these two scholars’ works are complementary,

not antithetical. It is not an issue about which book

is better or whose theory is superior. Both Deming

and Juran have made valuable contributions. Both

were ‘‘giants’’ in the field. Each emphasized different

issues. Indeed, there are few essential disagreements

between them. Among statisticians Deming is better

known, but Juran deserves equal recognition.

Deming’s and Juran’s teachings to Japanese execu-

tives, managers, and engineers were not one-way

communications. Both seemed to have learned from

their work in Japan. Both brought back many good

ideas. Surely the Japanese applied the ideas they

learned from their American teachers. But they also

made significant contributions themselves. By apply-

ing what they learned, the Japanese saw what

worked, what did not work, what shortcomings

existing approaches had, and came up with whole

new ideas of their own. Thus, many Japanese indus-

trialists, engineers, and scholars such as K. Ishikawa,

S. Toyoda, T. Ohno, S. Mizuno, H. Kume, N. Kano,

and G. Taguchi have made valuable contributions

to ‘‘the body of knowledge.’’ For an overview of

the history of modern quality management, see

Garvin (1988, chapter 1) and Juran (1995).

JURAN’S CONTRIBUTIONS TOQUALITY MANAGEMENT

Although his 1945 book discussed management

issues, Juran’s serious entry into the field of quality

management was the publication in 1951 of the first

edition of the Quality Control Handbook. This semi-

nal book, edited by Juran with numerous chapters

written by him, propelled Juran to the forefront of

the field and caught the attention of Japanese industri-

alists. Although there is a clear evolution and matur-

ing of the field over the years, the combined content

of the five editions of the Juran Quality Handbook

(Juran, 1962; 1947; 1988b) issued roughly every 10

years since 1951 constitute an essential source for

Juran’s thinking about quality management and for

the entire field. All five editions are significantly differ-

ent. Each contains different materials. All are worth-

while reviewing. As indicated above, Juran did not

write everything himself. However, he supposedly

exercised strong editorial control and was very

hands-on in the development of these monumental

books. The general trend from the first to the fifth

edition is a gradual change of focus from a technical

to a managerial. His book Managerial Breakthrough

( Juran, 1964) is also important. However, for a

modern presentation of his evolving view of what

he seemed to prefer to call ‘‘managing for quality’’

we recommend Juran on Leadership for Quality

( Juran, 1989). Other important books are Juran on

Quality by Design ( Juran, 1992) and Juran on Plan-

ning for Quality ( Juran, 1988a). His autobiography

( Juran, 2004) is also interesting but primarily about

his Horatio Alger–like personal life, struggles, and

successes. His book with longtime partner Dr. Frank

Gryna ( Juran and Gryna, 1993) constitutes a solid uni-

versity course textbook. Later, new or revised editions

of books organized by the Juran Institute but pub-

lished after Juran himself withdrew from daily

involvement seemed to have compromised the clarity

of Juran’s vision. Some of this seemed to have been in

an effort to keep up with the Six Sigma movement.

However, to us it seems that Six Sigma only ‘‘bor-

rowed’’ parts of Juran’s trilogy, to be discussed below.

Thus, we find that the best expression of Juran’s

mature views on quality management are to be found

in Juran (1989) or Juran and Godfrey (1999).

Juran’s main contribution to modern quality man-

agement was his concept of what is known as the

Juran Trilogy. This is a conceptualization of manag-

ing for quality consisting of three main functions:

quality planning, quality control, and quality

improvement. To fully appreciate this concept, we

first discuss Juran’s work on defining quality, related

economic models for quality, and his economic

perspective on the societal importance of quality.

We then discuss his management trilogy. We also

review his insight to how to implement quality

improvement at the top management level. Finally,

we discuss Juran’s work on developing precise

definitions, terminology and concepts, something

fundamental to any field of study.

Juran’s Definitions of Quality

Without well-defined terminology, discussions

about quality and quality management easily get

confused. Juran recognized that the word ‘‘quality’’

is not easily defined. Dictionaries provide many defi-

nitions, most of which are too vague or philosophical

S. Bisgaard 392

for technical use. However, Juran’s definition of

quality as fitness for use is widely recognized today

as one of the more useful. To appreciate the subtle-

ties of this seemingly awkward definition, consider

the following example. If a busy business traveler

needs a simple, clean, and safe accommodation for

the night, an expensive five-star hotel room may

not constitute ‘‘fitness for use’’ despite its luxury

delivered to perfection. However, the same person

may come back the next week for a vacation with

the family and find the same hotel a perfect ‘‘fit’’

for that use and therefore now be delighted with

the quality. In other words, it is the customer, not

the provider, who defines quality. Further, quality

depends on the circumstances and it is not ‘‘more

is better’’ but ‘‘fitness for use’’ that is the key issue.

Quality is a bundle of attributes timely delivered to

satisfaction that solves the customer’s problem. In

economic terms, quality is a non–price competitive

market signal. Quality does not thrive in monopol-

istic circumstances. Rather, in a competitive econ-

omy, customers have choices. There are typically

multiple alternative market offerings, but ultimately

customers will vote with their pocketbooks based

on a trade-off between cost and quality. Quality is

what appears appealing and fit to the customer

relative to alternative market offerings.

Older definitions of quality such as ‘‘conformance

to specifications’’ are no longer adequate at the high-

est conceptual level. Specifications may be useful

surrogates at lower conceptual levels—intrafirm

and operationally—but cannot serve as the predomi-

nant definition. Quality must be defined relative to

the customer’s needs and expectations.

Although ‘‘fitness for use’’ is the predominant

definition, Juran realized a need for further subsidiary

definitions, chiefly for economic reasons. On the one

hand, some people may argue that high quality ‘‘obvi-

ously’’will bemore costly and expensive. Advocates of

that viewpoint implicitly think of quality in terms of

more features. On the other hand, others may argue

that high quality is cheaper. This may seem contradic-

tory but is not. Advocates of that viewpoint consider

the cost of defects, delays, rework, and waste, or, more

broadly, deficiencies associated with poor quality.

They understand that first doing things incorrectly

and then having to fix it inevitably is expensive. Reduc-

ing deficiencies therefore reduces costs. Permanently

removing the causes of deficiencies is even better.

Clearly both viewpoints have merit. Juran resolved

the quandary by providing two subsidiary definitions

to quality: (a) features and (b) freedom from defi-

ciencies as illustrated in Figure 1. Features have to

do with the design of the product, process, or ser-

vice. It is what we intend to deliver. Deficiencies

have to do with the actual delivery. This profound

distinction is only slowly gaining popularity but is

important, especially in the context of Design for

Six Sigma. More importantly, the distinction is key

to understanding the economic reasons for pursuing

quality as a strategic objective, something Toyota

Motor Company, for example, consistently has done

with extraordinary success.

Admittedly, we have not been able to come up

with a better definition to replace Juran’s ‘‘fitness

for use.’’ However, Juran’s ‘‘features’’ and ‘‘freedom

from deficiencies’’ terminology may perhaps today

seem somewhat obsolete. Although we will continue

to use Juran’s terminology in this article devoted to a

discussion of his contributions, we find that ‘‘design

quality’’ and ‘‘delivery quality’’ better capture the

two subsidiary notions. Others may prefer to use

‘‘quality of design’’ and ‘‘quality of conformance.’’

The word ‘‘design’’ connotes what is desired and

something more general, strategic and important

than ‘‘features.’’ This term also seems better suited

when applied relative to a general market offering

whether a product or a service. Also, rather than

‘‘freedom from deficiencies’’ we find the word

‘‘delivery’’ more generally applicable to products

and services and more appealing than the authori-

tarian-sounding word ‘‘conformance.’’

Juran’s Economic Perspective

We are concerned about quality not necessarily

because we aspire to win quality awards. The real

reason is because striving for superior quality is

FIGURE 1 Quality defined by Juran as ‘‘fitness for use’’ and his

two subsidiary definitions as ‘‘features’’ and ‘‘deficiencies.’’

393 Quality Management and Juran’s Legacy

sound business economics. As Drucker (1973, p. 60)

noted, ‘‘Profit is not the explanation, cause, or ration-

ale of business behavior and business decisions, but

the test of their validity.’’ Milton Friedman has put it

more bluntly: ‘‘The business of business is business.’’

In today’s competitive business environment, quality

initiatives must justify themselves economically.

Quality is important because it delivers competitive

advantage and measurable, tangible economic bene-

fits in terms of reduced costs, better customer satis-

faction and improved bottom line profitability. The

opening chapter of the first edition of Juran’s Quality

Control Handbook is entitled ‘‘The Economics of

Quality.’’ In other words, in 1951 Juran was already

keen on the economic aspects of quality. Below

we discuss a firm-specific as well as a more global

societal perspective of the economics of quality.

