Why Meta and GA4 Metrics Never Match

A client running paid social pulled two reports for the same campaign, the same day, the same traffic. Meta Ads Manager showed 407 landing page views. GA4 showed 63. The gap looked like proof that something broke.

This gap shows up on a few paid social accounts we touch. We open both dashboards, sees numbers that refuse to reconcile, and assumes a tag fired wrong or a pixel dropped off the page. That assumption makes sense on its own terms. Consent mode is the actual mechanism, and once you see how each platform counts a visit, the gap is no longer a mystery.

Two Systems, Two Rules

Meta's Conversions API sends event data from the client's servers to Meta without routing through a browser. A landing page view fires the moment the page loads on the ad platform's side, before a visitor clicks through a cookie banner or makes any consent choice. Meta's count sits close to raw traffic volume because of this.

GA4 works under a different rule. Consent mode holds analytics tags until a visitor grants permission. If a visitor closes the banner without clicking accept, or the region requires an opt-in default rather than an opt-out one, GA4 suppresses the hit or files it under modeled data that never surfaces as a raw pageview. Every visitor who withholds consent disappears from GA4 while staying visible to Meta.

Geo-targeted consent banners widen the split further. A campaign running across the US and EU applies different consent defaults by region under CCPA and GDPR rules. A US visitor under a CCPA opt-out banner gets counted in GA4 unless they take action to decline. An EU visitor under a GDPR opt-in banner gets excluded from GA4 unless they take action to accept. The same campaign produces two consent baselines depending on where the click originated, and Meta's number never adjusts for either.

How We Traced the Gap

We didn't take either number at face value. Here's the sequence that isolated the cause, work that carried over directly into the GA4 consent mode debugging we later ran for another client.

1. Confirmed both the CAPI and GA4 tags were firing in the browser network panel, ruling out a broken implementation before looking anywhere else.

2. Pulled the consent mode configuration in Google Tag Manager and checked which tags sat gated behind analytics_storage versus which fired without restriction.

3. Segmented GA4 traffic by consent state and compared the granted-consent volume against Meta's reported views.

4. Found that the delta between 407 and 63 matched the volume of consent-denied and consent-pending sessions in GTM's own logs.

5. Cross-checked server-side CAPI logs against GA4's client-side hit count to confirm the server-side number reflected total sessions, not just consented ones.

Both platforms had reported a number correct for a different population of visitors.

What This Means for Budget Decisions

Treat Meta's number as your true reach and you overstate how many people saw the landing page. Treat GA4's number as your true reach and you undercount real traffic by the exact size of your consent-denied population, a population that shifts by region, by device, and by how a browser's default cookie settings interact with the banner in front of it.

Neither number stands alone for a budget call. The best figure is a blended one: platform-reported reach, cross-referenced against first-party CAPI data and GA4's consented volume, reconciled against the known consent rate for that traffic mix. Skip that step and you can expect panic over a GA4 dashboard showing a campaign underperforming when it isn't, or a budget gets set against a Meta number that never had the chance to reflect real site behavior.

The Diagnosis Behind the Optimization

Privacy regulation keeps expanding, and each new consent requirement widens the space between what Meta reports and what GA4 can see. An agency that treats this gap as a fire to put out will keep chasing phantom tracking bugs on every new account. An agency that treats it as a measurement architecture question can walk a client through exactly why the numbers diverge and exactly how much weight each one deserves.

That's the diagnosis we ran here, and it's the same instinct we bring to every account at Signal Lab. Before we touch a campaign's budget, we confirm what the campaign's numbers actually mean.