← All articles
Playbooks·11 min read·

The Meta Ads Audit Checklist: 12 Ways You Are Wasting Budget, and How to Fix Each

A senior media buyer's audit of Meta ad accounts. The 12 signals that quietly drain budget, the exact metric that reveals each one, and how to fix it.

Most Meta ad accounts are not broken. They are leaking. A campaign that looked healthy at launch slowly starts paying more for the same result, budget drifts toward the wrong ad set, a creative quietly stops working, and nobody notices until the monthly number comes in soft.

An audit is how you find those leaks before they compound. This is the checklist we run, in the order we run it. For each signal you get three things: what it is, the exact metric that reveals it, and how to fix it. No theory, no fluff.

A note before you start: pull at least the last 30 to 90 days of data, and remember that the trailing two or three days are still estimated and will move. Audit finalized data, act on trends, not on a single noisy day.

Part 1: Spend leaks

These are the fastest wins. Money is going somewhere it should not.

1. Budget trapped in a losing ad set

What it is. Inside one campaign, most of the spend sits on the ad set with the worst return, while a better ad set is starved.

How to spot it. At the ad set level, compare each ad set's share of campaign spend against its share of results. Sort by return on ad spend. If your highest-spending ad set is near the bottom on ROAS, you have a misallocation. Under Advantage Campaign Budget (formerly CBO), this should self-correct, so if it has not, something is forcing spend the wrong way, usually audience fragmentation or a bid setting.

How to fix it. Move budget out of the weak ad set in steps of roughly 20 percent so you do not reset the learning phase, or pause it and consolidate. If you run campaign budget optimization, fix the underlying cause: merge overlapping audiences so the algorithm stops splitting delivery.

2. Winners starved of budget

What it is. The mirror image of the last point. An ad set is quietly your best performer but only gets a sliver of the budget.

How to spot it. Look for ad sets with return well above the account average and a small budget share. That is unrealized profit.

How to fix it. Scale it gradually, about 20 percent every two to three days, and watch that the return holds as volume grows. Scaling too fast pushes it back into learning and the efficiency drops.

3. Cost per purchase outliers

What it is. One ad set or campaign pays far more per conversion than the rest of the account, dragging the blended cost up.

How to spot it. Compute the account median cost per purchase, then flag anything paying more than about 1.75 times that median with a meaningful number of conversions behind it. A single expensive conversion is noise; a pattern is a leak.

How to fix it. Compare the outlier's audience, placements, and creative against your efficient ad sets. Cap the cost with a Cost Cap bid if volume allows, or reallocate the budget toward ad sets near the median.

4. Placements that burn money

What it is. Spend flows to placements that do not convert, which is common with fully automatic placements and a weak creative set.

How to spot it. Break results down by publisher platform and platform position. Look for placements taking real spend with a cost per result well above the account norm.

How to fix it. You rarely want to hard-exclude placements, since that shrinks the auction. Instead, make sure you have the right creative format for each placement, and only trim a placement when it consistently spends without returning.

Part 2: Fatigue and decay

These leaks are about time. Something that used to work is wearing out.

5. Audience fatigue

What it is. The same people see your ads too many times. Costs rise, response falls.

How to spot it. Frequency is impressions divided by reach. On a prospecting ad set, a cumulative frequency climbing past 3 over a short window, paired with a falling return, is the classic fatigue signature. One important trap: frequency is not additive, so you cannot sum daily frequency to get a period figure. Read it over the whole window you care about.

How to fix it. Refresh the creative with new angles, expand or swap the audience, and exclude recent purchasers. For a fatigued retargeting pool, widen the entry window or cap frequency.

6. Creative decay

What it is. A specific ad has passed its peak. It still spends, but the click rate slides and the cost creeps up.

How to spot it. At the ad level, chart the link click-through rate over time. A steady decline alongside a rising cost per thousand impressions means the creative is tiring. Link click rate is a cleaner signal than the all-clicks rate because it reflects genuine interest.

How to fix it. Rotate in fresh creative built on a different hook. Do not just tweak colors; change the angle, the opening frame, or the offer framing. Retire the ad once a replacement is proven.

7. Quality and ranking drops

What it is. Meta scores each ad's quality, expected engagement, and expected conversion rate against competitors. A low ranking means you pay more to win the same impression.

How to spot it. At the ad level, read the three ad relevance diagnostics. Anything in the below-average bands, on an ad with real spend, is costing you in the auction. Note these only appear once an ad has enough impressions.

How to fix it. For low quality, remove clickbait, sensational text, and heavy image text, and lead with a clearer value proposition. For a low conversion-rate ranking, tighten the match between the ad promise and the landing page.

8. ROAS decay across the window

What it is. The whole campaign is trending down, even if no single day looks alarming.

How to spot it. Compare the average return in the first two weeks of your window against the most recent two. A drop of more than about 25 percent is a real decay curve, not noise.

How to fix it. Diagnose the driver. Rising frequency points to fatigue. A falling click rate points to creative. A stable click rate with worse conversions points to the landing page or the offer. Fix the weakest layer first.

Part 3: Structure and delivery

These are the quiet structural problems that cap performance.

9. Ad sets stuck in the learning phase

What it is. An ad set never gathers enough optimization events to deliver stably, so it stays expensive and volatile.

How to spot it. Meta reports a learning status on active ad sets. A failed learning state, or an ad set that keeps re-entering learning, is the flag. The rule of thumb is roughly 50 optimization events within about a week of the last significant edit.

How to fix it. Consolidate similar ad sets to pool conversions, raise the budget, or switch to a higher-volume optimization event. Above all, stop making significant edits, since each one restarts learning.

10. Disapproved or limited ads

What it is. An ad was rejected or limited in review. It cannot spend, and it can drag down the ad set's ability to exit learning.

How to spot it. Filter by effective status for disapproved and with-issues states. Read the specific policy reason attached to each.

How to fix it. Fix the offending element, usually text in the image or a restricted claim, and resubmit. If it keeps failing, duplicate it with a compliant creative rather than fighting the same ad.

11. Overlapping audiences competing with you

What it is. Two of your own active ad sets target the same people, so you bid against yourself. That raises costs and splits the data each ad set needs to learn.

How to spot it. Meta does not expose auction overlap through its API, so this one is inferred: look for active ad sets with matching or heavily overlapping targeting running at the same time.

How to fix it. Consolidate them into one, or add mutual exclusions so each reaches a distinct slice. Fewer, larger ad sets usually learn faster and cost less.

12. Fragmented testing that never converges

What it is. Too many tiny ad sets, each with a small budget, none of which ever gathers enough data to prove anything.

How to spot it. A campaign with many ad sets, each spending very little and each stuck in learning, is the pattern.

How to fix it. Consolidate. Give a smaller number of ad sets enough budget to exit learning and produce a clear signal, then scale the winners.

Turn the audit into a habit

The reason accounts leak is not that media buyers are careless. It is that a full audit is tedious, and by the time you finish one, the account has already changed. The signals above are the ones worth checking, but checking them once a quarter is not enough. Fatigue and decay move week to week.

That is exactly why we built FixAds. You connect your Meta account with read only access, and it runs this entire checklist automatically, quantifies the money at stake for each issue, and tells you the exact fix. It does not touch your budgets or creatives; every change stays in your hands.

If you want to see where your account is leaking right now, connect it and get your first finding free.

See where your budget is leaking

Connect your Meta account and get a full diagnostic in minutes. Read only, your first finding is free.

Connect Meta, free