Alert fatigue is the real risk
A channel that fires forty times a day is a channel people mute. At that point you have monitoring on paper and none in practice.
Fewer, better alerts beat comprehensive coverage nobody reads.
Three tiers
| Tier | Example | Delivery |
|---|---|---|
| Interrupt | Pipeline stopped, cost cap hit | Phone or direct message, immediately |
| Attention today | Quality drop, queue ageing | Email or channel, batched hourly |
| For information | Weekly volumes, drift trend | Weekly digest |
Most alerts belong in tier two and get built as tier one, which is exactly how a team learns to ignore them.
Route by who can act
- Cost alerts to whoever owns the budget
- Quality alerts to the business owner of the process
- Technical failures to whoever can restart it
- Queue ageing to whoever staffs the review
Copying everyone on everything means nobody owns anything.
Every alert names the next action
“Correction rate up 60% on invoices” is a fact. “Correction rate up 60% on invoices — check whether supplier X changed their template” is an alert.
The runbook entry belongs in the message, not in a document nobody opens at 8am.
Prune quarterly
Any alert that has fired repeatedly without anyone acting is either wrongly tuned or unnecessary. Fix it or delete it.
Ten minutes a quarter keeps the channel meaningful, which is what makes the important alert get read.