Project oriented cost of quality
Introduction to quality costs
Quality costs have been in use for almost 50 years. Many firms use quality costs, usually in some custom form modified for their goals and situation. The primary form is the PAF model, using a cost breakdown of prevention-appraisal-failure. Quality costs sorted into these three categories allow for a cause and effect examination of quality assurance efforts, and firms use this to gain a better grasp of the return on investment in quality processes.
The project environment
While quality costs have been applied in a wide variety of firms, it is in the area of mass production that their greatest application has been found. Mass production enterprises lend themselves to a gradual improvement, as the slow but steady alteration of the processes can be seen in a gradual change in the business results. The well-known PAF model works well in these mass production organizations. However, project-based work possesses several characteristics that make quality costs difficult to apply.
First, each project has a unique scope. While many projects may be similar, for a project manager each successive project will demand careful understanding of the cost and timing during planning and execution. From project to project, the scope, timing, size, and resources will change, and a good project manager needs to adapt their approach to each project.
Second, each project takes place in an altered environment. Outside forces are likely to vary significantly in each new project, making comparisons of one project to another more difficult. Different tasks may need to take place in each new project, so the quality cost measures in each project can be qualitatively and quantitatively different. While projects are still likely to fall into groups (in the context of quality costs) these groups may not be identifiable measuring the usual parameters.
Third, as a project starts, the project manager will have a different set of priorities than at the close of a project. The quality costs incurred at the beginning will be primarily prevention or appraisal costs, while failure costs will gather at the end of a project. This eliminates a steady-state analysis of quality costs, an approach that is used in mass production environments. Quite simply, at different points during a project, the ideal quality cost proportions of prevention to appraisal to failure will look different.
Fourth, a mass production system will tolerate a degree of failure. Every mass produced item has a tolerable level of poor quality. While that level varies for each product, it can be relatively high. For a project-based business, every project needs to meet its goals. A single failed project can devastate a project-based business, no matter how large the enterprise. In some cases, a larger enterprise can suffer more from a failed project due to the higher degree of visibility. A project manager needs a guide to understand how a current quality cost measure will reflect upon the outcomes of their project.
These characteristics operate together in a synergistic fashion to place quality costs in a different light for project-based businesses, as compared to mass production businesses. Such significantly different demands suggest a new model may be needed, something to better align the practice of quality costs to project work.
A new model of quality costs for project-based environments
Imagine a time chart of quality costs for a project. The chart begins at week 1, with some value for prevention costs. As time passes, the prevention costs peak and then decline. Appraisal costs begin as the project moves forward, again reaching a peak and dropping toward the end of the project. Failure costs begin to accumulate as the project reaches the end, perhaps reaching a peak just as the project winds down. Such a chart might look like this:
While each project would have a different chart, there are likely to be similarities. To compare different projects, these charts would have to be adjusted for project length and cost, to eliminate scope differences. More complex differences that remain will be due to fundamental characteristics of the project and the project environment. Comparison of differences between projects can identify the best level for prevention and appraisal, relative to controlling failure costs. Projects can be classified according to their quality cost characteristics or normalized quality cost curves, and future projects placed in these identified classes can be organized to achieve better outcomes.
In summary, it is likely that there are standard quality cost models, built around multiple project cost histories. The different standard models would help guide projects, providing a project manager with indicators for the best levels of appraisal and prevention activities.
Need for real world data
To develop this concept, a case study using data from multiple projects can help. There are several basic questions to be answered, and more questions are likely as cases are examined:
How does project scale affect the PAF data curves?
Are there different patterns for classes of projects?
What are the key drivers for different levels of prevention and/or appraisal costs?
Most importantly, real world data will help define how this tool for project control can be used.
The data will need to cover projects of varying sizes, and other data regarding the projects will help understanding. Information on the experience of project personnel, relative repeat-ability of the project, and other parameters determined by input from experienced project managers will aid in the model development.
Widespread benefits
This case study will help project-based organizations apply quality costs in a manner that fits their unique requirements. Projects small and large can benefit from a greater degree of managerial control; such a data driven method as outlined here will aid in reaching a higher level of project management excellence.
Case study proposal
A study is proposed to use data from multiple real projects supplemented by information from project managers. The data would include cost data, in the areas of prevention, appraisal, and failure costs. Other overall project cost and timing data is needed to provide parameter evaluation for size of project. Personnel background, project environment and other qualitative information will be needed from project management to understand the framework of the quantitative data.
Data from multiple projects would be compared, using statistical analyses to identify any similarities of cost curves. The projects would be classified by type so that the case study results would be used in future projects. Some additional data gathering is likely to be needed, as initial data analysis may identify key questions.
Data gathering, analysis and evaluation will vary depending on access to key personnel, data, and also upon the scope of the project.
The result of the analysis would be a case study, suitable for use by the client firm in guiding project management. The case study would also be published by DC Wood Consulting, providing other project management organizations the benefit of the results.
We invite your comments: contact information is available on the DC Wood Consulting Website.