How much testing is enough ? When to Stop Testing ?,Economics of Software Testing

How much testing is enough ?
Testing is a matter of judging risks against the cost of extra testing efforts.
Planning testing effort before you begin and setting completion criteria ensure
that testing is properly prioritized.

Factors deciding how much to test :

1.  Levels of Risk : Technical Risk / Business Product Risk / Project Risk.
2.  Project constraints : Time / Budget.
3.  Other factors : Complexity /
4. Criticality / Visibility /Reliability.

Economics of Testing

1. Complete testing is not possible ; and neither it is possible to get entirely defect free software So it is important to find the optimum level of testing to be conducted for a give project.
2. Referring to diagram on next page , the number of defects goes on decreasing as time elapses but at the same time, cost of testing goes on increasing exponentially. The crossover of these two graphs gives an optimum testing.
3. Once can stop testing at this point to balance a risk and cost of testing.
4. The cost of building a product and the number of defects in it increase steeply with the number of defects allowed to seep into the later phases.

When to Stop Testing ?

      1. Timeline / Project Deadlines : ex : Release Date, testing deadlines.

      2. Test Cases completed with certain percentage passed.

      3. Some cases as per management call – execution of on high priority test cases with all passed.

      4. Test budget has been depleted.

      5. Coverage of code, functionality or requirements reaches a specified point.

      6. Bug rate falls below a certain level

      7. Beta or Alpha testing period ends.

      8.Testing should focus on finding defects before customer find them