The short answer

Meta counts ad views within its own window; GA4 only counts the last click. A 20-40% gap is normal — neither number is wrong, and neither proves Meta ads are actually working. The only way to know that is to test it directly.

Key takeaways
  • Meta credits view-through conversions. GA4 doesn't. That's most of the gap.
  • 20-40% Meta-vs-GA4 gap is normal. Beyond ~50%, suspect a tracking fault.
  • Neither number proves causation. Test incrementality by cutting spend and watching topline revenue, not platform dashboards.

Why they disagree

Meta asks "did someone who saw or clicked my ad convert within my window?" GA4 asks "what was the last non-direct click before this conversion?" Different questions, different numbers — neither is broken.

META ATTRIBUTION WINDOW CONVERSION 1-day view 7-day click window GA4 ATTRIBUTION (LAST NON-DIRECT CLICK) Last click → Meta credits the impression on day 1. → GA4 credits nothing until a click — on day 6, if one ever happens.
Meta's default window credits a view within 1 day and a click within 7. GA4 only credits the last click — nothing before it.

What's normal

Meta reporting 20-40% more conversions than GA4 is typical for e-commerce. Beyond ~50%, look for duplicate events or broken tags before blaming attribution.

The real question: is it incremental?

Reconciling the two makes your reporting consistent — it doesn't prove Meta spend is driving revenue. Both platforms report correlated conversions, not caused ones.

The only real test: deliberately cut or raise Meta spend 30-50% for one to two weeks, and watch total business revenue — not either platform's dashboard. If revenue barely moves when spend is cut in half, most of those "conversions" were never incremental.

WEEKS 1–2 · NORMAL SPEND WEEKS 3–4 · SPEND CUT 40% Platform-reported conversions Actual topline revenue Conversions dropped sharply (illustrative: ~55%). Revenue barely moved (~9%). → Most of what Meta reported as "caused" was not incremental — those buyers converted anyway. This is the test; the platform dashboard is not.
An illustrative spend-down test: platform-reported conversions fell sharply, but topline revenue barely moved — meaning most of it wasn't incremental.

This is the test we run before recommending a client scale or cut a channel. Reconciliation tells you your reporting is sane. Only a real spend change, measured against topline, tells you the channel actually works.

Related service: Analytics & Attribution