The worst automation failure isn't the one that crashes loudly. It's the one that quietly stops, so a client finds out three days later that no follow-ups went out. Here's the setup we put on every client instance.

It's like a smoke alarm: useless if it's not wired in, and just as useless if it goes off every time someone makes toast.

1. One Error Workflow for the whole instance

Create a workflow that starts with the Error Trigger node, and set it as the error workflow in every production workflow's settings. It receives the failed workflow's name, the node that failed, the error message and a link to the execution.

2. Send it where someone will actually see it

Email gets ignored. We send a short WhatsApp or chat message: workflow name, failed node, the first line of the error, and the execution link. One line each, readable on a phone.

3. Don't alert on every failure

  • Use Retry On Fail on nodes that call outside APIs (two or three tries with a wait). Most API hiccups fix themselves.
  • Group repeats. If the same workflow fails five times in ten minutes, send one alert saying "failing repeatedly", not five.

4. Catch silent failures too

Some workflows "succeed" while doing nothing, such as an import that finds zero products because the supplier changed a field. Add a check at the end: if the count is zero (or far below normal), treat that as an error.

5. A heartbeat for scheduled workflows

Each important scheduled workflow writes "I ran at

6. Log, then fix the cause

Keep a table of every alert. Once a week, look at what fired most and fix the cause, so alerts stay rare and people keep reading them.

I run Arwahyn AI, an AI automation agency in Dhaka.