Software testing services
We are an independent testing company. We have tested other people's software since 2018, we are not the team that wrote yours, and we have no development contract to protect by staying quiet about what we find. 50+ engineers, in Cluj-Napoca, Romania.
This page is the index. The twelve service pages in the Services menu sit under it, along with four industry pages and a shelf of reference material. If you already know what you need, section 01 will get you there in one click. If you do not, read it anyway: the first thing to settle is the shape of the engagement, not the type of testing.
Two weeks, free if you are not happy with what we deliver. We invoice from week three.
Which of these are you
Pick the line that sounds like your situation. Most engagements start as one of these and turn into another, and the contract flexes when they do.
- You have no QA function and you are deciding whether to start one. Read QA outsourcing. It is the page that explains the model, what you hand over, and what it costs.
- You want the whole testing function run for you and reported back. Read managed testing services. We own the plan, the work, the reporting and the staffing, and we answer for the result.
- You have a QA lead and you need more hands under them. Read QA staffing. Your process, your lead, our engineers.
- You need a test suite built, or an abandoned one rescued. Read SDET engineers for a dedicated automation engineer, or test automation for the work itself.
- You need one specialist for a fixed piece of work. Performance, security, accessibility or API testing, for the few weeks a year when hiring full time makes no sense. Sections 04 and 05 route you.
- You are not sure whether you need one person or three, full time or part time. That is the normal position and it is a short conversation rather than a research project. Book a call and we will tell you what we would staff and why.
We take part-time engagements. Tell us the shape and we will tell you if it works.
Manual, automation, or both
We staff both, and almost every engagement is a mix. It is one of the first things people ask us, so here is the plain answer rather than a page you have to go and find.
Automation is worth it for the paths you will re-run on every release: the login, the checkout, the happy flow through the feature that pays for everything. Those are cheap to run a hundred times and expensive to check by hand. It is a poor investment for a screen that will be redesigned next month, for anything that needs a judgement about whether the result looks right, and for a product whose requirements are still moving.
So in practice the engineers test by hand where the product is new or changing and the question is whether the behaviour is correct, and they automate underneath that as it settles. Most of the automation is Playwright, Cypress, Appium, Selenium and Maestro, with JMeter for load and OWASP ZAP in the security work. The suites get maintained, which is the part that decides whether anybody still trusts them in six months.
Where the split lands changes what you pay. Section 03 says how.
What it costs
$25 to $45 an hour.
The rate moves with three things: the seniority of the engineer, how long you commit for, and the mix of manual, automation and specialist work. Longer commitments move down the range.
What the rate does not move with is whether we can find somebody cheaper. Every engineer put on your product has shipped commercial releases before. We would rather lose to a lower bid than win by staffing your product with people who have not seen your domain.
Included, with no separate licence cost:
- BugBoard turns a screenshot into a bug report and test cases in minutes, and holds the test management.
- Flows runs the automation and repairs its own tests when the interface changes underneath them.
- Auditi audits for GDPR, WCAG and FDA requirements.
- BetterFlow shows you what the hours went into, so the invoice is readable.
- The AI Security Toolkit runs a set of open-source security scanners in one pass and correlates their findings with checks we wrote ourselves, so what comes back is fixes rather than a list of CVEs.
We built all five because the off-the-shelf versions did not keep up with us. They are five QA tools, not a platform, and they come with the engineers rather than as a separate purchase.
What we test
- Functional testing. We become your users before they do. Every flow, every edge case, on real scenarios.
- Test automation. Suites built, or rescued and made trustworthy again. Playwright, Cypress, Appium, Selenium, wired into your CI.
- Mobile app testing. iOS and Android, native and hybrid, on real devices. Appium and Maestro for the automation.
- API testing. Payload modification, header manipulation, invalid response handling. The half of the product your users cannot see.
- Performance testing. Load testing with JMeter, and the question nobody asks until launch week: how does this behave on a bad connection in a lift.
- Security testing. Business logic flaws, API payload manipulation, MITM analysis. Scanners and hand testing together, and findings written as code fixes. There is also a SOC 2 readiness check that maps your security posture against the 61 Trust Service Criteria before you engage an auditor. It returns a gap report. SOC 2 itself is issued by a licensed auditor.
- Agentic QA testing. Tests generated from code diffs, run across environments, failures triaged automatically. A person signs off every release decision.
- Regression testing. Every release verified. Automation covers the happy paths so the engineers can spend their attention on what changed.
- Integration testing. Third-party APIs, microservices, payment gateways. The seams, which is where it breaks.
- Accessibility testing. WCAG 2.1 auditing, VPAT generation and screen reader testing, with Auditi doing the mechanical half.
