How to use the Meta ad set structure checker
The Meta ad set structure checker needs one export at ad set level and the number of days it covers. It reads spend and results from the export only, so it never needs access to your ad account. Run it again after you merge ad sets, once the new structure has had a full week to deliver, and compare the two scores.
- In Ads Manager: open the Ad sets tab with the columns Campaign name, Ad set name, Delivery, Amount spent and Results, the event each ad set optimizes for.
- Export the table data: then paste it or upload the file. “Load the example” shows made-up rows first.
- Enter the days in the export: results are scaled to a week, the window Meta’s learning phase uses.
- Read the result: a structure score, the spend in ad sets that can’t learn, and how many ad sets to merge away.

Under the result come three fixes to start with. For the example they are: merge ad sets that split the same audience, use a campaign budget, and go broader on targeting.

On a phone, the score sits under the export box.

To plan a budget that can feed every ad set you keep, use the ad budget calculator.
Why structure matters
Meta says an ad set generally needs about 50 optimization events within 7 days to exit the learning phase, when delivery is still unstable (Meta Business Help Center). Every ad set learns separately. Split the same budget across too many ad sets and none of them gets enough results.
How it works
- Each ad set’s results are scaled to a week (
results × 7 ÷ days in the export). 50 or more is “enough to learn”, 25–49 “nearly there”, under 25 “starved”. - Per campaign, the weekly results ÷ 50 is the number of ad sets it can actually feed (at least one). More ad sets than that, and the tool suggests merging down to that number.
- The score is
100 − 60 × share of spend in starved ad sets − 20 × share in nearly-there ones − 20 × share of campaigns with too many ad sets. 80+ is healthy, 50–79 too fragmented, under 50 starved. The weighting is ours.
If the whole account gets fewer than 50 results a week on its event, even a single ad set can’t exit learning, and the first fix is to optimize for a more frequent event until volume grows.
Getting the export
In Ads Manager, open the Ad sets tab for the last 7, 14 or 30 days, include the columns Campaign name, Ad set name, Delivery, Amount spent and Results, then export the table. Enter the number of days the export covers. “Results” must be the event each ad set optimizes for.
Limits
Results are counted per ad set as reported, without a breakdown by day, so a big change in the middle of the period isn’t visible. Ad sets with very different audiences or offers can have good reasons to stay separate; merging is for ad sets that split the same people.