Learning Limited means an ad set is unlikely to reach roughly 50 optimisation events in a week, so fix it only if results are poor, starting by combining ad sets or widening the audience before you raise the budget.
An ad set's delivery column changes from Learning to Learning Limited, and someone asks whether the campaign is broken. Usually it is not. The status says something narrower: Meta does not expect this ad set to collect enough results to finish learning.
What the status means
A new or heavily edited ad set enters the learning phase while Meta works out who responds. It generally leaves after about 50 optimisation events in a week. An optimisation event is whatever the ad set is set to get, such as a lead or a purchase.
Learning Limited appears when Meta predicts the ad set will not get there. The ads keep running. It is a warning about how much data the ad set is getting, not an error and not a penalty.
Learning Limited is a statement about volume. It says nothing yet about whether the leads are good.
When to leave it alone
Check the results before you touch anything. A small local business may never reach 50 leads a week, and it may not need to. If the cost per lead is steady and the sales team is happy with the leads, the label alone is no reason to change the ad set. Editing it can restart learning and do more harm than the status did.
What causes it
- Too many ad sets splitting a small budget, so none of them gets enough results.
- An audience too small to produce the volume.
- A budget too low for the cost of each result.
- A cost control or bid cap set below what results really cost.
- An optimisation event that happens rarely, such as a purchase on a low-traffic site.
A quick budget check
Multiply your usual cost per result by 50, then divide by 7. Illustrative example: at ₹300 a lead, 50 leads a week costs ₹15,000, or about ₹2,150 a day. An ad set on ₹500 a day was never going to reach that, and it would be unfair to blame the creative.
Fixes to try, in order
- Combine ad sets that target similar people, so one ad set gets all the results.
- Widen the audience or remove narrow targeting.
- Loosen or remove a cost control that is holding back delivery.
- Raise the budget if the check above shows it is short, in steps rather than one big jump.
- Optimise for a more frequent event, such as leads instead of a later stage, and judge lead quality separately.
Make one change and give it time. Several edits at once can restart learning repeatedly and leave you unable to tell which one helped.