How do you handle failed webhook retries in Bit Flows?

I am testing a client workflow where an external API occasionally times out. The workflow itself is fine, but I want a cleaner retry strategy instead of manually rerunning the whole automation.

How are you handling failed webhook or API steps in Bit Flows when the receiving service is temporarily unavailable? I would especially like to hear how you decide between retrying the failed step and restarting the workflow.