Not a replacement for manual testing: It's amazing how many times I hear people from various I.T. backgrounds talking about automating "one hundred percent" of their software testing. On the surface, it may sound like a sure fire way to increase your test coverage and find more defects, but in reality it seldom does. Many QA organizations who embark on this journey find themselves hit with a huge dose of reality when they realize that their investment in automation yielded very little improvement in quality, if any at all. Why is that? There's no simple answer, but I do have anecdotal evidence that would suggest that the displacement or reallocation of QA resources may have something to do with it. Test automation cannot account for the business context that is applied by an experienced tester, nor does automation have the ability to identify things that just don't seem to make sense, sometimes referred to as curio. Test automation can certainly help to produce quality software products, but it should not be the only weapon in the arsenal. It should be accompanied by disciplined use of robust QA processes that allow testers to apply their knowledge of the business environment to expose deficiencies in the application under test. Damico Nicome ...