$202,730top of the range in Washington · middle $104,300 / yr
High AI exposure
Software Testers in the United States earn a median of $104,300 a year. Pay starts near $61,440. Pay reaches $202,730 at the top of the range in Washington, the best-paying state for this work among those with at least 500 people in the job.
Source: U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2025 (Software Quality Assurance Analysts and Testers, SOC 15-1253). Last checked 9 September 2026.
Entry level
$61,440
Top of the range · Washington
$202,730
Education
Bachelor's degree in CS or IT
Wages — U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2025 (Software Quality Assurance Analysts and Testers). Top of the range is the highest state-level figure among states with at least 500 people in the job. AI-impact rating is PayCrunch's editorial assessment. Updated September 2026.
🆕 New & Trending AI Tools for Software TesterReviewed September 2026
We track new AI-tool launches every week and refresh this list — here’s what’s gaining traction for Software Tester work right now.
Claude CodeNEWFree / usage-based
Terminal coding agent that reads your repo, runs tests, and ships multi-file changes.
How a Software Tester uses it: describe a feature and let it implement and test it across the codebase
OpenAI CodexNEWIncl. w/ ChatGPT plans
Agent that runs longer, deterministic multi-step coding jobs on its own.
How a Software Tester uses it: delegate a well-defined build or migration and review the finished result
WindsurfNEWFree / $15 mo
Agentic IDE that keeps context across a whole project.
How a Software Tester uses it: make large, coordinated changes without losing track of the codebase
AWS KiroNEWPreview / see site
Spec-driven coding agent that turns written specs into working code.
How a Software Tester uses it: write the spec first and let it build to that spec
NotebookLMNEWFree / $7.99 mo
Google tool that answers questions grounded only in the documents you give it — with citations.
How a Software Tester uses it: load your own manuals, policies, or PDFs and ask questions that stay accurate to the source
CursorFree / $20 mo
AI-native code editor that edits across an entire project.
How a Software Tester uses it: describe a change in plain English and let it rewrite and refactor whole files
GitHub Copilot (Agent Mode)$10–19 mo
AI pair-programmer built into VS Code and GitHub that now completes multi-step tasks.
How a Software Tester uses it: hand off a task and have it plan, edit multiple files, and open a pull request
ChatGPTFree / $20 mo
The most-used AI assistant — writing, analysis, research, and images from a plain-language chat.
How a Software Tester uses it: draft emails and documents, summarize long files, and get instant answers to on-the-job questions
ClaudeFree / $20 mo
AI assistant known for careful writing, long-document analysis, and coding.
How a Software Tester uses it: analyze big reports or spreadsheets and turn messy notes into clean, finished writing
A build that looks fine until someone tries it
A software tester is the person who tries the product before a customer has to. A screen loads. A button saves the wrong record. A report that worked on Tuesday fails after a small change on Thursday. You notice, you write it so a developer can see it again, and you say how bad it is. The job is not cynicism. It is a careful habit of asking what a real person will do, including the messy things the happy path forgot.
The day moves between the product, the bug list, and the people who build it. You may follow a written set of checks for a release, then spend the rest of the afternoon exploring a feature nobody scripted. You talk to a product owner about what "done" means. You talk to a developer about a fix that created a new problem. You keep notes a stranger could follow. Teams that treat testing as a last-day surprise ship late. Teams that bring a tester in while the feature is still taking shape waste less.
Some testers stay close to manual exploration. Some grow into automation, writing checks that run whenever the code changes. Both belong in the craft. Automation is powerful on stable flows and blind on surprises. Exploration finds the surprises and does not scale by itself. A useful tester knows which mode the risk deserves. If you only click a script, you will miss the odd case. If you refuse every tool, you will burn the week repeating the same path.
Before anyone says the release is ready
Release week is where the title earns its keep. You look at what changed, what you could not cover, and what would hurt a customer if it broke. You name the risks in plain language. A crash on login is not the same kind of problem as a typo in a help link. You help the team decide with that difference visible. The decision to ship may sit with a manager. Your part is to make the decision informed, and to resist a vague "looks fine" when you have not looked.
The quiet work is setup. Environments that do not match production, test accounts that expired, and data that hides the bug are ordinary obstacles. You flag them early. You also learn the product's history: which areas break often, which integrations are fragile, which fix last month needs another look. A tester who arrives without that memory relies on luck. A tester who keeps a short record of past failures walks into the next release with a head start.
