Meta Ads and GA4 don’t match: when the gap is normal and when it’s lost data
If Meta Ads and GA4 show exactly the same number of sales, be suspicious. The two tools weren’t built to count the same way: each one answers a different question about the same purchase. So the useful question isn’t “why don’t they match?”. It’s “does this gap come from attribution, or am I losing

If Meta Ads and GA4 show exactly the same number of sales, be suspicious. The two tools weren’t built to count the same way: each one answers a different question about the same purchase.
So the useful question isn’t “why don’t they match?”. It’s “does this gap come from attribution, or am I losing data?”. Here’s how to tell them apart.
In short:
- Meta Ads and GA4 use different attribution windows and models. A gap between them is expected.
- The warning sign is when both fall below the store’s paid orders.
- What distorts the comparison most: UTMs, consent, deduplication between the pixel and the Conversions API, and purchase set as a key event in GA4.
Why Meta Ads and GA4 count differently
Meta Ads wants to know whether the ad took part in the sale. Meta’s Business Help Center explains that a purchase on the website is attributed to the ad if it happens within a certain number of days after someone viewed or clicked it. In the attribution settings of the standard model:
- Click: counts events that happened within 1 or 7 days after a link click on the ad.
- View: counts events that happened within 1 day after an ad impression, even without a click.
GA4 wants to know which channel brought the person to the purchase. It looks at the whole journey (ad, search, email, direct visit) and assigns credit using the attribution model chosen in the property: data-driven, last click, or paid and organic last click with Google paid channels. The default window is 30 days for acquisition key events and 90 days for all others.
The result: when comparing Meta Ads and GA4, the same purchase can belong to the Meta ad and to organic search in GA4. Neither one is wrong.

The five differences between Meta Ads and GA4
Meta Ads | GA4 | |
|---|---|---|
Who gets the credit | The ad, if the purchase falls in the window | One or more channels, depending on the attribution model |
Ad seen without a click | Can count (up to 1 day after the impression) | Doesn’t show up: GA4 doesn’t see a Meta ad impression, only the click that reaches the site |
Window | 1 or 7 days after the click | 30 days for acquisition, 90 for other key events |
How it knows the visit came from the ad | Through the pixel and the API linked to the ad account | Through the UTM parameters of the clicked link |
When data is missing | May use statistical modeling | May use consent mode modeling |
When comparing Meta Ads and GA4, the UTM row is usually the one that weighs most. GA4 only knows the visit came from a Meta campaign because the URL parameters arrive with the click. Without them, the same sale can land as referral or direct traffic.
How to set up UTMs so Meta Ads and GA4 speak the same language
The simplest path is to let Meta fill in the parameters. In the ad’s “URL parameters” field, Meta accepts dynamic parameters, which are automatically replaced with the ad’s information at click time. Meta’s Business Help Center lists, among others, {{campaign.name}}, {{adset.name}}, {{ad.name}}, {{placement}} and {{site_source_name}}, which shows where the ad ran (for example, fb for Facebook and ig for Instagram).
For Google Analytics, Meta recommends placing these values inside UTM parameters. A complete setup looks like this:
utm_source={{site_source_name}}&utm_medium=paid&utm_campaign={{campaign.name}}&utm_content={{ad.name}}
Three details make a difference when comparing Meta Ads and GA4:
- The name is the one at publication. According to Meta, name-based parameters keep the name the campaign, ad set or ad had when it was first published. Renaming it later doesn’t change what reaches GA4.
- Standardize before launching. If one campaign uses
utm_medium=paidand another usesutm_medium=cpc, GA4 may split the same kind of traffic into different rows. Set one standard for the whole account. - Not every format supports it. Meta states that dynamic parameters aren’t supported for collection ads in Instagram placements. In those cases, use fixed parameters.
With the right UTMs, GA4’s acquisition report shows campaign and ad with the same names as Ads Manager, and the comparison stops happening in the dark.
Where Meta Ads and GA4 really lose data
Beyond attribution, there’s the path data travels from the browser to Meta Ads and GA4. That’s where it disappears:
- Consent. When a visitor declines cookies, GA4 no longer has their behavioral data. Consent mode modeling only kicks in when the model is reliable; without enough traffic, those events aren’t reported.
- Blockers and browser restrictions. Extensions and browsers with tracking protection can stop the tag from loading or the event from being sent.
- Device switching. Without User-ID, GA4 identifies the visitor by device ID, which on websites comes from the cookie. Someone who clicks on their phone and buys on their computer becomes two people.
- Events that don’t fire. A thank-you page that doesn’t load, a payment confirmed later, a tag that breaks on a new theme.
There is one more detail when comparing Meta Ads and GA4: both tools may fill gaps with modeling. Meta states that it may use statistical models when conversion data is partial or missing, and GA4 uses consent mode behavioral modeling to estimate the behavior of users who decline cookies. Modeled numbers help read trends, but they are estimates, not orders. That is why the store remains the benchmark.
Attribution explains why the numbers differ. Data loss explains why Meta Ads and GA4 can both, together, fall below what the store sold.
Pixel and Conversions API: when Meta counts too much
The error can also go the other way. If you send the purchase through the pixel, in the browser, and through the Conversions API, on the server, you need to tell Meta that both sends are the same purchase. Otherwise, it may count two.
Meta’s documentation explains the rule: the pixel’s eventID must match the Conversions API’s event_id, and the event name must be the same in both (for example, Purchase). When both arrive identical, Meta generally keeps the one received first. There’s also a time limit: events are only deduplicated if they arrive within 48 hours of the first one.
In practice, the order number is the best event_id, because it’s unique and exists both on the thank-you page and on the store’s server. If Ads Manager shows more purchases than the store recorded, before blaming attribution, check in Events Manager whether the event_id arrives identical through both paths.
Purchase must be a key event in GA4
On the Google side, the detail most often missed is the event setup. Google Analytics Help explains that you choose an event that measures an important interaction and mark it as a key event. The lookback window and attribution reports work on key events.
So check in the GA4 Admin events list whether purchase is marked as a key event. A standard property accepts up to 30 key events, so it’s also worth checking that lower-value events, like page views, aren’t marked at the same level as the purchase. Without purchase as a key event, the comparison between Meta Ads and GA4 is skewed from the start.
The same care applies to the lookback window. Google Analytics Help explains that changes to the window apply from that moment on and show up in every report in the property. If someone changed it in the middle of the period you are analyzing, the comparison with Meta Ads mixes two rules.
What changed at Meta in 2026
On January 12, 2026, Meta stopped returning the 7-day and 28-day view windows through the Insights API (7d_view and 28d_view). The 1, 7 and 28-day click windows, the 1-day view and the 1-day engaged view remain available.
The effect shows up in dashboards and spreadsheets that pull Meta data through integrations. If an old report compared Meta Ads and GA4 using the 7-day view, it no longer has that number to show. Review which windows each dashboard requests before concluding that sales dropped.
How to tell attribution from data loss
To compare Meta Ads and GA4, Wascer recommends four steps, in this order:
- Use the store as the benchmark. Take the total paid orders in your e-commerce platform for the period. That’s the real number.
- In Meta Ads, compare attribution settings. In Columns, use “Compare attribution settings” and look at 7-day click only, without views. Part of the gap is usually there.
- In GA4, check the UTMs. See whether the ads carry
utm_sourceandutm_mediumand whether Meta traffic shows up as a campaign, not as referral. - Add up what each one saw. If Meta Ads and GA4 stay far below the store’s orders, even after adding the other channels, the problem isn’t attribution. It’s data that never arrived.
Want to send conversions from the server, without relying only on the browser? Create your free Wascer account.

