No WAIT. That is how human tests.

Image if you would test like this: Click the save button, wait two seconds, close the application. No way, you say. You would press the save button, wait till the changes are saved and then close the application. It is just, your test automation tool does not work like that.

How would you properly do it

Do not use any micro-specific timeouts and verify each action like a human would. But there must be a timeout, otherwise a broken test or a broken software might run forever. That is correct.

For our incident example from the top, define an overall timeout. How long does it usually take if a human operator enters a new incident with all information prepared? If your answer is 20 minutes, then you have your timeout. If the test finishes in five, that is fine. If it takes longer than 20 minutes some is so severely broken that a human would not accept it even if it would work from a functional point of view.

To identify correct timeouts, either ask you users or, depending on availability, use the KPIs for the underlying software.
This approach allows you to increase reliability, run testers faster through scaling and saves you time and costs because of the reduced manual interaction and fast results.

Try for Free