Project Management Institute

Applying Japanese total quality management to software project management

Issue Focus: Information Systems

Still another article applicable to all projects, not just software projects. It provides insights into how quality, and all its benefits, can be achieved on project-type work. “Quality is Free” is applicable to projects, too.

Lois Zells, Lois Zells & Associates, Scottsdale, Arizona

World-class rivals of U.S. companies have created a sudden and widespread outbreak of quality and management reform. Furthermore, the past practices that enabled U.S. organizations to achieve supremacy in international marketplaces are simply not working anymore. To compound matters, the Japanese now plan to develop prominence in the software exporting industry. If we are not careful, American software workers will lose employment to international competition just as, in recent years, U.S. workers have lost employment in our steel, automotive, and electronics industries.

Although they were originally considered applicable in only the manufacturing environment, Japanese quality concepts have caught on in all sectors of the economy. For example, many service organizations like health care and data processing firms are addressing initiatives in total quality management. The U.S. government is especially interested in this approach. To this end, each of the U.S. military organizations has major imperatives to implement the quality process.

Figure 1. The Model for S/TQM

Figure 1. The Model for S/TQM


Declaration of an organization's number one priority issue varies, depending on its unique goals or even the latest theories being touted in the press. Nevertheless, under the obvious and absolute superordinate goal of making a profit, there are three PREMIER organizational goals. They are: increasing quality, reducing costs, and/or reducing schedules [1].

These are the areas that are controllable and in which poor results are reversible! These elements are so closely tied that, if implemented correctly, improvements in any one of them always and automatically causes improvements in the others [2].

As an example, increasing quality will also cause a decrease in costs and in schedules. On the other hand, in order to decrease schedules, practitioners must look for the jobs (inside the schedule) that are taking the longest to complete—and shorten them. As part of reducing job durations, practitioners must learn to eliminate all need to repeat any incorrectly completed steps within the job: by ultimately doing each of the steps properly the first time. Of course, this causes corresponding increases in quality and decreases in costs.

In summary, each of the PREMIER goals is simply a different path to the same payoff: increased profits.


This issue is about uniting the finest of both American and Japanese quality approaches to bring about improvements in quality and productivity through Software Total Quality Management (S/TQM). (See Figure 1.)

A Total Quality Management (TQM) discipline embraces several arenas: Quality Function Deployment (QFD), Hoshin Strategic Planning (HSP), Project Management (PM), Statistical Process Control (SPC), Design of Experiments (DOE), Concurrent Engineering (CE), Productivity Gain Sharing (PGS), and Total Employee Involvement (TEI). In the software project management arena, these disciplines make more sense if they can be mapped into a paradigm like the model for S/TQM.


Total quality management means total; and, for most environments, this is one of the most difficult aspects of the entire process. Total means that all of the requirements are implemented by all of the people, all of the time. It does not mean that if an organization does not fancy one of the pieces they can simply disregard it and still expect to achieve the desired results. TQM does not mean that participation is inconsistently disseminated, voluntary, or part-time. If every individual in the organization cannot regularly describe his or her current, ongoing quality improvement initiatives, then tQm is simply not in effect!

Management Commitment

Working the model for S/TQM from the outside in, the most critical aspect of TQM is management's support. A high level of senior management commitment must be seen to demonstrate an uncompromising desire for success. Deming found this to be the single, most important step in the quality process [3]. The senior managers must be day-to-day participants, providing support, time, and resources. Furthermore, the process must be so important to them that their participation is never delegated. The effort should not even be considered without this commitment—and canceled if it disappears after the effort has been initiated.

Japanese Strategic Planning

When U.S. researchers went to Japan to learn about TQM, they found that the one thing that all world-class Japanese organizations had in common was their approach to, what in the West is called, strategic planning (Figure 1). The Japanese call this process Hoshin Kanri [4].

Starting at the top of the organization, a clearly defined vision is translated into a few achievable goals, the corporate goals. In turn, each level of the organization breaks these goals into their own goals. This process continues until everyone in the organization understands the role that they as individuals, play in achieving organizational objectives. Everyone in the organization manages their jobs as though they personally own the company; and every activity performed every day is related to how it affects the goals.

While it is possible to invoke a Hoshin-type planning process only at the software management level, to the degree that the plan may not truly reflect organization direction, its usefulness may be limited. But, regardless of having been implemented at the corporate level or the software management level, this is not a trivial exercise or one that is completed in a weekend! One of the reasons that the strategic planning process has been largely unsuccessful in the U.S. in the past is that it has been a rushed episode and implicitly viewed as a onetime exercise, to be shelved at its completion. Furthermore, the full benefits of Hoshin Kanri are not realized unless the total process has been implemented.


