Let WHAWIT propose it
Building a first on-call configuration by hand means inventing a rotation, a handoff hour, an escalation chain and a set of timeouts before you have any feel for how they behave. WHAWIT will propose the whole thing instead — escalation teams, schedules and escalation policies — and hand it to you to review before anything is created. Go to On Call → Configuration and start the guided setup. (Creating an individual on-call agent is a separate flow — New Oncall on the agents screen.)Step 1 — who is on the roster
Three ways to tell WHAWIT who is involved:From your directory
Uses the people already in your organization, both active members and those
with pending invitations. The fastest option if your team is already in
WHAWIT.
From a list of emails
Paste addresses. Anyone who is not yet an organization member is collected
into a list of people to invite, so nobody ends up on a rotation they cannot
log in to.
Describe it
Write what you want in plain language — who covers what, which hours matter,
how many tiers — and let WHAWIT work out the shape.
Whoever ends up in a rotation must be an organization member. The list of
emails mode surfaces the ones who are not, so you can invite them as part of
the same setup rather than discovering the gap when a page fails to route.
Step 2 — a few questions
Step 3 — review the proposal
WHAWIT returns a complete proposed configuration:- Escalation teams — the groups it thinks make sense from your roster.
- Schedules — layers, rotation, timezone and handoff, with coverage restrictions where the tier choice implies them.
- Escalation policies — ordered levels with targets, timeouts and channels, and the repeat you asked for.
- Users to invite — anyone on the roster who is not yet a member.
Step 4 — verify before you rely on it
This is the part worth not skipping. Four checks, in order of how often they catch something:Preview every policy, at an awkward hour
Preview every policy, at an awkward hour
Preview resolves a policy against the current rotations and overrides and
shows exactly who would be paged at each level. Run it at 03:00 on a weekend,
not during business hours — that is where the gaps are. A level that resolves
to nobody is the failure you are looking for.
Look at a full week of shifts
Look at a full week of shifts
View the upcoming shifts across seven days. If coverage restrictions across
your layers do not quite meet, there is a window where the schedule resolves
to nobody, and it will not be obvious from the configuration alone.
Check every user has a phone number
Check every user has a phone number
WhatsApp and SMS deliver to the number on the user’s profile. A user without
one is skipped for those channels while the escalation continues on the
others — so an SMS level can be silently useless. Fill them in.
Confirm one policy is the organization default
Confirm one policy is the organization default
The default is the fallback for any incident raised by something with no
policy of its own. Without it, such an incident has no escalation path at
all.
Adjusting later
Nothing here is permanent. Schedules, layers, rotations, restrictions, overrides, policies, levels and channels are all editable after the fact, and preview lets you check the effect of a change before an incident does. Most teams tighten their timeouts and add a secondary tier within the first month — that is the system working, not a sign the initial setup was wrong.Related
On-call features
The full capability list.
Schedules and rotations
The concepts behind the questions above.
Escalation policies
Levels, targets, timeouts and preview.
Migrating
Coming from PagerDuty or Opsgenie.