ChatGPT 
Claude 
Google Ads and GA4 report different conversions because Google Ads credits its own ad clicks and reports by click date, while GA4 uses cross-channel data-driven attribution and reports by event date. Consent Mode v2 adds a third difference: Google Ads can model conversions GA4 never records. A steady gap is normal. A growing one means something broke.
- 1Attribution modelsWho gets the credit for the same sale.
- 2Windows and datesWhen the credit lands in each report.
- 3Consent and modelingWhat each tool can see after Reject.
How do Google Ads and GA4 count a conversion?
The same sale can show up in different places and on different days, because each tool follows its own counting rules. Put them side by side before you call the gap a tracking fault.
- Credits
- Its own ad clicks, and eligible views
- Reports by
- The date of the ad click
- Window
- A conversion window per action, often 30 days
- After Reject
- Can model conversions when Consent Mode v2 sends the signals
- Credits
- Every channel, data-driven attribution by default
- Reports by
- The date of the key event
- Window
- Lookback windows set in Admin, Attribution settings
- After Reject
- No user-level record; modeled only if the property qualifies
If the gap is steady, it is usually a counting difference. If it grows after a site release, something is broken.
1. Attribution models: who gets the credit
Google Ads only sees its own clicks, so it can claim a conversion that GA4 gives partly to email or organic search. GA4 shares the credit across every channel the visitor touched, using data-driven attribution by default. Compare both using the same model before you treat the gap as a tracking fault.
2. Conversion windows and dates: when the credit lands
Google Ads uses a conversion window, often 30 days, and reports each conversion on the date of the click. GA4 reports by the date of the event. A campaign that ended last week can still gain conversions in Google Ads, while GA4 shows them in this week's numbers.
3. Consent and modeling: what each tool can see
When a visitor rejects cookies, GA4 may not record the visit at all, while Google Ads can still model the conversion if Consent Mode v2 sends the right signals. Advanced Consent Mode keeps that modeling possible; basic mode does not. The higher your denial rate, the wider this part of the gap, so review your cookie consent and Consent Mode v2 setup before you blame the tags.
Is your gap steady or growing?
The shape of the gap over time tells you which problem you have. A parallel gap is the three counting differences above. A gap that opens after a release is a break.
How to test yours
Four checks separate a counting difference from broken tracking. Run them in this order.
- Compare one campaign over a closed date range, with the same attribution model in both tools.
- Check that the conversion action in Google Ads imports the GA4 key event, or uses its own tag, but not both as primary.
- Open Tag Assistant on a fresh session and confirm the Consent Mode v2 default state loads before the first Google tag fires.
- Reconcile both against your CRM. Measured across Waftr client audits, 28-41% of CRM leads arrive with no GA4 client ID, so neither tool sees the whole picture.
If the four checks point to a break, a GA4 audit traces which tag, key event or consent setting changed in the release and ranks the fixes by revenue impact.