Quality engineer interview questions: what managers ask and how to answer each one

Quality engineer interviews test how you decide what to test, how you report and argue for bugs, and when you choose automation. Expect questions on risk-based test planning, a bug you found, automation choices, disagreements with developers, flaky tests, and testing a small feature on the spot. Answer each with one specific project, what you did, and the result.
By I start Monday team · Published November 5, 2026 · 8 min read
A quality engineer interview is a check on judgment. Nobody can test everything, so the interviewer wants to hear how you choose, how you explain a defect so a developer can fix it quickly, and whether you can disagree with a teammate and keep the working relationship. They also want evidence from real projects, which means every answer needs a specific bug, release, or test suite behind it. Your opening answer sets the tone, so have a short, practiced answer to "tell me about yourself" ready before the first technical question arrives.
A feature ships Friday and you have one day to test it. Where do you start?
Why they ask: This question shows how you prioritize when time is limited and whether you tell stakeholders what you left untested. The BLS Occupational Outlook Handbook describes software quality assurance analysts and testers as people who identify problems with applications or programs and report defects.
A strong answer: "I ask what hurts most if it breaks. For a checkout change that is payment, tax, and order confirmation, so I cover those first, including the two nastiest failure paths. Then I read the diff and check anything that touches shared code. Before release I write down what I skipped and tell the product owner, so the remaining risk is a decision someone made."
A weak answer: "I would open the test plan and work through the cases from the top until I run out of time."
Fix it in one line: Name the riskiest area first and say out loud what you will skip.
Tell me about a bug you found that mattered.
Why they ask: A story shows whether your testing changed an outcome for users, and whether your bug reports save developers time.
A strong answer: "On a billing project I noticed that customers who changed plans on the last day of a month were charged one day too much. I reproduced it with a three-row dataset, attached it to the ticket with the expected and actual amounts, and the developer fixed it that afternoon. Then I added the month-end boundary to our regression suite so it stays covered."
A weak answer: "I find bugs all the time. Mostly small UI things like misaligned buttons. I log them and the developers fix them."
Fix it in one line: Pick one bug, state who it affected, and describe how you made it easy to fix.
What do you automate, and what stays manual?
Why they ask: Automation costs maintenance time. They want to know you can tell which checks repay that cost.
A strong answer: "I automate checks that are stable, run every release, and are cheap to execute: API contract checks, core user paths, data validation. New features and anything that changes weekly get exploratory testing by hand until the design settles. Before automating, I ask whether the test will run often enough to pay back the time to write and maintain it."
A weak answer: "Automate everything. Manual testing is outdated and automation is always faster."
Fix it in one line: Give your criteria for automating a check, then say what you keep manual and why.
A developer closes your bug as "works as designed." What do you do?
Why they ask: Quality engineers and developers can disagree about whether a behavior is a defect. This question reveals whether you use evidence and respect who owns the decision.
A strong answer: "First I check whether I am wrong. I reread the requirement and ask the developer to walk me through the intended behavior. If the design still causes a real problem for users, I bring the reproduction and the user impact to the product owner and let them decide. If they accept the behavior, I close it and note the decision on the ticket."
A weak answer: "I reopen it until they fix it. If they keep refusing, I escalate to their manager."
Fix it in one line: Show that you recheck your own assumptions, then take evidence to whoever owns the decision.
How do you handle a flaky test?
Why they ask: Flaky tests erode trust in a pipeline. They want to see a method for finding causes and a way to contain the damage meanwhile. If the role involves testing machine learning systems, you may also want to skim AI engineer interview questions, since nondeterministic output raises similar issues.
A strong answer: "I treat it as a bug in the test or the product until shown otherwise. I rerun it locally and in CI to see the failure pattern, then check timing waits, shared test data, and environment differences. If I cannot fix it the same day, I quarantine it with a ticket and an owner so it stops blocking the pipeline without disappearing. A suite people ignore is worse than a smaller suite people trust."
