A dayparting schedule built from summer data is wrong by November. Not slightly wrong; shaped wrong. The holiday shopper keeps different hours, and the hours change again between the deal weekend and the December gift rush. A schedule that does not move with them either misses the best hours or pays for the worst ones.
This playbook covers how the shape of the day changes through Q4, what to change in the schedule for each window, and when to change it back. If dayparting itself is new, start with the dayparting guide; this is the seasonal layer on top.
What changes about the day
Mornings start earlier. Deal alerts, lightning deal launches and email promotions go out early, and shoppers act on them before work. Hours that are quiet in July can be some of the strongest of the day during event weeks.
Evenings run later. Holiday browsing happens after the day's obligations, and the late-evening window that tapers off in summer holds up into November and December.
Weekends strengthen. Saturday and Sunday mornings in particular, as gift shopping becomes a weekend task.
Overnight sometimes wakes up. During the biggest event days, a portion of shoppers buy in the small hours, and time zones spread demand across the night. This is the most category-dependent change, and the one to verify with data rather than assume.
Lunch hours matter more. Shopping at work rises as gift deadlines approach, and weekday midday conversion lifts in December.
Window 1: early October to mid-November
The research phase. Traffic climbs, conversion lags behind it. Shoppers are browsing gifts they will buy later.
Schedule change: small. Widen the evening edge by an hour if your data shows the edge hour converting near average. Keep overnight paused. This is a month to watch the edge hours rather than move aggressively, because the conversion that justifies opening them has not arrived yet.
Prime Big Deal Days falls in this window and is its own event. For the two days, open the early morning and widen evenings, then restore the day after. The Prime Big Deal Days checklist covers the event specifics.
Window 2: Black Friday through Cyber Monday
The peak. Everything widens.
Open early mornings fully. Deal drops make the first hours of the day some of the most valuable of the week.
Extend evenings to their latest point.
Test overnight. If last year's data shows event-night conversion, open those hours. If you have no data, open the hours closest to your current window first and watch them on day one, then decide for the remaining days. Leaving the deepest overnight hours paused even during the event is often still right.
Build this as an event schedule with a start and an end, not as edits to the standing schedule. The difference is that an event schedule restores itself. Event rules exist for this, and the BFCM post goes hour by hour through the weekend.
Window 3: the first week of December
The quiet week. Traffic falls from the peak but stays above October levels. This is where schedules most often go wrong, because the event schedule is still running.
Schedule change: restore to the window 1 shape, with the evening edge kept slightly wider. Close the overnight hours. If restoring is a manual task, it is the task most likely to be forgotten, and a week of event hours in a non-event week is expensive.
Window 4: mid-December to the shipping cutoff
The gift rush. Decided buyers with a deadline. Less deal-driven, more time-pressured.
Schedule change: widen again, differently. Weekday lunch hours open up. Evenings run late. Weekend mornings are strong. Early mornings matter less than during the deal weekend, because there are fewer deal drops to wake up for. The December playbook covers the buyer and budget side of this window.
On the Prime shipping cutoff day for your marketplace, close the schedule back down. The gift buyer leaves once delivery can no longer arrive in time, and the hours that converted the day before stop converting.
Window 5: after the cutoff
Restore to the narrowest schedule of the year. Late December and January are returns, gift card redemption and clearance browsing. Most hours convert below their annual average. The January recovery playbook covers the reset.
Building the windows in advance
The practical problem with holiday scheduling is not knowing what to do. It is doing five schedule changes on the right days in the busiest quarter of the year. The way around it is to build all five in October, while things are quiet, with dates attached:
Standing schedule, adjusted for window 1.
Event schedule for Prime Big Deal Days, with a restore date.
Event schedule for Black Friday through Cyber Monday, with a restore date.
Event schedule for mid-December through the cutoff, with a restore date.
A January schedule that takes over after the cutoff.
Each one is written once and fires on its date. The quarter then runs on what you decided in October, which is when you had the time to decide it well.
Reading it afterward
In January, pull the hourly data for each window separately and compare it to the schedule you ran. Which opened hours earned their cost? Which paused hours would have converted? That comparison is next year's schedule, and it is worth an hour of work while the data is fresh. The heatmap guide shows how to build the view.
Frequently asked questions
Should I turn off dayparting during Black Friday?
Widen it rather than turning it off. Deal traffic arrives early and runs late, so the active window should expand, and some overnight hours that are dead the rest of the year may convert during the event. Hours that still produce no orders under event traffic should stay paused.
When should holiday ad schedules start?
Shift gradually from early October as gift research begins, make the largest change for the Black Friday through Cyber Monday window, restore after it, and make a second smaller change for the mid-December gift rush.
How do I know which hours to open during the holidays?
Use your own account's hourly data from last Q4 if you have it. If not, use this October's data as the season starts and watch the hours just outside your current window. When those edge hours start converting near the daily average, widen the schedule to include them.
Off Hours lets you build all five holiday schedules in October, each with its own start and restore date, so Q4 runs on the plan instead of on reminders. Start a free 14-day trial.