On-call
on-KALL
A rotation where engineers are responsible for responding to production alerts and incidents outside business hours.
On-call is a rotation where an engineer carries a pager (usually their phone) and is responsible for responding to production alerts. When something breaks at 3 AM, the on-call engineer is the first responder. They investigate, triage, and either fix the problem or escalate to the right team.
Most teams rotate on-call weekly. One engineer is primary (gets paged first), another is secondary (gets paged if the primary does not respond within a few minutes). On-call schedules are managed through tools like PagerDuty, Opsgenie, or Grafana OnCall. Good on-call rotations are equitable: everyone on the team takes a turn, including senior engineers and managers.
On-call quality varies wildly between companies. At its best, on-call means getting paged once or twice a week for real issues, with clear runbooks and fast resolution. At its worst, it means getting paged 10 times a night for flapping alerts and poorly documented systems. The quality of on-call is a direct reflection of engineering investment in reliability, monitoring, and documentation.
Examples
An engineer starts their on-call rotation.
Monday at 9 AM, the engineer takes over as primary on-call. They review the handoff notes from the previous engineer: two ongoing issues being monitored, one alert that was silenced pending a fix, and a heads-up about a risky deploy scheduled for Wednesday. They verify their phone notifications are working and that they can access the VPN from home.
A team improves their on-call experience.
On-call engineers were getting paged 15 times per week, mostly for noisy alerts. The team spent two sprints reducing alert noise: deduplicating alerts, raising thresholds, and deleting monitors for decommissioned services. Pages dropped to 3 per week. On-call satisfaction scores went from 2/5 to 4/5.
A company introduces on-call compensation.
Engineers were burning out from unpaid on-call rotations. The company introduces $500 per week for carrying the pager, plus $200 per page outside business hours. On-call becomes a paid responsibility rather than an unspoken expectation. Retention improves and more engineers volunteer for rotations.
In practice
Read more on the blog
Frequently asked questions
How often should engineers be on-call?
A healthy rotation has each engineer on-call no more than one week every 4-6 weeks. More frequent than that leads to burnout. If your team is too small for this frequency, it is a signal to hire or to reduce the number of services the team owns. Some companies cap on-call at 25% of an engineer's time.
What is the difference between on-call and being available?
On-call means you have a formal obligation to respond to pages within a defined time window (usually 5-15 minutes). Being available means you might check Slack occasionally. On-call engineers carry pagers, follow escalation policies, and have explicit response time SLAs. The distinction matters for compensation, scheduling, and accountability.
Related terms
An unplanned event that disrupts a service or degrades it below its expected quality, requiring a coordinated response.
A classification system (SEV-1 through SEV-4) that ranks incidents by impact and urgency to determine response priority.
A written analysis of an incident: what happened, why, and what the team will do to prevent it from recurring.
The percentage of time a system is operational and accessible to users.
Service level objective: an internal reliability target for a service, like 99.9% availability or p99 latency under 200ms.
Want the complete playbook?
Picks and Shovels is the definitive guide to developer marketing. Amazon #1 bestseller with practical strategies from 30 years of marketing to developers.