Communication is half the role. A good report says what you did, what you expected, what happened, and how serious you think it is. A poor report says "it is broken" and stops. Developers trust people who can reproduce a problem and who will confirm the fix without drama. You will be wrong sometimes about the cause. Say so when the evidence changes. Credibility in this job is a long trail of clear notes, not a performance of catching people out.
No universal licence, and a certificate you may skip
There is no universal government licence for software testers. No state board has to grant you a card before you can test a product. A vendor certificate is optional. It can show you studied a tool or a testing syllabus a company cares about. It does not function as a licence, and it does not prove you can find a serious bug and explain it. If a posting names a certificate, read what that vendor actually grants. If the posting is silent, lead with your work.
People prepare by using software on purpose. A personal project, an internship, an open-source bug you reported with care, or a junior seat beside a senior tester all count. Learn how to write a clear report. Learn one automation tool well enough to be honest about its limits. Learn how teams track work. Skip the pile of badges that never touched a real release. In the interview, offer to test a small flow and talk through what you would try. That sample beats a certificate nobody requested.
Keep the optional credential in its place on the resume, under the products you have tested and the kinds of risk you have handled. Payments, privacy, mobile, data migration, and accessibility are different neighborhoods. Name the ones you know. If you hold a vendor certificate, say what you can do because of it, in a sentence about the work. A logo with no story attached will not move a skeptical lead. A bug report they can follow will.
What a test lead actually listens for
Hiring managers listen for how you think when time is short. Give them a release you joined late and the risk you chose to chase first. Give them a bug you almost missed and what changed your mind. Give them a disagreement with a developer that ended in a clearer fix, not a winner. They are listening for judgment and for tone. A tester who needs to be the smartest person in the room will exhaust a team by the second month.
Ask what you would own in the first season. A scripted regression seat, an embedded role on one squad, or a quality lead who coaches others are different jobs under similar titles. Ask how bugs are prioritized, who decides to ship, and whether automation is already in place or still a wish. Ask what "quality" means to them when a date slips. Those answers tell you whether you will be a partner or a late checkpoint. Bring examples that match their product. A game, a bank app, and a warehouse tool do not fail in the same way.
If you are new, say so and show practice that is real: a project, a class lab you label as a lab, a bug you filed in public. Do not inflate a class exercise into a production outage. Leads have heard that inflation. Specifics about what you tried, what broke, and what you wrote down will carry a junior resume farther than a list of every tool you once opened. Ask how the team treats a tester who blocks a release for a real risk, and how they treat one who raises noise. The answer tells you whether your judgment is welcome there, and whether the team can hear a hard finding without turning it into a contest.
Checker, owner, then the person who sets the bar
Many people start by executing checks someone else designed. That seat teaches the product and the pace. The next step is designing the checks, exploring without a script, and owning a feature's risk through a release. Later you may lead other testers, build the automation approach, or become the quality partner a director calls before a launch. Each step is a change in judgment, not only a change in title.
Pay tends to follow scope. A person who can frame risk for a whole product, mentor others, or keep an automation suite honest is doing different work from a person who runs a list. Some testers move into development, product, or site reliability. Those are exits, not required promotions. Stay if you like being the person who sees the failure early. Leave if you want to build the feature more than you want to interrogate it. Either way, keep a portfolio of reports and risks you can discuss without exposing a former employer's private data.
A lateral move can matter as much as a promotion. Testing a payments flow, a medical record, or a public-facing app teaches different risks than testing an internal form. When you change products, say what transfers: how you isolate a failure, how you write it down, how you talk about what you did not cover. Employers hire that habit. They cannot hire a memory of a codebase you will no longer touch. Keep the story portable, and keep private customer data out of it.
Wages in the quality-assurance series
Software testers share one survey series with software quality assurance analysts and testers. The figures are the May 2025 wage release. The middle of that series is $104,300. Early-career pay sits near $61,440, which is $42,860 under the middle. The upper published bound is $202,730 in Washington, and the distance from the national middle to that high end is $98,430. This is the one time to name the shared series. The rest of the pay talk uses the figures, not a second tour of the title.
