Test Automation

Software changes constantly. Every change can break something that used to work. Checking that by hand, every time, does not scale past a small codebase. Test automation is the practice of encoding those checks as code, so a machine re-runs them on every change instead of a person re-clicking through the application.

Automation does not replace testing as a discipline. It replaces the repetitive execution of tests that already have a known, checkable outcome. Exploratory testing, usability judgment, and deciding what to test in the first place stay human work. Test automation is how you make the checks you already know you want cheap enough to run constantly.

Why automate

A regression test re-run by a person costs the same effort the tenth time as the first. An automated test costs effort once, to write, and then runs at near-zero marginal cost. That shift changes what becomes economically possible: running the full check suite on every commit, not once before a release.

The Certified Tester Advanced Level Test Automation Engineering syllabus frames the case directly: automation buys speed, repeatability, and coverage that scales with the size of a test suite, at the cost of an upfront investment to build and an ongoing cost to maintain the automation itself. Automation is not free. Scripts rot as the application changes; a suite nobody maintains becomes a suite nobody trusts. The decision to automate a given check is a cost trade-off, not a default.