How Do You Build a Weekly Meta Ads Review?

A weekly review is the same short routine, run on the same day, that tells you whether to leave your campaigns alone or change one thing. It fits when campaigns are live and you are tempted to edit after every bad day. It does not fit a campaign that launched hours ago, where delivery and tracking need time before there is anything to judge.

Why review weekly and not daily?

Meta says day-to-day ROAS and cost fluctuate, and recommends weekly averages for its goal-based bid strategies. It also says ad sets are less stable during the learning phase and that editing during it resets learning. A weekly rhythm gives you a trend and fewer edits that restart learning.

What does the weekly routine look like?

1. Fix the view. Same ad account, same saved column preset, and a date range of the last full week next to the week before. Meta lists a date range that covers a time your ad was not running as one reason a table shows no data.
2. Check delivery. Read the Delivery column for campaigns, ad sets and ads.
3. Read the KPI. Look at your one primary number as a weekly average, under the same attribution setting as last week.
4. Scan the diagnostics. Find which of CPM, CTR, frequency or cost per click moved with the KPI.
5. Check the activity history. See what changed last week, and who or what changed it.
6. Decide and log. Choose at most one change, or none, and write down the date and the reason.

What should you check in the Delivery column?

Meta says a campaign, ad set and ad can show different statuses, for example an active campaign with an ad set in review, so check all three levels. Look for:

• Rejected: the ad cannot run because it does not comply with Meta’s advertising standards.
• Learning limited: the ad set did not generate enough results to exit the learning phase.
• Processing: the ad is being updated. Meta says this usually takes a few minutes, but some updates take up to 8 hours.
• Error or warning icons: Meta says an error means it cannot run at all and a warning means limited delivery. Hover over them for detail.

How should you treat ad sets in the learning phase?

Meta says to wait to edit an ad set until it is out of the learning phase. Ad sets usually exit after about 50 results in the week after the last significant edit. To see where an ad set stands, add the Last significant edit and Results columns.

For learning limited, Meta lists fixes to consider: combine ad sets and campaigns, expand the audience, raise the budget, raise a bid or cost control, or choose an optimization event that happens more often. Our suggestion is to pick one, not all.

Which diagnostics help, and which do not?

Meta says to use ad relevance diagnostics on underperforming ads, not to optimize ads already meeting your objectives, and that some high-performing ads rank below average. If an ad already meets your objectives, Meta says reviewing them may not be necessary.

How do you read a dash or a zero?

Meta says a dash means the metric does not apply to that ad, for example a video metric on an image ad. A zero means the metric applies but there are no results, which can also appear when the ad has not delivered, is awaiting approval or has a metrics delay.

How do you use the activity history?

Meta’s activity history shows who changed a campaign, ad set or ad, what changed and when, and it works on a computer only. Changes made by automated rules show the rule’s name. You can filter by budget, bid, audience, targeting, schedule or run status.

Check it before you explain a KPI move. An edit made mid-week may be the cause, and may have reset learning. Meta adds that people with an admin or advertiser role can make changes, so check permissions if you see an edit you do not recognize.

What do you record at the end?

Keep a short log: the KPI for the week, what moved, the decision and the date. Start next week’s review by checking whether last week’s decision did what you expected.

FAQ

Can I review daily instead?

Check delivery, spend and rejections daily if you like, because those need quick action. Meta says day-to-day ROAS and cost fluctuate, so leave performance decisions for the weekly read. Meta also says frequent budget changes, and edits that meaningfully change performance, can reset learning, so avoid edits without a clear reason.

What if a campaign is less than a week old?

Check delivery status and spend only. Meta says metrics can lag on a new ad and that ad sets usually exit the learning phase after about 50 results in the week after the last significant edit. Wait for a full week of data before you read the KPI or change anything.

Should I change something every week?

No. Meta says to edit only when you have reason to believe the edit should improve performance. A review that ends with “no change” is a valid result. Log it, and compare next week against the same KPI and view.