Six question sets for the owner or manager running the interview: core, manual QA, automation, a bug report exercise, a founder screen for interviewers who cannot read test code, and a scoring rubric with red flags. Download as DOCX.
The first time I sat across from a QA candidate, I had no idea how to grade the answers. I could tell that the person in front of me knew more testing vocabulary than I did, and that was the entire problem: I was interviewing for a skill I did not have, using questions I had copied off the internet an hour earlier. The hire worked out, but by luck rather than by process.
Most small companies hire their first tester exactly this way. There is no QA lead to run the loop, no test suite to compare a candidate against, and the person conducting the interview is the founder or an engineering manager with a full plate. At FirstHR, we build for that situation, so these six question sets are written for the interviewer, not the candidate.
Each set covers a different slice of the role: a core set for every candidate, manual test design, automation and continuous integration, a bug report work sample, a founder screen for interviewers who cannot read test code, and a scoring rubric with a seniority guide and red flags.
TL;DR
Interview a QA engineer on four things: test thinking, risk prioritization, bug reporting quality, and communication. Ask what they would test first and why, then run a short bug report exercise on your product and grade it against a checklist. Federal wage data puts the occupation median at about $104,300 a year.
What You Are Actually Interviewing For
You are interviewing for four behaviors: deciding what to test, finding the problems, writing them up so a developer can fix them, and protecting the next release from repeating the same failures. The title covers a wide range, from a manual tester running checks by hand to an automation engineer building test suites that run on every code change.
Decides what gets tested
Reads the requirement and asks what it should do
Ranks by business impact and recent change
Accepts that some risk goes untested
Finds the problems
Runs planned test cases before a release
Explores freely to find what nobody planned for
Checks the paths that touch money first
Writes the bug up
Reproduces it before filing anything
Steps, environment, expected versus actual
Proposes a severity with a business reason
Protects the next release
Retests fixes and runs regression checks
Automates the checks worth repeating
Tracks what escaped and why
That range matters for the interview, because the two profiles need different questions. Federal data groups the occupation with software developers: the Bureau of Labor Statistics projects employment for software developers, quality assurance analysts, and testers to grow 15 percent from 2024 to 2034, with about 129,200 openings a year on average. Write the role down before you interview. If you have not, the QA engineer job description templates split the role into manual, automation, lead, and junior versions, and picking one of those first makes the interview far easier to run.
Which Question Set Should You Use?
Pick the set that matches the role and your own comfort with testing. Every candidate gets the core set, and you add one specialty set on top. If you cannot evaluate test code yourself, add the founder screen and the bug exercise instead of guessing at technical depth.
Core QA Questions
Every candidate
The starting set for any QA hire: test thinking, bug life cycle, risk prioritization, and what they do when a developer disputes a bug. Start here.
Manual QA and Test Design
Tests by hand
For a hands-on tester: test case design, boundary values, exploratory versus scripted, cross-browser coverage, and working from thin requirements.
Automation and SDET
Writes test code
For an engineer who automates: frameworks, the test pyramid, flaky tests, CI pipelines, API testing, and a realistic first-90-days plan.
Bug Report Exercise
Work sample
A 20-minute practical exercise with a grading checklist. The most useful stage in a QA interview, and one you can grade without testing expertise.
Founder Screen
No test code needed
Culture and situational judgment questions for an owner or manager who is not a testing expert. Graded on judgment, honesty, and communication.
Rubric, Levels, Red Flags
Score and decide
A six-area scorecard, a junior, mid, and senior guide to what each level should sound like, and a red-flag checklist to compare candidates on evidence.
Match the Set to the Hire
First tester, hands on: Core plus Manual QA. Building automated coverage from nothing: Core plus Automation and SDET. You are not a testing expert: Core plus Founder Screen plus the Bug Report Exercise, with one engineer reviewing any code sample separately. Comparing finalists: the Rubric, Levels, and Red Flags set, used with any of the above. When in doubt, start with Core and the bug exercise.
6 Free QA Question Sets to Download
Download all six as a single Word document or copy individual sets. Each follows the same structure: when to use it, the questions with notes on what a good answer sounds like, a what-to-listen-for block, and a quick score at the end. Fill in the candidate details and use them as they are.