There are always two fundamental and underlying concepts of TQM. Permeated throughout every level and every position in the organization, there is:

  1. A passionate obsession with what it takes to make the customer truly happy (ensured by going directly to customers to discover just what it is that will make them supremely joyous); and
  2. A relentless focus on continuously improving every one of the processes that contribute to product completion.

Continuous Improvement

Total Quality Management is not a program and it is not a method of manufacturing. It is a management process for achieving continuous improvements in quality schedule reductions, higher productivity and lower costs.

The old adage of if it's not broke, don't fix it is abandoned. Since improvements do not remain static once a goal is achieved, the new emphasis stresses that-even if something appears to be working-it can always be made to work better. Continuous improvement means fervently, passionately, and forever challenging and upgrading the status quo. The daily focus is always on solving the challenge of how to make each process better and better.

In Japan, the word for continuous improvement is “Kaizen” [1]. For many workers, Kaizen is taken so seriously, it is almost like a second religion. Furthermore, every day in every way Kaizen drives the business decisions. This daily emphasis is another major difference between the U.S. and Japanese quality management styles.

Continuous Improvement and Software Total Quality Management

From a software management perspective, continuous improvement initiatives concentrate on the model for S/TQM. Every step of each component of the model (and there are hundreds, maybe thousands) is considered a process that is a candidate for contiguous process improvement. Obviously, achieving perfection of all of these steps is a sizable commitment. The Japanese approach is process-oriented, effort-driven, long-term, gradual, practically painless, emphasizes the importance of an organization's people, and works best when companies can accommodate slow-growth strategies. On the other hand, U.S. industry is usually results-oriented, performance-driven, short-term, innovative with great-leaps-forward, abrupt and often volatile, emphasizes the importance of an organization's technological/procedural capabilities, and operates under fast-growth expectations. It's going to take the West a long time to reconcile these differences and to select a decent subset of TQM processes that each organization will want to integrate into its culture.

Total Employee Involvement

Total Employee Involvement (TEI) (Figure 1) starts with the assumption that every employee comes to work every day wanting to do a good job. This is certainly true in the software engineering world—where most participants are highly motivated self-starters who take a great deal of pride in producing quality products. Furthermore, most software engineering professionals become very demoralized when they are forced to produce poor quality work.

Bill Stinnett describes TEI as “an ongoing, multifaceted approach that systematically involves all employees from all levels and functions within the organization in a process of continuous improvement. Decisions are made by the people with the expertise and not necessarily the people with the power” [5]. With TEI, project bottlenecks are easier to identify and, through systematic analysis, eliminate.

TEI deploys the responsibility for quality throughout the organization—by coordinating the skills to design and build products. Using the cross-functional team-building process, all of the people within (and without) the organization who can directly affect the product's outcome collaborate from the product's inception (Figure 1).

TEI is clearly and concisely linked to the business objectives of the organization; therefore, it is not a program whose intent is merely to enhance employee motivation. It is based on cross-functional product-group focus instead of the traditional vertical functional organization. The cross-functional product groups may be called work centers; and are composed of ten or fewer people whose output contributes to a finished product. Employee empowerment is an essential characteristic of TEI and work centers.

These work centers are self-regulating groups that:

  1. Identify a problem of special interest to their work area, learn problem solving and conflict resolution techniques, and apply those methods to solving the problem.
  2. Continually identify and document all conceivable answers to the questions: How can we be more productive? What can we do to increase the corporate bottom line? How can we make this a better place to work?

Finally people need to be taught and led in this endeavor. It does not happen by legislation. Other types of necessary training include team leadership, group facilitation, active listening, non-destructive confrontation, negotiation and persuasion, project planning, brainstorming, issue clarification, and decision making. Frequently, cross-functional training also plays an important part in the process.


The Model for Software/Total Quality Management

The kernel in the model for S/TQM consists of the software engineering techniques, the project management process, the quality assurance process, the methodologies, and the automated tools.

The software engineering techniques. Effective software development and maintenance requires expertise in the software engineering techniques that must be applied in order to produce quality systems: structured analysis, design, and programming; real-time analysis, design, and programming; information engineering; and/or object-oriented analysis, design, and programming.

The project management process. Project management expertise requires superior performance in a management process that is defined to include strategic systems planning, project planning (partitioning, estimating, scheduling), and project management (resource management, status reporting, managing to plans).

