Out of office after scheduling: a practical rescheduling checklist

Start with availability, not the event's title

A calendar entry named OOO is not necessarily the same thing as the provider's out-of-office event type. Google documents outOfOffice, focusTime, and workingLocation as distinct status events. Availability of these features depends on the account and they belong to primary calendars. Working location is not an absence. Inspect the event type and whether the time is busy; do not infer either from its color or title.

There is a useful technical distinction for whole-day absences. Google's API requires typed out-of-office events to have timed boundaries, even if the absence covers a whole day. Ordinary all-day busy events instead use date boundaries. Critical Path recognizes Google's outOfOffice type and also reads ordinary busy time, including all-day entries. These are two routes to representing unavailable time, not interchangeable API fields.

Keep the description restrained. A private absence can be called Unavailable; an automated planner does not need medical details, a travel itinerary, or family information to avoid its interval. The minimum useful evidence is calendar, start, end, timezone, and availability.

Sources: Google Calendar: status events and out-of-office representation

Confirm the planner can actually see it

A calendar application can display several calendars simultaneously while an integration reads a narrower scope. In the current Critical Path Google connection, availability is read from the connection's selected calendar, falling back to the primary calendar. Seeing a personal absence in Google's combined view is therefore not enough to prove that this scheduler received it.

Before changing any tasks, open Settings → Connections and identify the connected account and the calendar used for writes. Confirm that the absence is represented in the availability source. If your employer permits it, a private busy hold on the relevant work calendar can represent a personal absence without disclosing its reason. Avoid copying confidential details or adding guests unnecessarily.

Also check authorization health. A disconnected or attention-required connection cannot supply dependable current availability. Reconnect the intended account when asked, then inspect the next sync result. A failed read must not be interpreted as an empty day. Critical Path preserves existing protected state when availability cannot be obtained; that is safer than moving work using an invented free calendar.

Worked example: an afternoon absence arrives late

Consider a fictional Monday plan in one timezone. A meeting occupies 09:00–10:00. A 90-minute analysis block is at 10:15–11:45, a partner call is at 12:00–12:30, and a 60-minute writing block is at 13:00–14:00. At 08:30, an absence is added for 12:00–17:00. The analysis still fits with 15 minutes before it and before the absence. The writing block does not.

If Tuesday has a meeting at 09:00–09:45 and no later constraints, 10:00–11:00 is a candidate for the writing block. It is not a promise of the selected placement: other tasks, earlier valid reservations, its due date, and current availability can change the result. The partner call remains a human coordination issue; the planner should not silently move an external organizer's meeting.

Understand what the next reconciliation can do

Critical Path is configured to attempt calendar reconciliation on a five-minute schedule. It reads provider availability, assesses task blocks, and attempts the relevant calendar changes. This cadence is a background-work policy, not a guarantee that every external edit appears within exactly five minutes. Authorization failures, unavailable providers, write errors, and work still in progress can delay completion.

The current scheduling policy uses weekday 09:00–17:00 windows in the workspace user's timezone, a minimum 15-minute separation, and sessions up to two hours. It keeps placements that remain valid before finding new ones. Increasing weekly capacity does not expand these daily windows. If your absence removes the only suitable time, a bigger capacity number will not create a valid slot.

Treat provider event identity and confirmed sync state as stronger evidence than an invitation email. A sent email and an updated calendar event are different observations. Refreshing the application may reveal a newer saved state, but repeatedly clicking scheduling controls while an operation is unresolved makes diagnosis harder. Record the visible status and inspect one affected task first.

Verify one safe example before trusting the whole week

Use a low-stakes task in your own calendar for a controlled check. Creating or moving an event can send notifications, so choose the test deliberately and avoid adding colleagues. Record what you observe rather than treating the checklist below as results already obtained.

1. Record the task's allocation, due date, current block boundaries, provider calendar, and timezone.
2. Add the intended absence in the calendar source the connection reads. Check busy status and exact start and end.
3. Allow a reconciliation attempt, then inspect Connections for authorization or synchronization errors.
4. Open the task and Google Calendar. Compare the replacement start and end, not just the notification subject.
5. Inspect the preceding and following event boundaries. There must be at least 15 minutes around the action block under the application's policy.
6. Check whether the obsolete app-owned placement remains. If it does, capture private diagnostic details for support; do not bulk-delete by color or sender.
7. Repeat separately for a whole-day busy absence and a deadline with insufficient remaining capacity. Record failures honestly.

Make the shortage visible instead of hiding it

Suppose the writing task is due Monday and the absence leaves no complete 60-minute session with the required gaps. The correct decision is not to shorten its recorded effort without changing its scope. Choose among reducing the deliverable, assigning a qualified owner who has time, negotiating the deadline, or explicitly cancelling the commitment. Write down which decision was made.

The scheduler aims to place the full allocation; an unplaced task is an unresolved planning exception. Do not read it as completed, dismissed, or safe to forget. A weekly review should surface these exceptions next to the outcomes they threaten. If a deadline is immovable, its consequence belongs in the decision conversation rather than in an invisible scheduling workaround.

Finally, check the first working day after returning. A calendar that avoids the absence but packs every following minute can still be a poor operating plan. Leave recovery space for messages and changed priorities, review owners, and confirm that a rescheduled session still advances a useful result. Automatic movement protects geometry; you remain responsible for whether the work should happen.

Common questions