The right choice when
- Releases keep breaking features that used to work.
- Developers test their own work and nobody else does.
- A launch is coming and you want an outside check first.
Software quality assurance
Find the bugs before your users do. Manual and automated testing.
We test web and mobile software against what it is meant to do: by hand where judgement matters, and with automated checks where the same test must run on every release.
- A change is ready
A feature or fix, with what it is meant to do written down.
- It is tested
Automated checks run on every build. A person tests what a script cannot judge.
- It ships, or goes back
With a clear report: what passed, what failed and how to reproduce it.
Testing shows the bugs it finds. It cannot prove there are none.
01 / Is this the one
Is separate QA what you need?
Every team tests. The question is whether you need people whose only job is to break things.
Look elsewhere when
- The product is a prototype that changes weekly. A heavy test suite would be rewritten each sprint.
- You want someone to sign off that the software has no defects. No tester can honestly do that.
- The real problem is unclear requirements. Testing cannot fix what nobody has defined.
02 / What the work covers
From a test plan to checks that run themselves.
We start with the journeys that would hurt most if they broke, then widen.
- 01
Test plan
Which journeys matter most, on which devices and browsers, and what counts as a pass.
- 02
Manual testing
Functional, usability and exploratory testing by a person, on real devices where it matters.
- 03
Automated tests
Repeatable checks for web and mobile, with tools such as Selenium and Appium, run on every build in your CI pipeline.
- 04
Performance and security checks
Load tests on the slow paths, and checks on sign-in, access rules and common web weaknesses.
- 05
Reports
Each defect logged with steps to reproduce, severity and evidence, in the tracker your team uses.
03 / Limits and risks
What testing will not do.
It cannot prove there are no bugs
Testing lowers risk. It does not remove it. We report what was tested and what was not, so you know where the risk sits.
Automated tests need upkeep
A test suite is code. When the product changes, the tests change too. Budget for that.
Late testing is expensive testing
A defect found the week before launch costs far more to fix than one found in the sprint that caused it.
04 / What we can show
How we think about testing, in writing.
The clearest thing we can show is how we approach testing, and software of our own that we release.
What this does not show
A testing-only engagement for a named client, or defect and coverage figures. We have published neither, so we quote none.
05 / What to do next
Price it, compare it or talk it through.
The first step is a free call about your product and your last few releases. We then send a short test plan and a written estimate.
A little clarity
Before testing begins.
How does NSDBytes approach software quality assurance?
We start with a test plan: the journeys that matter most, the devices and browsers to cover, and what counts as a pass. Then we test by hand, automate the checks that must run on every release, and report each defect with steps to reproduce it.
What types of testing does NSDBytes perform as part of software quality assurance?
Functional, regression, usability, performance and basic security testing, on web and mobile. Manual testing covers what needs judgement. Automated tests cover what must be repeated on every build.
Can NSDBytes integrate software quality assurance into ongoing development projects?
Yes. A tester can join your sprints, test each change as it is finished and add automated checks to your CI pipeline. It works with your developers or ours.
Can testing promise that our software has no defects?
No, and we would distrust anyone who says it can. Testing finds defects and lowers the risk of a release. We tell you what was tested, what was not and what we found, so you can decide when to ship.
How much does QA cost?
It depends on the size of the product, the devices to cover and how much you want automated. The price is agreed after a free scoping call and a short test plan. If testing is part of a build, the project cost estimator gives a range for the whole project.
Start wherever you are
Bring the release that worries you.
We’ll say what we would test first, what to automate and what it would take.