One thing I’ve learned when designing systems:
I don’t design to avoid failure.
I design to control how the system fails.
That might sound strange at first.
But think about vehicles.
Recently, one of my rear lights went out and suddenly my car started automatically triggering the hazard lights.
At first I was annoyed.
I thought the car was malfunctioning.
Then I looked into it and realized something clever:
The engineers knew the rear light failing was dangerous.
So instead of allowing the system to fail silently…
they designed the vehicle to fail visibly.
That’s smart system design.
And honestly, business systems are no different.
Because failures are going to happen.
APIs fail.
People miss steps.
Data gets corrupted.
Automations break.
The question isn’t whether a system will fail.
The question is:
When it fails… what happens next?
Does the issue get exposed immediately?
Does the process recover safely?
Or does the problem quietly spread through the business unnoticed?
That’s why good architecture matters.
The best systems aren’t the ones pretending failure doesn’t exist.
They’re the ones designed to handle failure intelligently.
