Budget
Budget pacing on TikTok: ABO, CBO and mid-flight moves
Daily totals tell you what happened. The hourly curve tells you what is about to happen — and it is usually where the cheapest optimization on the account is hiding.
The pattern almost every account has
Open any ad group's hourly spend chart and you will usually see one of three shapes:
- Front-loaded: most of the budget is gone before midday, often in the cheapest inventory hours. The evening — frequently the best converting window for consumer offers — is funded with whatever is left, which is nothing.
- Flat: spend is spread evenly. Comfortable, but it means the ad group is treating a 9 a.m. impression and an 8 p.m. impression as equally valuable, which they rarely are.
- Back-loaded: delivery struggles early and catches up late. Usually a signal of a bid or audience constraint rather than a pacing problem.
None of these is automatically wrong. What is wrong is not knowing which one you have.
Why front-loading happens
Delivery systems buy the cheapest impressions that satisfy the objective. Early in the day, the auction is often less competitive, so the same budget wins more impressions. If those impressions convert at a lower rate, you get a familiar and frustrating outcome: high impression volume, decent CPM, weak CPA, and a converting window that never receives budget because the daily cap is already reached.
The diagnostic: plot conversions by hour next to spend by hour. If the shapes do not match, your budget and your results live in different parts of the day. That gap is recoverable money.
What to do about it
1. Dayparting, carefully
Scheduling can force budget into the hours that actually convert. The cost is less data and a smaller auction pool, which can make delivery less stable — especially for ad groups that are still learning. Use it when the hourly pattern is strong and consistent over several weeks, not after a single good Tuesday.
2. Budget shape instead of budget size
Often the fix is not more budget but a different split. Two ad groups sharing an objective rarely have the same marginal cost per result. Moving budget from the ad group whose next dollar is expensive to the one whose next dollar is cheap improves the account without increasing spend.
3. Watch the ad groups that are capped and profitable
An ad group that hits its daily budget every day while holding CPA below target is telling you something plainly: it wants more money. These are the cheapest scaling decisions available, and they are easy to miss when you only look at accounts one at a time.
ABO and CBO pace differently
| Ad group budget (ABO) | Campaign budget (CBO) | |
|---|---|---|
| Control | You decide exactly what each ad group can spend | The system decides, based on where it expects results |
| Best for | Testing structured hypotheses — audiences, offers, placements | Scaling a set of ad groups that have already proven they work |
| Failure mode | You keep funding an ad group that has stopped earning it | The campaign concentrates on one ad group and starves the rest, including ones that were working |
| Pacing signal to watch | Hourly spend vs. hourly conversions per ad group | Share of campaign spend per ad group over time |
A practical middle path many teams land on: test in ABO where control matters, consolidate proven ad groups into CBO for scale, and monitor the concentration of spend inside the CBO campaign so a single ad group cannot quietly become the whole campaign.
Moving budget mid-flight without breaking delivery
Every budget change is a signal to the delivery system. Large, frequent changes make performance harder to read and can push an ad group back into an unstable state. What tends to work:
- Change budgets in moderate steps rather than doubling — 20–30% at a time is a common working range
- Leave time between changes so you can attribute the effect to something
- Avoid stacking a budget change, a bid change and a creative swap on the same day; you will not know which one moved the number
- Make increases when the ad group is performing, not as a rescue attempt for one that is not
- Set a floor: never cut budget on an ad group that is comfortably beating target just because a single day looked odd
The rule we run ourselves
Rule: protect CPA on scaling ad groups
A pacing-protective rule with guardrails and a dry run — illustrative sample data.
Note the conditions that are not about performance at all: a minimum spend so the rule cannot react to noise, a check that the ad group is out of the learning phase, a cap of one change per ad group per day, and a protected class of ad groups that rules may never touch. Those four guardrails prevent most of the damage that automation does to accounts.
What to check tomorrow morning
- Which ad groups spent more than 60% of their daily budget before midday?
- Which ad groups are budget-capped while beating target CPA?
- In each CBO campaign, what share of spend went to the single largest ad group?
- Where do hourly spend and hourly conversions disagree most?
Four questions, and in most accounts at least one of them has an expensive answer.
In Freshlytics: hourly pacing, projected end-of-day spend, budget headroom and reallocation suggestions are computed for every connected ad account, with alerts when an ad group breaks its own pattern. Request beta access.