8 kinds of test cases should not be automated

Tram Ho

Every day, we see thousands of new products and applications on the market. A lot of testing has to be done before releasing it to the public. Automated testing can speed up testing processes and can be seen as an alternative to manual testing. However, manual inspection still plays an important role in the Quality Assurance process. By using only automated testing, testers will lose many opportunities to improve system quality. Performing manual inspection with automatic inspection will result in much higher product quality and stability. The main reasons behind this are:

Can’t reach 100% of goals with automated testing:

It is not possible to achieve 100% of the target using automated testing because some checks cannot be automated as simple as a window or dialog background check. The automation script will have limited scope based on what is being tested or written in which is not suitable for the human eye. The human eye can identify potential problems within seconds. Things like discovery testing, interface, user related testing, for example if the user wants to see if they can do a task easily, etc. should be done manually.

The automatic test cannot solve all problems:

In some cases, automated testing cannot solve all the problems, but creates new ones. The automated test package timing becomes obsolete when other changes made to the application and test maintenance are ignored. The main focus of organizations is maintaining existing test cases and creating new test cases. Debt checking is often not their highest priority.

Several test areas need to be investigated. Some testers are “false negatives” which means testing time has expired due to some technical reason in the environment. This process will be more time consuming than saving it. The most challenging aspect of any automated system is understanding the limits of automated systems and setting realistic expectations to avoid any disappointment.

Types of checks that need to be done manually:

1. Subjective Validation:

Subject validation protects the validity of words, statements or initials. This also includes cognitive testing and testing. This is the stage where humans can be quickly discovered and provide feedback rather than automated systems that require multiple steps to write. So this kind of test is highly appreciated by hand.

2. New functions (New Functionalities):

Performing automated testing for applications under development is not a good technique. It will take a lot of time and resources to keep the automation tests up to date and maintained. Automated testing will often fail and need to be updated frequently as applications change according to new requirements.

3. Strategic Development:

Certain applications require specific attention and expertise on the topic. Manual testing is more suitable for these types of testing than automated testing.

For example, if there are business functions that require special attention, the tester should focus on those areas with greater concentration and attention. Detailed test cases to be followed cover every aspect. This usually covers the most important part of the application.

4. User Experience (User Experience): Basically, the success or negligence of an application depends on its usefulness. When it comes to user experience, nothing can compete with the human eye. It can detect any problem in seconds by looking at a picture like language, resolution, durability and format, etc.While automated testing will require a large amount of time. Therefore, we propose to attribute the user experience of any system to the human eye instead of an automated one.

5. Complex Function (Complex Functionality):

Some modules have complex functions for which automated systems cannot function effectively. So it is better to automate those operations manually as is the case with the Mobile Device Test case. Mobile test requires test by exiting and reconnecting Wi-Fi, Run apps concurrently, authorize device, receive and make calls.

6. Quality Control:

If the overall quality of the finished application needs to be checked, it is preferred manually over automatic. Test automation can only test specific output generated in test case while human can navigate whole system and some kind of workflow, happy and sad path, success and fail of some criteria quickly.

Automated systems cannot generate the original thought. They will only perform certain pre-programmed tasks. Nor are they able to generate the effective feedback a user can provide. So you should do Quality Control manually.

7. Functions with low return on investment:

The main purpose of the system tests is to save time and money. If the test cases cannot provide these two then it is useless. Some applications are simpler and contain fewer modules. Manually testing these applications can be done by dedicating fewer resources to the test processes.

For example, you are testing a simple form with small content, and the business doesn’t care if it’s there or not. Another example might be an application part there but no one is using it.

8. Check installation and setup (Installation and setup testing):

During the setup and installation checks, the system needs to check with various hardware and software such as loading CD-ROM, memory disk and Tape. Such types of systems also require manual testing.


Today, testing is the most important part of software development. No software can be kept. Certified testing sites and Quality Assurance groups are a gateway to your product stability. There are many benefits to doing manual or automated testing. Therefore, it is important to consider both approaches while designing your QA strategy. It doesn’t matter how well automated testing is developed. There will always be a place for manual testing in software applications.

Translation from: https://www.softwaretestingmaterial.com/test-cases-not-to-be-automated/

Share the news now

Source : Viblo