Economic Model

Juran’s definition of quality and the two subsidiary

definitions as features and deficiencies provide the

basis for understanding the business economics of

quality. Typically, improving (design) quality in

terms of features will increase the cost of producing

a product or service. However, it will also allow the

company to charge a higher price and may increase

sales volume. In accounting parlance, added or

improved features have a beneficial top-line effect.

On the other hand, improving (delivery) quality by

reducing the number of errors and deficiencies will

usually dramatically reduce costs. Thus, it has a

‘‘middle line’’ cost-saving impact that immediately

trickles down to the ‘‘bottom line’’ as increased

profit. In the long run, reducing deficiencies will also

improve a company’s market reputation and its

brand. Therefore, reduced deficiencies may allow

for charging a premium price, increase the market

share or both. Figure 2 provides a graphical summary

of the economic relations. Given these simple econ-

omic relations, it can be seen that the common but

naı̈ve perception that ‘‘high quality costs more’’ is

not necessarily true. In fact, when it comes to defi-

ciencies the contrary is usually the case: high quality

costs less! This is why top executives ought to be

keenly interested in quality.

Cost of Poor Quality

Juran is credited for having developed most of the

fundamental concepts and theory behind what is call

cost-of-poor-quality (COPQ). Feigenbaum (1961)

and Crosby (1980) have also made important contri-

butions in this conceptual area, but Juran’s efforts

were early, comprehensive, and have had lasting

impact. Rather than the confusing notion of ‘‘cost

of quality,’’ used by many authors, Juran defined

the concept of ‘‘cost of poor quality’’ (COPQ) as

‘‘the sum of all costs that would disappear if there

were no quality problems.’’ Cost-of-poor-quality

concepts are accounting tools adapted to assess the

economic consequences of poor quality; that is, defi-

ciencies. Juran recognized that upper management

primarily understands and deals with money-related

issues. Thus, for communication reasons, he noted

that it makes sense to try to communicate in monet-

ary terms the impact of poor quality and use these

concepts to justify quality programs, select projects

and account for actual results. For more on this,

see Campanella (1990) and Bisgaard and Freiesleben

(2000).

Societal Impact of Quality

Juran has also addressed the larger societal

perspective of the importance of quality. Several edi-

tions of Juran’s Quality Handbook elaborate on these

themes. However, we briefly mention that Juran

pointed out that in an increasingly sophisticated

industrialized society, quality in a general sense is

imperative. We have all become dependent on

reliable products, services, and systems. We have

FIGURE 2 A graphical summary of the economic relations of

quality defined as features and deficiencies. Improving features

primarily impact the ‘‘top line.’’ Reducing deficiencies impacts

the ‘‘middle lines.’’ Both contribute to improving the ‘‘bottom

line.’’

S. Bisgaard 394

concerns about protection against power outages,

clean water, environmental catastrophes, safe drugs,

safe products and services in general, safe medical

care, reliable communication channels, healthy food,

etc. As Juran explained, ‘‘like the Dutch who have

reclaimed so much land from the sea, we secure

the benefits of technology. However, we need pro-

tective dikes in the form of good quality to shield

society against service interruptions and to guard

against disasters.’’ Juran calls this ‘‘life behind the

quality dikes.’’ Quality assurance is important from

a societal perspective even if rarely yet discussed

by economists.

A Unified Quality ManagementFramework: The Juran Trilogy

One of the key contributions by Juran is his uni-

fied concept of the Juran Trilogy, a comprehensive

framework and set of principles for organizing qual-

ity within an organization. This concept was first

articulated in Juran (1986). Based on his experience

at the Hawthorne Works of the Western Electric

Company, Juran rejected the notion early that quality

was only an inspection function; see Juran (1993).

He further objected to the notion that quality should

be the sole responsibility of the quality inspection

department. In his view, the responsibility should

remain with the operating departments; those that

make the defects have the responsibility for them,

not the inspectors. Any other allocation of responsi-

bility will have disastrous consequences for quality

and operational costs. Moreover, he was keen on

breaking down institutional barriers that prevented

quality improvement initiatives. As he learned as a

young engineer, in a traditional management

environment, ‘‘production was the job of one unit,

quality of another unit, and no one was in charge

of process improvement’’; see Juran (1993, p. 40).

Juran suggested that the financial function pro-

vides a useful managerial model to emulate for the

quality function both in terms of job description

and organization. It is the operating department’s

responsibility to produce financial results, not the

finance department’s. Likewise, it should be the qual-

ity function’s responsibly to coordinate activities rel-

evant to quality. Continuing the analogy, Juran

pointed out that the financial management function

consists of (a) budgeting, (b) budget control, and

(c) cost reduction. In generic terms, these three func-

tions are concerned with (a) planning, (b) control,

and (c) improvement. Juran suggested that quality

management likewise should be organized into three

equally important functions, (a) quality planning, (b)

quality control, and (c) quality improvement. Table 1,

based on Juran (1989) but adapted and modified,

outlines the tasks and responsibilities of these three

functions.

It may seem logical to implement quality planning

before engaging in quality control and quality

improvement. However, Juran suggested that it is

more pragmatic to start with quality improvement.

Any existing organization will be able to make sub-

stantial improvements right away with a small

upfront investment. This will help establish quick

wins and early employee buy-in. This is psychologi-

cally important for any change management pro-

gram; see Kotter (1995). Nevertheless, we discuss

below Juran’s Trilogy in the logical order of

planning, control, and improvement.

Quality Planning

To stay competitive, we must do more than just

remove deficiencies. We must develop new products

and services with new features that appeal to an

evolving customer taste and continue to be better

than the competitor’s offerings. Moreover, we must

do so without repeating the mistakes of the past

and without designing deficiencies into the product

(i.e., a product can also be a service or process).

Juran motivated this, by the analogy of an alligator

hatchery. If we are up to our waist in alligators, it

is not enough to kill the alligators around us. We

must also go after the production of new alligators,

the alligator hatchery. In other words, at the product

design stage we must proactively try to prevent the

need for subsequent quality improvement.

Quality planning is the process of preparing the

launch of new competitive products, services, and

processes that meet customers’ needs and expecta-

tions, minimize product and service dissatisfaction,

avoid costly deficiencies, optimize company per-

formance, and provide participation from those

affected by the product or service. Quality planning

is essentially a marriage of the traditional marketing

function with the research and development or

engineering design function assisted by the quality

395 Quality Management and Juran’s Legacy

function to provide tools, formal standards, measure-

ments, and data on performance. In this view, Juran

is even today at the cutting edge if not ahead of mod-

ern thinking relative to innovation and commerciali-

zation of new products; see, for example, Kotler

(2003).

Typically, quality planning involves developing

new or updating existing products to meet evolving

market demands or take advantage of new or emerg-

ing technologies. Table 1 as well as Figure 3 summar-

ize the steps of the quality planning processes.

Quality planning starts with establishing a team pro-

ject. Next, the cross-functional team needs to identify

the customers, discover the customers’ needs,

develop the product or service, develop the process

for the delivery of the product or service, develop

the controls, and transfer to operations. Upper

management must take responsibility for initiating,

supporting, provide resources and monitor the

quality planning process.

Quality Control: Managing the Control

Function

The second function of Juran’s Trilogy is quality

control. Although control is related to the original

inspection function and widely discussed in the stat-

istical literature since Shewhart (1931), Juran has con-

tributed profound managerial ideas to this area that

we now review. The basic tasks of the control func-

tion are summarized in the center column of Table 1.

Figure 4 provides a systems diagram of the uni-

versal control function. A sensor evaluates actual

performance. The actual performance is reported

back to the umpire. The umpire compares the actual

performance to the target (goal). If the difference is

significant (i.e., larger than noise), the umpire orders

a control action. The actuator makes the necessary

changes to bring the process back on target (goal).

This idea is applied universally at all levels ofFIGURE 3 Juran’s quality planning roadmap. Adapted and

modified from Juran (1992, p. 20).

