Escalation Policies That Page the Right Person

Tiered escalation chains with automatic advancement, acknowledgment, and repeat cycling so no alert ever goes unanswered.

Ensure Critical Alerts Reach the Right People

Tiered escalation chains with automatic advancement and acknowledgment. No alert goes unanswered.

metrictower.com/settings/escalation-policies

Production Critical

3 steps -- repeat enabled -- 4 integrations

Active
Edit
1
Immediate Wait 5 min before step 2
#ops-alerts (Slack) On-call engineer (Email)
2
After 5 minutes Wait 15 min before step 3
PagerDuty (on-call)
3
After 15 minutes Final step -- then repeat
PagerDuty (engineering lead) CTO (SMS)

Alerts That Climb the Chain

Define who gets notified, in what order, and how long to wait before escalating.

Multi-Step Escalation

Configure as many steps as you need. Each step defines which channels to notify and how long to wait before advancing to the next one.

Acknowledgment

Team members can acknowledge alerts to stop escalation. Once acknowledged, the chain halts and the alert is owned by whoever claimed it.

Repeat Cycling

Optionally restart from step 1 after exhausting all steps. Critical incidents keep escalating until someone acknowledges, so nothing falls through the cracks.

Flexible Routing, Complete Control

Integration-Based Routing

Each escalation step can notify any combination of your configured integrations -- Slack channels, PagerDuty services, email addresses, SMS numbers, or webhooks. Mix and match channels per step to reach the right people at the right time.

Configurable Wait Times

Set the wait time between each step independently. First responder gets 5 minutes, the on-call lead gets 15, the VP of Engineering gets 30. Match your escalation timing to your organization's response expectations.

Severity Filtering

Assign escalation policies to specific alert severities or categories. Critical production outages follow a fast-track 3-step policy while medium-severity warnings use a gentler single-step notification. Different problems get different response workflows.

Escalation Audit Trail

Every escalation step is logged with timestamps, delivery status, and acknowledgment events. Review the full lifecycle of any incident -- when it was raised, who was notified, and when it was claimed. Essential for post-incident reviews.

How Escalation Works

1

Alert Fires

An uptime check fails, a cron job misses, or a vulnerability scan finds a critical issue. The alert is routed to the matching escalation policy.

2

Step 1 Notifies

The first step fires immediately, notifying the configured channels. The timer starts for the next step.

3

Escalate or Acknowledge

If no one acknowledges within the wait time, the next step fires. Acknowledging at any point halts the chain.

4

Repeat or Resolve

After all steps, the policy optionally cycles back to step 1. The cycle continues until acknowledgment or resolution.

Escalation Policies That Never Let Alerts Slip

Define multi-step escalation chains that automatically advance through your team hierarchy. Acknowledgment stops the chain, repeat cycling ensures nothing is missed, and a full audit trail covers every incident lifecycle.