How do you decide which test cases to automate and which to keep manual?

Automate test cases that are repetitive, stable, high-risk, or data-driven, and those with a high return on investment. Keep manual tests for complex scenarios, UI/UX checks, exploratory testing, and cases that require human judgment or are infrequent and costly to automate.

Deciding which test cases to automate and which to keep manual involves considering factors such as: 1.Frequency of execution⏰:Automate repetitive tests that require frequent execution, while keeping ad-hoc or one-time tests manual. 2.Complexity and criticality🔍:Automate tests that are complex or critical to ensure consistent and thorough validation, while leaving simpler tests manual. 3.ROI and time constraints💰:Automate tests with high ROI, where automation reduces time and effort significantly, while manual testing may suffice for low-impact scenarios. 4.Stability and predictability📊 5.User interface and human judgment 👨💻

There are numerous benefits to automating software testing, improving the overall efficiency and quality of your product. To name a few: 1. Increased Efficiency and Speed 2. Enhanced Accuracy 3. Improved Coverage 4. Reduced Costs 5. Early Bug Detection

Deciding which test cases to automate involves evaluating several factors: 1. Automate tests that are executed frequently such as regression tests or smoke tests. 2. Automate tests that involve repetitive actions. 3. Automate tests for features that don’t change often and key features of the product. 4. Automate tests that save time and effort and cover critical functionalities.

Deciding which test cases to automate versus keeping manual involves considering factors like test frequency, stability of application features, and ROI. Automate repetitive, time-consuming, and critical scenarios for efficiency. Keep exploratory and ad-hoc tests manual, as they require human intuition. Prioritize tests with a high likelihood of regression issues for automation. Striking a balance based on project needs and resource constraints ensures an effective testing strategy that optimally utilizes both automated and manual testing approaches.

There are different criterias which we need to look upon to understand the testcase need to be automate as follows: - Is the task is going to repeat again - Is the test are stable enough to consider manual to automation - Test which are simple and quick to complete E2E - Test has subject to human error - Test which are low risk and complex

Decide based on factors like frequency of execution, complexity, and ROI. Automate repetitive, critical, and stable tests, while keeping exploratory and ad-hoc cases manual.

Automation is challenging due to certain reasons like: 1. In one word, everything cannot be automated like captcha code, Gmail, etc. They are impossible or extremely hard to be automated. Such test cases have to be tested manually. 2. If there are updates in the AUT, the automation code also has to be updated due to change in element locators. 3. Mastering the automation tool is another challenge. 4. Time taking for generating automation code. If client is waiting or something has to be reported urgently after testing, better to test manually unless your automation code is ready and updated (refer to point 2). 5. Lack of technology required.

Automation is just like any development/coding project it needs maintenance and time to keep it updated. Other challenges are: 1. Test data management 2. Debugging failing tests 3. Not having a good report integration making it difficult to identify reasons of failure

Some of the most common challenges of automation: - Demands specialized technical expertise of a tool to build, run, troubleshoot and maintain the automated tests - Initial challenges with the significant amount of time and money to setup scripts, frameworks and tools - Get ready to face technical challenges caused by system updates, compatibility concerns, or infrastructure limitations - Automation also has its limits, lacking the intuition, creativity, and judgment that only humans can bring to exploratory or usability testing - As the product or application evolves test scripts will need continuous maintenance to stay up-to-date.

->Focus on ROI- Prioritize automating tests that, once automated, will be run many times throughout their lifecycle, thereby providing a high return on investment. These often include high-risk areas where frequent testing is crucial. -> Tests that are run frequently, such as regression tests or those that are part of a daily or weekly test suite, are ideal candidates for automation. Automating these tests saves time and resources in the long run.

Test cases are selected for automation based on factors such as repetition, frequency, and impact. Repetitive tests or those executed frequently across builds are ideal for automation, as they save time and ensure consistency. Tests that involve human judgment, exploratory scenarios or one-off validations are better suited for manual execution. The decision is driven by balancing efficiency, resource investment, and the need for reliable test coverage.