TABLE 1 The Juran Trilogy Consisting of the Three Functions of Quality Planning, Quality Control and Quality Improvement. This Table

is Based on Juran (1989, p. 22) but Modified and Adapted

Quality management: Juran’s Trilogy

Quality planning Quality control Quality improvement

Determine who the customers are;

classify customer segments

Planning for control: Establish infrastructure for

improvement

Determine what the needs of each

customer segment are

Develop an understanding of

what needs to be controlled

relative to customer needs

Identify improvement projects

Design products with features and

specifications that satisfy the needs

of the customer segments

Develop a process flow diagram

Establish improvement teams

Develop products and processes that are

capable of delivering the product or

service

Choose what and where to

control; control points

Provide teams with resources,

training, and motivation:

Develop metrics and control mechanisms

for monitoring and control

Establish measures

Diagnose root causes

Provide training in the delivery processes

Establish goals and standards of

performance

Find remedies; improve

Executing control:

Establish controls to institutionalize

and hold on to the gains

Evaluate actual outcomes

Compare actual outcome to goals

Disband the team

Take action on the difference

S. Bisgaard 396

management. All employees, from the shop floor

worker to the CEO, exercise control. The difference

is the subject and scope of control. Workers typically

control product and process features. Executives

control budgets, sales, etc.

The above concepts are basic engineering control

theory. Nothing new here! Juran’s contribution has to

do with the management of the control function, the

distribution of authority, and responsibility, a topic

rarely discussed in the statistical process control

literature. We already mentioned that traditional

quality control by inspection carried out by a separ-

ate inspection department tends to develop an

unhealthy transfer of the responsibility from the

producer to the inspector. That approach has proven

ineffective. The issue of separation of authority from

responsibility however applies more generally, not

just to the shop floor inspection function. Juran

pointed out that necessary criteria for what he

termed ‘‘controllability’’ of a process are (1) knowing

what the goals are; (2) the ability to know what the

actual performance is; and (3) having means for

and authority to change the performance when the

process does not conform to goals or standards.

Without these three criteria fulfilled, a person cannot

fairly be said to be in control. Ideally, responsibility

for control should be assigned to individuals and

combined with authority. Indeed, to hold someone

responsible in the absence of controllability is bad

management. A test for completeness of planning

for control is to check whether the three criteria for

controllability are met.

It is interesting to note that Juran’s principle for

controllability is parallel to management expert

Peter Drucker’s notion of self-control; see Drucker

(1954, pp. 130–132). Drucker explains that control

can mean (1) ability to direct oneself and one’s

work or (2) domination of one person by another.

The objective of management control should be

the ability to direct oneself and one’s work! Drucker

claims that it is a major objective of modern man-

agement to substitute management by domination

with management by self-control. However, to

make management by self-control a reality requires

more than saying that it is desirable. It requires new

tools and far-reaching changes in traditional think-

ing and practices. A manager needs to know his

goals, be able to measure his performance and

results against goals, and have the authority to

make changes. Each manager should have the

information needed for control and receive it soon

enough to be able to make necessary changes for

achieving the necessary results. The information

should go to the manager himself, not to his

superior. Measurements should be a means of

self-control, not a tool of control from above. If

information technology is abused to impose control

on managers from above, it will inflict harm by

demoralizing management and by lowering the

effectiveness of managers. Enlightened management

will primarily rely on self-control and personal

responsibility. Thus, Drucker’s notion of self-control

is parallel to Juran’s controllability.

It is also interesting to note that the idea of

controllability is related to Shewhart’s and Deming’s

notions of special causes and common or systems

causes. Special causes are those the operator can

control, is responsible for, and have the authority

to change. Systems causes are causes inherent to

the system that only management can control and

FIGURE 4 The concept of process control.

FIGURE 5 A graphical depiction of the Juran trilogy, adapted

from Juran (1989).

397 Quality Management and Juran’s Legacy

have the authority to change. Hence, systems pro-

blems should be the responsibility of upper manage-

ment, not lower level workers. Anything else is

unfair, unreasonable, and counterproductive; see

Deming (1986).

Quality Improvement

We now consider Juran’s insight to how to orga-

nize and implement quality improvement. In Juran

(1993), he explained how he personally came to rea-

lize that quality control without quality improvement

was ineffective, if not futile. He related the story

about a Western Electric production process of a cer-

tain circuit breaker produced in large volume where

the defect rate was 15%. A statistical investigation

revealed that the copper wire exhibited excessive

variability from coil to coil, causing many out of spec

products. A remedy was subsequently developed to

compensate for the excessive variability. Eventually

this problem was permanently removed and the

defect rate reduced virtually to zero. Thus, rather

than a chronic waste of 15% of labor cost, materials,

and costly inspection, the cost of poor quality was

essentially permanently eliminated. Guided by this

watershed experience, in 1954 Juran explained to

Japanese executives that if they were serious about

quality, they should not just rely on inspection and

quality control. Rather, they should aggressively

pursue a strategy of quality improvement by

permanently removing chronic problems and waste.

Such a strategy would have an extraordinary return

on the investment, he promised. However, it would

require the executives to be involved, make quality

a strategic issue, break down barriers between

departments, and make quality a company-wide

effort.

It is likely that Juran was not the first to

recognize that quality improvement needs to be

done via special projects. However, his declaration

that ‘‘Quality is improved project-by-project and in

no other way’’ sums up his point. He has more

succinctly than anyone formulated the basic princi-

ples for how to implement and organize for quality

improvement.

Juran outlined a universal roadmap for quality

improvement that is analogous to detective work.

The initial steps are first to gather information on

needs, for example, on a cost-of-poor-quality basis,

then proceed to identify potential projects and select

projects, for example, based on a cost-of-poor-

quality ranking, and finally organize project teams.

Once the project teams are formed, they are required

to follow a six-step road map: (1) review that the

chosen project is important, (2) define the project

mission and objective, (3) diagnose root cause(s),

(4) develop a remedy and verify its effectiveness,

(5) deal with organizational resistance to change,

and (6) institute controls to hold the gains. Anyone

familiar with Six Sigma will recognize this as a

generic version of the Define, Measure, Analyze,

Improve, and Control (DMAIC) framework.

Upper Management Responsibilities

Like many other consultants in quality manage-

ment, Juran no doubt repeatedly experienced that

unless upper management were fully involved and

onboard, any long-term effort is futile. For executives

to just pronounce support and walk away is not

enough. Upper management must be thoroughly

engaged throughout the journey. But what does that

mean? Unlike many quality consultants who just

voiced frustration, Juran proceeded to be prescrip-

tive and explain in detail what top management

involvement ought to mean.

The first step is to establish a quality council. A

quality council is a group of top executives and

upper managers that develops the quality strategy

and guides and supports the implementation. The

responsibility of the quality council is to launch,

coordinate, and institutionalize annual quality

improvement goals and plans. The council formu-

lates policies and improvement priorities, establishes

metrics, establishes a project nomination and selec-

tion process, establishes a team selection process,

provides resources, assures implementation,

establishes needed benchmarks to gauge progress,

establishes a progress review process, faces up

to employee apprehension from workers made

superfluous, retrains or reassigns workers, provides

recognition, and revises the reward system to accom-

modate for quality.

Councils may be established at several organiza-

tional levels. Large companies may establish councils

on division as well as at corporate levels. At any

level, the membership should consist of upper man-

agers from line and staff. Members of higher levels

S. Bisgaard 398

often chair lower level councils. Senior manager

membership is a must. Otherwise, only ‘‘useful

many’’ type problems are solved, not the ‘‘vital

few’’ that produce the greatest return.

According to Juran, the chairperson of the coun-

cils should be the manager with overall responsibility

and authority for the unit. One member of the coun-

cil should be the director of quality. However, this

structure, although logically sound, may need modi-

fication. For example, General Electric (GE) has

found that senior managers may not necessarily have

sufficient skills and background in quality manage-

ment to effectively head a quality council. Instead,

GE has successfully experimented with having pro-

fessionally trained quality leaders head the councils.

However, the senior manager’s membership, pres-

ence, and support on the council were found to be

critical for success. This modified structure is not

unlike what is common practice in financial manage-

ment where it is typically the CFO and not the CEO

who heads up the finance committee.

Terminology

A more subtle but important contribution to qual-

ity management is Juran’s work on definitions and

terminology. Any scientific field requires its own

precise and well-defined terminology. Terminology

and definitions are fundamental to any science. With-

out it, confusion prevails. For example, a lay person

will not be particularly careful about using ‘‘heat’’

and ‘‘temperature’’ synonymously. However, a

well-trained physicist, chemist, or engineer would

be abhorred. For them these terms have precise

and very different meanings.

In the honorable scientific tradition of the French

chemist Lavoisier and the English physicist Faraday,

Juran recognized that for the nascent science of

quality management to become on a sound footing,

he needed to develop terms and define new con-

cepts. For example, as we already alluded to Juran

pointed out that ‘‘cost of quality’’ was an ambiguous

term. The cost-of-quality concept is too confusing

and difficult to narrow down. It fails to distinguish

between the cost of providing quality features and

the cost of deficiencies. Cost of poor quality, on the

other hand, can be precisely defined as the sum of

all costs that will disappear if the deficiencies are

removed.