State medians are a different statistic from the high end of the range. Middle wages in this note run Virginia $121,590, then Massachusetts $122,210, Colorado $123,690, Washington $128,480, and California $128,740. California's middle wage is the highest of those. The lowest median is Oklahoma at $77,700. The gap between those two medians is $51,040. From the national median up to California's median is $24,440.
Different statistics, different places
Washington holds the high end, $202,730. California holds the highest median, $128,740. Those are different statistics and different places. Washington's median of $128,480 is a third figure, the middle in the state that holds the high end.
How to talk about an offer
Set an offer beside $61,440 and $104,300 before you reach for the high end. Near the early figure, the national middle is $42,860 away. Ask what moves pay across that gap: owning risk for a feature, automation that the team actually runs, or a lead duty. Near $104,300, you match the country. If you work where the state median is higher, use that median as the local comparison. California's median of $128,740 is $24,440 above the national middle. The ordered medians for Virginia, Massachusetts, Colorado, and Washington also sit above the national middle, ending with California.
Use $202,730 only as the upper published bound in Washington. Do not treat it as California's typical wage, and do not treat it as Washington's median. Washington's middle wage is $128,480. California's middle wage is $128,740. The high end and those medians measure different facts. A recruiter who folds them into one "top pay" sentence is mixing a range top in one place with a median in another. Correct it calmly and point at the statistic you mean.
Oklahoma's median of $77,700 sits between the early national figure and the national middle. The gap from that median to California's median is $51,040. Use the gap if you are comparing places, and pair it with the kind of product, because a regulated app and a small internal tool are different weeks even at the same median. The distance from the national middle to Washington's high end is $98,430. That climb belongs in a talk about senior scope in a market that already pays at the top of the published range, not in a first offer.
A vendor certificate stays optional unless the employer named it. If they want one, ask whether pay moves when you earn it, and do not assume the move reaches $202,730. Write down the offer, the national median of $104,300, and the state median if you have it from the list above. Add Washington's high end only to stop someone from using it as a typical wage. Then decide on the work: what you will own, who ships, and whether your judgment is welcome before the last day of the release.
The top of Software Tester pay — and how to get there with AI
$202,730what Software Tester pay reaches in Washington
Highest state-level top-of-range annual wage for Software Quality Assurance Analysts and Testers, among states with at least 500 people in the job. U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2025.
And the role it leads to — Software Developers — reaches $272,670 in California.
$61,440entry$104,300middle$202,730top end
Anyone can be trained to execute a written test case, so what puts a software tester at the top of this range is owning a corner nobody can reproduce in a lab, usually pre-release evaluation at real customer sites where the data, the hardware and the habits do not resemble the staging build.
Visiting beta testing sites to evaluate software performance sits in this occupation's task list and is skipped almost everywhere because it looks awkward and unbillable, which is precisely what makes it scarce. Scripted checking has become largely automatic, with assistants generating cases from a specification and keeping automated test scripts current, so value moved to findings no script would ever produce: the customer whose dataset is ten times larger, the connection that drops mid-transaction, the workflow nobody wrote down. A tester running that programme holds real influence over whether a release goes out, and can give developers feedback on usability and functionality that comes from watching people rather than imagining them.
Your playbook, by where you are now
Just startingGet good at the failure trail
Trace a fault to its source yourself, through configuration files, logs and code, before you open a ticket about it.
Test system modifications against data shaped like a customer's rather than the tidy fixture set everyone uses.
Follow every bug you raise until it closes, and record how long resolution genuinely took.
Ask Gemini to propose edge cases from a requirements document, then throw out the ones that could not occur in your product.
What proves it: A defect history where your reports name causes rather than symptoms.
Realistic span: the first two years
A few years inTake the beta programme
Offer to run pre-release evaluation with two or three real customers, and go to their sites rather than screen-sharing.
Record how their data volume, hardware and daily workflow differ from your environment, and write those differences into the test strategy.
Plan test schedules around the beta window so findings arrive while there is still time to act on them.
Check the product documentation that beta customers actually read, and close the gaps you find there yourself.
Keep field notes in Evernote or Airtable so patterns across customers become visible over several releases.
What proves it: A beta programme you ran, with defects found at customer sites that staging never surfaced.
Realistic span: years three through six
ExperiencedBe the release decision
Sit in product design reviews and raise field failure modes while the design is still cheap to change.
Set the entry and exit criteria for the beta phase, then hold them when the delivery date starts pushing.
