In the crush of meeting deadlines, schedules, and release dates, it’s tempting to postpone anything that seems to halt forward progress. Some might feel that regular functional testing falls squarely into that category. “Let’s just get to this next stopping point” is a tempting avoidance tactic. And so it is easy to go farther and farther through the development process having skipped over the all-important testing step.
What happens if you don’t test? Ask the U.S. government.
We’ve all read about the massive failure of healthcare site, and have heard from a variety of sources that the software wasn’t adequately tested. With a huge financial outlay for development, facing the prospect of going back, finding and fixing issues that may be anywhere in the software will dramatically drive up the final the cost of that project.
This embarrassing, politically charged and professionally damaging situation would have been so easily avoided by implementing best practices into the workflow. With costs estimated as ranging between $150 million and $300 million for the portal and the data hub, and three years in the making, the excuse that “we didn’t test enough” strikes us as inexcusable.
The question is never about what and whether you are going to test, it is always about when and how much you are going to test. Like with software coding defects, it is least expensive to find a defect as early in the development process. The same applies to testing - it should be done in parallel and in-sync with development and in a continuous fashion. Leaving testing to be done in the end will always lead to not enough testing and then you find the hard-way the high cost testing and loss of reputation for not testing enough.