We already discussed Juran’s definition of quality

and its two subsidiary definitions. However, he

defined many other important terms. For example,

he defined the meaning of customers, processors,

and suppliers and explained how these three com-

bined play a universal role in any process. Thus,

he called this triple role ‘‘Triprol.’’ For more, see

Juran and Godfrey (1999).

CONCLUSION: AN APPRECIATION OFJURAN’S IMPACT NOW AND IN THE

FUTURE

According to the economist Schumpeter (1950),

in a free market, economic reality is distinguished

by competition from new commodities, new

technologies, new sources of supply, new types of

organization—competition that commands a decis-

ive cost or quality advantage. Innovation-based

competition is extremely effective and strikes not

at the margins of existing firms but at their founda-

tions and threatens their survival. New innovations

render older innovations obsolete. Schumpeter

(1950) referred to this as ‘‘the perennial gale of

creative destruction.’’

Quality is about innovation—innovation of better

products, better services, better processes, and better

organizational structures. We used to think of quality

as only related to deficiencies and only related to

production floor problems. Modern quality manage-

ment defines quality more broadly as ‘‘fitness for

use’’ with the subsidiary meaning of features and

deficiencies. This expanded definition puts the cus-

tomer front and center and implies the need for

developing innovative ways to retain existing and

attract new customers with competitive market offer-

ings. Improvements aimed at eliminating chronic

sources of deficiencies from products and processes

are innovations that reduce cost and improve our

competitive position. But we cannot only rely on

reducing deficiencies. We must also compete on pro-

duct innovations that involve new features—design

and develop new products or services that provide

better value to the customers—market offerings that

better solve the customer’s problems. Juran (1989)

called the process of innovating new market offer-

ings ‘‘quality planning.’’ In Six Sigma terminology

this is called ‘‘Design for Six Sigma’’ (DFSS), but the

399 Quality Management and Juran’s Legacy

concept is the same. From this perspective, quality

management is a systematic and well-organized

approach to managing the process of designing

and developing process and product innovations.

Rather than hoping for haphazard occurrences, qual-

ity management puts innovation on a schedule and

provides the organizational framework for managing

innovations, large and small.

Juran’s Trilogy is a system for managing not just

quality but more general for managing innovation.

His teaching provides us with a practical and com-

prehensive framework for managing innovations.

Juran tells us what to do. His writing provides

hands-on operational information about how to go

about organizing and implementing a quality

management program within an organization. In

many aspects, Juran was ahead of his time.

If we should be critical of Juran it would be

relative to his impact. Many of his ideas are still

seldom used and are waiting to be ‘‘discovered.’’

Although he wrote many books and publications,

we attribute his lack of widespread popularity

and impact especially among managers to his

somewhat arduous engineering style of writing.

Juran is not easy reading, but patience makes it

highly rewarding. The irony is that what has made

his contributions to quality management so impor-

tant, namely his precise and detailed writing style,

has perhaps also turned off the larger population

of managers for whom quality management is so

essential. We even suspect that few Six Sigma prac-

titioners are familiar with his work. Indeed, in

some areas, Six Sigma is still trying to catch up

with or reinvent Juran’s ideas. For example, Six

Sigma does not yet adequately account for the

quality control function. Juran’s concept of quality

planning also appears to be much more compre-

hensive than the current design for Six Sigma con-

cept. Moreover, we find that in most areas of

quality management Juran is far better articulated

about fundamental principles.

Another criticism that arguably may not be widely

share is that Juran never took a strong position

relative to the ISO 9000 standardization movement.

We share with Juran a sincere appreciation for the

importance of standardization in general. We also

appreciate that in his role as the Nestor of the

quality movement, he may not have wanted to be

perceived as causing dissention. We also feel

confident that ISO 9000 initially was well intended.

However, in our judgment the effect has been a

diversion away from a forward looking business

focus—improving quality and satisfy customers—

to a defensive focus on satisfying self-appointed

quality auditors to pass a certification with what

sometimes more looked like Potemkin village

contrivances. Indeed, we feel that ISO 9000 has

had a corrupting effect on the quality movement,

especially in Europe.

In fairness Juran (1995, p. 595) did write that:

The ISO standards have a degree of merit. The criteriadefine a comprehensive quality control system. The certi-fication process may well get rid of the plague of multipleassessments which have burdened suppliers in the past.However, the criteria fail to include some of the essentialsneeded to attain world-class quality, such as personalleadership by the upper managers; training the hierarchyin managing for quality; quality goals in the business plan;maintaining a revolutionary rate of quality improvement;participation and empowerment of the workforce.

All in all, there is a risk that European companies are infor a massive let down. They are getting certified to ISO9000, but this alone will not enable them to attain qualityleadership.

We wish Juran would have been willing to be

more outspoken and publicly against ISO 9000 from

the start. It will take time to recover. Six Sigma, with

its business focus, is a good start. Lean Six Sigma is

perhaps what will help the quality movement regain

credibility among upper managers especially in

Europe.

To sum up, quality engineers and statisticians

involved with quality will be more effective if they

also concern themselves with the managerial

environment in which they operate and apply their

tools. There may be alternative systems for quality

management, but Juran’s trilogy is in our opinion

comprehensive and effective. Moreover, his hand-

on advice about going about organizing the quality

function and how to involve upper management is

unsurpassed. We hope this article will inspire

quality professionals to revisit the extensive Juran

literature.

ACKNOWLEDGMENTS

The author thanks Professor Xavier Tort and two

anonymous referees for very useful comments that

helped improve this article. The work on this article

S. Bisgaard 400

was supported by the Isenberg Program for Tech-

nology Management, the Isenberg School of

Management, University of Massachusetts Amherst.

ABOUT THE AUTHOR

Dr. Søren Bisgaard is the Eugene M. Isenberg

Professor of techonology management at the Univer-

sity of Massachusetts, Amherst, and Professor of

business and industrial statistics at the University of

Amsterdam, The Netherlands. He is a fellow of the

American Society for Quality and of the American

Statistical Association and has received numerous

awards, including the Shewhart Medal, the Wilcoxon

Prize, the Shewell Award and the Brumbaugh Award.

He serves on the editorial board for the Journal of

Quality Technology, Quality Engineering and is on

the management board of Technometrics.

REFERENCES

Bisgaard, S., Freiesleben, J. (2004). Six Sigma and the bottom line. QualityProgress, 37(9):57–62.

Campanella, J., Ed. (1990). Principles of Quality Costs, 2nd ed.Milwaukee, WI: ASQC Quality Press.

Crosby, P. B. (1980). Quality Is Free. New York: Penguin.Deming, W. E. (1986). Out of the Crisis. Cambridge, MA: MIT Center for

Advanced Educational Services.Drucker, P. F. (1974). Management: Task, Responsibilities and Practices.

New York: Harper & Row, Publishers.Feigenbaum, A. V. (1951). Total Quality Control. New York: McGraw-Hill.

Garvin, D. (1988). Managing Quality. New York: The Free Press.Kotler, P. (2002). Marketing Management, 11th ed. Upper Saddle River,

NJ: Prentice Hall.Kotter, J. P. (1995). Leading change. Harvard Business Review. 73(2).Juran, J. M. (1945). Management of Inspection and Quality Control. New

York: Harper and Brothers Publishers.Juran, J. M., Ed. (1951). Quality Control Handbook. New York:

McGraw-Hill.Juran, J. M., Ed. (1962). Quality Control Handbook, 2nd ed. New York:

McGraw-Hill.Juran, J. M. (1964). Managerial Breakthrough. New York: McGraw-Hill.Juran, J. M., Ed. (1974). Quality Control Handbook, 3rd ed. New York:

McGraw-Hill.Juran, J. M. (1986). The quality trilogy: A universal approach to managing

for quality. Quality Progress, 19–24.Juran, J. M. (1988a). Juran on Planning for Quality. New York: The Free

Press.Juran, J. M., Ed. (1988b). Quality Control Handbook, 4th ed. New York:

McGraw-Hill.Juran, J. M. (1989). Juran on Leadership for Quality. New York: The Free

Press.Juran, J. M. (1992). Juran on Quality by Design. New York: The Free

Press.Juran, J. M. (1993). Made in USA: A renaissance in quality. Harvard

Business Review, 42–50.Juran, J. M., Ed. (1995). A History of Managing for Quality. Milwaukee,

WI: ASQ Press.Juran, J. M. (2004). Architect of Quality: The Autobiography of Dr. Joseph,

M. Juran. New York: McGraw-Hill.Juran, J. M., Godfrey, A. B. (1999). Juran’s Quality Handbook, 5th ed.

New York: McGraw-Hill.Juran, J. M., Gryna, F. M. (1993). Quality Planning and Analysis, 3rd ed.

