Why Your GA4 Numbers Don't Match Meta (And Which to Trust)

1. Why the numbers were never going to match

One dashboard asks whether an ad interaction happened before a conversion inside its allowed window. The other assigns credit among session-based touchpoints across channels. Those are different definitions, so a perfect reconciliation is the wrong target.

1. The attribution windows do not match. Meta's default is 7-day click + 1-day view, with an optional 1-day engaged-view setting for video, according to Jon Loomer's attribution guide. The longest available view-through window is now 1 day; Meta removed the old 7-day view and 28-day view options. GA4 commonly uses a 30-day lookback for acquisition events and up to 90 days for other conversions, according to Adamigo. A person can therefore fall inside one system's eligible window and outside the comparable report in the other.
2. One includes view-through credit; the other cannot see the impression. Under Meta's 1-day view rule, someone can see an ad, avoid clicking, then buy later that day. Ads Manager may credit the campaign. GA4 has no Meta impression tied to a website session, so it cannot assign that same view-through credit. Adamigo's reporting comparison confirms that GA4 excludes view-through conversions while the platform includes them by default.
3. The credit logic answers a different question. Meta's standard reporting gives its ad interaction credit when the defined conversion occurs within the selected window. GA4's default standard-report model is data-driven attribution, which distributes credit across eligible session touchpoints. Adamigo documents that contrast. A paid-social click followed by an email visit and a purchase can be credited wholly to Meta in one report and split across channels in the other.

The distinction survives recent product changes. In March 2026, Meta narrowed click-through attribution so an actual link click counts, rather than any social engagement. theOptimizer.io's 2026 attribution analysis explains that the change reduced one source of difference, but it did not align the windows, view-through rules, or credit models.

2. Meta Ads Manager vs. GA4 default reporting

This table compares the reporting defaults, not every custom configuration an advertiser can build.

Changing a report to click-only can narrow the gap. It cannot turn the systems into identical measurement tools.

3. How big is the gap, in practice?

Start with a working range, not a promised correction factor. Under Meta's default 7-day click + 1-day view setting versus GA4's data-driven attribution, Ads Manager commonly reports 20–50% more conversions for the same campaigns. theOptimizer.io reports that range, while many practitioners use 20–40% as a rough heuristic. These are field ranges, not universal constants.

View-through credit can explain a large part of the difference. In large-spend accounts, 1-day view-through conversions can represent 20–40% of Meta's reported total, according to the attribution-setting analyses from Jon Loomer and theOptimizer.io. GA4 cannot see that impression as a session touchpoint. Removing view-through from a comparison may make the totals closer, but it does not prove that those influenced sales were worthless.

The click window alone can move the platform result before GA4 enters the discussion. On Meta, 7-day click reporting typically shows 30–50% higher conversions or ROAS than 1-day click, because the longer window includes delayed converters. That range comes from theOptimizer.io's 2026 analysis. The extra credit may include legitimate delayed demand; the longer setting simply answers a broader question.

Privacy loss pulls in the opposite direction. ATT opt-outs cause Meta to underreport true iOS conversions by roughly 15–30% in many accounts, and sometimes 30–50% for iOS-heavy audiences with weak server-side setups, according to Cometly's 2026 iOS tracking guide. Ads Manager can therefore claim more conversions than analytics because of its attribution rules while still missing real iOS conversions. Both conditions can exist at once.

GA4 has its own missing-data problem. Cookie rejection and blocking are commonly cited as causing it to undercount paid traffic by roughly 18–35%, as summarized in Adamigo's reporting analysis. It is session- and cookie-dependent, so it is not a complete ledger of causal impact.

Do not add or subtract these percentages to manufacture a “true” dashboard value. The ranges overlap, account conditions vary, and the tools lose or assign credit for different reasons. Use them as diagnostic bounds. If your gap is far outside the common 20–50% Meta-over-GA4 range reported by theOptimizer.io, inspect event duplication, missing tags, UTMs, time zones, event definitions, and consent behavior before making a media decision.

4. Which number to trust

The rule is simple: use actual CRM, order, or payment-system revenue as the source of truth for budget decisions. Use GA4 as the next-best cross-channel measurement layer when the business record is unavailable or when you need to inspect the measured journey. Use Ads Manager as an optimization signal inside Meta's own system.