Download All 6 QA Engineer Question Sets
Core, manual QA, automation and SDET, the bug report exercise, a founder screen, and a scoring rubric with red flags. All in one DOCX.
Set 1: Core QA Engineer Questions
The starting set for every QA candidate, manual or automation, junior or senior: test thinking, the bug life cycle, risk prioritization, and what they do when a developer disputes a bug. Start here.
Core QA Engineer Interview Questions
CORE QA ENGINEER INTERVIEW QUESTIONS
Candidate: __
Company: __
Interviewer: __
Date: _
HOW TO USE THIS SET
This is the starting set for every QA candidate: manual or automation, junior or
senior. Ask 7 to 9 of these in a 45-minute conversation. The notes in parentheses
describe what a good answer sounds like, so you can judge the response without
being a testing expert. Score at the end while the conversation is fresh, and ask
the same questions of every candidate for the same role.
CORE QUESTIONS
1. Walk me through how you would test a feature you have never seen before.
(Good answer: reads the requirement, asks what the feature is supposed to do,
covers the main path first, then edges, errors, and bad input. Asks questions
before testing rather than guessing.)
2. What is the difference between a test case and a test plan?
(Good answer: a test case is one specific check with steps and an expected
result. A plan is the strategy: what gets tested, by whom, in what order.)
3. Describe the bug life cycle you have worked in, from found to closed.
(Good answer: found, logged with steps, triaged and prioritized, fixed,
retested, closed, plus a regression check. Names the states they actually used.)
4. There is not enough time to test everything before a release. What do you test?
(Good answer: risk based. Highest business impact and most-changed code first.
Checkout and signup before cosmetic screens.)
5. Tell me about a serious bug that reached customers. What changed afterward?
(Good answer: owns their part of it and describes a process change, not just
what a developer did wrong.)
6. How do you write a bug report a developer can act on immediately?
(Good answer: title, environment, steps to reproduce, expected versus actual,
evidence, severity. Reproduces it before filing.)
7. A developer says your bug is not a bug, it is expected behavior. What now?
(Good answer: goes back to the requirement, brings evidence, escalates on
facts. Comfortable being wrong when the requirement says so.)
8. What is regression testing, and when do you run it?
9. What would you want to know about our product before you started testing it?
(Good answer: asks about users, the paths that make money, known weak spots,
and release cadence. A candidate with no questions is a warning sign.)
WHAT TO LISTEN FOR
•Curiosity: asks what a feature is supposed to do before testing it
•Risk thinking: can rank what matters instead of testing everything equally
•Reproducibility: specific steps, environments, and evidence, not "it broke"
•Ownership: talks about fixing the process, not about blaming developers
•Plain language: explains testing to a non-engineer without hiding behind jargon
QUICK SCORE (1 = no evidence, 5 = strong, specific evidence)
Test thinking and coverage [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ]
For a hands-on tester: designing test cases on the spot, boundary values, exploratory versus scripted testing, cross-browser coverage, and working from thin requirements that were never written down anywhere.
Manual QA and Test Design Questions
MANUAL QA AND TEST DESIGN QUESTIONS
Candidate: __
Company: __
Interviewer: __
WHEN TO USE THIS SET
Use this set when the hire will test by hand: exploring the product, designing
test cases, running them before each release, and filing the bugs. Most first QA
hires at a small company are manual testers, because there is no test suite yet
and someone has to find out what is actually broken. Pair with the core set.
QUESTIONS
1. Design five test cases for a login screen, out loud, right now.
For an engineer who writes test code: frameworks and why they chose them, the test pyramid, flaky tests, continuous integration, API testing, and a realistic first-90-days plan for a codebase with zero tests.
QA Automation and SDET Questions
QA AUTOMATION AND SDET INTERVIEW QUESTIONS
Candidate: __
Company: __
Interviewer: __
WHEN TO USE THIS SET
Use this set when the hire will write code that tests your product: automated UI
tests, API tests, and checks that run on every code change. This is a higher pay
band than manual QA and a different skill set. If nobody on your side can read
test code, use the founder screen set as well and bring in a technical reviewer.