New York: McGraw-Hill.Millman, S., Ed. (1984). A History of Engineering & Science in the Bell

System. Indianapolis, IN: AT & T Bell Laboratories.Schumpeter, J. A. (1950). Capitalism, Socialism and Democracy, 3rd ed.

New York: Harper and Row.Shewhart, W. A. (1931). The Economic Control of the Quality of

Manufactured Product. New York: Macmillan.

401 Quality Management and Juran’s Legacy

Control (design + material + product -h process)^ Costs (inspection + rejects)

X CustoiTier satisfaction

TOTALQUALITY

CONTROLBy Armand V. Feigenbaum

To design, process, and sell products com-petitively in the 1956 market place, Americanbusinessmen must take full account of thesecrucial trends;

e Customers — both industrial and consumer— have been increasing their quality requirementsvery sharply in recent years. This tendency islikely to be greatly amplified by the intense com-petition that seems inevitable in the near future.

For example, the electrical relay that could com-mand the lion's share of the 1950 industrial mar-ket is no longer acceptable for 1956 operatingneeds. Consumers are progressively more minutein their examination of the finish of appliances,or in their judgment of the tone of a radio or tele-vision set. Even for military products on whichquality has always been the major consideration —e.g., jet engines, airborne electronics, and ordnance— specifications are continually being made morerigorous.

tí As a result of this i^icreased customer demandfor higher quality products, present in-plant qual-ity practices and techniques are now, or soon willbe, outmoded.

Thus, the machined part that could once bechecked with a pocket scale or a pair of micrometersmust now be carefully measured with an air^auge; and material that could once be visuallyiccepted if it were "reddish brown and shiny"nust now be carefully analyzed both chemically

and physically to assure that it is beryllium cop-per instead of phosphor bronze. At the same time,automation, in which rapid quality evaluation is apivotal point, has magnified the need for mecha-nization of inspection and test equipment — muchof which is now in the hand-tool stage. Indeed,the qualit}' control content of the manufacturingequipment investment dollar may well double inthe next decade to purchase the benefit of thismechanization.

€ Quality costs have become very high. Formany companies they may he much too high ifthese companies are to maintain and improve theircompetitive position over the long run.

In fact, quality costs (inspection, testing, labora-tory checks, scrap, rework, customer complaints,and similar expenses) have crept up to become amultimillion-dollar item. For many businesses theyare comparable in degree with total direct labordollars, with distribution dollars, or with purchasedmaterial dollars! While I can find no documentedresearch on the subject, evidence points stronglyto the fact that many businesses have quality-costexpenditures representing 7%, 8%, 10%, andeven more of their cost of sales!

e Taken together, these three trends spell outthe twin quality objective that 1956 competitiveconditions present to American business manage-ment: (a) considerable improvement in the qualityof many products and many quality practices, and,at the same time, (b) substantial reductions in theover-all costs of maintainijtg quality.

93

94 Harvard Business ReviewUnder these conditions, if quahty must be not

only maintained but upgraded, the wave of thefuture looks like an expensive one to ride. Howmany of the frailer business craft will be able toavoid getting swamped?

Broad Scope

Fortunately, there is a way out of the di-lemma imposed on businessmen by increasinglydemanding customers and by ever-spiraling costsof quality. This "way out" seems to lie in a newkind of quality control, which might be called"total quality control."

The underlying principle of this total qualityview — and its basic difference from all otherconcepts — is that, to provide genuine effective-ness, control must start with the design of theproduct and end only when the product hasbeen placed in the hands of a customer who re-mains satisfied.

The reason for this breadth of scope is thatthe quality of any product is affected at manystages of the industrial cycle:

(1) Marketing evaluates the level of qualitywhich customers want and for which they are will-ing to pay.

(2) Engineering reduces this marketing evalu-ation to exact specifications.

(3) Purchasing chooses, contracts with, and re-tains vendors for parts and materials.

(4) Manufacturing engineering selects the jigs,tools, and processes for relay production.

(5) Manufacturing supervision and shop oper-ators exert a major quality influence during partsmaking, subassembly, and final assembly.

(6) Mechanical inspection and functional testcheck conformance to specifications.

(7) Shipping influences the caliber of the pack-aging and transportation.

In other words, the determination both ofquality and of quality cost actually takes placethroughout the entire industrial cycle. This isthe reason why real quality control cannot beaccomplished by concentrating on inspectionalone, or design alone, or reject trouble-shootingalone, or operator education alone, or statisticalanalysis alone — important as each of these in-dividual elements is.

The breadth of the job makes quality controla new and important business managementfunction. Just as the theme of the historical in-spection activity was "they (i.e., bad parts) shall

not pass," the theme of this new approach is"make them right the first time." Emphasis ison defect prevention so that routine inspectionwill not be needed to as large an extent. Theburden of quality proof rests, not with inspec-tion, but with the makers of the part — ma-chinist, assembly foreman, vendor, as the casemay be.

Like traditional inspection, the quality controlfunction in this total quality view is still re-sponsible for assurance of the quality of prod-ucts shipped, but its broader scope places a majoraddition on this responsibility. Quality controlbecomes responsible for quality assurance at op-tiiiiuni qziality costs.

The total quality view sees the prototype qual-ity control man, not as an inspector, but as aquality control engineer — with an adequatebackground of the applicable product technol-ogy and with training in statistical methods, ininspection techniques, and in other useful toolsfor improving and controlling product quality.

Compared With Other ViewsIt may serve further to clarify the character

of the total quality view if we compare it withother quality control concepts. Actually, therehave been and are today a great many differentconcepts both of the meaning of the term "qual-ity control" and of what the principal elementsof the quality control activity are. The two mostwidely accepted of these concepts may be de-scribed as the "modern inspection view" and the"statistical view."

Historically, quality control meant nothingmore than the activity of traditional factory in-spection, which was intended to protect the cus-tomer by screening bad material from goodprior to shipment. In the modem inspectionview, quality control means this traditional in-spection function updated and made more ef-ficient by the use of certain statistical methodsand work-in-process inspection routines. Thus:

e Statistically verified sampling plans assure thequality of outgoing lots better and more econom-ically than do the older ioo% inspection or hit-or-miss spot check procedures.

C In-process sampling inspections detect qualityerrors before too many bad parts have been pro-duced, and are consequently more effective thanconcentration on final inspection with its risk ofproducing a large number of bad parts.

An impressive weight of dollars-and-cents evi-dence demonstrates that such techniques repre-

sent a great improvement over old-fashionedpractices. Probably most systems of quality con-trol in American business today are examples ofthis modern inspection point of view. That is,they see the prototype quality control man as awell-grounded inspection specialist who has badtraining in useful statistical methods.

The statistical view, in turn, reflects the majorand increasing contributions which probabilitymethods are making to the improvement of in-dustrial decision making. It is a view that pre-dominates both in the literature of quality con-trol and in professional meetings on the subject.It sees the prototype quality control man as anindustrial statistician, who works on problemsranging from the design of laboratory experi-ments, through the establishment of controlcharts for production processes, on to the analy-sis of manufacturing rejects. It sees him, inother words, as capable of making contributionsin fields not directly connected with productquality, such as time study and safety.

These modern inspection and statistical con-cepts of quality control have been and are highlyuseful in the areas of product quality whichthey cover. But, compared with total qualitycontrol, their scope is much too limited; they areable to provide only a partial grasp of the over-allquality problem that faces American business-men. They simply are not geared to the factthat quality considerations are involved in everyphase of industrial activity, and are not equippedto keep over-all costs of quality at a minimum.

Effect of Cost AccountingIf the burden and sharp upward trend of

these quality costs — and the need for genu-inely broad quality control effort — are onlynow becoming recognized in some businesses,part of the reason must be ascribed to traditionalindustrial cost accounting practices. Cost ac-counting methods often have not identified andgrouped quality costs in a form suitable for thedevelopment of adequate controls. The magni-tude of the quality cost sum has tended to beobscured by the piecemeal identification of cer-tain individual quality cost elements: for exam-ple, scrap and spoilage, or field complaint ex-pense. Most often, quality cost has been con-ceived as tbe cost of the company inspection ac-tivity — actually just a fraction of over-all qual-jl-y (.Q5(- and controls have been establishedon tbis fragmentary basis, ^ ^ ^ ^ ^

Regardless of the source of the fault, the only

Total Quality Control 95clear answer to the quality cost problem seemsto lie in the new concept of total quality control.

