Overview of Testing - Software Testing: Testing Across the Entire Software Development Life Cycle - page 13

It is important for the tester to understand the different kinds of business losses in
order to identify the most appropriate kinds of testing that can mitigate the losses.

Different Kinds of Business Losses

• revenue or profit

• testing resources (skills, tools, and equipment)

• customers

litigation
One of the fi rst measures used by a business to put boundaries around testing is the
cost of testing. Regardless of the size of the risk to be reduced by testing, there is a
cost associated with performing the tests. Testing does not contribute directly to the
bottom-line of a business. Spending $5M on more car testing does not result in an
offsetting $5M in increased car sales; therefore, regardless of how well planned and
executed the tests are, testing reduces the total profi t of the fi nal product.
Unfortunately for project budgets, the cost of testing goes beyond the immedi-
ate testing efforts. As the authors of this textbook advocate, good testers need good
training, good tools, and good testing environments. These resources are not one-
time expenses. Most of these costs are ongoing.
The fi nal two kinds of business losses (customer and litigation) typically represent
the highest risk because the cost to the company cannot be forecast as accurately as
tester salaries, tools, and facilities. The loss of customers due to perceived issues of
poor quality or unsafe products can directly affect the bottom-line of the company,
but how many customers will be lost as a result? Part of the answer lies in how
the customer developed the negative perception, that is, by trade journal, magazine,
newspaper, or TV news commentator, to mention a few ways. To complete the loss
cycle, if enough customers develop a negative perception, then large numbers of
individual lawsuits or class action suits might result. The loss from litigation might
be beyond anyone's ability to imagine, much less to forecast. Finally, at some level
the tester must realize that, for test planning purposes, an unhappy customer can do
a company as much fi nancial damage as an injured customer.

•

1.3.3 Basili and Boehm's Rule of Exponentially
Increasing Costs to Correct New Software

Managers and executives of companies that develop computer software have per-
petuated the myth that quality can be tested into a software product at the end of the
development cycle. Quality in this context usually means software that exhibits zero
defects when used by a customer. It is an expedient myth from a business planning
perspective, but it ignores two truths: (1) Testing must be started as early as possible
in the software development process to have the greatest positive impact on the qual-
ity of the product and (2) You can not test in quality … period!
The reluctance of many managers to include testing early in the development
cycle comes from the perception of testing as a “watchdog” or “policeman” ready
to pounce on the tiniest product fl aw and cause expensive delays in making the