Most Meta ad accounts we take over aren't broken because of bad targeting or a weak product. They're broken because nobody has looked under the hood in months. Budgets get increased on gut feel, creative keeps running past the point of fatigue, and the tracking setup that was "good enough" at launch quietly stopped being accurate somewhere along the way.
This is the audit we run on every account before we touch a single dollar of budget — in the order we actually run it, so you can do the same thing in about 30 minutes with your own account open next to this page.
01Check Event Match Quality first
Before you look at a single campaign, open Events Manager → Data Sources → your pixel and check the Event Match Quality (EMQ) score on your Purchase event. This number tells you how much of your reported performance you can actually trust.
A score below 6 means Meta is working with incomplete customer data — email, phone, and other matching parameters aren't reaching the platform reliably. Everything downstream (ROAS, cost per purchase, lookalike audience quality) is only as good as this number.
- EMQ score is 7 or higher on your primary Purchase event
- Conversions API (CAPI) is connected alongside the browser pixel, not instead of it
- Deduplication between pixel and CAPI events is confirmed (check the "Deduplicated Events" column)
02Look for creative fatigue before blaming targeting
The single most common misdiagnosis we see: an account owner assumes their audience is "tapped out" when the real problem is that their top three creatives have been running unchanged for six or eight weeks. Frequency and CTR decline are almost always a creative signal before they're an audience signal.
| Signal | Healthy Range | Warning Sign |
|---|---|---|
| Frequency (cold audience, 7-day) | Under 2.0 | Above 2.5 |
| CTR trend (week over week) | Stable or rising | Declining 20%+ for 2+ weeks |
| Hook rate (3-sec views ÷ impressions) | 25%+ | Below 15% |
If frequency is climbing on a cold-audience ad set while CTR falls, the fix is new creative angles — not a bigger budget or a narrower audience. Refreshing the hook and first three seconds of the video usually resolves this faster than any targeting change.
03Check for audience overlap between ad sets
Open Ads Manager → Audiences → Overlap Tool and compare your active ad sets against each other. Meaningful overlap (above roughly 20-25%) means you're bidding against your own campaigns in the auction — Meta's system ends up splitting spend inefficiently between two ad sets chasing the same person.
This is especially common when an account has both a broad Advantage+ campaign and several manual interest-based ad sets running simultaneously. In the post-Andromeda era, the fix is usually consolidation — fewer, broader ad sets with distinct creative angles, rather than narrow audience stacking.
04Confirm your attribution window matches your sales cycle
Default attribution windows (typically 7-day click, 1-day view) work fine for impulse-purchase products under $50. They systematically undercount performance for considered purchases — anything with a multi-day decision process, higher price point, or B2B sales cycle.
Cross-reference Meta's reported conversions against your actual order data in Shopify, GA4, or your CRM for the same date range. A gap of more than 15-20% usually points to an attribution mismatch, not a tracking failure — which changes how you should read every ROAS number in the account going forward.
05Review the three settings most accounts get wrong
- Campaign budget optimization left on for a testing campaign — CBO redistributes spend toward whichever ad set is winning early, which quietly starves new creative of the impressions it needs to prove itself.
- Placements left fully automatic without a placement-level performance check — Audience Network and some Facebook placements can drag down blended CPA for certain products even when the campaign-level number looks fine.
- No exclusion audience for recent purchasers — without this, you're paying to re-acquire someone who already bought in the last purchase cycle, inflating both spend and frequency for no incremental revenue.
06What to do with what you find
Not every issue above needs fixing today. Prioritize in this order: tracking accuracy first (EMQ, attribution), because every other number in the account depends on it being right. Then creative fatigue, because it's usually the fastest lever to pull. Structural issues — overlap, CBO, exclusions — come last, since they require more careful testing to change without disrupting an account's existing learning phase.
An audit like this doesn't replace ongoing management, but it tells you within half an hour whether an account's underperformance is a strategy problem or a hygiene problem — and those two things get fixed in completely different ways.