Operation of the Function

The work of this total quality control func-tion may be classified into four broad categories,as follows (see EXHIBIT I for elaboration):

1. New design control, or the planning of controlsfor new or modified products prior to the startof production.

2. Incoming material control, or the control ofincoming purchased parts and materials.

3. Product control, or the shop floor control ofmaterials, parts, and batches from machines,processes, and assembly lines.

4. Special process studies, or the conducting ofspecial analyses of factory and processingproblems.

In this work, the two basic responsibilities ofthe quality control function are: ( i ) to providequality assurance for the company's product —i.e,, simply, to be sure that the products shippedare right; and (2) to assist in assuring optimumquality costs for those products. It fulfills theseresponsibilities through its suhfunctions: qual-ity control engineering, inspection, and test,which operate a continuous feedback cycle:

(1) Quality control engineering does qualityplanning; this establishes the basic framework ofthe quality control system for the firm's products.

(2) Inspection and test do quality measuring;this determines, in accordance with the qualityplan, the conformance and performance of partsand products with engineering specifications,

(3) There is rapid feedback to quality controlengineering for quality analysis, which results innew quality planning, thus completing the cycle.(This analysis also fosters corrective action forproduct quality deviations.)

Engineering Component

The true nerve center of the total qualitycontrol function is the engineering component.Its activities in each of the four broad qualitycontrol jobs deserve examination in some detail.

New Design ControlIn this area, quality control engineering pro-

vides three main activities:

1. Preproduction service to design engineer-ing and inannfacturing engineering in analyzing

96 Harvard Business Review

EXHIBIT I. QUALITY CONTROL IN THE QUALITY ACTIVITY CYCLE

ACTIVITIES

Rectivinq a Inap«c1inqQuality Moteriot

INCOMING MATERIAL CONTROLManufacturing Quality

Por)8 a Productj

Infptctinq BTntlnq

Ouaftty Product«

PRODUCT CONTROL

SPECIAL PROCESS STUDIES

OuoljtyProducti'

the quality-ability of new products and produc-tion processes, and in de-bugging quality prob-lems — This assures a product which will be asdefeet-free as possible prior to the start of pro-duction. Among the new technical tools whichthe quality control engineer brings to this effortare process quality capability studies, toleranceanalysis technique, pilot run practice, and awide variety of statistical methods.

2. Planning of inspections and tests to becarried on when production is under way on thenew product — This is to establish continuouscontrol of in-process quality. It involves deter-mining the following:

• Dimensions and characteristics of the partsto be checked.

• Degree to which they are to be checked.• In-process and final production points at which

checks are required.• Methods and procedures to be used — includ-

ing statistical sampling plans, control charts,and so on.

• Personnel who will make the checks — thatis, production operators or people from theinspection and test subfunctions.

3. Design of genuinely modern inspectionand testing equipment, which, to the fullest pos-sible extent, is physically integrated with manu-facturing equipment to permit the viachine to

check its own work — The aim of this activityis economical investment expenditures, maxi-mum equipment utilization, and fullest practicalmechanization and automation both of opera-tions and of quality control paper work.

Incoming Material ControlIn this area, quality control engineering must

assist in the establishment of good quality rela-tionships with suppliers. It contributes to thisobjective in the following ways:

« By planning the periodic rating of the qualityperformance of present suppliers, it provides factswhich assist the purchasing function in quicklybringing satisfactory or unsatisfactory quality per-formance to the attention of vendors.

C By evaluating the quality capabihty of poten-tial suppliers, it provides facts which assist pur-chasing to select good quality vendors.

e By working with the vendors, it assists themin understanding the quality control requirementsof the purchase contracts they have won.

e By establishing quality certification programs,it places the burden of quality proof on the vendorrather than on an extensive, expensive in-plant in-coming inspection effort.

Product ControlIn this area, quality control engineering car-

ries on the cost measurement and quality costreduction project activity required for over-all

quality cost control and reduction, and it worksclosely with the inspection and test subfunctionswhich perform the actual measuring work. Italso performs process quality capability studiesto determine the quality limits within which amachine or process may be expected to operate.The aim is to make sure that parts will be routedto equipment which is capable of economicallymaintaining engineering specifications.

Special Process StudiesIn this area, the job of quality control engi-

neering is to analyze complex, in-process qualityproblems which have been fed back to it by in-spection and test. These studies are directedboth to the elimination of defects and to the de-velopment of possible improvements in presentquality levels.

Specialized Activity

Certain elements of this quality control engi-neering work have previously been performedon a sporadic or divided basis. But the qualitycontrol engineer himself is something new underthe sun. For quality control engineering is notmerely a new label for the inspection planningpackage, nor a fresh designation for the testequipment engineer, nor yet a technologicallyfiavored title for the industrial statistician. Itis, instead, a specialized activity with a charac-ter all its own, calling for a unique combinationof skills.

Quality control engineering work is the prod-uct of the cross-fertilization of modern develop-ments in several fields — in statistical method-ology, in fast-response high-precision inspectionand testing equipment, in management under-standing of the nature of the control functionin modern business. Altogether, it has the at-tributes of a genuinely new sector of the engi-neering profession.

In experience, education, aptitude, and atti-tude, the man entering quality control engineer-ing work today is, in fact, not very different fromthe man entering other longer established majortechnical fields as, for example, product engi-neering or manufacturing engineering. He mustpossess, or have the capacity to acquire, thenecessary product and process background. Hemust have the personal characteristics to workeffectively in a dynamic atmosphere with peopleof diverse interests. He must possess the tech-nical background which will enable him to ac-

Total Quality Control 97quire, if he does not already have it, a growingbody of quality control engineering knowledge.Finally, he needs the analytical ability to usethis knowledge in solving new and differentquality problems.

Inspection & Test

The planning and analysis work of the qual-ity control engineer makes a new, more positivetype of inspection and test both possible andnecessary in the modern quality control func-tion. Instead of policing the manufacturingprocess, this type of inspection and test becomesa direct part of that process by assisting in theproduction of good quality products.

Thus, during incoming material control andproduct control, the inspection and test sub-functions are responsible not only for fully estab-lishing that the materials received and the prod-ucts in-process and shipped are of the specifiedquality, but also for thoroughly and promptlyfeeding back facts for preventing the purchaseand production of poor quality material in thefuture.

This positive quality measuring requires onlya very minimum of routine hand-sorting inspec-tion and test. In product control. For example,this result is made possible through a continu-ous sequence of engineering work to assure thatwith the facilities provided production operatorscan make parts right the first time, know thatthey can, and have the necessary equipment andgauges to check their own work. On this basis,then, inspection and test can provide genuineassistance in the production of the right qual-ity by:

• Becoming auditors of the good quality prac-tices that have been established.

• Providing as much as possible on-the-spot,shop floor analysis of defects.

• Feeding back facts about these defects forcorrective analysis and action elsewhere.

Such quality assurance efFort has been termedcontrohaudit inspection and test. Inevitably itmeans the upgrading of traditional inspectionand test; it requires considerably fewer but muchmore highly qualified and more specialized per-sonnel — those who have genuine ability to behelpful in making the right quality. An in-stance in point is the arc-welding inspector whonow not only knows whether or not weld pene-tration on a part is satisfactory but also, in thecase of defective welds, may be able to counsel

98 Harvard Business Review

the shop on the reasons why the penetration hasbeen unsatisfactory.

Organizational Problem

In organizing a modern quality control func-tion, the first principle to recognize is that qual-ity is everybody's job.

In defiance of this principle, there have beenmany business experiments over the years whichhave attempted to make the quality activitycycle less of a decentralized, Tinkers-to-Evers-to-Chance sequence. Often these attempts havetaken the form of centralizing all quality re-sponsibility by organizing a component whosejob was handsomely described as "responsibilit)'for all factors affecting product quality."

These experiments have had a life span of aslong as six months — when the job incumbenthad the advantage of a strong stomach, a rhi-noceros hide, and a well-spent and sober boy-hood. Others not similarly endowed did notlast even that long.

The simple fact of the matter is that themarketing man is in the best position to evalu-ate adequately customer's quality preferences;the design engineer is the only man who caneffectively establish specification quality levels;the shop supervisor is the individual who canbest concentrate on the building of quality.Total quality control programs therefore require,as a first step, top management's re-emphasison the responsibility and accountability of allcompany employees in new design control, in-coming material control, product control, andspecial process studies.

