Rule schedules and dayparting in AdOps

AdOps controls timing in two layers. The schedule decides when a rule becomes due, either as a check interval between 15 minutes and 72 hours or as a weekly timetable of days and hours. Time conditions decide whether an action may fire once the rule is running, comparing the current hour and weekday against values you set.

Which scheduling mode should a rule use?#

Interval mode is the default shape and answers the question how often. Ten intervals are selectable: 15 minutes, 30 minutes, then 1, 2, 3, 6, 12, 24, 36 and 72 hours. A rule that somehow carries no explicit interval falls back to 60 minutes.

Timetable mode answers the question when instead. The builder shows a grid with columns Week, Su, Mo, Tu, We, Th, Fr and Sa against 24 hour rows, and you click cells to activate them. The Week column toggles that hour across all seven days at once, which turns a daily 06:00 run into one click rather than seven.

When a timetable rule finishes, the producer looks for the next active slot after the current time. If none remains today it rolls forward day by day, up to 7 days ahead, so a rule whose only slot is Sunday at 22:00 simply becomes due next Sunday.

Either mode can be fenced with an optional date range. The picker defaults to today through 7 days out and will not accept a start date in the past. On a saved rule, an info alert shows the exact Next Scheduled Execution timestamp, so you never have to infer when it will next fire.

How is dayparting expressed?#

Not as a separate settings panel. Dayparting in AdOps is a condition like any other, which means it combines with performance conditions in the same task under the same AND or OR operator.

Two metrics carry it:

Metric	Key	Behaviour
Time greater than	time_greater_than	Passes when the current hour is later than the chosen hour
Time less than	time_less_than	Passes when the current hour is earlier than the chosen hour

Each takes an hour from a 24-option picker labelled 12:00 AM through 11:00 PM, and each can carry a day-of-week multi-select running Sunday, Monday, Tuesday, Wednesday, Thursday, Friday, Saturday. If today’s weekday is not in that list, the condition fails and nothing fires.

Both are point-in-time metrics resolved without a Meta Insights call, alongside daily budget, remaining budget, lifetime budget, hours since creation and custom metrics. Adding a dayparting condition costs no extra Meta request.

What does a business-hours rule look like?#

Combine the two time conditions to bracket a window. A task with Time greater than 09:00 AM and Time less than 06:00 PM, both restricted to Monday through Friday, and both joined by AND to a performance condition, will only act during the working week when someone is around to notice.

That pattern matters most on aggressive actions. Scaling a budget at 03:00 on a Sunday because a small hour of data looked good is how automation loses trust. Fencing the action to hours a human reviews is cheap and it is one condition.

The night budget reset#

The classic pattern runs in the opposite direction: let a daytime rule scale winners, then put the budget back before the next day starts.

The shape is one rule with a Set budget action, a time condition placing it late in the day, and whatever pre-filters scope it to the campaigns your scaling rule touches. Because Set budget writes an absolute figure rather than a delta, the baseline is restored exactly, no matter how many increases fired during the day.

Two supporting details make it reliable. Set budget accepts an action frequency, so you can cap it at once a day and know it will not fight a scaling rule that runs every 15 minutes. And AdOps re-reads the campaign’s live daily budget from Meta immediately before writing, so the change is applied against the real current value.

AdOps ships Night Budget Reset as one of its 10 rule templates, in the budget category, so the shape can be imported and edited rather than assembled from scratch.

What is worth knowing about timing accuracy?#

The dispatcher checks for due work every 10 seconds, which is how quickly a rule becomes eligible after its interval elapses. It is not a guarantee about completion time: a sweep processes due rules one after another, and a rule covering many ad accounts takes as long as Meta takes to answer.

Treat the interval as a floor on frequency. If a rule must act inside a tight window, give it a time condition that defines the window explicitly rather than relying on the schedule alone to land there.