GA4 and Google Ads almost never agree on conversion counts. Most accounts see 10-30% variance, and the gap is rarely a tracking bug. It is usually a combination of attribution model differences, signal loss, timing windows, and deduplication issues. This post walks through the four real causes and how to close each one. The standard remediation is server-side tracking paired with offline conversion tracking.
- Typical variance
- 10-30% between GA4 and Google Ads in most accounts
- Biggest single cause
- Attribution model differences (GA4 last-non-direct vs Google Ads data-driven)
- Second biggest
- Signal loss from ad blockers, Safari ITP, and iOS attribution caps
- Standard fix
- Server-side tracking plus offline conversion tracking, paired
- Acceptable variance
- Under 10% after a clean tracking implementation
GA4 vs Google Ads at a Glance
Why GA4 conversions don't match Google Ads
Four causes account for nearly all GA4-to-Google-Ads variance. In order of typical impact:
01Attribution model differences
GA4 defaults to last-non-direct click attribution. Google Ads uses data-driven attribution. The two will not agree even with perfect tracking because they are answering different questions. GA4 credits the last non-direct channel; Google Ads credits the click that played a measurable role in the conversion path, weighted by its model. The variance is real but expected.
02Signal loss
Ad blockers strip the Google Ads pixel before it fires. Safari ITP truncates first-party cookies, breaking cross-session attribution. iOS attribution caps limit how long Meta and other in-app advertisers can credit conversions. Every dropped event is a conversion GA4 sees but Google Ads does not (or vice versa).
03Timing windows
Google Ads conversions are credited by click date. GA4 conversions are credited by conversion date. A conversion that happens today from an ad click two weeks ago lands in today's GA4 report and the two-week-old date in Google Ads. Over short reporting windows, the two will look misaligned even when nothing is broken.
04Deduplication issues
If the same event fires from both a web GTM tag and a server-side container without proper deduplication keys, conversions get double-counted. If deduplication is too aggressive, real conversions get suppressed. Either way, the platform-by-platform numbers stop matching.
What “normal” variance looks like
| Variance | Diagnosis | Likely Cause |
|---|---|---|
| Under 10% | Healthy | Attribution model + timing only |
| 10-20% | Acceptable | Mild signal loss, no tracking gaps |
| 20-30% | Investigate | Significant signal loss; server-side recommended |
| Over 30% | Broken | Tracking gap, deduplication bug, or both |
How to fix GA4 and Google Ads variance
The standard remediation runs in this order. Each step closes a specific category of variance:
- Reconcile. Pull 30 days of GA4 conversion data and 30 days of Google Ads conversion data. Match them by date. Categorize each gap into one of the four causes above. The size of the variance and its source determines the fix.
- Install server-side tracking. Move conversion events onto a first-party server endpoint so ad blockers, Safari ITP, and iOS attribution caps stop stripping signal. This typically recovers 20-35% of conversions.
- Add offline conversion tracking. Capture GCLID at lead time, store against the CRM contact, push closed-won back to Google Ads via Enhanced Conversions for Leads. Aligns Google Ads bidding with the same revenue events the CRM is recording.
- Align attribution models. If GA4 and Google Ads still disagree after signal recovery, switch GA4 to data-driven attribution to match Google Ads. The numbers will not match exactly but the gap shrinks to a single-digit percentage in most accounts.
Is some variance between GA4 and Google Ads normal?
Yes. GA4 and Google Ads use different attribution models, different timing conventions, and different signal sources. Even with perfect tracking, expect single-digit percentage variance. The benchmark for a clean implementation is under 10%. Variance above 15-30% usually means signal loss or a tracking gap that can be closed.
Should I trust GA4 or Google Ads for conversion reporting?
Trust the system that owns the conversion definition. For ecommerce, trust your store or CRM. For B2B, trust the CRM. GA4 and Google Ads are downstream reporting layers. They should both reconcile back to the same source of truth. If they disagree with that source by more than 10%, the tracking needs work.
Will server-side tracking alone fix the GA4 vs Google Ads mismatch?
It closes the signal-loss gap, which is usually the biggest single component of the variance. It does not fix attribution model differences or timing windows. Most accounts that install server-side tracking see GA4-to-Google-Ads variance drop from 20-30% to under 10%, but it will not fall to zero without also aligning attribution settings.
Ready to close the variance in your account?
15 minutes is enough to forecast how much of your GA4-to-Google Ads gap is signal loss versus model difference.