| Check | Finding (before) | Fix | Status |
|---|---|---|---|
| Meta Pixel | Missing on the order confirmation page | Installed on every page | Fixed |
| Conversions API (CAPI) | Not connected | Server events connected, deduplication set up | Fixed |
| Purchase event value | Recorded as $0 | Real order value passed through | Fixed |
| Domain verification | Not completed | Verified | Fixed |
| Campaign structure | Traffic and conversion goals mixed in one campaign | Separate campaign per objective | Fixed |
| Ad set audience overlap | 38% overlap between ad sets | Exclusions added (now 9%) | Fixed |
| Exclude recent buyers | Not set | Buyers from the last 30 days excluded | Fixed |
| Google Ads conversion tracking | Duplicated with GA4 import, counting conversions twice | Consolidated to one primary conversion | Fixed |
| UTM rules | Different on every channel | Standard naming rules documented | Fixed |
| Budget allocation | 42% of budget on the weakest ad set | Reallocated by performance | Monitoring |
If conversions aren't tracked correctly, ad platforms learn from bad data. In this account, purchases were recorded at $0 and Google counted every conversion twice, so results looked better than they really were. The numbers have to be right before any optimization means anything.
This is the audit format I actually use when taking over an account. Detailed figures have been replaced with sample data to protect the client.