Automatic testing is like a superpower — we can record scripts of specific tests and have them run on a regular schedule or after any server /​content change. This detect problems in large applications well before a person could catch them.

Automation testing is seriously like magic – especially when dealing with a large application or website with multiple variables that need to be tested.

Basically, we create Selenium scripts of specific functional tests being completed. The scripts are then grouped into suites to target different environments, from development, staging and production, to ensure consistency and integrity. Moving forward, at each release or sprint, we have the server automatically tested for regression to ensure that the addition did not affect the existing scripts negatively. This type of implementation will save you hours of testing, and help velocity as well as give you the ability to test continuously, from different browsers, devices, screen sizes and parts of the world.

We recommend automatic testing if we are doing frequent releases on a given complex application, otherwise the manual testing burden is quite high. Getting the first set of tests established does take time, but they pay back that investment quite quickly. Imagine having two hours of manual testing in fifteen minutes at each deployment.

Moreover, it's nice that the tests can be run holistically for the entire application at once, meaning that if some code accidently impacts another region of the app, we can quickly find it and recover that problem.