QUESTIONS
1. What have you automated, in what language and framework, and why that one?
(Good answer: names a real stack such as Playwright, Cypress, Selenium, or
pytest, and gives a reason tied to the product, not to fashion.)
2. What should never be automated?
(Good answer: one-off checks, anything changing weekly, usability and visual
judgment, and tests that cost more to maintain than they save. A candidate who
says "automate everything" has not maintained a suite.)
3. Describe the test pyramid and where you would put our effort first.
(Good answer: many fast unit tests, fewer service or API tests, few slow UI
tests. Starts with API-level coverage of the paths that make money.)
4. How do you handle a flaky test that fails maybe one run in ten?
(Good answer: investigates the root cause, fixes waits and test data, and
quarantines rather than adds a retry and moves on. Treats flakiness as a bug.)
5. How do you manage test data and keep tests independent of each other?
6. How would your tests run automatically on every code change here?
(Good answer: describes a CI pipeline, what runs on a pull request versus
nightly, and how long the suite is allowed to take.)
7. How do you test an API directly rather than through the interface?
8. We have zero automated tests today. What do you build in your first 90 days?
(Good answer: a small, reliable smoke suite over the critical paths first, then
API coverage, then breadth. Not a rewrite, not a 500-test plan.)
WHAT TO LISTEN FOR
•Names a specific stack and can justify the choice
•Knows the limits of automation, not just the promise
•Treats flaky tests as defects rather than noise to retry away
•Has a realistic, staged first-90-days plan for a codebase with no tests
•Talks about maintenance cost, not only about coverage numbers
QUICK SCORE (1 = no evidence, 5 = strong, specific evidence)
A 20-minute work sample with a candidate brief and a grading checklist. This is the single most useful stage in a QA interview, and you can grade it fairly without any testing expertise of your own.
Bug Report Writing Exercise
BUG REPORT WRITING EXERCISE (WORK SAMPLE)
Candidate: __
Company: __
Interviewer: __
Time allowed: 20 to 30 minutes
WHY THIS EXERCISE
A QA engineer writes bug reports every single day, and the quality of those
reports decides how fast your developers can fix things. This is the single most
useful stage in a QA interview, and you can grade it without knowing how to test.
Give the same brief to every candidate and grade against the same checklist.
THE BRIEF (GIVE THIS TO THE CANDIDATE)
"Here is our product on a test account. Spend 20 minutes on [name a real area:
signup, checkout, search, the profile page]. Write up the two most important
problems you find as bug reports we could hand straight to a developer. If you
find nothing broken, write up the two riskiest things you would test next and
why."
Give them: a test account, the environment, and a link. Give them nothing else.
GRADING CHECKLIST (SCORE EACH REPORT)
[ ] Title states the problem in one line, not "bug on page"
[ ] Severity or priority proposed, with a business reason
[ ] The bug is real and reproducible, not a misunderstanding of the feature
[ ] They found something that genuinely matters, not only a typo
WHAT SEPARATES A STRONG SUBMISSION
•Ranks the two issues by business impact, not by which was found first
•Reproduces the bug before writing it up and says so
•Flags an assumption where the requirement was unclear
•Suggests what else nearby is probably broken for the same reason
•Written in language a non-engineer can follow
FAIR-PROCESS NOTES
•Keep the exercise short and unpaid, or pay for anything longer than 30 minutes
•Use a test environment and a test account, never live customer data
•Give every candidate for the role the identical brief and the identical time
•Grade the artifact, not the person, and keep the completed checklist on file
QUICK SCORE (1 = no evidence, 5 = strong, specific evidence)
Reproducibility of the steps [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ]
Clarity and structure of report [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ]
Severity judgment [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ]
Quality of what they found [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ]
Writing for a non-engineer [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ]
Total: ______ / 25
NOTES: __
__
Set 5: Founder Screen, Culture and Judgment
Situational and culture questions for an owner or manager who cannot evaluate test code: release pressure, disagreeing with engineering, admitting a miss, and explaining technical work in plain language.