Testing for your industry
Domain knowledge changes the test plan before it changes anything else. Four industries have their own page, written for the people who do that job:
- Financial services. Payments, core banking and trading. A screen that looks right is not evidence of anything when a figure that is wrong by four pence looks identical to one that is right.
- Insurance. Policy administration, rating and claims. Promises about money that have to hold for years and be recalculable on any date in between.
- Retail. Ecommerce, omnichannel and store systems, against a peak date nobody in the business controls.
- Embedded and connected devices. Firmware and medical devices, where the failure happens somewhere you cannot reach and nobody sends you a stack trace.
We also test healthcare and medical device software, for AdviNow Medical and CardiaSync among others, and SaaS platforms including Collaboard in Switzerland and ContextQA. Construction technology, travel and mobility too. ISO 13485:2016 is the certificate medical device manufacturers ask a testing supplier for, and we hold it.
If your industry is not on the list, that is usually fine and occasionally a reason to say no. Section 06 is the honest version.
When we are not the right answer
We would rather say this here than on the third call.
Where the contract requires staff on site, or nationals only. Common in defence work and in parts of healthcare. We are in Romania and we work remotely, so that is a straight no rather than a negotiation.
Where testing is genuinely core to how your company competes. If the testing is the product, or the thing that differentiates it, you should own that capability and not rent it.
If you want somebody to write the features as well. We test software we did not write, and that independence is the whole reason to hire us. A supplier who builds the thing and then certifies it is not giving you a second opinion.
If you want the tools without the people. The five tools come with an engagement today. They are not sold as a subscription, and we would rather say so than take a call about it.
We used to add "and not below about five developers". That is no longer true. A small team shipping with coding agents produces several times the code it did two years ago, the defects scale with the output rather than with the headcount, and nobody added review capacity to match. Team size stopped predicting how much testing you need.
Credentials, and what each one buys you
- Four ISO certificates, all current to 05.12.2027. ISO/IEC 27001:2022 covers how we handle your data, credentials and test environments. ISO 9001:2015 governs how an engagement runs. ISO 13485:2016 is the one a medical device manufacturer needs, because their own quality system has to account for who tested their software. ISO 14001:2015 is environmental. We send the certificates with their scope statement, so your procurement team reads the scope first-hand instead of taking our word for it.
- A NATO procurement agreement. The NATO NCIA Basic Ordering Agreement, NCAGE 1JGAL, reference NCIA/BOA/42405442. It is a signed ordering agreement, which is the instrument NATO issues to suppliers, so any NATO body and the government agencies of member nations can order from us on terms already signed and a project starts without a contract negotiation. It is not a certification and it is not a clearance.
- Anthropic. Anthropic approached BetterQA and we provided our security testing work.
- 4.9 on Clutch, from more than 65 reviews. The one item here written by clients rather than by an auditor or a procurement office.
- Specialists you would not hire full time. A performance engineer or a security tester can join your account for a fortnight without us recruiting for it, which is most of the argument for buying testing rather than building it.
Common questions
Do you have automation engineers as well as manual testers?
Yes, both, and most engagements use both. Section 02 says where we draw the line.
Can you work with the QA team we already have?
Most of our work is exactly this. Your lead keeps the process and we add capacity or specialists underneath it.
How fast can you start?
About two weeks to have a team working on your product. Signing the paperwork usually takes longer than staffing the work.
What if it does not work out?
You do not pay for the first two weeks if you are not happy with what we deliver, and after that you are not locked into a long contract to find out. When an engagement ends you keep the test suites, the documentation and the tooling configuration.
How do you handle our data?
NDAs, and we work inside your VPN and your tooling where you want us to. The handling is covered by our ISO 27001 management system. For a regulated product, ask for the scope statement rather than the certificate.
Reference material
Twenty-three pages under this one are reference rather than sales: definitions, levels, techniques and test work products, written for people who want the vocabulary rather than a supplier. They are grouped here so they are findable and so nobody mistakes them for a service.
What clients say
Three of the teams we test for, in their own words.
"BetterQA's ability to provide comprehensive, high-quality testing across multiple platforms stands out."
"They have a holistic testing approach and are good at finding bugs."
"They were so awesome to work with."
We are rated 4.9 on Clutch, from more than 65 reviews. See our client projects for what the work looked like.
If you are still comparing providers, our guide to choosing software quality assurance services covers pricing models and what to ask.
Talk to us
Tell us what you ship, how often, and what is going wrong. We will tell you which of the pages above you actually need, and if the answer is that you do not need us, we will tell you that instead.