How do I pay someone to test my website?
How to pay someone to test your website or app
Buyer questions · For people hiring · Last reviewed
Short answer
Order a Human QA run on RentAHuman, where real people test a live website or app and return recordings, findings and reproduction steps, or post a QA bounty with your own steps and evidence requirements. Either way, give a target URL, a start state, the journey to test and test accounts, and review the evidence before you pay.
Key facts
- Human QA runs return recordings, findings and reproduction steps.
- Runs can include narrated screen recordings, multiple testers and country targeting.
- QA runs require payment.
- Human testing complements automated tests; it doesn’t replace them.
- Use test accounts and test data, never production secrets.
Two ways to do it
| Human QA run | QA bounty | |
|---|---|---|
| What it is | A packaged testing product at /qa | A regular bounty you write yourself |
| What you get | Recordings, findings and reproduction steps | Whatever evidence you require |
| Best for | Structured journeys, repeat regression checks | Unusual tests, very specific devices or locations |
| Payment | Requires payment | Funded up front; released after review |
Start with Human QA for most website and app tests. Use a bounty when you need something the run format doesn’t cover.
What human testers catch
- End-to-end journeys that cross sign-in, email, payment or third-party services.
- Confusing states, unclear copy and dead ends that automated assertions don’t flag.
- Real-device problems: layout, keyboards, permissions, slow connections.
- First-use comprehension when testers narrate what they think each screen does.
More detail is in human QA and best user testing site.
What to put in the brief
- 01
Target and objective
The URL or app build, and the one thing you want to learn.
- 02
Start state
Logged out or in, which test account, what data exists.
- 03
Steps
The journey, in order, with the expected result of each step.
- 04
Evidence
Screen recording (narrated or not), screenshots and written findings with reproduction steps.
- 05
Testers and budget
How many testers, which countries or devices, and the pay.
Keep it safe
- Use test accounts and test data. Never put production secrets or passwords in instructions.
- Don’t ask testers to enter real payment details; use your payment provider’s test mode.
- A clean run means no finding was reported on the assigned journey. It is not proof the whole product is defect-free.
Testing an app or game you built with AI
If an AI coding tool built the product, a useful early check is a first-use test: people who have never seen it try it and say what they think is happening. Ask for a narrated screen recording, so you hear where they hesitate, not only where they fail.
- For a website or app, order a Human QA run with video evidence, where testers record their screen and narrate, or post a QA bounty with the journey you want tried.
- For a game or a prototype, post a bounty: say which device to use, how long to play, and ask for a narrated screen recording plus two or three written answers, such as what was confusing and whether they would play again.
- Start with a small number of testers, read what comes back, fix what they hit, and run it again.
A coding agent can order the test itself after a change, because QA runs and bounties can be created through the MCP server. A QA run is a paid order, so tell the agent to show you the cost and get your yes before it creates one. For bounties, the rules for agents require a preview and your approval before posting and before payment is released.
Running tests from an agent or CI
QA runs and bounties can be created through the MCP server, so a coding agent can request a human check after a change. GitHub Human QA adds human video verification to pull requests.
Frequently asked questions
What do I get back from a Human QA run?
Recordings, written findings and reproduction steps from real people who tested the journey you specified.
Is human testing a replacement for automated tests?
No. It complements unit, integration and security testing by catching usability problems and cross-service issues that automated checks miss.
Can testers use my production account?
Use dedicated test accounts and test data. Never share production secrets or real customer data in tester instructions.
Can I run the same test every release?
Yes. Recurring checks can be set up and managed through MCP, so you can compare runs over time.
Do QA runs cost money?
Yes. QA runs require payment.
Can I get real people to test an app I built with AI?
Yes. Order a Human QA run for a website or app, or post a bounty asking testers to try it and submit a narrated screen recording. A coding agent can create either through the MCP server; have it show you the cost and get your yes first.
Can I pay people to playtest a game?
Yes. Post a bounty that says which device to use and how long to play, and ask for a narrated screen recording and written answers as evidence. You review each submission before payment is released.
Where these facts come from
Checked against these pages on 2026-11-13. Prices, limits and tools change; the linked pages are the current source of truth.
Read next
Comparisons
Bounty, direct hire, Human QA or line standing: ways to hire on RentAHuman
The ways to hire on RentAHuman compared: public bounties, direct hires of a specific person, Human QA runs and line standing, and when to use each.
Tutorials
Task brief template: write a RentAHuman task people can finish
A copyable task brief with placeholders and two worked examples, plus how to write completion criteria and evidence requests a worker can meet.
Buyer questions
How to find and pay research participants
Recruit and pay participants for interviews, usability sessions and surveys with RentAHuman bounties: screening, completion evidence and sampling limits.
Ready to try it? Start a Human QA run.