Founder Screen: Culture and Judgment
FOUNDER SCREEN: CULTURE AND SITUATIONAL JUDGMENT
For interviewers who cannot evaluate test code themselves
Candidate: __
Company: __
Interviewer: __
WHEN TO USE THIS SET
Use this set if you are the owner or manager running the interview and you are
not a testing expert. Every question here is graded on judgment, honesty, and
communication, which you can assess perfectly well. Run this alongside the bug
report exercise, and have a technical reviewer check any test code separately.
SITUATIONAL JUDGMENT QUESTIONS
1. We ship on Friday. On Thursday you find a bug. Walk me through what you do.
(Good answer: assesses impact and who is affected, tells the decision maker
early with options, does not silently sit on it or unilaterally block release.)
2. Engineering says a release is ready. You disagree. How do you make the case?
(Good answer: frames it as risk to customers and revenue with evidence, gives
a recommendation, then accepts a business decision without sulking.)
3. You are the only QA person here. How do you decide what is not worth testing?
(Good answer: explicitly accepts that some risk goes untested and explains how
they choose. An answer that claims full coverage is not credible.)
4. Tell me about a time you missed something important. What did you learn?
(Good answer: a real, specific miss with a concrete process change. No example
at all usually means no reflection, or no ownership.)
5. How do you keep a good relationship with developers whose work you criticize
all day?
(Good answer: talks about the bug and not the person, shares findings early,
celebrates fixes. Views QA as part of the team, not a gate.)
6. What would you need from us in the first month to be effective?
(Good answer: product context, access, a decision maker for priorities, and a
clear definition of done. Realistic, not a wish list.)
7. Explain a technical problem you solved to me as if I know nothing technical.
(Good answer: plain language, no jargon, checks that you followed. This is the
single best predictor of whether you two will work well together.)
WHAT TO LISTEN FOR
•Business judgment: weighs customer and revenue impact, not just correctness
•Honesty: has a real example of a miss and what changed after it
•Escalation instinct: raises problems early with options attached
•Collaboration: describes developers as teammates, not adversaries
•Translation skill: explains technical work to you clearly and patiently
QUICK SCORE (1 = no evidence, 5 = strong, specific evidence)
Business and risk judgment [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ]
Set 6: Scoring Rubric, Seniority Guide, and Red Flags
A six-area scorecard, a guide to what junior, mid, and senior candidates should sound like, and a red-flag checklist, so you compare candidates on evidence instead of on which conversation felt best.
Scoring Rubric, Seniority Guide, and Red Flags
QA ENGINEER SCORING RUBRIC, SENIORITY GUIDE, AND RED FLAGS
Candidate: __
Company: __
Interviewer: __
Date: _
HOW TO SCORE
Fill this in within an hour of the interview, while it is fresh. Anchor every
score to something the candidate actually said or did. If more than one person
interviews, each scores independently first, then compares. Use the same rubric
for every candidate for the role. Evidence-based scoring is fairer, easier to
defend, and produces better hires than comparing gut feelings.
Rating scale:
5 = Strong, specific evidence 4 = Solid evidence 3 = Some evidence
2 = Weak or mixed evidence 1 = No evidence or red flags
SCORING AREAS
Test thinking: designs coverage, generates edge and negative cases
Score [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ]
Evidence: ______
Risk prioritization: decides what to test first and what to skip
Score [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ]
Evidence: ______
Bug reporting: reproducible steps, clear expected versus actual, evidence
Score [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ]
Evidence: ______
Technical depth: tools, automation, APIs, at the level the role requires
Score [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ]
Evidence: ______
Communication: explains testing plainly, raises problems early
Score [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ]
Evidence: ______
Ownership: takes responsibility, improves the process after a miss
Score [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ]
Evidence: ______
SENIORITY GUIDE (WHAT EACH LEVEL SHOULD SOUND LIKE)
JUNIOR
•Executes test cases someone else wrote and files clean bug reports
•Needs the priorities set for them
•Ask: core set questions 1, 2, 3, 6, 8, plus the bug report exercise
MID-LEVEL
•Designs their own test cases and owns a feature area end to end
•Prioritizes within a release without hand-holding
•Ask: full core set, plus manual or automation set for the specialty
SENIOR OR LEAD
•Sets the testing strategy, builds the process, and decides what to automate
•Can be the only QA person in the company and still make good calls
•Ask: full core set, automation set, founder screen, plus a strategy discussion
RED FLAGS (WEIGH CAREFULLY)
[ ] Cannot describe a single bug they found in specific detail
[ ] Claims full test coverage or says they never miss anything
[ ] Blames developers for every escaped defect, owns none of it
[ ] Says "automate everything" with no view on maintenance cost
[ ] Cannot explain a technical concept without jargon
[ ] No questions for you about the product or the users
[ ] Bug report from the exercise is not reproducible by a stranger
[ ] Treats QA as a gate that blocks releases rather than as a teammate
DECISION
Total score: ______ / 30
Level assessed: [ ] Junior [ ] Mid [ ] Senior or Lead
You do not need testing expertise to run a good QA interview; you need to grade the things a non-expert can see. Judgment, reproducibility, ownership, and communication are all visible to anyone paying attention, and they predict performance better than a memorized definition of black-box testing.
