Google Ads Conversions vs GA4: Why They Never Match

Neither number is the real one, because they are answers to two different questions. Google Ads counts actions taken by people who interacted with a Google ad, filed under the date of the click. Google Analytics counts the action whenever it happens, from any source, filed under the date it happened. Use the Google Ads number for anything about bidding, use the Analytics number for anything comparing channels, and stop trying to reconcile them to zero.

The useful work is not deciding who wins. It is subtracting the part of the gap that is explainable, so you can see whether anything is left over. Most of the time nothing is, and the reconciliation takes about twenty minutes. When something is left over, you now have a real problem with a much smaller search space.

Google renames and re-defaults both products regularly, so every product-specific statement below is attributed to a Google Help page we read on 2026-09-07 and dated to that day. Where the vocabulary itself is the confusion, key events, conversions and the words GA4 keeps renaming covers the naming in full and this page assumes it.

Subtract the explainable gap in this order

Work down the list. Each step removes a portion of the difference that was never a fault. Do not skip to the interesting one.

Order What to align Why it comes here
1 The clock Costs nothing, often removes the largest single chunk
2 The date range and time zone Two accounts can be set to two zones and nobody notices for a year
3 The population Restrict Analytics to paid Google traffic before comparing
4 The counting rule The two products have separate settings with separate defaults
5 The crediting rule Different models, different windows
6 Whatever is left This, and only this, is worth investigating

1. Align the clock first

Google Ads files its primary conversion columns against the click, not the conversion. Its own wording: "The primary conversion columns mentioned above are calculated based on the time of the click, not the time of the conversion", and if the ad was clicked last week and converted this week, "both the click and the conversion are reported back to last week" (Google Ads Help, "Understand your conversion tracking data", checked 2026-09-07).

Analytics reports the event on the day it fired. So a comparison of the same seven days is comparing two different sets of events. The same Help page provides columns labeled "by conv. time", which report on the date the conversion occurred. Switch to those before anything else. On a business with a long decision cycle this step alone can account for most of the difference, and it also explains why last week's Google Ads numbers keep growing after you exported them.

2. Check the date range and the time zone

Trivial, and it defeats people for months. Google Ads and Google Analytics each carry their own time zone setting, and they do not have to agree. A few hours of offset moves conversions across day boundaries and makes a small gap look like a persistent bias. Confirm both settings before you compare a single day, and prefer a full week over a single day for anything you intend to act on.

3. Restrict Analytics to the same population

Google Ads counts an action only when the person had an ad interaction. Analytics counts everyone. Comparing an all-traffic key event total to a Google Ads conversion total is comparing a pond to a bucket that was dipped into it.

Filter the Analytics side down to Google paid traffic before you compare. If that filtered number is far below the Google Ads number, you have a link problem rather than a counting problem, and the causes sit in the gclid parameter and the redirect that quietly kills attribution: paid visits arriving without their click identifier stop looking like paid visits, so the population you filtered to is smaller than the population that actually arrived.

4. Compare the two counting settings, which have opposite defaults

Both products decide how many times a single person can count, and they use different words and different defaults for it.

Setting What it does Default applies to
Google Analytics Once per event "Counts a key event every time it's triggered" All key events except those migrated from Universal Analytics goals
Google Analytics Once per session "Counts a key event only once within a particular session" Key events migrated from Universal Analytics goals
Google Ads Every "Google Ads counts every conversion (per tracked conversion action) that happens after an ad interaction" Website conversions, in-app actions, Analytics transactions, imported conversion actions
Google Ads One "Google Ads counts only one conversion per ad click" Call-from-ads conversions, Analytics goals, calls to website numbers, app installs

Sources: Analytics Help, "Change the counting method of key events", and Google Ads Help, "About conversion counting options", both checked 2026-09-07. Google's own worked example is that five triggers in one session produce five key events under once per event and one under once per session.

Read the table and the trap is obvious. A form somebody submits twice can be one number in one product and two in the other, purely from defaults nobody chose. This is also the setting that most often makes an older property disagree with a newer one, because a property carrying migrated goals is still counting on the old rule.