Give developers written recommendations on usability and functionality grounded in what you watched customers do.
Teach newer testers to run site visits properly, since a corner is only valuable while the company keeps it.
Note where this work pays most, Washington leading for software testers, and that the developer track sits directly above.
What proves it: Named authority over release readiness, backed by evidence gathered in the field.
Realistic span: seven years and beyond
The next 90 days
Get yourself in front of one real customer within ninety days. Ask which account has the largest dataset, the slowest connection or the most unusual configuration, arrange to sit with them while they use the current build, and watch without correcting them. Note every point where the product behaved differently than it does in your environment, every workaround they have invented, and every step that took longer than anyone assumed. Then write it up as a test strategy change rather than a list of complaints. That single visit typically produces findings the whole test suite has been missing, and it is the beginning of a software tester owning a corner that cannot be handed to a script.
Wage figures: BLS OEWS, May 2025. The playbook is PayCrunch editorial guidance, not a guarantee of pay or placement.
Every figure is the national median from the U.S. Bureau of Labor Statistics (OEWS) shown on that role’s own page.
Never used AI before? Start here (2 minutes).
Start by pairing an AI coding assistant with a real automation framework. Open GitHub Copilot (or Cursor or Claude) alongside Playwright and have it scaffold your first end-to-end test from a described user flow. You'll write working automation faster than you thought possible, and every test you build is a step from manual tester toward SDET pay.
For learning, use Claude or ChatGPT to explain a framework, debug an error, or review your test code - pasting only non-confidential snippets. Keep production data, secrets, and proprietary source out of consumer AI; use your team's approved private tooling for real code. AI is the pair-programmer that writes the first draft; you are the engineer who makes it correct, stable, and trustworthy.
The one rule, forever: Never paste production data, customer PII, credentials, API keys, or proprietary source code into a consumer AI tool - use synthetic data and your organization's approved, private AI tooling for code. AI-generated test code is often flaky or falsely green (hardcoded waits, assertions that never fail): review and run it before trusting it. Run performance and security tests only against authorized environments. You own the quality gate.
The plays — exact steps, exact prompts
Do these in order. Each one is copy-paste ready. You do not need to know anything about AI going in.
1
Write automation fast with AI-assisted coding
Why this pays: Coded automation is what separates a six-figure SDET from a manual tester. An AI pair-programmer lets you build and maintain Playwright or Selenium suites at speed even while your coding is still leveling up - the skill jump that moves pay toward $202,730.
GitHub CopilotPlaywrightCursor
1
Use GitHub Copilot or Cursor alongside Playwright to scaffold page objects and end-to-end tests from a described flow, then refine the code yourself.
2
Prompt for a solid first draft.
Copy-paste this prompt
Write a Playwright test in TypeScript using the Page Object Model for this flow: [log in, search for a product, add to cart, check out with a test card, assert the order confirmation number appears]. Use stable locators (prefer role or text over brittle CSS), explicit waits (no hardcoded sleeps), and clear assertions at each step. Use [environment-variable placeholders] for credentials.
Never hardcode real credentials or run against production. Review the generated code - AI loves hardcoded waits and brittle selectors that cause flakiness.
3
Refactor AI output into a maintainable framework (fixtures, config, reporting) so the suite scales.
What you'll haveCoded automation suites built at speed - the SDET-grade skill that lifts your pay out of manual testing.
2
Generate unit and integration test coverage with AI
Why this pays: Testers who can raise code coverage at the unit level operate at developer-adjacent value - and pay. AI unit-test generators produce coverage fast, making you the person who closes the gaps developers leave.
Qodo (CodiumAI)Diffblue CoverGitHub Copilot
1
Use Qodo (formerly CodiumAI) or Diffblue Cover (for Java) to auto-generate unit tests that capture current behavior and expose edge cases in a codebase.
2
Direct and review the generated tests.
Copy-paste this prompt
Analyze this function and generate a thorough set of unit tests: cover the happy path, boundary values, null and empty inputs, and error conditions. Point out any behavior that looks like a bug rather than intended logic. Use [test framework]. Here is the non-proprietary code snippet: [paste].
Only paste non-confidential code, or use your org's private AI. AI-generated unit tests can lock in existing bugs as 'expected' - review each assertion against intended behavior.
3
Wire the generated tests into CI and track coverage so gaps stay closed.
