It’s easy to find articles discussing the importance of test automation and the death of manual testing.
So easy, in fact, that it might discourage many great testers or future testers.
The truth is: manual testing is still important and will continue to be important.
While we continue to pull release cycles in and trim development time down, the temptation is to say that test automation will do the same on the testing side. This can be true in some cases, but not in all cases. To write really strong test automation, we must first develop better coding standards, such as creating unit and integration tests along with data hooks for our automation to work with. This is a time-consuming process and many companies have not gotten there yet.
“The Answer to the Great Question… Of Life”
Let’s also remember that test automation is written by humans and is only as good as the information provided. If the test cases are written too vaguely, then what are we automating? What are the correct expectations? It is very similar to the Great Question from “The Hitchhiker’s Guide to the Galaxy”. If you don’t have a clear definition of what needs to be automated or well-written instructions for testing, then how will you know you have accomplished anything? The idea of automating everything falls flat in cases where we don’t create a plan for it.
What does a 500-pound canary say walking down the street?
“Here kitty, kitty.” There is a 500 pound canary in the room and that is a set of test cases that might not be available to automate. Whether because of complexity, scope, or legacy systems, these test cases need to be run manually. We always need to evaluate the worth of test cases as well as the availability for automating. If we have high value and low availability then we have manual as our testing approach.
The more you know
Manual testing is about more than “testing”. Yes, testing is a large part of it, but your testing team is a knowledge base of how your system works, from the user’s perspective. This knowledge is powerful and is not something that you will get from test automation results. That is an awesome starting point, but not an ending point. As a tester, you are a knowledge worker. So expand it into other areas. Here is a link to an article I found that shows some excellent areas to grow in.
Asking questions is important. It is so important that we balance development by asking what they are doing. Asking about the areas of impact, the scope of changes, and asking what the product owner really means/wants. I have seen many cases where development and product thought they were on the same page, only to find out at the end they were in different books. In other cases, they missed or forgot about critical functionality. This is where testers shine.
It’s about time and value
In the end, it’s about timing and speed to market. We will need test automation, but that does not end manual testing. Testers are valuable resources that find issues as users and act as a balance to the development and product teams. Working with all groups (product, development, and the automation team) as a knowledge resource, will keep manual testers valuable for a long time. All testing (unit, integration, automated, manual) is valuable and should be treated as such. Remember, the real question is not about the replacement of manual testing, but about the value provided.
Nicholas is a software and technology enthusiast that specializes in quality. He is based in the Greater Los Angeles area, where he is currently employed as a QA Manager. His views and comments are his own and do not reflect his employer’s views.
This guest post originally appeared on LinkedIn.