Automation Tester Interview Questions and Answers

Questions and model answers

9 questions you are likely to hear.

1. 1

   Explain the difference between implicit, explicit and fluent waits in Selenium.

   Why they ask: Wait strategy is central to stable Selenium tests.

   Sample answer: An implicit wait sets a global timeout for finding elements, so every findElement call retries until the element appears or time runs out. An explicit wait, with WebDriverWait and ExpectedConditions, waits for a specific condition like an element being clickable. A fluent wait is an explicit wait with configurable polling interval and ignored exceptions. I avoid mixing implicit and explicit waits because it can cause unpredictable delays. In my framework, I removed implicit waits and used explicit waits in page objects, which cut flaky failures from around 15 percent of runs to under 2 percent.

2. 2

   How do you design an automation framework from scratch?

   Why they ask: Interviewers want to see structure, maintainability and scalability.

   Sample answer: I start with the language and tools that match the team, for example Java with Selenium, TestNG and Maven. I use the Page Object Model so locators and actions live in page classes, keep tests readable and free of locators, and put utilities like waits, screenshots and config readers in separate packages. Test data comes from JSON or Excel files, or APIs for setup. Config handles environments and browsers. Reports use Allure with screenshots on failure. Tests run in parallel on Selenium Grid in Docker through a Jenkins pipeline triggered on each build.

3. 3

   How do you handle flaky tests?

   Why they ask: Flakiness destroys trust in automation, so this is a key senior skill.

   Sample answer: I treat flaky tests as defects. First I track them: which tests fail intermittently and with what errors. Common causes are timing issues, shared test data, dependencies between tests and unstable environments. I fix timing with proper explicit waits, make each test create its own data, and remove order dependencies. If a test cannot be fixed quickly, I quarantine it so it does not block the pipeline while I work on it. Retries are a last resort, not a fix. In one project, these steps brought the pass rate of our nightly suite from about 80 to 98 percent.

4. 4

   Write an XPath to find a button with text Submit inside a form with id checkout.

   Why they ask: Locator skills are tested in almost every automation interview.

   Sample answer: I would write //form[@id='checkout']//button[normalize-space()='Submit']. The double slash allows the button to be nested at any level inside the form, and normalize-space handles extra whitespace in the button text. An equivalent CSS selector cannot match by text, so I would use form#checkout button[type='submit'] if the type is reliable. In practice, I prefer asking developers to add stable data-testid attributes, then use a selector like [data-testid='checkout-submit'], which survives UI text and layout changes much better than text-based XPath.

5. 5

   Selenium or Playwright: which would you choose for a new project?

   Why they ask: This tests up-to-date tool judgement rather than loyalty.

   Sample answer: For a new web project in 2026, I would usually choose Playwright: it has auto-waiting, built-in parallelism, network interception, multiple browser contexts, tracing for debugging and a built-in API testing client, which reduces flakiness and setup. Selenium remains a strong choice when the team has an existing Java framework, needs wide language support or older browser coverage, or uses Appium for mobile in the same stack. On a recent project, migrating a small, flaky Selenium suite to Playwright with TypeScript reduced run time by about half and made failures much easier to debug with traces.

6. 6

   How do you automate API tests, and how do they fit with UI tests?

   Why they ask: Modern automation strategies rely heavily on API tests.

   Sample answer: I use Rest Assured in Java or Playwright's request context for API tests, validating status codes, response schemas, field values and error cases. API tests are faster and more stable than UI tests, so following the test pyramid, I cover most business rules at the API level and keep UI tests for critical user journeys. I also use APIs for test setup, such as creating a user and order before a UI test, which cuts UI test time significantly. In my last team, moving data setup to APIs reduced our UI suite run time from 90 to about 35 minutes.

7. 7

   How do you run automation tests in a CI/CD pipeline?

   Why they ask: Automation only creates value when it runs continuously.

   Sample answer: In Jenkins or GitHub Actions, I trigger a smoke suite on every pull request and a full regression nightly or before releases. Tests run in parallel in Docker containers with Selenium Grid or Playwright's built-in sharding. The pipeline publishes Allure reports, saves screenshots, videos or traces on failure and posts results to Slack or Teams. Environment details and credentials come from pipeline secrets. A failing smoke test blocks the merge. This setup meant developers saw test feedback within about 15 minutes of raising a pull request.

8. 8

   How do you handle test data in automation?

   Why they ask: Poor test data management is a common root cause of flaky and blocked tests.

   Sample answer: I make each test independent by creating its own data, usually via APIs or database scripts in setup, and cleaning up after if needed. Unique values like timestamps or UUIDs avoid collisions in parallel runs. For data-driven tests, I keep input combinations in JSON or CSV files and use TestNG data providers. Sensitive data is masked or synthetic. In a banking project, shared test accounts caused tests to block each other; switching to API-created accounts per test removed that whole class of failures.

9. 9

   Tell me about the biggest impact your automation work had.

   Why they ask: This shows the business value of your automation, not just scripts written.

   Sample answer: At an insurance client, regression for each monthly release took a team of four testers about five days manually. I built a Selenium and TestNG framework with page objects and API-based data setup, automated around 400 high-priority scenarios and integrated them into Jenkins with parallel execution on a Docker grid. Full regression dropped to about three hours, and the manual team shifted to exploratory testing of new features. The client moved to fortnightly releases, and production defects from regression issues dropped noticeably over the next two quarters.

Keep going.

• QA Engineer Interview Questions and Answers
• Java Developer Interview Questions and Answers
• Selenium Job Support Online for Automation Testers
• QA Testing Job Support Online for Software Testers
• Coding Interview Tips for Developers
• Coding interview help