This was originally posted on my personal blog: nicolalindgren.com
While it’s important to make sure your test automation has good coverage, I think it’s also important to know under which circumstances you should not write automated tests.
In my opinion, when it comes to test automation, more isn’t necessarily better.
More tests don’t necessarily result in better coverage.
More tests don’t necessarily increase your chance of finding regression bugs.
More tests is just that. More tests.
When deciding what not to write automated tests for, I think it’s important to ask yourself a few things.
1. What is your goal with your test automation?
Be specific and realistic about your expectations.
Come up with some specific examples of the types of regression issues that could come up and would be found from the test automation you have in place.
The best part of having a clear goal for your test automation is that it will guide you in knowing when and when not to write tests.
2. When deciding to add tests weigh the benefits of having the test against the costs
Is your test likely to find issues? Do you already have a test/ tests in place that would catch the issue(s) your possible test is designed to find?
3. Is this area of the app/website looking to change soon? If so, is it worthwhile writing tests for this area now or waiting a little bit?
A factor to consider here is how considerable the changes are looking to be and how far in the future changes for this area of the app/website will be.
Do you also write test automation? What makes you decide whether a test is or is not worth adding to your test automation suite?
If you want to read more interesting articles on testing, development, agile and automation, you can follow me on Twitter @nicolalindgren 👋
If you want to read more about when you shouldn’t write Test Automation along with tips on how to get started learning test automation, check out my upcoming eBook: Starting Your Software Testing Career – due out in January 2022.
Top comments (0)