Ad scheduling, often called dayparting, means setting which days and hours your campaign runs and how bids change in those slots. The logic is simple: demand is not evenly distributed through the day, so budget need not be either. The practice is less simple; switching off the wrong hour often loses more than it saves, because conversion lag and Smart Bidding's learning both get involved.
Why does timing change performance?
The same person behaves differently at different hours. They research on the commute in the morning and decide at home in the evening. That behavioral gap shows up directly in conversion rate: even with a stable click cost, acquisition cost can double in a low-converting hour. The job of scheduling is to reflect that gap in how budget is distributed.
How do you read hourly data correctly?
The most common mistake is deciding from a few days of data. Sample size collapses fast in an hourly breakdown: in an account getting 20 conversions a day, a single hour slot may not even land one conversion per day. Collect at least four weeks and evaluate weekdays separately; Saturday morning and Tuesday morning are the same clock but not the same audience.
How do you set up scheduling?
The right order is gradual intervention. Start with bid adjustments; cutting the bid in weak hours lowers cost without losing the valuable clicks that still arrive then. Switching off is the last resort.
- Collect data: at least four weeks of conversion and cost data broken down by hour and day.
- Confirm the pattern: does the same weakness repeat every week, or was it a one-off?
- Adjust gently: lower bids step by step in weak slots and raise them in strong ones.
- Measure the effect: if total conversion volume fell, you produced a loss rather than a saving.
At which hour does your budget evaporate?
Ads Sensor analyzes your campaigns with AI and shows when and where spend goes unrewarded, with the reasoning attached.
Do you need scheduling with Smart Bidding?
Usually not. Strategies like tCPA and tROAS already weigh time of day, device and user signals in every auction, and bid low in bad hours. Manually blocking hours cuts the data flow and weakens learning. The one case where scheduling still makes sense alongside Smart Bidding is an operational constraint: if nobody answers the phone at night, collecting leads then is pointless anyway. For the full bidding picture see tCPA versus tROAS.
When is scheduling an operational requirement?
In some businesses scheduling is not an optimization but a necessity. A company that sells by phone should not run search ads when nobody picks up. The same holds for messaging campaigns, where late replies burn budget directly, as we detailed in our WhatsApp ads guide. Matching the schedule to team capacity matters more than any performance tuning here.
Scheduling is a small but steady lever; overdone, it kills volume. To catch hour and day level waste automatically, join the Ads Sensor beta.