Why GA4, Google Ads and your shop admin show different conversions
Your shop admin shows 120 orders, GA4 shows 98 and Google Ads 74? That is not necessarily an error. Three systems count in three ways. Here is which differences are normal, which point to a fault, and how to tell them apart.
In short
- The shop admin counts every order, GA4 only the ones it measured, and Google Ads only the ones it credits to ads. A perfect match does not exist and is not the goal.
- Normal differences come from attribution, declined cookies, modelled conversions, refunds, time zones and the conversion window.
- Faults show up as duplicates (more conversions than orders) and sharp drops after a banner or theme change.
Why three numbers never match
The shop admin is the source of truth for orders: it records every one, including those with no ad visit and no cookie consent. GA4 records only purchases it managed to measure, and assigns them to sources. Google Ads records only conversions it credits to its own ads under its own rules.
So a difference is not an error as long as you can explain it. What you can judge is the relationship: is the ratio stable over time, or did it jump after some change? The causes below are ordered from the most common.
Normal causes of differences
- Cookie consent. Visitors who declined are missing from GA4. The share of declines differs from site to site and can be noticeable.
- Attribution. An order from a customer who came from a newsletter is assigned to e-mail in GA4, not to an ad. Google Ads only knows the clicks and views of its own ads.
- Different attribution models. The default model in both GA4 and Google Ads is now data-driven, but each system calculates it from different data. Older accounts may have different settings.
- Conversion window. Google Ads counts a conversion only if it happened within the set window after the click (the default length is in the conversion action settings). A purchase after a longer time is not counted.
- Date of record. By default Google Ads credits a conversion to the day of the ad click, GA4 to the day of purchase. Over a week or a month the numbers differ even with perfect measurement.
- Time zones. The GA4 account, Google Ads account and shop admin may use different time zones. Orders just before midnight then land on different days.
- Cancellations and refunds. The admin can subtract them. GA4 and Google Ads do not know about them without extra setup, so the purchase stays counted.
- Different definitions. In the admin an “order” exists when it is submitted. In GA4
purchasefires when the thank-you page loads. An unfinished card payment may appear in one system and not in the other.
Modelled conversions
Where the visitor gave no consent, Google Ads and GA4 can fill in part of the missing data with a model. This works with Consent Mode v2 in advanced mode: Google tags send anonymous signals without consent, and Google estimates part of the conversions from them. The model needs enough traffic to start working and fills only part of the gap, never all of it.
As a result, after you introduce a banner or Consent Mode v2, the number of measured conversions usually drops and modelled values arrive with a delay. In reports, never compare the new period with the old one without a note about the change. How to set up and check Consent Mode is in Consent Mode v2: how to check it on your shop.
Causes that mean a fault
- Duplicates. GA4 or the Google Ads conversion runs twice, for example in the platform integration and in GTM at once. There are more conversions than orders.
- Conversion fires wrongly. It is measured when the thank-you page is displayed, so reloading the page or going back counts again.
- Measurement broken by a change. A theme update, a new payment gateway or a new banner, and the conversions vanish. You see a sharp break in the chart.
- Wrong value. The conversion carries a price without VAT, or with shipping, while the admin counts differently.
- Missing link. GA4 is not linked to Google Ads, or Ads uses a different conversion action than GA4.
- Tags that ignore consent. The numbers look good, but they are data that should not have been collected.
How to compare them in practice
Pick a closed period
Take the whole past week or month, not today, so that report delays have passed. Set the same time zone in all three systems, or at least find out what each one uses.
Subtract cancellations in the admin
For comparison use the number and value of orders under the same definition as the measurement: submitted, not cancelled. Note the number of cancelled orders as well.
Calculate the GA4 to admin ratio
The ratio should be stable over time and below 100%. Look for sudden changes around the last change to the shop or the banner. More than 100% is almost certainly a duplicate.
Compare Google Ads with GA4
In Google Ads compare the Conversions column with the Conversions by conversion time column. If the numbers differ, it is a difference in the date of record. Also check the conversion window and whether only the goals you want are counted.
Check one purchase
Place a test order and watch whether
purchaseappears once in GA4 DebugView. Then look for it in Google Ads too.
How a customer journey creates the differences
Picture a customer who clicks an ad on Monday, puts the purchase off and orders on Friday from a newsletter. The admin has one order. GA4 assigns it to e-mail, because that is the last source. Google Ads may credit it to the Monday click if it fits the conversion window and the model attributes it to the ad. Nobody is wrong; each system answers a different question.
That is why we recommend watching trends, not single numbers. If Google Ads and GA4 rise and fall together in relation to the admin, measurement works. If they diverge over time, look for a change on the shop.
What to do about it
The goal is not for the numbers to be equal. The goal is for each system to measure consistently and for you to know why they differ. A reasonable state: the admin is the truth about sales, GA4 about behaviour and sources, and Google Ads about ad performance, always knowing how they differ.
If you find duplicates, gaps or conversions that ignore consent, that is a technical fix we do on the conversion tracking and GA4 page from €100 excl. VAT. Verification against the shop admin is part of it.
Frequently asked questions
What difference between GA4 and the admin is still normal?
We cannot give a fixed limit: it depends on the share of declined cookies, traffic sources and cancellations. What matters is that the ratio is stable and GA4 never shows more orders than the admin.
Why does Google Ads show more conversions than GA4?
Google Ads may count conversions from a different window, credit them to the day of the click and add modelled ones. Often it is double counting, so check that the same conversion action is not counted twice.
Can cancellations be subtracted?
Yes, but not automatically. The fix is to send refunds to GA4 or import offline conversions into Google Ads. For an overview it is often simpler to compare with orders after cancellations.
Order a free audit
Enter the shop address and your e-mail. We send the report in two working days. We use your details only to answer this request.
- No access and nothing to install
- The report is yours even if you order nothing
- A person answers, not a bot
Thank you, we have your request
We will send the report to your e-mail in two working days.
Sending did not work
Please send us the request by e-mail. The text is ready.
info@lyneon.eu
More guides
- Consent Mode v2: how to check it on your shop5 min read
- How to set up DMARC (and SPF and DKIM) for your shop5 min read
- European Accessibility Act: what it means for online shops5 min read
- How to check whether Meta Pixel fires before consent6 min read
- Disapproved products in Google Shopping: the most common reasons5 min read
- Core Web Vitals for online shops: LCP, CLS and INP explained5 min read
- Why Google does not index your shop pages and what to do6 min read
- Why e-mails from your shop go to spam6 min read