Build multi-step alert chains that escalate through your team until someone acknowledges the incident.

Escalation Policies

Escalation policies require a Pro plan or above.

An escalation policy defines what happens after a monitor fails and an incident is created. Instead of sending a single alert and hoping someone sees it, escalation policies send alerts through a series of steps -- each one reaching a wider or different group of people -- until someone acknowledges the incident.

How escalation works

Think of an escalation policy as a chain of increasingly urgent notifications:

Step 0 fires immediately when the incident is created. This typically goes to the primary on-call person or the most relevant channel.
Step 1 fires after a configurable delay (e.g., 5 minutes) if no one has acknowledged the incident. This might go to a backup contact or a broader channel.
Step 2, 3, ... continue the pattern with additional contacts and delays.

If someone acknowledges the incident at any point, the escalation stops. No further steps will fire.

Creating an escalation policy

Name the policy

Navigate to Operations > Incidents > Escalation Policies and click New Policy. Give it a clear, descriptive name -- for example, "Production API Escalation" or "Database Team Chain."

Add Step 0 (immediate)

Step 0 fires the moment an incident is created. Configure who gets notified:

Notification channels -- Select one or more channels (Slack, Discord, email, etc.)
On-call schedule -- Select an on-call schedule to notify whoever is currently on-call (see On-Call Schedules)

You can combine both -- for example, notify the on-call person via their personal notification rules and also post to a Slack channel.

Add Step 1 (delayed)

Click Add Step to create the next escalation level. Configure:

Delay -- How many minutes to wait after Step 0 before firing this step (e.g., 5 minutes)
Notification channels and/or on-call schedule -- Who to alert if the incident is still unacknowledged

For example, Step 1 might notify the engineering lead and post to a wider team channel after 5 minutes of no response.

Add more steps as needed

Continue adding steps for your escalation chain. Each step can have a different delay and different contacts. Common patterns:

Step 0: On-call engineer (immediate)
Step 1: Engineering lead + team Slack channel (5 minutes)
Step 2: Engineering manager + SMS (15 minutes)
Step 3: CTO + all-hands channel (30 minutes)

Configure escalation repeat

Enable Escalation Repeat to loop back to Step 0 after all steps have fired without acknowledgment. This ensures that alerts keep going until someone responds, rather than stopping after the last step.

When repeat is enabled, the entire escalation chain restarts from Step 0 once the last step fires without acknowledgment.

Alert grouping

Alert grouping controls how Seenty creates incidents when multiple monitor checks fail:

Mode	Behavior
Never	Each failure threshold crossing creates a separate incident. Use this when every failure needs individual tracking.
By monitor	Only one active incident per monitor at a time. If a monitor is already in an active incident, new failures are grouped into the existing incident rather than creating a new one.
By time window	Failures that occur within a configurable time window (e.g., 5 minutes) are grouped into a single incident. Useful for reducing noise during cascading failures.

The default is By monitor, which prevents duplicate incidents for the same monitor.

Owner fallback

Every escalation policy includes an automatic safety net: if all escalation steps fire and no one acknowledges the incident, Seenty sends an email to the organization owner as a last resort.

This fallback cannot be disabled. It ensures that critical incidents are never silently ignored, even if the configured contacts are unavailable.

Enabling and disabling policies

You can disable an escalation policy without deleting it. A disabled policy will not create incidents or send notifications when a linked monitor fails. This is useful for planned maintenance windows or when you temporarily want to silence alerts for a specific service.

Monitors linked to a disabled policy will fall back to direct notification behavior until the policy is re-enabled.

Linking policies to monitors

Each monitor can be assigned to one escalation policy. To link a policy:

Open the monitor configuration
In the Alert mode section, select Escalation Policy
Choose the policy from the dropdown
Save the monitor

Multiple monitors can share the same escalation policy. For example, you might have a single "Production Services" policy used by all your production monitors.

Policy limits

Plan	Escalation policies
Pro	Up to 5
Ultra	Up to 20
Enterprise	Unlimited

Escalation Policies

On this page