5. Accept that crediting will not align

Analytics offers three attribution models and can move credit after the fact: "Conversions can be reattributed for up to 7 days after the conversion" (Analytics Help, attribution and attribution modeling, checked 2026-09-07). Google Ads applies its own crediting inside its own conversion window. Two crediting systems over two populations will not produce the same allocation, and no setting reconciles them. If the model choice is itself the question, attribution models in plain language covers which model changes a decision and which just changes a chart.

6. What is left is the actual problem

After steps one to five, you have a residual. Judge it on direction, not size.

Google Ads shows conversions, Analytics shows almost nothing from Google paid. Look at the arrival of the click, not the tag: redirects, consent handling, a landing page that strips parameters.

Analytics shows the action, Google Ads shows zero. The action is being recorded but is not reaching the ad account, or is not eligible. That has its own ordered checklist in why a GA4 conversion is not showing in Google Ads, and it is a different failure from a gap.

Both move together, but one is consistently higher by a stable proportion. That is the normal state. Stop. Write down what the proportion is for your account and use it as your baseline.

Which number to actually use

Pick per decision, before the meeting, not during it.

  • Bidding, budget and paid efficiency: the Google Ads number. This is not a preference. Automated bidding optimizes toward the conversions it can see, so a conversion the ad platform does not count is invisible to the system making the decisions. If bidding appears to behave erratically, the volume question in what Smart Bidding does when there are not enough conversions usually comes first and the counting rule above comes second, because counting Every rather than One changes how much apparent data the system has.
  • Channel comparison and on-site behavior: the Analytics number. It is the only one that sees non-paid sources on the same terms.
  • Money: neither. Both are steering instruments. Revenue comes from the system that issues invoices.

Label the source and the clock on every slide. "Conversions, Google Ads, click date" next to "Key events, GA4, event date" ends the standing argument for the price of a footnote.

The percentage we are not going to give you

The most common thing people want from this query is a number: an acceptable variance, a threshold above which to worry. Ranges circulate widely, and most of the pages ranking here are published by companies selling an attribution product.

We could not attribute an expected discrepancy figure to a Google documentation page on 2026-09-07, so this page prints none. Use your own instead. Run the reconciliation above for three or four consecutive weeks while nothing in the account is changing, always in the same columns, and record the residual. That number is real, it belongs to your account, and unlike a borrowed benchmark it will tell you when something has genuinely changed.

We also have no client audit or test to cite here. Everything above is mechanism, sourced to Google's own documentation and dated. Because both products change defaults and wording on a schedule nobody outside Google controls, treat this page as accurate on 2026-09-07 and check the linked Help pages before acting on it a year from now.

FAQ

Why does Google Ads show more conversions than GA4? Most often the clock and the crediting. Google Ads files conversions on the click date and takes credit for any conversion following an ad interaction, including across devices where it can match the user. Switch to the "by conv. time" columns and filter Analytics to Google paid traffic before concluding anything.

Why does GA4 show more than Google Ads? Usually population. Key events count everyone, including organic, direct and referral visitors, unless you filter to paid Google traffic. A counting method set to once per event on one side and One per click on the other will widen it further.

Should I change a setting to make them match? No. Aligning the counting methods can narrow the gap, but the population and the clock are definitional and cannot be aligned. Choosing settings to make two dashboards agree usually means choosing a setting that is wrong for at least one of the decisions those dashboards support.

Which number should automated bidding use? The one in the ad account. Bidding cannot optimize toward data it never receives, so the conversion action in Google Ads, and its counting option, is what the system is actually learning from.

How long should I wait before comparing? Give conversions time to land. Because Google Ads keeps filling in past days, a comparison of the last two or three days will always look wrong. Compare complete weeks that are at least a few days old.

Sources, read 2026-09-07

Google Ads Help: "Understand your conversion tracking data" (support.google.com/google-ads/answer/6270625), "About conversion counting options" (answer/3438531). Analytics Help: "Change the counting method of key events" (support.google.com/analytics/answer/13366706), attribution and attribution modeling (answer/10596866).

Both products rename settings and change defaults often. This page carries a six month review date and describes behavior rather than menu paths for that reason.

Similar Posts