How Wascer helps
With server-side tracking, the purchase goes from the server to Meta’s Conversions API and to GA4, with the same order ID in both. That way, Meta Ads and GA4 receive the same order. Attribution stays different, but browser losses no longer explain the gap. Wascer is the platform that spins up and monitors a server-side Google Tag Manager container in one click, with no DevOps and no command line, and with its own tags for Meta, GA4 and Google Ads. Learn what Wascer is and how it works.
Frequently asked questions

Is it normal for Meta Ads to show more sales than GA4?
Yes. Meta counts purchases after views without a click and credits the ad on its own; GA4 splits the credit among the channels in the journey.
Can Meta Ads and GA4 match 100%?
No, because they use different attribution windows and models. The goal is for both to receive every event, not to show the same number.
Which one is right?
Both, each for its own question. To know how much the store sold, the source is your e-commerce platform.
Do I need UTMs if I already use the Meta pixel?
Yes. The pixel informs Meta; GA4 only knows the visit came from the ad through the URL parameters.
Why does Meta Ads show more purchases than the store sold?
A common cause is the pixel and the Conversions API sending the same purchase without the same event_id, which prevents deduplication.
Conclusion
A gap between Meta Ads and GA4 isn’t a sign of broken tracking. It’s a sign that each tool measures something different. What calls for action is when both stay below what the store sold: then the problem is lost data, and budget decisions made on top of it.
Want your conversions to go from the server to Meta and GA4? Create your free Wascer account.

By Wascer
Wascer spins up and monitors a server-side Google Tag Manager container in one click: no DevOps, no Docker and no command line. Technical content on tracking, Meta, GA4 and Google Ads for people who decide with data. wascer.com
Sources
- Meta, about Facebook’s attribution system: https://www.facebook.com/business/help/458681590974355
- Meta, about attribution models and attribution settings: https://www.facebook.com/business/help/2198119873776795
- Meta, why results are sometimes calculated differently (statistical modeling): https://www.facebook.com/business/help/1329822420714248
- Meta, dynamic URL parameter specifications: https://www.facebook.com/business/help/2360940870872492
- Meta, deduplicate pixel and server events: https://developers.facebook.com/docs/marketing-api/conversions-api/deduplicate-pixel-and-server-events/
- Meta, Ads Insights API metric availability updates (01/12/2026): https://developers.facebook.com/blog/post/2025/10/16/ads-insights-api-metric-availability-updates/
- Google Analytics, mark events as key events: https://support.google.com/analytics/answer/13128484?hl=en
- Google Analytics, attribution models: https://support.google.com/analytics/answer/10596866?hl=en
- Google Analytics, key event lookback window: https://support.google.com/analytics/answer/10597962?hl=en
- Google Analytics, collect campaign data with custom URLs (UTM): https://support.google.com/analytics/answer/10917952?hl=en
- Google Analytics, behavioral modeling for consent mode: https://support.google.com/analytics/answer/11161109?hl=en
- Google Analytics, reporting identity: https://support.google.com/analytics/answer/10976610?hl=en
Adam Yoshikai · Published on 30 Oct 2026 · updated on 6 Nov 2026