What you'll haveFast, meaningful unit and integration coverage - the developer-adjacent skill that commands top testing pay.
3
Kill flaky maintenance with self-maintaining AI tests
Why this pays: Flaky, high-maintenance suites are where automation ROI and tester credibility die. AI self-healing and record-based tools slash maintenance, letting you own more coverage with less upkeep - the efficiency that makes you the automation lead.
MeticuloustestRigorMabl
1
Use Meticulous to auto-generate and maintain end-to-end tests from real usage, or testRigor or Mabl for self-healing tests that survive UI changes.
2
Triage flaky failures with AI instead of by hand.
Copy-paste this prompt
Here are the last 20 CI runs of an automated test suite with pass/fail status and error messages: [paste de-identified summary]. Identify which failures are likely flaky (intermittent, timing or environment) versus real regressions, group them by root cause, and recommend fixes to stabilize the suite. Reasoning only.
Confirm each 'flaky' verdict - dismissing a real intermittent bug as flakiness is how defects ship. You decide what's a real failure.
3
Set a flake-rate budget and use AI triage to keep the suite trustworthy in CI.
What you'll haveStable, low-maintenance suites that engineers trust - the reliability that earns the automation-lead role and pay.
4
Specialize in API and performance testing with AI
Why this pays: API and performance testing are higher-paid specialties because they demand skills manual testers lack. AI that scripts API tests and load scenarios lets you claim that specialty faster - a clear step up the pay band.
Postman (Postbot)k6JMeter
1
Use Postman's Postbot to generate API tests and assertions, and k6 (or JMeter) for load and performance tests written as code.
2
Draft a load-test script with AI.
Copy-paste this prompt
Write a k6 load-test script in JavaScript for this API scenario: ramp to [500] virtual users over [2 minutes], sustain for [5 minutes], hit endpoints [list], with thresholds that fail the test if p95 latency exceeds [500ms] or error rate exceeds [1%]. Include realistic think-time and use [environment-variable placeholders] for tokens.
Run load tests only against authorized, non-production (or sanctioned) environments - never blindly against live systems. Validate the thresholds match real SLAs.
3
Report performance findings (latency percentiles, breaking points) in a way that drives engineering decisions.
What you'll haveA genuine API and performance specialty - the higher-value skill set that moves pay toward the top of the range.
5
Own the CI/CD quality gate
Why this pays: The engineer who owns quality in the pipeline is indispensable - and paid accordingly. AI helps you build and maintain the CI quality gate and triage its failures, the pipeline ownership that leads to senior SDET and lead roles.
GitHub ActionsClaudePlaywright
1
Integrate your automated suites into GitHub Actions (or Jenkins or GitLab CI) so tests gate every merge, with parallelization and reporting.
2
Build the pipeline config and gates with AI help.
Copy-paste this prompt
Write a GitHub Actions workflow that runs a Playwright test suite on every pull request: install dependencies, run tests sharded across [4] parallel jobs, upload the HTML report and traces as artifacts, and fail the PR if any test fails. Add a nightly full-regression run. Explain each step. Use repository secrets - no secrets in the file.
Never commit secrets - use the CI secret store. Review the workflow; a misconfigured gate that passes on failure is worse than no gate.
3
Define entry and exit criteria and a flake policy so the gate is trusted, not bypassed.
What you'll haveOwnership of the pipeline quality gate - the indispensable role that leads to senior SDET and lead pay.
6
Add a security-testing edge with AI
Why this pays: Security testing is a scarce, well-paid specialty most testers never touch. AI that helps you run and interpret security scans lets you add that edge - a premium niche that distinguishes your pay from the pack.
OWASP ZAPBurp SuiteClaude
1
Learn the basics of OWASP ZAP or Burp Suite for authorized security testing, and use AI to explain findings and prioritize them.
2
Interpret scan results with AI.
Copy-paste this prompt
Explain these de-identified web-app security scan findings in plain terms: [paste finding types, e.g., reflected XSS, missing security headers, insecure cookie flags]. For each, describe the risk, how to safely reproduce or confirm it, the likely fix, and a severity ranking. Educational security guidance only.
Only test systems you're authorized to test. AI helps you understand findings, but confirm each one - scanners produce false positives, and acting on a wrong result wastes engineering trust.
3
Report confirmed vulnerabilities with clear reproduction and remediation guidance to developers.
