Budgets: who decides where the money goes
Week 1 gave you the foundation. Week 2 goes deep on the controls behind the three levers — five lessons, five settings clusters. First: budget - and the single biggest structural choice, who allocates it across your ad sets, you or the algorithm.
Daily vs lifetime is a detail. The real choice is where the budget sits - at the campaign level (algorithm allocates) or the ad-set level (you allocate). That single switch shapes how the whole campaign behaves.
1The big one: CBO vs ABO
2Daily vs Lifetime budget
Now the small decision, cleared in 60 seconds.
- Daily Meta can spend up to 75% more than your daily budget on a high-opportunity day, then balances it out so a calendar week never exceeds 7× your daily budget. Best default: simple, continuous, easy to scale. (Meta's documented daily-budget pacing rule, as of mid-2026 — check current limits in Help Center.)
- Lifetime a total for the whole run; Meta paces it across the dates. Needed for scheduling/dayparting, but less flexible day to day.
For most always-on acquisition, use daily. Use lifetime when you have a fixed total and a fixed window (a launch, a sale).
3When to use which
The modern default leans heavily CBO/Advantage+ - it aligns with the "fewer, broader, let-AI-decide" direction of the whole platform. Why? The system reallocates in real time on auction-level conversion signals you never see in your reports; you reallocate once a day at best, using yesterday's data. Connect it back: CBO helps an ad set cross the ~50-event threshold (Day 5) by concentrating spend, instead of three ad sets each stuck at 20.
Worked example · illustrative, your results will vary. Same €60/day, 3 ad sets:
| Setup | How the €60 splits | What happens |
|---|---|---|
| ABO | €20 each, fixed | ~20 events/week each → all three stuck Learning Limited |
| CBO | algorithm shifts ~€40 to the leader | leader crosses ~50 events → exits learning, CPA stabilizes |
ABO = you hand each rep an identical, fixed stack of leads and refuse to change it, even if one is clearly closing more. Fair test, but you knowingly waste leads on weaker reps. CBO = you funnel more leads to whoever's hottest today, automatically. More deals closed, less fairness - exactly right once you're past testing and into scaling.
The toggle is a switch on the campaign setup screen - flipping it later is a significant edit that can reset learning, so decide up front.
(1) ABO with tiny per-ad-set budgets across many ad sets - the fastest route to Learning Limited everywhere. (2) Editing CBO budgets in big jumps, resetting learning each time. (3) Believing more control = better results. Against the algorithm, your manual allocation is usually worse. Control is for testing fairness, not for out-spending the machine.
Open Ads Manager and build a draft campaign so you see exactly where budget control lives. Nothing below spends a cent - just don't press Publish.
Recap - 30 seconds
- Daily budget for always-on; Lifetime for fixed window/scheduling.
- CBO = one budget at campaign level, algorithm allocates. Best for scaling.
- ABO = separate budget per ad set, you allocate. Best for fair testing.
- CBO helps ad sets cross the ~50-event threshold by concentrating spend.
- More manual control usually means worse allocation - reserve ABO for when test fairness matters.