Your conversion numbers never match because each platform uses different tracking methods, attribution windows, and modeling. Meta over-reports by ~26% due to modeled conversions and view-through attribution. Google Ads over-attributes by 15-20% with Enhanced Conversions. GA4 underreports by 18-35% because it only counts users who accepted cookies. Only 31% of users accept tracking—the rest are estimated or invisible.
The Attribution Gap Problem
Every marketer has experienced this: you check Meta Ads Manager and see 100 purchases. You open Google Ads—85 conversions. You pull up GA4—65. Your finance team asks which number is real. The answer? All of them, and none of them.
Since iOS 14.5 introduced App Tracking Transparency (ATT), roughly 70% of users opt out of tracking. This created a massive data gap that platforms now fill with statistical modeling.
What Are Modeled Conversions?
Modeled conversions are statistical estimates. When a user blocks cookies or opts out of tracking, platforms can't directly observe their behavior. Instead, they use machine learning to predict what likely happened based on patterns from users who DID consent.
Google explains it simply: "Modeled conversions use data that doesn't identify individual users to estimate conversions that Google is unable to observe directly."
"Google uses techniques like holdback validation to check accuracy of modeling—they hold back a portion of observed conversions and model for that slice, then compare the modeled results to actual observed conversions."— Google Ads Help
Why Each Platform Shows Different Numbers
Meta Ads (+26% vs Analytics)
Meta counts modeled conversions AND includes view-through attribution by default (someone saw your ad, didn't click, but converted within 1 day). It also uses a 7-day click window. Result: the highest numbers.
Google Ads (+15-20% with Enhanced Conversions)
Google Ads models conversions for users who rejected Consent Mode and adds Enhanced Conversions data (hashed first-party data matching). It typically uses last-click attribution. Result: higher than GA4, lower than Meta.
GA4 (-18-35% vs Platform Data)
GA4 only shows observed conversions from users who accepted cookies. No modeling, no view-through. It uses data-driven attribution across all traffic sources. Result: the most conservative (and often incomplete) picture.
Which Numbers Should You Trust?
The uncomfortable truth: none of them are "correct" in an absolute sense. Each serves a different purpose:
- GA4: Your baseline for observed user behavior. Underreports paid, but consistent methodology.
- Platform data (Meta/Google): Best for optimizing campaigns within that platform. The algorithms use this data.
- Incrementality testing: The only way to measure TRUE causal impact (see our incrementality guide).
How to Reduce the Gap
You can't eliminate the gap entirely, but you can narrow it:
- Implement server-side tracking: Meta CAPI + Google Enhanced Conversions improve match rates by 10-40%.
- Use consistent attribution windows: Align Meta (7-day click, 1-day view) with Google (30-day default) where possible.
- Create a "blended" metric: Many teams use a weighted average: (GA4 × 1.2) or (Platform × 0.85) as a middle ground.
- Run incrementality tests: Periodically validate platform claims with geo-lift experiments.
The Bottom Line
Your conversion numbers will never match across platforms. That's not a bug—it's a feature of how modern privacy-first measurement works. The real question isn't "which number is right?" but "what decision does each number help me make?"
Use platform data to optimize campaigns. Use GA4 to understand user journeys. Use incrementality to prove actual business impact. Together, they give you a more complete picture than any single source.