The first alert is only step one. Escalation policies define what happens when nobody acknowledges an incident within a set time — notify the backup engineer, then the manager, then SMS everyone.
Why escalation matters
Small teams take holidays. On-call rotations slip. Escalation is the safety net that prevents a missed phone notification from becoming a twelve-hour outage.
Design principles
- Time-based steps — e.g. email at 0 min, Telegram at 5 min, SMS at 15 min
- Clear ownership — Each step names a person or role, not "the team"
- Stop when acknowledged — No continued paging after someone takes the incident
- Document runbooks — Link from alert templates to "what to check first"
Keep policies maintainable
When staff change, update escalation lists the same day access is revoked. Quarterly drills (fire a test alert) catch broken phone numbers and expired API tokens.
Fit policy to business hours
A bakery website might escalate slowly overnight but fast during opening hours. Use quiet hours and severity to tune behaviour without turning monitoring off.
UpMonix escalation settings integrate with the communication platform so policies apply across monitors — website, domain, SSL and API.