Google Ads conversions not matching GA4
Some difference is expected, because the two systems are built to answer different questions and count differently by design. The more useful question is not which dashboard to believe, but whether either number matches what your business actually received.
Five reasons the numbers differ even when nothing is wrong
Work through these before assuming a fault. If your gap can be accounted for by the rows below, you probably do not need anybody to audit anything.
How much difference is normal?
There is no honest universal percentage, and anybody offering one is guessing. What separates an ordinary difference from a real fault is whether the gap can be explained by the known differences in how the two systems count.
The signals worth investigating are more specific than size. A gap that changes suddenly without a corresponding change in the account. A gap running in the opposite direction to what the attribution models would predict. Or the two systems disagreeing about which campaigns produced the results, rather than merely how many.
That last one is the most informative and the least examined. Two systems that differ by twenty per cent but agree on the ranking of your campaigns are telling you roughly the same story. Two systems that agree on the total and disagree about which campaigns earned it are not.
What this cannot tell you Reconciling a specific gap needs access to both properties and their configuration. An external check can see whether campaign information survives the journey through the site, which is one common cause, but not how either system is configured.
If Google Ads and GA4 agree, does that mean both are right?
No, and this is the assumption most worth abandoning. Two systems reading the same incorrect signal will agree with each other precisely.
If a page view is being counted as a conversion, both platforms will report it. Their agreement is evidence that they are receiving the same data, not evidence that the data means anything.
This is why reconciling dashboards against each other is a weaker check than most people assume. The only reconciliation that establishes anything is against your own records: enquiries received, calls answered, appointments booked, invoices raised.
Which number should I use for reporting?
Whichever one has been checked against something that happened in your business. That sounds like an evasion and it is the actual answer: the choice between dashboards matters far less than whether either corresponds to real enquiries, calls and sales.
For decisions about advertising spend, use the Google Ads figures, because that is the system making the bidding decisions and it can only act on what it can see.
But reconcile them against your own records before treating either as a business result. If your reports say forty and your team remembers twelve, no amount of choosing between dashboards resolves that.
Check both against something real.
The free check looks at what your website actually sends, which is upstream of both dashboards and is where most discrepancies begin.
Why do Google Ads and GA4 report different conversion numbers?
Because they are built to answer different questions and they count differently by design. Google Ads attributes a conversion back to the date of the click; GA4 records it on the date it happened. Their attribution models, lookback windows and handling of multiple conversions per visitor also differ.
So some difference between the two is expected and correct. A gap on its own is not evidence of a fault.
Should we import GA4 conversions into Google Ads instead?
It is a legitimate setup and it does not by itself make anything more accurate. Importing moves where the conversion is defined; it does not change whether the underlying action represents a customer.
It also introduces a specific risk worth knowing about. An analytics event built to measure engagement, such as time on page or scroll depth, is easy to import and looks like a conversion once it arrives. If both an imported conversion and a direct one exist for the same action, the same customer can be counted twice.
Why does GA4 show more paid sessions than Google Ads shows clicks?
Because a click and a session are different units. One person clicking your ad, leaving the tab, and returning half an hour later is one click and can be two sessions.
Other ordinary causes include Google Ads filtering invalid clicks that analytics still records, and a redirect on your landing page starting a new session. Small differences here are expected. A large or sudden one is more often a website or measurement change than an advertising one.
Does our cookie banner affect the two systems differently?
It can, which is one reason the gap between them is rarely stable across regions. The two systems handle a declined or absent consent choice differently, and the modelling each applies to fill the gap is not the same.
The practical consequence is that comparing a European gap against a North American one tells you very little, and a change in the gap that coincides with a banner change is more likely to be about consent than about campaign performance.
Which system should the campaigns actually optimise toward?
Whichever holds the conversion definition you have verified. In practice that means picking one as the source of truth for bidding rather than maintaining both and hoping they agree.
The failure mode worth avoiding is running direct conversions and imported analytics conversions for the same action at once. That is how the same enquiry is counted twice, and double-counted conversions are worse than either system alone, because the bidding is now optimising toward an inflated signal.