E2E (end-to-end testing)

E2E tests (end-to-end) are automated tests that verify an application as a coherent whole by simulating real user behaviour. Instead of checking individual functions in isolation, an E2E test reproduces a complete scenario — for example registering an account, adding a product to the cart and completing checkout — passing through every layer of the system: the interface, server logic, the database and external integrations.

How E2E tests work

An E2E test typically runs in a real or headless browser driven by an automation tool. The test script clicks buttons, fills in forms and reads page content exactly as a human would, then asserts that the application responded correctly. In the classic testing pyramid, E2E tests sit at the very top — there are fewest of them, because they:

• run slower than unit tests and integration tests, since they exercise the entire stack,
• can be flaky — sensitive to network latency and interface asynchronicity,
• require an environment that closely mirrors production.

Popular tools include Playwright, Cypress and Selenium. A good practice is to cover the critical business paths with E2E tests and push logic details down to cheaper, lower-level tests.

E2E testing in practice

E2E tests protect an application's most important flows against regression. They run inside a CI/CD pipeline before each deployment to catch cases where a new change has broken login, payments or order placement. Combined with a TDD approach and a suite of regression tests, they form a safety net that lets a team ship changes faster and with greater confidence that users won't hit a critical bug.