Make your alerts say what broke, not just that something did

A useful alert tells you exactly what went wrong and where. A vague one just says something failed, and leaves you to go dig through the whole automation to find out what actually happened.

This matters because time spent figuring out what an alert even means is time you could have spent fixing the actual problem. An alert that says "step failed" sends you hunting through every step one at a time. An alert that says which step failed, with which piece of information, and why, sends you straight to the fix. The difference between those two alerts is the difference between minutes and a much longer, more frustrating search. I rewrite my own alerts with that question in mind, since I've read plenty that told me something failed without ever saying what.

Whenever you build an alert, ask what someone reading it cold, with no other context, would need to know to act on it right away.

Try this today: look at your last alert message and rewrite it so it names the exact step and reason, not just the fact that something failed.