The quality assurance process. The implementation of software engineering process improvement addresses verification and validation; requirements traceability; walkthroughs; inspections; phase reviews; independent testing; configuration management (versions of source code and corresponding test cases, versions of documentation, change management, release management); quality and risk assessments; the measurement process; the process improvement approach; and consulting assistance.

The methodology. It is essential to have written procedures: job aids; technical papers; roadmaps for development alternatives (large projects, small projects, rapid development, prototyping, software selection, etc.); the identification of phase deliverables, phase completion and success criteria; task steps and operational definitions; check lists; and template project plans—for the techniques, skills, and tools that support the project management process, the software engineering process, and the quality assurance process.

The automated tools. The kernel of the model for S/TQM is completed by providing automated tools for doing analysis, design, data base management, prograrnming, application generation, estimating, project scheduling, and project management.

The Importance of The Kernel

Careful consideration shows that each of the above five components are very closely intertwined. Furthermore, the S/TQM process demands that the organization focus on continuously improving and achieving excellence in every step of all of the components of this entire kernel!

Yet, many companies fall prey to the misguided belief that, if they can develop expertise in one area, it will negate the need to attend to the others. Unfortunately this is a fallacy. As a case in point, automated tools are productivity aids; they do not replace the need for expertise in the software development techniques. For example:

  • Providing analysts with the latest state-of-the-art CASE tool will not do very much good if the analyst does not know how to do analysis.
  • Providing planners with project scheduling software does very little to improve the group's ability to meet its target dates if the planners do not know what jobs to enter into the software system.


Organizations that are looking for the quick fix should not consider embracing S/TQM. People are often surprised at how hard this type of effort actually is and they are astonished at its scope. Perseverance, organizational commitment, personal commitment, vision, and courage are required as each new challenge arises. Most importantly, successful institutionalization of a quality process requires top management leadership. Ultimately accountable for the effective implementation of any quality process, they must provide visible, demonstrable, and sustained support of the program.

Failure by management to understand the fundamental nature of the change process, the implications of the change, and the scope of its impact will surely undermine the success of an otherwise well-intentioned project.

Similarly, if the organizational leaders are unable or unwilling to take the necessary actions to secure resources, or cannot fulfill their role requirements, they must be reeducated or replaced. Otherwise, the quality initiative is likely to fail to meet its stated objectives.

Thus, all levels of management must demonstrate an uncompromising desire for success. They share in the training, and become day-to-day participants in the process. Drop-in membership is not deemed appropriate.

Additionally nothing is likely to happen until the organization gets enough champions—intense advocates of the faith. These believers are the ones that start breaking down the barriers to success.

In the same vein, it is improbable that workers will be willing to change if it is not clear that the new behaviors will be valued and rewarded by their managers. For example, management must agree not to send out paradoxical messages that only serve to confuse and frustrate their employees. As an illustration, a directorate memo followed by noticeable absence from the program, or by behaviors that send contradictory and conflicting messages, will simply degrade the introduction into a meaningless exercise, leaving participants feeling frustrated and hopeless. The bottom line is that management commitment must be both real in substance and perceived as real by the employees.


1. Masaaki Imai. 1986. Kaizen: The Key To Japan's Competitive Success. McGraw-Hill.

2. W. Edwards Deming. 1986. Out of the Crisis. MIT Center for Advanced Engineering.

3. Ibid.

4. B. King. 1989. Hoshin Planning. Massachusetts: GOAL/QPC.

5. W.D. Stinnett and R.G. Hanson. 1990. Corporate Madness. Phoenix: Leadership Press.


Lois Zells is an international author, lecturer, and business consultant, specializing in software engineering consulting. She has authored the best seller, Managing Software Projects published by QED Information Sciences, Inc., and has written many articles in the major periodicals of the industry. Her most recent efforts include the new Total Quality Management seminar series “Excellence Through Performance” and “Excellence Through Total Quality Management.” She has also authored the popular, totally-integrated, three-tier learning program on software engineering project management called Successful Projects: The Common Sense Approach. Ms. Zells has written the introductory chapter for Total Quality Management for Software. a text soon to be published by Van Nostrand Reinhold. She is now also working on her second book: Applying Japanese Quality Management in U.S. Software Engineering ing.

Ms. Zells graduated Summa Cum Laude in data processing management from the University of Baltimore and did her masters studies in computer sciences at Johns Hopkins University

This material has been reproduced with the permission of the copyright owner. Unauthorized reproduction of this material is strictly prohibited. For permission to reproduce this material, please contact PMI.

MAY 1992 pm network



Related Content