Meta Ads Audit Checklist: The 12 Leaks We Find Most Often

A Meta ads audit should check twelve things: creative fatigue, audience overlap, pixel/CAPI signal loss, budget fragmentation, frequency burnout, broad-vs-interest waste, landing page mismatch, dayparting waste, placement waste, stale negative audiences, attribution double-counting, and unspent learning-phase churn. Most accounts we look at are actively leaking on four or five of these at once — not one dramatic mistake, a dozen small ones compounding.

What actually counts as a "leak" in a Meta ads account?

A leak is spend that's technically "working" — the campaign is live, it's spending budget, Ads Manager shows a plausible ROAS — but isn't converting as efficiently as the same budget would if the account weren't fighting itself. Leaks rarely show up as an obvious red number. They show up as a ROAS that's fine instead of good, and they compound quietly because nothing in Ads Manager flags "this campaign is competing with your other campaign for the same buyer."

The 12 leaks we find most often

1. Creative fatigue. The same creative running past its useful life. CTR erodes, frequency climbs, and the algorithm compensates by paying more per result — you're not seeing a spend problem, you're seeing a creative-age problem wearing a spend costume.
2. Audience overlap. Two or more ad sets targeting overlapping audiences, forcing Meta's delivery system to auction against your own budget. Nobody set out to do this; it accumulates as campaigns get added over time without anyone checking overlap.
3. Pixel/CAPI signal loss. Broken or partial server-side event coverage (iOS 14.5+ made this structural, not optional) means the algorithm is optimizing on incomplete data — it can't find more of your best buyers if it can't reliably see who they were.
4. Budget fragmentation. Budget split across too many ad sets, each one starved below the volume it needs to exit learning and optimize properly. Consolidation, not more budget, is usually the fix.
5. Frequency burnout. The same person seeing the same ad too many times in a short window. Diminishing returns kick in well before most accounts refresh creative or expand audiences.
6. Broad-vs-interest waste. Running narrow interest targeting alongside Advantage+ broad campaigns for the same product, when broad — given enough signal — usually outperforms manual interest stacks today.
7. Landing page mismatch. The ad promises one thing, the landing page delivers another (different offer, different price, slower load, no mobile optimization). Meta can hand you a perfect click and still lose the sale in the two seconds after it lands.
8. Dayparting waste. Spending flat across 24 hours when conversion rate by hour is anything but flat — most accounts never look at hour-of-day performance at all.
9. Placement waste. Automatic placements quietly burning budget on a placement (certain Audience Network or low-intent Reels slots) with a materially worse conversion rate than the account average, unexamined because "Advantage+ placements" is the default and nobody checked the breakdown.
10. Stale negative audiences. Exclusion lists (past purchasers, existing customers) that were never updated, either suppressing people who should be back in the funnel or, worse, missing entirely and letting retargeting spend chase people who already converted.
11. Attribution double-counting. The same conversion getting credited across two campaigns or two platforms because attribution windows overlap, inflating reported ROAS and hiding which spend actually drove the result.
12. Unspent learning-phase churn. Ad sets that never leave the learning phase because they're constantly edited, paused, or under-funded — every restart resets the clock, and the algorithm never gets the volume it needs to stabilize.

How do you actually run through this list without a tool?

Pull three reports before you touch a single campaign: the account-level breakdown by ad set for the last 30 days, the frequency and reach breakdown for the same window, and your pixel/CAPI event-match quality score from Events Manager. Most of the 12 leaks show up in the gaps between those three reports, not inside any single one of them — a campaign that looks fine in isolation can be quietly cannibalizing another campaign's audience, and you only see that by cross-referencing overlap, not by reading either campaign's dashboard alone. Budget half your audit time on the account-structure leaks (fragmentation, overlap, learning-phase churn) and the other half on execution leaks (creative fatigue, landing mismatch, dayparting) — the two categories fail independently, and fixing one doesn't fix the other.

Which of these should you check first?

Start with the ones that are structural, not tactical: pixel/CAPI signal loss and budget fragmentation. Both quietly cap every other lever — a fragmented budget can't generate the signal volume the algorithm needs, and a broken pixel means whatever signal does exist is wrong. Fix those two before touching creative or targeting, because optimizing on top of bad signal just optimizes the wrong thing faster.

How often should you run this audit?

Monthly, at minimum, and after any account-level change — new campaign structure, iOS/tracking update, creative refresh cadence change. Leaks like frequency burnout and creative fatigue develop on a scale of weeks, not quarters, so a quarterly audit cadence means you're always finding problems after they've already cost real budget. Pair a Meta audit with a look at your AI visibility too — paid waste and AI invisibility are usually the two biggest, least-visible drains on the same growth budget.