The findings below are representative of what we see in a typical ecommerce account. Names and numbers are illustrative. Your report will carry your data, your screenshots and your recoverable spend.
Section 1
Overall and per layer, out of 100. The one number to track after fixes.
Section 2
Half a page for the founder or CMO: what is broken, what it costs, what to do first.
Section 3
Every issue with evidence, impact and fix, ranked critical to medium.
Section 4
How much of your budget is being optimized on wrong signals, and how much a fix recovers.
Section 5
Ordered by impact, with effort estimates and who can do it.
Each check is weighted by how much spend it touches. A duplicated purchase event weighs more than a missing referral exclusion. The score is comparable across audits, so you know where you stand and how far a fix moves you.
0-49
Optimizing blind
50-79
Leaking budget
80-100
Trustworthy
Overall: 41 / 100
21 of 67 checks failed. Estimated 22% of monthly spend optimized on wrong signals.
| Layer | Score | Failed |
|---|---|---|
| Meta Pixel + CAPI | 28 | 5 of 12 |
| Google Ads | 55 | 4 of 12 |
| GA4 + GTM | 50 | 5 of 14 |
| CRM + offline | 20 | 4 of 10 |
| Consent Mode v2 | 100 | 0 of 9 |
| Ecommerce platform | 70 | 3 of 10 |
Where it is, how we know, what it costs, how to fix it. Nothing a developer has to guess at.
Evidence
Events Manager shows 2 purchase events per order; deduplication rate 0%.
Impact
Reported ROAS roughly doubled. Campaign budgets scaled on conversions that did not happen.
Fix
Generate one event_id per order on the server and pass it to both the pixel and the CAPI call.
Evidence
DebugView shows repeated purchase with the same transaction_id on reload; 214% revenue vs store.
Impact
GA4 and the Google Ads import overstate revenue; tROAS bidding is chasing a number that does not exist.
Fix
Fire purchase only on first load of the confirmation page and pass transaction_id so GA4 can deduplicate.
Evidence
Conversion action diagnostics: 'Enhanced conversions not set up'.
Impact
Conversions lost to cookie restrictions are never recovered; Smart Bidding is under-informed on Safari and iOS.
Fix
Enable Enhanced Conversions and pass hashed email and phone from the confirmation page via GTM.
Evidence
Sample of 200 leads: 0 have click ids stored; UTMs present on 31%.
Impact
Won deals cannot be uploaded as offline conversions. Platforms optimize for cheap leads, not paying customers.
Fix
Capture click ids and UTMs in hidden form fields and store them on the lead. Then send Won events back with value.
Evidence
Razorpay appears as a top 'referral' source for purchases.
Impact
Attribution for a share of paid purchases is credited to the payment gateway instead of the campaign.
Fix
Add the gateway domains to the referral exclusion list and confirm cross-domain settings.
Evidence
Default denied in EEA, update fires on choice, tags react. Modeling active in Google Ads.
Impact
No action needed.
Fix
Keep the CMP updated and re-test after template changes.
Ordered by recoverable spend, not by how interesting the bug is. Most accounts recover the majority of the leak in the first two weeks of fixes.
| # | Fix | Effort | Recovers |
|---|---|---|---|
| 1 | Shared event_id for pixel + CAPI purchase | 3 h | ~9% of spend |
| 2 | Fire GA4 purchase once, pass transaction_id | 2 h | ~6% of spend |
| 3 | Capture click ids + UTMs on every lead, upload Won as offline conversions | 1 to 2 days | ~4% of spend |
| 4 | Enable Enhanced Conversions with hashed email and phone | 3 h | ~2% of spend |
| 5 | Referral exclusions for payment gateways | 30 min | Attribution accuracy |
We never change anything during an audit, so viewer-level access is all it takes. We send exact instructions after you submit the request.
Read permission on the web (and server) container.
Viewer on the property.
Analyst on the pixel or dataset, and ad account read access.
Read-only access to the account.
A recent orders or leads export, or a 15-minute screen-share.
Nothing needed. We test it from the front end.
Request the free audit, grant read-only access, and get a health score, ranked findings and a roadmap you can act on immediately.
Free. No obligation. You keep the report either way.