There is not enough time to test everything. What do you test?
Strong answer: Ranks by risk: what would hurt customers or revenue most, and what changed most in this release. Names concrete areas, such as checkout and signup before a settings screen, and says out loud what will not be covered so the decision is yours to accept.
Weak answer: A weak answer promises to test everything, or lists steps without ever saying what matters more than what.
How do you write a bug report a developer can act on immediately?
Strong answer: A clear one-line title, the environment, numbered steps a stranger could follow, expected result and actual result written separately, evidence attached, and a proposed severity with a reason. Mentions reproducing the bug before filing it.
Weak answer: A weak answer stops at describing the symptom, or treats the bug tracker as a place to dump screenshots without steps.
Tell me about a serious bug that reached customers.
Strong answer: A specific incident, honest about their own part in it, and a concrete change that followed: a new regression check, an added test, a different release process. Focus is on the system, not on assigning blame.
Weak answer: A weak answer has no example at all, or blames a developer or a deadline and describes nothing that changed afterward.
Three moves carry most of the weight. Ask the candidate to rank what they would test first on your actual product and explain the reasoning. Have them write up two real bugs after a short hands-on session, then grade the write-ups against a checklist. Ask them to explain a technical problem to you in plain words, which doubles as a preview of the working relationship. If the role involves automated tests, ask one engineer or a trusted contractor to review a code sample separately, on its own, rather than trying to fake technical depth in the room.
Manual and Automation Ask Different Things
Manual QA and automation are related but distinct hires, and interviewing them with the same questions is the most common mistake I see. Both need test design and risk thinking. Only one needs to write maintainable code, run a continuous integration pipeline, and live with the maintenance cost of a suite.
What you are testing for
Manual QA
Automation / SDET
Designs test cases and edge cases
Prioritizes by business risk
Writes and maintains test code
Sets up tests to run on every code change
Explores freely to find unplanned defects
Judges usability and visual quality
The practical rule for a small company: hire manual first unless you already have a product stable enough that repeating the same checks is wasting real hours. A first automation hire who arrives before the product has settled will spend months rewriting tests against a moving target. If you are hiring both a tester and developers, the software engineer job description templates help you keep the two roles distinct on paper.
Companies Using FirstHR Onboard 3x Faster
Join hundreds of small businesses who transformed their new hire experience.
Give every QA candidate the same short work sample: 20 minutes on a named area of your real product on a test account, then two written bug reports you could hand straight to a developer. Nothing else in the interview tells you as much, and you can grade it without knowing how to test.
Grading item
What a passing report shows
Title
States the problem in one line, not 'bug on page'
Environment
Browser, device, operating system, account, build
Steps to reproduce
Numbered, and a stranger could follow them exactly
Expected vs actual
Written as two separate statements, not implied
Evidence
Screenshot, recording, error text, or request details
Severity
Proposed, with a business reason attached
Keep it fair and keep it short. Use a test environment and a test account, never live customer data. Give every candidate for the role the identical brief and the identical time limit, pay for anything longer than 30 minutes, and grade the artifact rather than the person. Run it after a phone screen so you are not asking unqualified candidates to spend time on it.
Red Flags in a QA Interview
The most reliable warning sign is vagueness about their own work: a candidate who cannot describe one specific bug they found, in detail, has usually watched a QA process rather than run one. The second is overclaiming, because nobody tests everything and a candidate who says they never miss anything is not being straight with you about tradeoffs.
