There are cases where even if an action fails, the process should continue. We want to be able to manage the failure within the process and handle it proactively instead of reactively (only after it has failed).
There are different examples and scenarios where this can be useful, but this is the most recent one we are dealing with:
we have a web service call to an external platform to send data from Nintex. There are cases were, for no specific reason, this action fails and the workflow stops.
Manually restarting the workflow seems to overcome the issue, but this has to be done manually.
If the process could proceed even after this action fails, we could insert a pause and a loop that allows us to retry the action again, allowing the process to complete successfully without human intervention.