What you'll haveA security-testing capability few testers have - the scarce specialty that pushes your pay above the pack.
Your 12-month sequence to the top of the range
How the plays above stack into a path from median pay toward the $202,730 tier.
Month 1
Pair GitHub Copilot with Playwright and build your first coded end-to-end tests; refactor them yourself.
Months 2-3
Add AI unit and integration test generation (Qodo, Diffblue) and wire the tests into CI.
Months 3-6
Cut maintenance with self-healing and record-based AI (Meticulous, testRigor) and master AI-assisted flake triage.
Months 6-9
Specialize in API and performance testing (Postbot, k6) and own the CI/CD quality gate.
Months 9-12
Add a security-testing edge (ZAP, Burp Suite) - the scarce specialty that tops out the pay band.
Gear for this job
As an Amazon Associate, PayCrunch earns from qualifying purchases. Links to books and tools are for the job on this page; we only recommend what we’d use in the work.
Same live Jossey-Bass 3rd already on high-school-teacher / middle-school-teacher / math-teacher / test-prep-instructor / substitute-teacher / science-teacher / music-teacher / drama-teacher / adult-education-teacher / corporate-trainer / instructional-designer / stem-teacher / pe-teacher / speech-teacher / curriculum-developer / education-consultant / college-professor / assistant-principal / financial-literacy-educator / school-principal / vice-principal / homeschool-consultant / school-administrator / edtech-specialist / education-administrator / distance-learning-coordinator / capitol-police-officer / tsa-agent / piano-tuner / birth-doula / dive-master / translator / voice-over-director / wordpress-developer / balloon-artist / circus-performer / nutritionist / academic-advisor / dermatologist / train-conductor / calligrapher / choreographer / motivational-speaker / marble-polisher / compensation-analyst / fleet-manager / music-producer / iot-engineer / it-director / media-buyer / hospital-administrator / ship-broker / dean / clinical-pharmacist / dental-surgeon / casino-dealer / coroner / digital-transformation-consultant / sheriff / financial-crime-investigator / emergency-medical-dispatcher / railroad-engineer / correctional-officer / healthcare-consultant / compliance-officer / organ-transplant-coordinator / dispatcher / county-clerk / parole-officer / customs-officer / census-taker / patent-attorney / quantum-computing-researcher / regulatory-affairs-specialist / game-designer / dental-therapist / recruiter / web-content-manager / magistrate / bailiff / financial-aid-counselor / immunologist / nuclear-physicist / compliance-analyst / escrow-officer / geriatrician / oral-surgeon / orthodontist / pediatrician / psychiatrist / financial-examiner / orthopedic-surgeon / pain-management-specialist / pathologist / zoological-veterinarian / physician / clinical-research-coordinator / pulmonologist / rheumatologist / child-life-specialist / forensic-pathologist / marine-surveyor / color-consultant / sound-engineer / crisis-counselor (ASIN 1119712610). This leftover page is BLS Software Quality Assurance Analysts and Testers (SOC 15-1253); title is Own the Beta Site Corner; H1 is The software tester who tests where customers are; just-starting track is Get good at the failure trail; few-years track is Take the beta programme; experienced track is Be the release decision; the playbook says to teach newer testers to run site visits properly, since a corner is only valuable while the company keeps it; start-here is Start by pairing an AI coding assistant with a real automation framework; one-rule is Never paste production data, customer PII, credentials, API keys, or proprietary source code into a consumer AI tool. This instructional-technique guide directly supports that explicit Teach-newer method instruction with a teaching rationale. Classroom technique for leftover instructional work — not leftover Wong as the lead (that is convention-planner / wedding-planner / magazine-editor / public-relations-manager / venture-capital-analyst) and not leftover Praxis as a dump. Confirm 1119712610. Live page HTTP 200, no PC_GEAR / amazon.com/dp / tag=paycrunch-20 at 2026-09-18 8:28:08 AM PT. Source page: drama-teacher.
Next steps for a Software Tester
Some links below are affiliate or partner links. PayCrunch may earn a commission if you enroll or subscribe through them, at no extra cost to you. Wage figures on this page still come from the Bureau of Labor Statistics, not from these programs.