No specific bug they can describe
A real tester remembers the interesting ones. If every example is generic, they may have watched a QA process rather than run one.
Claims total coverage
Nobody tests everything. A candidate who never misses anything is either inexperienced or not being straight with you about tradeoffs.
Blames developers for everything
Escaped defects are a system problem. Listen for someone who changes the process, not someone who builds a case against a colleague.
Wants to automate everything
A candidate who has maintained a real suite knows the maintenance cost and can name what should stay manual. Enthusiasm without that view is a risk.
Two more worth watching: no questions for you about the product or its users, and a bug report from the exercise that you cannot reproduce from the steps given. None of these disqualifies a candidate on its own, but two or three together are a pattern. Balance the negatives against evidence, and treat a reference check as the place to confirm anything that felt off rather than as a formality.
Keep Every Question Job-Related
The legal caution here is the same one that applies to any interview: keep questions tied to the job and ask them consistently. The EEOC prohibits employment decisions based on protected characteristics such as race, color, religion, sex, national origin, age, or disability, which means no questions about family plans, health, or where someone is originally from. Using the same set and the same scorecard for every candidate is itself a safeguard. This is general information, not legal advice.
Scoring the Interview
Fill in the scorecard within an hour of each interview, while it is fresh, and anchor every rating to something the candidate actually said or did. Score six areas from 1 to 5 for every candidate for the role, then compare the numbers rather than your memory of which conversation felt best.
Scoring area
What a 5 looks like
Test thinking
Generates edge and negative cases without prompting
Risk prioritization
Ranks by business impact and says what will not be covered
Bug reporting
Reproducible steps, expected vs actual, evidence, severity
Technical depth
Tools and automation at the level this role requires
Communication
Explains testing plainly and raises problems early
Ownership
Owns a real miss and names the process change that followed
If more than one person interviews, each should score independently before comparing, so the loudest opinion in the debrief does not set the tone. The same questions plus the same rubric for every candidate is the core of a structured interview, and the completed scores make the interview feedback step fast and specific.
A reusable interview evaluation form keeps the format consistent across every role you hire for. That consistency matters more than it sounds once you are comparing a candidate you met on Monday against one you met three weeks earlier.
QA Engineer Pay and Classification
QA pay varies enormously by specialty, seniority, and metro market, so start from federal survey data and adjust from there. The occupation spans junior manual testers and senior automation engineers, which is why the published range is so wide.
Median $104,300 a Year (BLS OEWS, May 2025)
According to the Bureau of Labor Statistics Occupational Employment and Wage Statistics survey (May 2025), software quality assurance analysts and testers had a median annual wage of $104,300, a mean of $111,490, with the lowest 10 percent under $61,440 and the highest 10 percent over $167,010. The occupation held about 186,740 jobs (U.S. Bureau of Labor Statistics).
Budget toward the lower half for a first manual QA hire in a lower-cost market and toward the upper half for an automation engineer building your test infrastructure from nothing. Add employer payroll taxes, benefits, and equipment on top of base pay when you compare an employee against a contractor arrangement, and read the employee versus contractor rules before you decide.
Classification matters here too. The Fair Labor Standards Act computer employee exemption can cover skilled workers whose primary duty includes the design, development, documentation, analysis, creation, testing, or modification of computer systems or programs, when they are paid on a salary or fee basis at not less than $684 per week or hourly at not less than $27.63 an hour. A senior automation engineer often fits; a manual tester executing cases someone else wrote may not. Job titles do not decide it, and the exempt versus non-exempt analysis turns on actual duties. State rules can be stricter. This is general information, not legal advice.
Hiring QA at a Small Business
A large company hires testers through a QA lead who already knows what good looks like. A small business hires through a founder who does not, and that difference drives most of the avoidable mistakes. Here is how the constraint actually plays out, and what to do about each part of it. Applicant tracking is coming soon to FirstHR, so for now a shared folder and a consistent scorecard will carry the process.
