Why 80% of GA4 implementations are misconfigured
GA4 is not GA Universal. The paradigm shift—from sessions to events—brought with it a wave of rushed implementations where events are duplicated, conversions miscounted, or ecommerce data simply missing.
A GA4 audit is not a technical luxury. It's the difference between making decisions on real data or on statistical noise.
GA4 audit checklist: the 8 critical points
This is the checklist that we apply before any performance analysis for a client. If you detect errors in more than 3 points, your current data is not reliable for investment decisions.
- 1. Does the GA4 tag fire exactly once per page? (check in DebugView without duplicates)
- 2. Do the conversion events (purchase, generate_lead, etc.) have the correct parameters? (currency, value, transaction_id)
- 3. Is your device's internal traffic excluded by IP or network range?
- 4. Do custom events follow GA4's snake_case convention?
- 5. Do sessions with duration 0 represent less than 15% of the total? (if it is higher, there is a filter or bot traffic problem)
- 6. Do GA4 conversions match ±10% with those from the CRM or payment gateway?
- 7. Is enhanced measurement configured correctly for your site type? (avoid false scroll events in SPAs)
- 8. Are cross domains (subdomain, external checkout) configured for session maintenance?
How to identify duplicate events in GA4
The most common mistake: firing the same event from GTM and from hardcoded code simultaneously. The result is conversion metrics double what is real.
To detect it: open the GA4 DebugView, navigate through the purchase or lead flow and count how many times the same event appears per user action. If it appears two or more times, you have duplicates.
The solution depends on the source of the duplicate event. If it's GTM + code, remove one. If there are two tags in GTM, review the trigger conditions and consolidate.
Does this apply to your company?
In 30 minutes we'll tell you if we have a solution for your case.
Validate conversions: the junction that matters most
The conversion metric in GA4 only has value if it reasonably matches the source of truth for your business: the CRM, payment gateway or order database.
A simple cross-reference: take GA4 conversions from the last month and compare them to actual orders from the same period. If the difference exceeds 20%, you have a measurement problem that invalidates any ROAS or CPA analysis.
| Difference GA4 vs reality | Probable diagnosis | Action |
|---|---|---|
| < 10% | Correct implementation | Monitor monthly |
| 10% – 30% | Data loss (ad blockers, SSR, cookieless) | Implement server-side tagging |
| > 30% | Critical configuration error | Complete audit + urgent fix |
| GA4 > reality (>10%) | Duplicate events | Remove duplicates in GTM/code |
The traffic problem (not set) and how to solve it
If a significant portion of your traffic appears as "(not set)" in the channel or source, you are losing valuable attribution information. The most common causes are: malformed UTM parameters, redirects that eliminate UTMs, or unidentified direct traffic.
The underlying solution is to implement custom Channel Grouping in GA4 that maps your specific traffic sources. At Alaz we use this to correctly separate traffic from WhatsApp, email prospecting and referrals.
If you need help implementing or auditing GA4 in your company, our team technology consulting can make the diagnosis in 48 hours.
When to do a GA4 audit
Don't wait until the data is obviously wrong. These are the times when an audit is mandatory:
- Before making Google Ads or Meta investment decisions based on ROAS
- When you migrate the site, change the CMS or redesign the checkout
- When you detect unexplained drops in conversions without an equivalent drop in traffic
- When you incorporate a new channel (WhatsApp, email, affiliates) and you need to measure its real impact
- Every 6 months as preventive maintenance
