Run free audit
Troubleshooting · part of Fix GA4

Your GA4 and Google Ads conversions do not match

They are not supposed to. Three documented differences guarantee a gap before anything is broken at all.

Which makes this a harder question than it looks, because the useful skill is not closing the gap, it is telling the designed part from the broken part. Most of what is written about this online comes from vendors, and they contradict each other on the basic numbers. So everything in the first half of this page is from Google’s own documentation, and where Google does not state something, we say so rather than filling it in.

Three reasons they can never match

  1. They date the same conversion to different days

    Google Ads credits a conversion to the ad interaction that earned it, not to the moment it happened. A click on Monday that becomes a purchase on Wednesday is reported by Ads against Monday and by GA4 against Wednesday. Worth knowing that Google's own pages word this differently in different places, one saying the impression date and another the click, but the principle is the same: the two tools are answering different questions about when.

    What to do about it: Comparing a single day, or a week during a campaign ramp, is close to meaningless. Compare a full month, after it has closed.

  2. The default windows are different numbers out of the box

    Google Ads attribution reports use a 30 day lookback by default, which you can change to 60 or 90. GA4's key event lookback defaults to 90 days for most key events, and 30 for acquisition events like first_visit. On top of that sits the Google Ads conversion window, a separate setting between 1 and 90 days, which Google's documentation explicitly flags as different from the lookback window and which people routinely confuse with it.

    What to do about it: Two tools with different defaults will disagree even with flawless tracking. Before investigating anything, check what each is actually set to.

  3. They credit different things

    Google Ads only ever credits its own ad interactions. GA4 looks across every channel, so a sale it attributes to organic search or direct simply does not appear in the Ads number at all. The attribution models are configured separately in each tool as well, and Google notes that switching models changes the conversion data you see.

    What to do about it: The Ads number is a subset by definition. If it matched the GA4 total, that would be the thing worth investigating.

Five counting rules that move the number

Beyond the three above, Google lists several settings and behaviours that change what gets counted. None of them are faults, and all of them are worth checking before you go hunting for one.

When the gap is actually a fault

Everything above produces a gap that is stable. It sits at roughly the same proportion month after month, and it moves only when you change a setting. A gap that is unstable, or that appeared suddenly, or that runs in the wrong direction with Ads reporting more conversions than GA4, is a different animal. These are the five faults worth checking, in the order they turn up.

1. The same action is counted twice in Ads

A Google Ads conversion tag firing on the thank-you page, and the same action imported from GA4 as a key event. Both land in Ads. This is the most common real fault behind a gap that runs the wrong way, with Ads higher than GA4.

2. Enhanced Measurement is duplicating a key event

If the duplicated event is marked as a key event and imported into Ads, the inflation is carried straight into bidding. The six collisions, and the parameter signature that tells the two sources apart, are covered in why GA4 events fire twice.

3. Something trivial is marked as a key event

page_view marked as a key event, then imported, produces a conversion count that is really a pageview count. It is usually obvious from the magnitude: conversions roughly equal to sessions.

4. Consent denial is handled differently by each tool

Where a visitor declines, GA4 may model the conversion while Ads counts what its own tags could observe. A consent setup that changed recently will move the gap without anything else changing, which is why the gap's stability matters more than its size.

5. The session that earned the conversion is being split

A payment gateway or login provider that is not excluded starts a new session on return. GA4 then credits the conversion to that referral while Ads still sees its own click, so the two diverge on exactly the conversions that matter most.

How to compare them properly

  • One action, not totals. Compare a single conversion action against its GA4 key event. Totals hide which one is wrong.
  • A closed month, not a day. The date difference alone makes short ranges unusable.
  • Match the windows first. Set the Ads attribution lookback and the GA4 key event lookback to the same number before you compare anything.
  • Wait. Google asks for 24 to 48 hours for synchronisation, and conversions can arrive up to 90 days after the click.
  • Judge stability, not size. Write the ratio down each month. A steady ratio is design. A moving one is a fault.

Four of the five faults above are things an audit finds without you looking: duplicate conversion counting, trivial key events, consent gaps and sessions split by a payment gateway. Tracking Auditor checks all of them against your live setup, scored A–F. A free audit every month.

Run a free audit →

Go deeper

The full process for verifying every conversion end to end is in the conversion tracking audit guide. If purchases specifically are missing rather than mismatched, see GA4 purchase event not firing. If a payment gateway is splitting the session, the free referral exclusion list generator builds the list to paste in, and consent differences are covered in the consent mode audit guide. More symptoms are indexed at Fix GA4.

FAQs

Why don't my GA4 and Google Ads conversions match?

Mostly because they are not supposed to. Google Ads dates a conversion to the ad interaction that earned it while GA4 dates it to when it happened, the two tools use different default lookback windows (30 days in Ads attribution reports, 90 days for most GA4 key events), and Ads only credits its own ad interactions while GA4 credits every channel. On top of those design differences sit real faults, the most common being the same action counted twice in Ads because a conversion tag and a GA4 import are both running.

What size of discrepancy is normal?

Google does not publish a figure, and the numbers circulating online come from vendors rather than from any study, so we are not going to repeat one. The useful test is not size but stability: a gap that holds steady month to month is almost always the design differences above, while a gap that moves, or that appeared suddenly, points at something broken.

Should I trust Google Ads or GA4 for conversions?

Use Ads to make bidding decisions, because that is the number the bidding algorithm optimises against, and use GA4 to understand the whole journey across channels. Asking which is correct is the wrong question: they measure different things on purpose. What does need to be correct is that each one counts each action exactly once.

How long should I wait before comparing the numbers?

Google's own guidance is to allow 24 to 48 hours for synchronisation between Analytics and Ads before comparing, and it notes that conversions can continue to be reported up to 90 days after the click, and can take up to 7 days to appear fully after an account linking change. Comparing yesterday to yesterday will always look broken.

How do I compare them properly?

Compare one conversion action rather than totals, over a closed month rather than a day, after checking that the lookback windows in each tool are set to the same value. If the gap survives all of that and is not stable, start looking for duplicate counting.

The attribution dates, lookback window defaults, conversion window range, counting settings and synchronisation guidance above are from Google’s own help documentation on Ads data discrepancies, Ads attribution reports, GA4 attribution settings and the GA4 key event lookback window, checked on 20 October 2026. Where Google publishes no figure, such as a normal size of discrepancy, none is given here.