Software Tester work is specific enough that a stamped 'check out these courses' block would be noise. BLS files this work as Software Quality Assurance Analysts and Testers (SOC 15-1253). O*NET Job Zone 4 is typical: a bachelor's degree, so the honest next credential is a professional certificate or bachelor's-level coursework — not a random catalog dump.
The occupation's listed knowledge areas include Engineering and Technology and Design; the links search those subjects, not a generic 'career courses' list.
Software Testers in this dataset list AJAX among the tools in use, so a program that names that stack is a better fit than a survey course.
Coursera search for engineering and technology — a professional certificate or bachelor's-level coursework that lines up with computing, not a generic professional-development aisle.
FlexJobs screens remote, hybrid, freelance, and flexible listings so you are not wading through unverified ads. This is a job-board search for Software Tester work, not a claim that they list a counted SOC 15-1253 inventory.
Write a Software Tester resume, or one aimed at Software Developers, instead of a blank template. Resume Now is a resume builder; we are not claiming a counted template set for this SOC.
A Software Tester resume that names the actual tasks on this page, or the step-up title Software Developers, beats a blank template when you apply.
What Software Testers earn by state
These are the Bureau of Labor Statistics’ own figures for Software Quality Assurance Analysts and Testers, state by state — not a cost-of-living adjustment applied to the national number. Only states employing at least 500 people in the occupation are shown, because a state median drawn from a handful of workers is noise rather than a signal.
California
$128,740
highest of them · +23% vs the national median
Oklahoma
$77,700
lowest of the 35 states and D.C. that qualify · -26% vs the national median
The same job pays $51,040 more a year at the median in California than in Oklahoma — 66% higher. That gap is what the Bureau measured, before any question of what it costs to live in either place. The top-of-range figure quoted at the head of this page, $202,730, is a different statistic in a different place: it is the 90th-percentile wage in Washington. The state that pays the typical worker most and the state where the best-paid go highest are not always the same one.
Source: U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2025, SOC 15-1253. 35 states and D.C. clear the 500-employee reporting floor for this occupation; those below it are left out rather than shown with a wide error band.
Free data. Use any of it.
PayCrunch publishes verified, BLS-sourced salary + AI-playbook data on 1,000+ professions — free, no signup.
Manual, repetitive testing is being automated hard - AI now writes tests, maintains them, and generates coverage, so pure manual testing is one of the most exposed roles in tech. But automation engineering is going the other way: someone has to architect frameworks, judge what to test, own the CI gate, and do the performance and security work AI can't own. Testers who use AI to become SDETs move toward developer pay; those who stay manual-only are the most at risk on this list.
Can I trust AI-generated test code?
Only after you review and run it. AI writes tests fast but loves hardcoded waits, brittle selectors, and assertions that never actually fail - a suite that's green for the wrong reasons. And AI-generated unit tests can enshrine existing bugs as 'expected behavior.' Treat AI code as a first draft: read every assertion, run it, and confirm it fails when the behavior breaks.
How does AI move a software tester's pay toward the top?
By accelerating the jump from manual testing to automation engineering. AI helps you write coded automation, generate unit and API coverage, cut flaky maintenance, own the CI gate, and add performance or security specialties - the exact skills that define an SDET. SDET and automation-lead roles pay at the top of the testing band ($202,730) and edge toward developer comp.
Is it safe to use AI coding tools for testing?
For non-confidential code, learning, and scaffolding - yes, and they're a big accelerator. But never paste production data, secrets, API keys, or proprietary source into consumer AI; use your organization's approved private tooling for real code. Run performance and security tests only against authorized environments. The quality gate and the data handling are your responsibility.
Which AI skill should I build first?
AI-assisted coded automation - GitHub Copilot or Cursor with Playwright - because coded automation is the single skill that separates a manual tester's pay from an SDET's, and AI lets you build it while your coding levels up. Start there this month, then add AI unit-test generation and CI integration. That sequence moves you from tester toward automation engineer fastest.
Methodology & sources
Salary (median, 10th, top of the range) — U.S. Bureau of Labor Statistics, OEWS.
By state — the Bureau of Labor Statistics’ own state medians, limited to states employing at least 500 people in the occupation. No cost-of-living arithmetic is applied to a wage anywhere on this page.
The plays — PayCrunch's own step-by-step guidance using publicly available AI tools. Tool names/URLs are real and current as of August 2026; prompts written to work as-is. Verify any professional output before relying on it.