8.6 PSPSM Breakeven Point Methodology

8.2 PSP ^SM Benefit Methodology

The benefit methodology for the Personal Software Process ^SM is a procedure to measure, quantify, and analyze the amount of money returned. The Personal Software Process ^SM eliminates defects early and efficiently , when they are least expensive, resulting in no maintenance costs. Benefit is the economic value of using the Personal Software Process ^SM to create a new and improved software process. Its benefit methodology is a four-part process that consists of estimating baseline software costs, test costs, total life cycle costs of testing, and benefits. Its benefit methodology consists of a variety of defect models used in combination. Key elements include the test cost model and the total life cycle cost model. These are used to compare the total life cycle costs of software testing to those of the Personal Software Process ^SM . (Since the Personal Software Proces ^SM results in no estimated maintenance costs, the custom software cost model serves as a custom total life cycle cost model. The third substep is not necessary, but appears here to illustrate and highlight the process of estimating benefits.) Figure 34 illustrates the benefit methodology for the Personal Software Process ^SM .

Figure 34: Benefit Methodology for Personal Software Process ^SM

Estimate software cost (baseline) for PSP ^SM : The objective of this activity is to estimate typical costs of software analysis, design, and implementation. This substep includes: estimate software cost (Boehm) for PSP ^SM , estimate software cost (Walston/Felix) for PSP ^SM , estimate software cost (Bailey/Basili) for PSP ^SM , estimate software cost (Doty) for PSP ^SM , and estimate software cost (average) for PSP ^SM .

In this case, the outputs of the software cost models by Boehm, Walston/Felix, Bailey/Basili, and Doty were averaged together. This was done as sort of a Delphi method to arrive at an average software cost. The reader is free to substitute contemporary cost models, such as COCOMO II, PRICE-S , SLIM , Knowledge Plan , or any others, in order to accurately estimate costs. The formulas for these models are illustrated in Chapter 5.

Estimate test cost (baseline) for PSP ^SM : The objective of this activity is to estimate the cost of removing the maximum number of software defects using software testing. This substep includes: estimate post-test defects (baseline) for PSP ^SM and estimate baseline test cost ( projected ) for PSP ^SM .

Estimate total life cycle cost (test) for PSP ^SM : The objective of this activity is to estimate software development and maintenance costs associated with using software testing. This substep includes: estimate total software cost (test) for PSP ^SM , estimate total test cost (test) for PSP ^SM , and subtract total test from software cost for PSP ^SM .

Estimate benefits (projected) for PSP ^SM : The objective of this activity is to compare development and maintenance costs of testing and PSP ^SM . This substep includes: subtract total life cycle cost of PSP ^SM from test for PSP ^SM .