Criteria for automation in software testing include repetitive tasks, stable functionality, and high ROI. Tasks that are time-consuming and repetitive are ideal candidates for automation as they can be executed efficiently and consistently by automated tools. Additionally, automation is suitable for testing stable functionalities where the expected behavior remains unchanged across multiple test cycles. Moreover, automation is justified when it offers a high return on investment by significantly reducing testing time, effort, and resources while improving test coverage and accuracy. Selecting automation criteria ensures optimal utilization of resources and maximizes the effectiveness of software testing processes.

I believe we should try and automate most of the tests we have identified and follow something called as Test Pyramid approach. It recommends having tests at different layers. When it comes to UI tests to be automated we should try and automate: 1. Complex scenarios 2. High frequency functionality 3. Important/Critical functionality 4. ROI for the test

Another criterion I would add is to start with the simplest tests first. The simpler the manual test, the easier it will be to automate. This approach allows for automating more complex tests over time, focusing on stability and risk, as the article suggests.

Here are some key best practices to consider for an automation project: 1. Develop a clear automation strategy that aligns with your project's goals, requirements, and limitations. 2. Select the right automation tools and frameworks that suit your product, platform, and team's expertise. 3. Ensure test scripts are modular, reusable and follows best coding practices for maintainability. 4. Organize your test scripts, test data and test results to keep everything easily traceable. 5. Regularly monitor the performance, quality, and ROI of your automation efforts, and make necessary adjustments as needed to optimize the expectations and outcome.

Automation cannot replace the knowledge and intelligence of a QA. We need to do manual testing in following areas: 1. Exploratory testing 2. User Experience testing 3. Usability testing 4. Accessibility testing

Testing matters! complete reliance on an automation tool is not a recommended approach, why? Testing provides a critical human insights that's essential for evaluating new unstable features, handling complex workflows that require logic and intuition and thinking from business context. When it comes to assessing user experience, interface elements, human emotions and perceptions are key factors. So Testing(I don't like to call "Manual Testing") always matters!!

A automação de testes emerge como uma ferramenta valiosa para otimizar a eficiência e a cobertura em projetos de software. Ao acelerar a execução de testes repetitivos, proporciona rapidez e consistência. Contudo, os testes manuais conservam sua importância vital na validação de aspectos complexos, exploratórios e sensíveis à experiência do usuário, onde a intuição humana se destaca. A combinação estratégica de ambas as abordagens cria um equilíbrio sinérgico, maximizando a eficácia da garantia de qualidade e assegurando a entrega de software robusto e adaptável.

Automation is good to rush after but, one mus'nt forget that manual testing still matters. Why? 1st thing, everything cannot be automated. So even if you try to automate a scenario fully, somethings have to be manually tested like captcha code, gmail, SMS, etc. 2nd thing, we are testing for end-users who will use the AUT manually and not with automation. That is why manual testing must be done to get real-user experience. Automating the AUT may seem to work without error but when you do it manually, unknown issues are found. Eg: I have a form to fill. A user may click TAB to go from current field to next one. To automate, I use the locator & it works, unknowing that the user clicks tab and it goes to another field actually. BOOM!

While automation is great in enhancing efficiency and repeatability of QA processes, manual testing is helpful in a few specific circumstances. Manual testing is only suitable if one needs to perform ad-hoc tests—especially in the two scenarios of tight deadlines or testing of new functionalities that have not been automated yet. Combining the automation and the manual testing approach will assist QA teams in ensuring that all the test suites are covered in order to deliver quality software according to user expectations.

For Automation, the test case preference should be to repetitive, high-risk and time-consuming situations in which uniformity and speed are beneficial. Retain exploratory, usability and one-time tests for manual process which allows a QA Engineer to think out of the box. Automated testing increases productivity, whereas manual approach helps with reasoning and flexibility.

Automation is powerful, but it has its own limitations. It's limited to handling repetitive, well-defined, and predictable tasks. To achieve comprehensive testing, it must be balanced with human efforts, especially in areas where human judgment and intuition are crucial. Depending solely on a tool can miss important aspects of the software that require human attention. So, yes automation has it's own benefits but it is not the one stop solution.