Open Google Ads and look at your conversions for last month. Now open Google Analytics 4 (GA4) and look at the exact same conversion event for the exact same period. The numbers do not match. In fact, they are almost never even close, and Google Ads frequently reports 15% to 40% more conversions than GA4.
This widespread discrepancy causes endless boardroom friction, but it is not a bug. Both platforms are working exactly as they were engineered to. They simply see the internet through entirely different mathematical lenses.
The fundamental difference comes down to which interactions each platform is even allowed to look at.
Google Ads asks: 'Did my ads play a role in this conversion?' It only ever measures interactions with Google ads. If a user clicks an ad, browses your site, leaves, and returns via organic search three days later to buy, Google Ads still attributes that conversion to the ad click, because the click happened within its conversion window and no competing Google ad interaction took the credit.
GA4 asks: 'What was the complete multi-channel journey, and how should credit be distributed?' By default, GA4 uses cross-channel Data-Driven Attribution (DDA). It looks at organic search, social media, direct traffic, and email alongside paid. In that exact same journey, GA4 slices up the conversion credit, giving a fraction to the Google ad and a fraction to organic search. When you look at a channel report in GA4, you only see the fraction assigned to paid search, instantly producing a lower number than Google Ads.
It is worth noting that Google Ads also uses data-driven attribution today, but only across Google ad interactions. The key divide is scope: Google Ads measures its own clicks, while GA4 weighs every channel against each other.
This is the single most overlooked reason for rolling data discrepancies, especially when looking at shorter timeframes like the past week.
Google Ads stamps a conversion on the date and time of the ad click. GA4 stamps the conversion on the date and time the purchase actually occurred.
Imagine a user clicks your Google ad on a Monday, adds an item to their cart, but does not buy until Thursday. Google Ads backdates the conversion to Monday. GA4 records it on Thursday. If you pull a report on Tuesday, Google Ads will show a conversion while GA4 will show zero. The numbers drift closer as the attribution window closes, but they will never perfectly align.
With Consent Mode v2 and the strict data controls in force across the EU, privacy rejections routinely block traditional tracking cookies. To combat this, both platforms use machine learning to fill in the blanks, but they do it independently and to different degrees.
When a user denies consent, Google Ads relies on Advanced Consent Mode to read anonymous 'cookieless pings' and model the likelihood of a conversion, keeping its bidding algorithms fed. GA4 behavioral modeling is more conservative, requiring data volume thresholds (historically around 1,000 daily users meeting consent criteria over a sustained period) before it will model missing events. Because Google Ads models more readily while GA4 often leaves a blank space, the gap between the two widens in privacy-regulated markets.
The platforms handle user identity differently. Google Ads leverages its vast network of signed-in Google Accounts, via features like Enhanced Conversions, to link a mobile ad click to a desktop purchase. GA4 has cross-device capability too, but it relies on your site's User-ID setup or device-level signals, which are more vulnerable to browser tracking blocks.
Their counting logic also varies by default settings. Google Ads can be configured to count 'Every' conversion (if a user places three separate orders after one click, it records three) or just 'One' per click, which suits lead generation. GA4 counts occurrences of the conversion event but resolves credit through its cross-channel model, so individual touchpoints are calculated differently depending on how your conversion events are scoped.
A baseline variance of 10% to 20% between these platforms is normal. It reflects structural differences in the math, not broken code. Use Google Ads numbers to optimise your bidding and ad budgets, and use GA4 to evaluate your holistic multi-channel marketing mix.
However, if your numbers are off by more than 30%, stop blaming attribution models. A divergence that large usually means your tracking infrastructure needs a comprehensive review, likely due to broken cross-domain parameters, duplicate tag firing, or a misconfigured consent management platform blocking your data tags.
Neither is 'more correct'. Use Google Ads conversions to optimise bids and budgets within your paid campaigns, and use GA4 to understand the full multi-channel journey. They answer different questions.
Google Ads attributes the full conversion to its own ad click and backdates it to the click date. GA4 splits credit across every channel and reports on the date the conversion happened, so paid search ends up with a smaller share.
A 10% to 20% variance is expected and healthy. If the two platforms differ by more than 30%, it usually points to a tracking issue rather than attribution differences.
No. Because the platforms use different attribution models, counting methods, dating logic, and modeling, perfect alignment is not possible by design.