--- title: On-Call Scheduling & Escalation Policies description: On-call rotations, overrides and escalation policies built into WatchFor. Confirmed critical incidents page the person on call until someone acknowledges. canonical: https://watchfor.io/on-call --- # On-Call Scheduling and Escalation, Built Into Your Monitoring A message in a team channel at 3 AM is a hope, not a response. WatchFor pages the person on call right now, on their own phone, until they acknowledge. [Set up on-call](/auth/sign-up)[Plans with on-call](/pricing) ## Everything On-Call, in One Place From “who's on this week?” to “who got paged and when?” ### Rotation Schedules Daily, weekly or every N days (1 to 90), with a handoff time, in the schedule's own timezone so handoffs stay put across daylight-saving changes. Up to five layers per schedule. Presets build weekly, every-2-weeks, monthly (30-day) and weekdays/weekends rotations in one click. ### Live Rotation Preview The editor draws the next four weeks as you change the rotation, so you see every handoff before you save. ### Month Calendar & Fair Totals A week or month view of every shift, with time on call, share of the period and shift count per person. A shift counts in the period its handoff falls in, so a week ending on the 1st belongs to the new month. Print or save the view as PDF, totals included. ### Overrides & Swaps Click a shift on the calendar to hand it to someone else for vacation cover or a swap. Overlapping overrides are refused, and so is an override that gives a shift to the person already on call. An override in the past corrects the record and the totals. ### Permanent Replacements Replace someone in every layer, immediately or from the next shift. The newcomer takes the same slot, so nobody else's turns move. ### Escalation Policies Up to ten steps. Each step pages the current on-call of a schedule, specific people or a notification channel, then waits from 0 minutes to 24 hours. Repeat the whole chain if nobody has acknowledged. Acknowledge or resolve the incident and the next step does not fire. ### Personal Paging Methods Each person adds their own email, Telegram, ntfy, Pushover, Gotify or Pushbullet, and every enabled method is paged. Someone on a schedule with no method gets a warning in the editor, and the dashboard says when the person on call right now can't be paged. ### The Full Notification Flow The Flow button on a monitor or contact group shows the whole path: which channels fire immediately, then each escalation step and who it would page right now. ## Who is on call, and how they get paged The editor previews four weeks and flags anyone who can't be paged. Channels are notified first, then three escalation levels if no one acks. Each person adds their own paging methods: email, Telegram, ntfy, Pushover. ## Why On-Call Belongs Next to the Monitors A separate paging tool means a second user directory and a webhook bridge between the two, and that bridge is one more thing that can break during an outage. In WatchFor the escalation starts from the incident itself. Only confirmed critical incidents start a chain. A flapping, muted or maintenance-window monitor pages nobody. Acknowledging in the dashboard or over the API stops the chain, because each step re-reads the incident before it fires. [Start monitoring now](/auth/sign-up) Only confirmed critical incidents page; warnings go to channels only Team channels still get every notification while the chain runs Acknowledge or resolve and the next step never fires Muted monitors, maintenance windows and flapping incidents never start a chain Pending chains are stored in the database, so a restart doesn't drop a page Every escalation step that fires is written to the incident timeline Schedule, override and policy changes land in the activity log GET /v1/on-call tells scripts and AI agents who is on call right now ## How On-Call Works Four pieces that click together 1 ### Build a Rotation Pick the people and a pattern, or start from a preset, and watch the four-week preview. The calendar then shows who owns every shift. 2 ### Define the Escalation Create a policy: step 1 pages the on-call person, later steps page backups or a channel, with a wait between steps and an optional repeat. 3 ### Attach It to a Contact Group Select the policy on a contact group. Every monitor that targets that group now escalates its critical incidents through it, and the group's channels keep working as before. 4 ### Get Paged, Acknowledge, Done A confirmed critical incident starts the chain. The on-call engineer is paged on their personal methods, acknowledges, and the escalation stops. ## On-Call Use Cases For teams that put a name on every incident ### Small Teams Rotating Fairly Three to six engineers on a weekly rotation. Per-person totals and shift counts for the month make on-call pay a lookup instead of a spreadsheet. ### Office Hours and After Hours A base layer covers nights and weekends; a higher layer restricted to 09:00–17:00, Monday to Friday, covers office hours. Overnight ranges such as 22:00–06:00 wrap past midnight. ### Escalation Safety Net Step 1 pages the primary. With no acknowledgement five minutes later, step 2 pages the backup, and step 3 posts to the team's PagerDuty service or Slack room. ### Replacing a Departing Teammate Swap a person out of the rotation permanently without moving anyone else's turns, and cover the handover week with a one-click override. ## Frequently asked questions Everything you need to know about on-call in WatchFor [All questions→](/faq)[Ask us directly](/contact) What starts an escalation? A confirmed critical incident on a monitor whose contact group has an escalation policy. Warnings, single failed checks, flapping incidents and monitors that are muted or in a maintenance window never start a chain. Do team channels still get notified when escalation is on? How do rotations work? How do overrides work? What does an escalation policy look like? What stops an escalation chain? How do people get paged personally? Can I see who gets paged before an incident happens? How is on-call time counted for compensation? Is on-call a separate product or plan? ## From our blog Guides, deep-dives, and best practices from the WatchFor team [Reliability ### On-call scheduling without a second tool: rotations, escalations and fair shift math Aug 23, 2026·6 min readRead article](/blog/on-call-scheduling-in-watchfor)[Reliability ### On-Call Best Practices: a rotation that doesn't burn people out May 27, 2026·4 min readRead article](/blog/on-call-best-practices)[Reliability ### Alert Fatigue: why your team stopped reading alerts (and how to fix it) Jun 25, 2026·5 min readRead article](/blog/alert-fatigue) ## Put a Name on Every Critical Incident Build a rotation, attach an escalation policy to a contact group, and confirmed critical incidents page the right person until someone acknowledges. On-call is included from the Starter plan; creating the account is free and needs no credit card. [Set up on-call](/auth/sign-up)[View documentation](/docs/alerting/on-call) --- Canonical page: https://watchfor.io/on-call · Site guide: https://watchfor.io/llms.txt