What is automatic testing?
A plain-language introduction to automatic testing, and how it differs from and complements manual testing.
Manual testing vs. automatic testing
Manual testing is a person sitting down at a keyboard, clicking through an application, and checking with their own eyes that it behaves correctly: does the login form accept valid credentials, does the search box return results, does the checkout button actually charge the right amount. It is slow, it does not scale, and it depends on that person's attention and memory every single time.
Automatic testing (also called automated testing) is a program — written once by a person — that does the same checks itself: it opens the application (or calls a function, or drives a real browser), performs some actions, and asserts that the result matches what was expected. Once written, that program can run in seconds, as often as anyone likes, without ever getting tired, distracted, or bored of clicking the same button for the thousandth time.
The two are not rivals so much as complements. Automatic tests are extremely good at repeating exactly the same check, forever, cheaply. Manual testing — in particular exploratory testing, where a skilled person deliberately pokes at an application looking for the unexpected — is good at finding the kinds of problems nobody thought to write a check for. Mature teams use both: automation for the checks that must never regress, and human judgement for everything a checklist cannot anticipate.
Examples you can ask AI
- What's the practical difference between manual testing and automatic testing, and why would a team bother writing a script instead of just clicking through the app by hand?
- My team does everything by manual testing right now — how do we decide which checks are worth automating first?
- How do experienced teams decide what to deliberately keep as manual, exploratory testing even after most of their checks are automated?