You are hiring your first QA engineer and cannot test the tester
At a small company the person running the QA interview is usually the founder or an engineering manager who has never worked in test. That is uncomfortable, and it pushes people toward trivia questions they found online, which measure memorization rather than skill. The way out is to change what you grade. Judgment, reproducibility, and communication are all visible to a non-expert, and they predict on-the-job performance better than a definition of black-box testing does. Ask the candidate to rank what they would test first and explain why, ask them to write up a real bug, and ask them to explain something technical to you in plain words. Then have one engineer review any test code separately.
Your first QA hire will be alone, so judgment matters more than tooling
In a larger company a QA engineer inherits a strategy, a suite, and a lead who sets priorities. At a small company the first QA hire walks into none of that: no test plan, no automated coverage, requirements that live in someone's head, and a release schedule that does not pause for them. That makes independence and prioritization the top two hiring criteria, well ahead of familiarity with any particular framework. A candidate who can only work inside an existing process will stall in the first month, while a candidate with strong judgment will produce a usable smoke suite over your critical paths and start finding real problems almost immediately.
Running a fair process without a recruiting team
You do not need recruiting software to run a defensible interview, but you do need consistency: the same questions, the same exercise brief, the same time limit, and the same scorecard for every candidate for the role. That consistency is what makes a structured interview more predictive, and it is also what you would point to if a hiring decision were ever questioned. Keep the completed scorecards and the graded exercises on file. Once you pick someone, the work shifts to the offer and onboarding, which is where FirstHR fits: e-signature for the offer and confidentiality agreement, document management for signed paperwork, and task workflows for access and policy sign-off. Applicant tracking is coming soon to FirstHR.
One more thing worth planning: this hire will probably be alone. Ask directly how they would work without a QA lead, and listen for someone who sets their own priorities and reports risk upward rather than waiting for instructions. Mixing in a few situational interview questions is the fastest way to surface that, since the answers are about judgment rather than recall.
From Interview to Onboarding
The interview is step one. A QA hire needs access to environments, a bug tracker, source control, and staging accounts before the first morning, plus a signed offer and a confidentiality agreement covering the product internals they are about to see. Set the first real test assignment before they arrive.
Send the offer and the NDA
Confirm title, level, pay, and classification in writing, and have the new tester sign a confidentiality agreement before they see the product internals.
Set up access on day one
Test environments, the bug tracker, source control, staging accounts, and any device or browser tooling, ready before the first morning.
Give them a real first task
A defined feature area to test in week one beats a week of reading. Pair it with a named person who answers product questions.
Store the records
Keep the signed offer, NDA, I-9, W-4, and policy acknowledgments organized, plus the interview scorecards for the role.
Alongside that, the standard I-9, W-4, and the rest of the new hire paperwork still have to be collected and stored properly, and a busy first week is exactly when a form goes missing.
An onboarding template keeps the sequence consistent so nothing gets skipped. It also gives the new tester a predictable first few days instead of a laptop, a pile of forms, and a vague instruction to start looking around.
FirstHR connects the offer, the confidentiality agreement, e-signatures, the new-hire documents, and the access-and-policy checklist in one place, so a small business can onboard a QA engineer without a dedicated HR person tracking it all by hand. FirstHR is an onboarding and HR platform, not a payroll provider, so pair it with your payroll system. Applicant tracking is coming soon to FirstHR.
Key Takeaways
Interview a QA engineer on test thinking, risk prioritization, bug reporting quality, and communication, not on testing trivia.
Decide manual or automation before you interview, because the two need different question sets and different pay bands.
Run a 20-minute bug report exercise on your real product; it is the clearest signal you will get and a non-expert can grade it.
If you cannot read test code, grade judgment and communication yourself and have one engineer review a code sample separately.
Watch for candidates who cannot describe a specific bug, claim total coverage, or blame developers for every escaped defect.
Federal survey data (BLS OEWS, May 2025) puts the occupation median at $104,300 a year, with a very wide range by specialty and market.
Frequently Asked Questions
What questions should I ask a QA engineer in an interview?
Ask questions that test judgment rather than definitions. The highest-value core questions are: walk me through how you would test a feature you have never seen; there is not enough time to test everything before a release, what do you test; describe the bug life cycle you have worked in; how do you write a bug report a developer can act on immediately; and tell me about a serious bug that reached customers and what changed afterward. Add manual test design questions for a hands-on tester, or automation and CI questions for someone who writes test code. Close with a short bug report exercise on your real product, because how someone documents a defect is the clearest signal you will get. This page includes six ready-to-use sets plus a scorecard, each with notes on what a good answer sounds like.
