As one of our recent posts attests, the automation of quality assurance should be approached cautiously and critically. It’s common sense that you should never just assume that you can automate everything.
Based on a risk and cost analysis, you will first want to have a good idea of which parts of your software are candidates for automation, and to what depth you want to automate. Even then, however, you need to have the flexibility to update your assessment as you proceed.
During one recent project, we discovered something after we had created all the support libraries and functions. It turned out that the simplicity of the application’s functionality, combined with the power of the tool we were using, allowed us to easily and quickly script a greater proportion of test cases than we had originally intended. Admittedly, this scenario does not happen very often. In this specific case, however, the client trusted us enough to allow us to continue with the automation beyond the original scope.
More often, though, it is the opposite that occurs; some previously unknown variable or difficulty presents itself and makes achieving the original scope more costly than anticipated. In those situations, rather than blindly insisting on completing the original scope, it is always better to step back and conduct a logical re-evaluation. Doing so sometimes means halting work and taking a loss.
But to our way of thinking, this is much better than compounding the problem by continuing a project that has no chance of providing a positive ROI. The lesson here is not to allow your pride to get in the way of making a rational decision. Everyone begins an automation project with lofty expectations of saving time and money down the road. However, you must be open to modifying your approach when encountering previously unforeseen variables.
For reliable and affordable software quality assurance and beta testing right here in the United States, contact the team at Beta Breakers today.