Regression Testing: How It Works

What is regression testing?

Regression testing is an important phase of the ETL testing process. It is a testing technique to make sure that the existing product works fine with new functionality, bug fixes, or any change in the existing feature. Regression testing goes underway after functionality checklist of the new changes is performed.

All the test cases are re-executed to cross-check whether the existing functionality of the application is working fine, and the new changes have not produced any bugs. Take a look at the regression testing cycle:

When can we perform regression testing?

We perform regression testing whenever the production code is changed. It should be taken up:

1.  When new functionality comes in.

For example, a mobile application has a login functionality which allows users to login ONLY with email. Now providing a new feature to login with email or mobile number.

2. When there is a change request (that is, changes in existing functionality).

For example, removing the existing “Remember Password” functionality from login screen.

3. When the defect is fixed or when the code is refactored.

For example, assume that login functionality is not working on the webpage and a tester reports the bug. Once the bug is fixed by developers, the tester tests it to make sure the login functionality is working as per the expected result. Simultaneously, tester tests other functionalities like forgot password, login on mobile, login via SSO are also working fine (which are related to the login).

4. When there is a performance issue fix.

For example, loading the screen on the application that has training videos takes 7 seconds, reducing the load time to 3 seconds.

5. When there is an environment change.

Usually, five environments are used in SDLC: (1) dev instance, (2) stage instance (3) alpha instance (4) beta instance (5) production instance. Whenever the product code is deployed on the next instance, regression testing assures that the software product performs flawlessly on each environment.
For example, when we release the product, it is important to test the functionality at production instance as well.

What should be a Regression Testing Strategy?

Regression testing can be performed manually or can be automated to run daily. Prioritizing the regression test cases depending on critical and frequently used functionalities. In the case of frequent regression testing, it is more effective to automate the test cases.

Manual regression

Manual regression testing is the most diligent form of precaution testing you can apply before you move on to the next phase of your product’s journey.

(a) Strategy for performing manual regression

• Have a defined/prepared functionality checklist in place. Checklists can be multiple depending on webpage/mobile/screen-wise/feature-wise/etc.
• Gather all test cases you intend to execute, including but not limited to:

1. All integration test cases
2. All complex test cases
3. Boundary value test cases
4. A sample of successful test cases
5. A sample of failure test cases

• Prioritize the test cases. Selecting test cases based on priority will greatly reduce the regression test suite.

Automated regression

The automation team at Nitor Infotech progresses on automating test cases on web and mobile applications. Working in agile paradigms, releases, and software revisions are frequent. With this, a rapid regression testing process is required, using which product teams can receive more informative feedback and respond instantly. You might need to re-run regression tests. Automated tools allow you to do so quickly and efficiently.

(a) Strategy for performing automated regression

Automating regression means to automate the application of workflows, plans, scripts, and other activities.

• Outline what can be automated and how.
• Gather all test cases you intend to execute, including but not limited to:

1. All integration test cases
2. All complex test cases
3. Boundary value test cases
4. A sample of successful test cases
5. A sample of failure test cases

• Gauge the time that would be needed for the execution of the test cases.
• Focus on improvements that can be made to the test cases.

(b) Business benefits of automating the regression test cases

• Faster than manual operations: Automated jobs are often completed in a short period and with superior results compared to manual operations.
• Cost benefit: Use of open source will cost no penny. Also, the time required to complete all the tests can reduce the costly project delays.
• Shorter regression cycles: Manual execution is always on the higher side, however, automating the same gradually decreases the minor maintenance. And automation truly has reduced the resource efforts of executing the regression test cycle by 70-80%.
• Extended test coverage: As per our case study, the automation regression coverage was five times more than the manual regression coverage after 4 weeks of using the automation tool.

Why is regression testing important?

Functionality testing only inspects the behavior of the new features and functionalities. However, it still needs to confirm their compatibility with the existing ones. Therefore, without regression testing, it is more difficult and time-consuming to investigate the root cause and the architecture of the product.

That’s why developers and testers must not leave even the smallest stone unturned to check for what is NOT supposed to go out.

Send us an email at Nitor Infotech with your feedback about this blog or if you’d like to learn more about our quality engineering services. You might also like to read this blog that will tell you all you need to know about user acceptance testing.