- What Assessments Do
- The Four Principal Functions of Assessments
- The Analytical Function of Assessments
- Assessments Function as Fulcrums of Positive Change
- Assessments Transform Organizations by the Way They Work
- Assessments Educate as They Analyze, Motivate, and Transform
- Why Gaming the Results of an Assessment Doesn't Help (Though Many Try)
- Can Assessments Really Change an Organization? A Preview of an Extended Case History to Be Found in Chapter 12
- Bottom-Line Profit and Cost Numbers: Assessments Pay
This chapter is from the book
This chapter is from the book
This chapter is from the book
1.9 Bottom-Line Profit and Cost Numbers: Assessments Pay
How much did Organization Z save by this process of reiterated assessments?
At its most basic level, process maturity creates an organizational discipline that finds and fixes defects early and then ultimately prevents them. The cost benefits of this kind of process maturity improvement may be simply calculated:
A typical software development project injects approximately 100–250 defects per thousand lines of code [Humphrey 02].
1 defect found early in development costs $100 to $200 to fix.
1 defect found in the testing stage costs $1,000 to $2,000 to fix.
1 defect found during operations costs $10,000 to $20,000 to fix. [Bush 02]
Today's development expectations for a Level 1 organization count on at least three to six operational defects per thousand lines of code. The fixing of these alone translates into unnecessary costs of between $30,000 and $120,00 per thousand lines of code.
Not finding or preventing defects before testing, though, can be almost as costly. A Level 1 organization is lucky to find 25% of defects before test, whereas a Level 3 organization typically finds 50% of defects before test, and a Level 5 organization finds between 75% and 80%. A Level 3 organization can thus easily save $30,000 per thousand lines of code, and a Level 5 organization can save $60,000 per thousand lines of code.
Many projects now produce programs consisting of 250,000 to one million lines of code, and the programs are getting more complex all the time.
Bottom line: Process maturity can translate into hundreds of millions of dollars of preventable costs.
Case in point—When Organization Z progressed over four years from a Level 2 to a Level 5, it reported, among other benefits, a gain in software productivity costs of 47% and an increase in customer satisfaction of 9% per year. These percentages can be easily translated into dollars by looking at the organization's history of finding and fixing defects.
At Level 2, Organization Z found 50% of its defects before testing, but after four years of assessments and improvements, that figure rose to 75%. As a result, Organization Z saved $4,542,000 in development costs over these four years, not including the positive ripple effect into other departments.
This was of course not all profit. Organization Z spent a total of $3,138,000 for software improvement costs over the four years, including assessments but also including training and new software improvement practices. ($1,956,000 of this in fact went for the cost of administering Fagan Inspections.)
The results are still startling. The numbers work out to a total return on investment of software improvement costs of $1,404,000 over four years—$350,100 per year—or in percentage terms, a return of 44.6% per year.
In a larger picture, as a percentage of Organization Z's total software development budget of $14,000,000 per year, the same figure represents a net increase in profit of 2.5%.
Nor were Organization Z's bottom line numbers extraordinary. Organization Z was a better-than-average company, producing high-quality software to its customers, though sometimes behind schedule and over budget. Companies with more problematic records benefit even more dramatically from software improvement efforts.
Interested? Read on.
Footnotes
For an exploration of the value of assessments, see Why Do Organizations Have Assessments? Do They Pay Off? [Dunaway 99].
It is true that the CMM and CMMI are associated with auditing procedures for the cases that arise when an outside organization wants to evaluate the organization being audited for the purpose of, for example, deciding whether to use the organization as a contractor. These audits are called Software Capability Evaluations (SCEs) and SCAMPI/SCEs respectively and are treated briefly in Chapter 2. This book, however, is principally about assessments and will mention these other procedures only in passing.