The Impact of Regression Testing: Six Do's and Don't

• “Automated testing is easy using our tool!”
• “You just record your scripts and away you go! No coding is necessary."
• “Using this tool to automate your tests will reduce your costs and head count."

How many times have we sat across the table from tool vendors telling us this?! The truth is, automated regression testing is not easy, it takes the right strategy and planning to give you a fighting chance of getting it right. In our experience of working with clients who have tried to automate their testing, there are a number of common factors that caused them to fail to meet their intended objectives or give them any cost savings. 

We've come up with a list of the top six reasons why automated regression testing fails to save money and the top six things to do to ensure success and put together a useful white paper. Here's an extract:

Six Reasons why automated testing fails to save money

1. ‘Record and Playback’ Automated Regression Tests easily and frequently break and are time consuming to maintain. Changes to the application under test quickly render the tests out of date and useless or otherwise an expensive overhead.
2. Automated tests fail to handle errors correctly. It is difficult to identify the root cause of failure and often the execution run stops, requiring reexecution of all test cases once the problems are investigated. This results in a lot of unnecessary, wasted time. 
3. It takes too long to automate functional tests ‘In sprint’ and the automated test coverage lags behind the system under test. Automated tests retrace the same steps over and over, and don’t find new bugs. All this results in application code that is not fully tested being released.