Key takeaways
- Testing means checking software against what it should do and reporting what you find, so the team can decide what to do about it.
- It reduces risk. It can show bugs are present, but never prove there are none.
- Testing covers more than "does the button work": speed, security, usability and more. And it starts before any code exists, because the earlier a problem is found, the cheaper it is to fix.
What testing actually is
Software testing is the work of running a program, or reviewing it, to find out whether it behaves as expected.
A tester compares two things: what the software should do and what it actually does. Every difference is worth a closer look.
That's the whole idea. Everything else (techniques, tools, automation) helps you do it faster and more thoroughly.
A simple example
Imagine a sign-up form with a rule: "The password must have at least 8 characters."
A tester doesn't just type a good password and move on. They try 7 characters, exactly 8, an empty field, 200 characters, spaces only, and emojis. Each try asks a question: does the software do the right thing here? If 7 characters are accepted, that's a bug, and the tester reports it.
What testing is not
A few common misconceptions:
- "Testing is just clicking around." Good testing is planned. You decide what to check and why.
- "Testing proves the software works." It can't. It shows where it fails, under the conditions you tried.
- "Testers are responsible for quality." Quality is the whole team's job. Testers give the team the information to act on.
- "Testing happens at the end." The most valuable testing often happens early, on requirements and designs.
Why it matters
Bugs cost money, time and trust. Some cost far more.
About 37 seconds after launch, the first Ariane 5 rocket veered off course and was destroyed.
The cause was software reused from Ariane 4. It converted a 64-bit number into a 16-bit integer. Ariane 5 flew faster, the number got too big, and the conversion failed.
The code had worked for years. It had never been tested against the new rocket's flight conditions.
Lesson: "it worked before" is not evidence. Test the software in the conditions it will really run in.
Most bugs are less dramatic: a checkout that fails on one browser, an invoice with the wrong tax, a password reset email that never arrives. Each one costs you users, support time or money.
What testing checks
"Does it work?" is only the first question. Testers look at several kinds of quality:
| Question | What it's called | Example |
|---|---|---|
| Does it do the right thing? | Functional testing | The discount is applied correctly. |
| Is it fast enough? | Performance testing | The page loads in under 2 seconds with 1,000 users. |
| Is it safe? | Security testing | One user can't see another user's orders. |
| Is it easy to use? | Usability testing | New users find the "Pay" button without help. |
| Can everyone use it? | Accessibility testing | It works with a screen reader and a keyboard. |
| Does it work everywhere it should? | Compatibility testing | It works on Chrome, Safari, Android and iPhone. |
You don't need to master all of these at once. Most testers start with functional testing and add the others over time. The article on functional vs non-functional testing goes deeper.
Two ways to test: looking and running
- Static testing examines something without running it: reading requirements, reviewing designs, reviewing code. It finds problems before they're built. See static testing and reviews.
- Dynamic testing runs the software and observes what happens. This is what most people picture when they think of testing.
Both matter. A misunderstanding caught in a requirement review never becomes a bug at all.
The earlier, the cheaper
A typo in a requirements document takes a minute to fix. The same mistake found after release can mean a hotfix, support tickets and an apology to customers.
That's why testing isn't only something that happens at the end. Testers review requirements, ask questions about designs and test small pieces as soon as they exist.
Who tests?
Testing isn't done only by people called "tester":
- Developers write unit tests for their own code.
- Testers / QA engineers design and run tests across features, explore for unexpected problems, and often build automated test suites.
- Product owners and users check that what was built is what they actually need, in acceptance testing.
A tester's special value is a different point of view: the developer asks "how do I make it work?", the tester asks "how could it fail?". The article on what it means to be a software tester describes the job in detail.
What a tester does day to day
- Understand what the feature should do: read requirements, talk to the team.
- Decide what to check: the important flows and the risky edge cases.
- Run the checks, by hand or with automated tests.
- Report what you found, clearly enough that someone else can reproduce it.
- Recheck once it's fixed, and make sure nothing else broke.
Where to go next
If you're new, read these in order: the 7 principles of testing, then what it means to be a software tester, then error, defect, bug, failure. If your goal is a job in testing, the career roadmap shows the whole path.
Check yourself
A feature passed all its tests. Can you say it has no bugs?
Show answer
No. Passing tests show the feature works for the cases you tested. Bugs can still hide in cases nobody tried. Testing reduces risk; it can't remove it.