The second principle of total quality controlorganization is a corollary to the first one. It isthat because quality is everybody's job, it maybecome nobody's job. Thus the second majorstep required in total quality programs becomesclear. Top management must recognize thatthe many individual responsibilities for qualitywill be exercised most effectively when they arebuttressed and serviced by a well-organized,genuinely modern management function whoseonly area of specialization is product quality,and whose only area of operation is in the qual-ity control job.

Location of the FunctionIn view of these two organizational principles,

where should the quality control function beplaced in the larger structure of company or-

ganization? Should it be part of marketing, ofengineering, of manufacturing? Should it re-port direct to general management?

While these are crucial questions, they are notsusceptible to categorical answers. Certainly,quality control in any company should reporthigh enough so that it can implement its respon-sibilities for quality assurance at optimum costs.Certainly, also, it should be close enough to thefiring line so that it will be able to fulfill itstechnological role. However, companies varywidely in their objectives, their character, theirphilosophy of organization structure, and theirtechnology. The answer to the question ofwhere to locate quality control will necessarilyvary also.

It may be worthwhile, however, to report onefirm's approach to this issue. In the GeneralElectric Company's product departments, eachof which operates as a decentralized businesswith profit and loss accountability reposing withthe department general manager, the cycle ofbasic quality responsibility is as follows:

« The marketing component is responsible forevaluating customers' quality preferences and de-termining the prices these customers are willingto pay for the various quality levels.

e The engineering component is responsible fortranslating marketing's requirements into exactdrawings and specifications.

e The manufacturing component is responsiblefor building products to these drawings, and forknowing that it has done so.

Within this structure of responsibility, qual-ity control clearly emerges as a manufacturingfunction. Thus, in the General Electric productdepartment, the quality control manager reportsto the chief manufacturing executive in that de-partment — the manufacturing manager — andoperates at the same organization level as theproduction superintendents, the managers ofmaterials, and the managers of manufacturingengineering.

The Results

Experience in an increasing number of com-panies shows that operation of a total qualitycontrol program has paid off in six ways:

1. Improved product quality.2. Reduced scrap, complaint, inspection, and

other quality costs.

3. Better product design.4. Ehmination of many production bottlenecks.5. improved processing methods.6. Development of a greater spirit of quality-

mindedness on the production shop floor.

Certainly our experience with this programhas been highly satisfactory in the Ceneral Elec-tric Company, where we have been developingthe concept and procedure for some years.

Total quality control has thus, in actual prac-tice, been successful in meeting the dual objec-tive of better quality at lower quality costs. Thereason for the satisfactory better-quality resultis fairly clear from the very nature of the pre-vention-centered, step-by-step, technically thor-ough program. But the explanation may notbe nearly so obvious for the accompanying by-product of lower over-all quality cost. Thisneeds to be spelled out, especially since it in-cludes, in the long run, lower expenses for thequality control activities themselves as comparedwith the costs of traditional inspection and test.

Costs of Quality

The reason for the favorable cost result oftotal quality control is that it cuts the two majorcost segments of quality (which might be calledfailure and appraisal costs) by means of muchsmaller increases in the third and smallest seg-ment (prevention costs). Why this is possiblecan be seen as soon as the character of thesethree categories is considered:

(1) Eailure costs are caused by defective mate-rials and products that do not meet company qual-ity specifications. They include such loss elementsas scrap, spoilage, rework, field complaints, etc.

(2) Appraisal costs include the expenses formaintaining company quality levels by means offormal evaluations of product quality. This in-volves such cost elements as inspection, test, qual-ity audits, laboratory acceptance examinations, andoutside endorsements.

(3) Prevention costs are for the purpose ofkeeping defects from occurring in the first place.Included here are such costs as quality control en-gineering, employee quality training, and the qual-ity maintenance of patterns and tools.

In the absence of formal nationwide studiesof quality costs in various businesses, it is im-possible to generalize with any authority aboutthe relative magnitude of these three elementsof quality cost. However, I believe it would not

Total Quality Control 99be far wrong to assert that failure costs mayrepresent from one-half to three-quarters oftotal quality costs, while appraisal costs prob-ably range in the neighborhood of one-quarterof this total. Prevention costs, on the otherhand, probably do not exceed one-tenth of thequality cost total in most businesses. Out of thisi o % , usually S%-g% are directed into suchtraditional channels as pattern and tool main-tenance and the specification-changing or inter-preting work of product engineering. This leavesonly 1% or 2% that is spent for elements ofquality control engineering work.

It is a significant fact that, historically, underthe traditional inspection function, failure andappraisal costs have tended to move upward to-gether, and it has been difficult to pull themdown once they have started to rise. The reasonfor this relationship is that:

C As defects increase — thus pushing up failurecosts — the number of inspectors has been in-creased to maintain the "they shall not pass" screento protect the customer. This, in turn, has pushedup appraisal costs.

C For the reasons mentioned earlier in this ar-ticle, screening inspection does not have mucheffect in eliminating the defects, nor can it com-pletely prevent some of the defective products fromleaving the plant and going into the hands of com-plaining customers.

€ Appraisal costs thus stay up as long as failurecosts remain high. In fact, the higher these fail-ure and appraisal costs go, the higher they arelikely to go without successful preventive activity.

Once these two main elements of quality costhave started to rise — as they seem to havethroughout industry generally today — the onebest hope for pulling them to earth again seemsto be spending more on the third and smallestelement, namely, prevention cost. The i o %now spent may well need to be doubled, muchof the increase going for quality control engi-neering as well as for improved methods of in-spection and test equipment automation.

At first glance such increases in preventioncosts may not seem to be in the interest of qual-ity cost improvement, but this objection is rap-idly dispelled as soon as results are considered.Translated into quality cost terms, the operationof total quality control has the following se-quence of results:

(1) A substantial cut in failure costs — whichhas the highest cost reduction potential of all qual-

100 Harvard Business Reviewity cost elements — occurs because of the reducednumber of defects and the improvements in prod-uct quality brought about by modern quality con-trol practice.

(2) Fewer defects mean somewhat less need forroutine inspection and test, causing a reduction inappraisal costs.

(3) Better inspection and test equipment andpractices, and the replacement of many routineoperators by less numerous but more effective con-trol audit inspectors and testers, bring about addi-tional reductions in appraisal costs,

(4) Because the new cojitrol-andit inspectionand test is effective in preventing defects, appraisaldollars for the first time begin to exercise a posi-tive downward pull on failure costs.

The ultimate end result is that total qualitycontrol brings about a sizable reduction in over-all quality costs, and a major alteration in theproportions of the three cost segments. Nolarge, long-term increase in the size of the qual-ity control function is required as a necessarycondition for quality cost improvement. Instead,quality control expense, as a proportion of totalcompany expense, will be down in the long run.Improvements of one-third or more in over-allquality costs are not unusual.

Quality Dollar BudgetingIt is worth noting that this identification and

analysis of quality costs permits a major forwardstep in the business budgeting process. That is,it makes feasible determining the dollars neededfor quality control, not on the basis of historicalinspection cost experience, but on the basis ofcurrent company objectives in product qualityand quality costs.

Quality needs and problems differ so mucbfrom company to company that it is not realisticto generalize about the specific mix of qualitycosts that should be budgeted under total qual-ity control. But the direction of budgetarytrends may be suggested by an example, whichembodies current industrial experience in tbis

area. EXHIBIT II shows how one company ex-pects to cut its quality costs by switching froma mild version of the inspection view to totalquality control. The company anticipates thattotal quality expenses will drop from the currenthigh of 7% of sales to 5%, with declinesachieved both in failure and appraisal costswhile prevention costs increase from only 0.70%of sales to a still modest 1.25%. In this ex-ample, the cost savings budgeted are rather mod-erate, owing to the presence of complicatingfactors such as the following:

• An anticipated 50% increase in sales over thenext five years — from $50 million to $75million.

• Planned additions to a product line that isalready highly technical and diversified, henceaccompanied by major quality problems.

Such a planned 30% improvement in qualitycost ratios is feasible, indeed conservative, witha successful total quality control program —even with a 50% business expansion and evenwith counterbalancing quality eost increasesbrought about by the introduction of new prod-uets. While the company in the example maynot be typical (probably there is no such thingas a typical business enterprise), it is at leastillustrative of the good results that can beachieved even when circumstances pose un-usual difficulties.

ConclusionTotal quality control thus represents another

forward step in management science. Its inte-gration of design-througb-shipment control ofthe many elements in the quahty picture makesit much more effective than the unlinked frag-mentary controls of the past. As a major newbusiness management activity, it provides pro-fessional effort in meeting the objective of as-sured product quality at minimum quality cost.