A weak answer: "I just rerun the pipeline until it goes green. Or I add a longer sleep to the test."
Fix it in one line: Describe how you find the cause and how you contain the test while you work on it.
How would you test a password reset feature?
Why they ask: Interviewers use small features to watch how you organize test ideas in real time. The order of your thinking matters as much as the list.
A strong answer: "I would start with the main flow: request a reset, receive the email, set a new password, log in. Then edge cases: an expired link, a link used twice, an unknown email address, a weak new password, two requests in a row. Then security: the response should not reveal whether an account exists, and the old password must stop working. Last, I would check the email on a phone and how repeated requests are limited."
A weak answer: "I would enter a valid email and check that I get an email. If it works, I mark it as passed."
Fix it in one line: Cover the main flow, then edge cases, then security, and say each group out loud.
CALCULATE IT
Defect escape rate for one release = bugs found in production ÷ (bugs found before release + bugs found in production). For example, 9 found before release and 1 found by users gives 1 ÷ 10 = 0.1, which is 10 percent. Work it out for the last release you tested, using your tracker. Use the result in your bug story and your automation answer, and say what you changed after seeing it.
| Question | What the interviewer is really asking | A strong answer in one line | A red flag |
|---|---|---|---|
| Where do you start when time is short? | Do you rank work by risk? | Test the costliest failure first and report what was skipped. | Working through the test plan top to bottom. |
| Tell me about a bug that mattered. | Can you show real user impact? | One specific bug, its impact, and how you made it easy to fix. | Vague claims about finding lots of bugs. |
| What do you automate and what stays manual? | Do you know when automation pays back? | Automate stable, repeated checks; explore new features by hand. | Automate everything, or automate nothing. |
| A developer closes your bug as works as designed. Now what? | Can you disagree with evidence and stay professional? | Recheck the requirement, then bring user impact to the decision owner. | Reopening repeatedly or escalating to a manager. |
| How do you handle a flaky test? | Do you find causes or hide symptoms? | Find the cause, quarantine with an owner, fix it soon. | Rerunning until it passes or adding sleeps. |
| How would you test a password reset? | Can you structure test ideas out loud? | Main flow, then edge cases, then security checks. | One happy-path check and a pass mark. |
The night before
- Pick three real stories: a bug that mattered, a disagreement with a developer, and a release where something slipped through. Write the defect escape rate or another number from your own work next to each.
- Reread the job posting and list every tool and language it names. Be ready to say how you have used each one, or the closest one you know.
- Say the password reset answer out loud in two minutes. Spoken answers expose gaps that silent reading hides.
- Prepare two questions about the team's release process and who owns quality when a release is at risk.
- Work through the interview preparation checklist for logistics, and keep the interview cheat sheet open for a final scan.
Try it
Search across every ATS I start Monday tracks in one pass, and build a tailored resume and cover letter for the roles that match. Start on I start Monday
Frequently asked questions
Will a quality engineer interview include coding?
The format depends on the employer. Ask the recruiter whether to expect a live coding exercise, a test-design exercise, or a code review. Read the posting for languages and tools, such as a scripting language or a test framework, and practice reading and writing small scripts in the one it names.
I have only done manual testing. Can I still apply?
Yes, if the posting's requirements match your skills. Lead with your test design, bug reports, and risk judgment, and describe any scripting you have done, even small. Be honest about gaps and say what you are learning. A clear explanation of how you decide what to test is worth including alongside any tools you list.
How many stories should I prepare?
Three to five is enough. Choose a bug that mattered, a disagreement with a developer, a test you automated or retired, and a release where something slipped through. Each one can answer several questions if you can state the situation, your action, and the result in about a minute.
More on Tech and IT
- Cybersecurity analyst interview questions: what managers ask and how to answer each one
- AI engineer interview questions: what managers ask and how to answer
- Network administrator resume examples: rewrite yours line by line, with before and after