That hierarchy matches what each record can prove. An order system can show that revenue happened. GA4 can show measured sessions and distribute channel credit. Meta can show which ads its delivery system associates with conversions under its attribution setting. None proves the same thing as the others.

1. Deciding whether to increase the total Meta budget. Start with actual order or CRM revenue, gross margin, refunds, lead quality, and total spend for a stable period. Use GA4 to compare paid social with other measured acquisition paths. Then check whether Meta's internal trends support the change. Do not raise the budget only because platform ROAS looks strong under 7-day click + 1-day view; that window includes credit GA4 does not observe, as Jon Loomer explains.
2. Choosing which Meta campaign to scale within an already approved budget. Compare campaigns using the same event, window, and attribution setting in Ads Manager. That is where the platform signal is most useful: all candidates are judged inside the system that controls delivery. Check CRM or order outcomes as a guardrail, especially when view-through is a large share of results.
3. Deciding whether a campaign is actually profitable. Use booked and collected revenue, or qualified pipeline tied to the campaign when the sale happens offline. Subtract the real costs required by the business. GA4 helps explain the post-click path; Meta's result can indicate influence. Neither dashboard should replace the commercial record.
4. Testing a new creative or audience. Use Ads Manager to read relative movement among ads delivered under the same conditions. Its attribution setting affects both reporting and optimization, according to Jon Loomer's guide. Confirm that the apparent winner also produces acceptable downstream lead or order quality before moving more budget behind it.

A practical weekly view has three columns: business outcomes, GA4-measured channel contribution, and Meta-attributed results. Watch direction and ratios over time. Do not force the columns to equal one another. The useful question is whether revenue, measured acquisition, and the platform signal are moving together closely enough to support the next decision.

Build a comparison that supports a decision

Before comparing totals, write down the decision you are making. “Which ad should keep delivering?” is a platform-optimization question. “Should paid social receive more of the company’s acquisition budget?” is a business-allocation question. The first can lean on Ads Manager. The second must start with revenue or qualified pipeline outside the ad platform.

Next, freeze the definitions. Use the same date range, time zone, currency, conversion event, and account scope. Record the Meta attribution setting beside the result. In GA4, record the report, attribution model, and source or medium filter. This does not make the totals equal; it removes avoidable configuration noise so the remaining difference is interpretable.

Then separate level from direction. The level is the reported total in each system. The direction is whether that total rose or fell against a comparable prior period. A stable difference can still support useful trend analysis. A sudden change in the ratio is a measurement alarm: inspect releases, consent behavior, event firing, deduplication, UTMs, and checkout changes before blaming campaign delivery.

Finally, document the action and the evidence that authorized it. A budget increase should point to the business outcome that justified more spend. An ad-level shift should point to the consistent in-platform comparison that justified reallocating delivery. This short record prevents a later analyst from treating platform-attributed revenue as booked revenue or treating a cookie-dependent analytics total as a complete count of influence.

5. What this doesn't mean

This is not a reason to distrust Meta's platform entirely. Ads Manager has information GA4 does not: impressions, logged-in platform activity, delivery behavior, and modeled attribution within its system. View-through credit may reflect real influence when a person sees an ad and returns another way. It may also claim demand that would have converted without that impression. A standard attribution report does not settle causality.

It is also not proof that GA4 is “more accurate” in an absolute sense. Analytics is better suited to measured, session-based cross-channel comparison. Yet cookie rejection can produce an estimated 18–35% paid-traffic undercount, according to Adamigo. A smaller number is not automatically a truer number.

Nor should you assume the higher Meta total is always inflated. ATT-related loss can leave the platform underreporting true iOS conversions by roughly 15–30% in many accounts, with 30–50% possible for iOS-heavy audiences and weak server-side tracking, according to Cometly. Modeled credit and missing observations can sit in the same report.

Both tools can be working correctly while showing different totals. Meta is an attribution-based platform report. GA4 is a session- and cookie-dependent analytics report. Treat each as directionally useful for its assigned job, keep the business record above both, and investigate sudden changes in the gap instead of trying to reconcile it to zero.

Why don't my GA4 and Meta dashboard numbers match?

Read next

• Server-side tracking: what it fixes, what it costs, one structural way to recover conversion signals that browser-side tracking misses.
• Tracking and Measurement Audit, get a written diagnosis of event loss, duplication, attribution settings, and decision-risk in your stack.
• Field Notes, more practical guides on paid media and measurement.