With this concept, inspection and test have

EXHIBIT II. BUDGETED QUALITY COSTS AND SAVINGS UNDER TOTAL QUALITY CONTROL(Assuming sales increase from $50 million to $75 million)

Qualitycost element

FailureAppraisalPrevention

Total

$ 2

$3

Total

Present

,275,000875,000350,000

,500,000

dollars

S-year goal

$2,062,500750,000937.500

$3,750,000

Per

Present

4-55%1-750.70

7.00%

cent of sales

S-year goal

2.75%1.001.25

5.00%

Per centquality

Present

65%2510

roo%

of totalcost

S-year goal

55%20

25

100%

a chance to develop in the direction that con-scientious inspectors and testers have alwayssought; that is, into an activity with a positiverole in assisting other members of the manufac-turing, engineering, and marketing team towardquality improvement and quality cost control.No longer are inspection and test confined toessentially a negative, fist-shaking role in sortinghad parts from good, a role placing them con-tinually on the defensive and evoking the hos-tility of other managers.

Further, those tools of statistical methodologythat have proved practical and useful can nowbe brought to their fullest effectiveness. Withthe quality control engineers as tool builders,and the control audit inspectors as tool users,statistical techniques can be put to work in adown-to-earth fashion that welds them into regu-lar day-by-day controls. No longer will thesetechniques be treated — as too often in the past— merely as curiosities, to he employed in spe-cial situations on a pinch-hitting basis.

With equipment for inspection and test adirect and major responsibility of the quality

Total Quality Control 101control function, more use can he made ofequipment specialists who wish to concentratetheir skills on the great needs, opportunities, andunique complexities of today's quality controlequipment field.

Total quality control thus welds this newtechnology into a strong yet flexible weapon,capable of successfully coming to grips with thethree major quality problems that modern busi-ness must face and solve: the upward customerpressure on quality levels, the resulting rapid ob-solescence of quality praetices, the very highlevel of quality costs.

M^hile delivery and other factors may sell aproduct the first time, it is usually quahty whichkeeps the product sold and which keeps the cus-tomer coming back a second and a third time.Quality cost — perhaps 8% or more of eost ofsales — is one of the major elements of productcost that must be made right to permit the set-ting of the right price to the customer. Helpingto assure this right quality at this right qualitycost is the way the new total tjuality control canserve its company in the years ahead.

FOREMEN recognized that in the last analysis quality performance is a mat-ter of attitudes, of both the individual worker and the group as a whole.

Foremen who were prone to consider a lack of interest in quality as someinnate attribute of workers were also likely to be the ones who tried to dictatea "you-do-it-this-way-or-else" policy. Those foremen who assumed that theaverage man wanted to do a good job, who gave careful training to the newoperator, and who helped outfall of their operators when they needed it weremore likely to he rewarded with good quality work.

The present study lends support to the administrative principle of delegat-ing responsibility to hourly wage earners and especially to utility men. Thoseforemen most highly thought of by management were those who freely ad-mitted that they could not possibly check on all the quality problems through-out the workin'a day. It was their conviction that a foreman must dependupon his whole"team — his utility man and his other operators — to speakup on quality problems and so prevent bad work from going through to theinspector Quality consciousness, it may be inferred, is not somethmg de-veloped solely between a foreman and the individual worker. The men musttake on a self-disciplinary s,roiip responsibility for doing a good )ob and doingit for a foreman who is their leader in fact as well as m tttle.Charles R. Walker, Robert H. Guest, and Arthur N. Turner, The Foreman

on tke Assembly Line , , , . . ^ /-Cambridge, Harvard University Press, 1956, p. 75.

öer Knowing the cause

is half the battleSymptoms can be signposts on the route to curing problems

^ A ^ V ^ ^ ^ V hen we come across a prob-^Ê^m ^Ê^m 1cm, sümedmes we know^ W ^ W more about the problem

itself than we do about the cause of the prob-lem. In manu facruring, there may be manyvariables and contributing factors. The morecomplex the production process, the moredifficult it may be to sort out the problem'scause.

Our mission needs to be to find andcure the causes using the symptoms as sign-

posts. Luckily, we haveSeven Quality Tools in( •iir "quaiit)' toolbox" toliL-lp us identify causes.

The cause and effectJnii;ram was developedi)\ Kaoru lshikawa tosorr out possible causes

This exercise stirs the thought processes ofthe group's members and extracts core processknowledge to identify likely contributing fac-tors. Along the way, each person in the groupasks and answers questions about possiblerelationships of the five M categories.

Other tools may also be used in order totest causes and provide solutions. Check sheetsmay be used to determine fretjuency of variouscause occurrences. But how do we use a list orcheck sheet to capture information?

ID (Daceor part tt)

1/1/08

1/2/08

1/3/08

Defect 1:too short

3

2

1

Defect 2:too long

4

2

S

Defect 3:aborted cycle

3

3

3

Defect 4:bad stock

1

3

0

Defect 5:other

0

1

2

Figure 2-Check Sheet for Part Defects

E/izahel/i Mazc-Eiuerv

associated with producdon problems.This method is commonly referred to as:i iishbone diagram because of how thesordng process is illustrated.

The diagram is best constructed by agroup that is knowledgeable about theprocess at hand. The group starts with

the fishbone framework composed of the fiveATs: Materials, Machines, Measurements, Man,and Methods. The group brainstorms whatvariables in each category may contribute tothe current problem.

For example, what if we examining a pileof defective parts? We found that there wereseveral key defects. This check sheet allows usto capture that informadon and quantify thefrequency of these defects.

We can total the frequency of each defectand generate a picture of the greatest defectproblem. In this exampie, we use the data col-lected on the check sheet by plotting the totalsin descending order of occurrence. This is thePareto chart shown in Figure 3.

Pareto charts, histograms, and control

\ . Materials \ ^ Machines \

\ \

M a n ^ / ^/Methods

Measurements\

\ Effect\ or

\ Problem

Figure I-Diagram of Five M's Cause Categories andResulting Effect

Freq

uenc

y

12

10

8

6

4

2

Defect 2;too long

Defect 3:aborted

cycle

Defect 1:too short

Defect 4:bad stock

Defect 5:other

2 8 November/December 2008

Figure 3-Pareto Chart of Part Defects

www.ToolingAndProduction.com

quality

Machine Parts Process Flow

DrawingReview

MachineSetup

StockInspection

MachineParts

Defective

Good

CheckSheet

Figure 4—Flowcharts

chart use may also point to signiricantcauses. Flowcharts can help map theprocess to establish where factors areintroduced.

X

Upper Control Limit

Lower Control Limit

Parts & Size

0.90

0.91

0.920.930.940.950,960.970.980.991.001.011.021.031.041.051.061.071.081.091.10

1 2 3

f[

4 5

1

6 7 8

- -

9 11 12

i"

Figure 6-Controt Chart

What is a histogram? How does thattool work? It works by showing therelative bel] curve of data distribution.Figure 5 is an example. The part sizeis to be 1.0". An inspector at the endof the process has collected data.

Once key contributing factors havebeen identified, causes and possiblesolutions can be reviewed for feasibil-ity, tested for validitj', and posted astroubleshooting fixes.

The control chart helps to iden-tify out-of-control points and assign-able causes. First something must beknown about the normal in-specifica-tion manufacturing process. The pro-cess average, or in-specification mean,is plotted as the centerline. Upper andlower control limits can be calculatedand plotted.

This process average can be consid-ered the centerline of a road and the

control limits the edges ofthe toad that we must staybetween to avoid runningoff the road surface. Bv

plotting new data within these limits,we can monitor and respond to trendsor out'Of-control conditions.

A scatter diagram shows nonlinearreiationships that result in scattereddata piots. If variables have relation-ships, these will show clear correla-tions. This scatter diagram is fromWikipedia, and indicates the relation-ship between duration of an eruptionof Old Faithful geyser and the lengthbetween eruptions.

The discipline of total quality' con-trol uses Seven Quality Tools to iden-tify focus for continuous improve-ment efforts in manufacturing. Overthe years, c|ualit)' experts realized thatmany quality-related problems couldbe solved with the above tools.

Elii^abeth Ma^e-Emery is a quality profes-sional from Di/yton, OH. For more infor-mation iihout quality tools or related topics,e-mail her at ema^eemery@jiu. edu.

Old Faithful Eruptions

oO)

oOD

O

oCD

Eruption Duration (Min)

\ I i I 1 I 1 r1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

Figure 5-Histogram Figure 7-Scatter Diagram

www.ToolingAndProduction.com November/December 2008 I S ° t i r i 5 Ä „