How do I interview a QA engineer if I am not technical?
You can assess most of what matters without reading a line of test code. Judgment, reproducibility, ownership, and communication are all visible to a non-expert, and they predict on-the-job performance better than testing trivia does. Ask the candidate to rank what they would test first on your product and explain why, then have them write up two real bugs from a short hands-on session and grade the write-up against a checklist: title, environment, numbered steps, expected versus actual, evidence, and proposed severity. Ask them to explain a technical problem to you in plain language, which also tells you how the working relationship will feel. If the role involves writing automated tests, ask one engineer or a trusted contractor to review a code sample separately. The founder screen set on this page is built for exactly this situation.
What is the difference between a manual QA and a QA automation engineer?
A manual QA engineer tests the product by hand: designing test cases, exploring for defects nobody planned for, running checks before each release, and filing bugs. A QA automation engineer, sometimes titled SDET, writes code that tests the product automatically, building UI and API test suites that run on every code change in a CI pipeline. The skills overlap in test design and risk thinking but diverge sharply on tooling, and automation typically commands a higher pay band because it requires real programming ability. Most first QA hires at a small company are manual, because there is no suite to maintain yet and someone has to find out what is actually broken. Interview them with different question sets, and never assume a strong manual tester automatically writes maintainable test code, or the reverse.
Should I give a QA candidate a test or work sample?
Yes, and a bug report exercise is the best one for this role. Give every candidate the same brief: here is a test account, spend 20 minutes on a named area of the product, and write up the two most important problems you find as reports we could hand to a developer. Grade the submissions against an identical checklist covering reproducible steps, environment details, expected versus actual results, evidence, and a proposed severity with a business reason. Keep the exercise to 30 minutes or less, or pay for the candidate's time if it runs longer. Use a test environment and a test account, never live customer data. Give every candidate for the role the same brief and the same time limit, and keep the completed checklists on file so your decision rests on comparable evidence.
How much does a QA engineer cost to hire?
According to the Bureau of Labor Statistics Occupational Employment and Wage Statistics survey (May 2025), software quality assurance analysts and testers had a median annual wage of $104,300, with the lowest 10 percent under $61,440 and the highest 10 percent over $167,010. The mean annual wage was $111,490 and the occupation held about 186,740 jobs. That spread is wide because it covers junior manual testers through senior automation engineers in expensive metro markets. Budget toward the lower half for a first manual QA hire in a lower-cost market, and toward the upper half for an automation engineer who will build your test infrastructure from nothing. Adjust for your local market and add employer payroll taxes, benefits, and equipment on top of base pay. This is general information, not financial advice.
Is a QA engineer exempt or non-exempt under the FLSA?
It depends on actual duties and pay, not on the job title. The Fair Labor Standards Act computer employee exemption covers systems analysts, programmers, software engineers, and similarly skilled workers whose primary duty involves systems analysis, or the design, development, documentation, analysis, creation, testing, or modification of computer systems or programs, and who are paid on a salary or fee basis at not less than $684 per week, or hourly at not less than $27.63 an hour. A senior automation engineer designing and writing test code often fits. A manual tester who executes test cases someone else wrote may not, and misclassifying that person creates real overtime liability. Review the duties against the Department of Labor regulations, and check your state rules, which can be stricter. This is general information, not legal advice.
What are the red flags when hiring a QA engineer?
The clearest warning signs are a candidate who cannot describe a single specific bug they found, one who claims complete test coverage or says they never miss anything, and one who blames developers for every escaped defect without describing a process change. Watch also for someone who wants to automate everything with no view on maintenance cost, who cannot explain a technical idea without jargon, or who asks you no questions at all about the product and its users. In the bug report exercise, the fatal flag is a write-up a stranger could not reproduce from the steps given. None of these is automatically disqualifying on its own, but two or three together usually mean the candidate has watched a QA process rather than run one. This page includes a full red-flag checklist you can score against.