Data Discrepancy
A data discrepancy is a mismatch between the same metric reported by two different systems, such as Google Analytics 4 showing 400 conversions while Meta Ads Manager shows 620 for the same campaign and the same dates. Discrepancies exist because every platform counts with its own rules: different units, different attribution windows, different time zones, and different amounts of data loss. Most discrepancies are methodology, not error.
Why Analytics Numbers Never Match Exactly
No two measurement systems count the same thing, so their totals cannot agree. GA4 measures sessions and key events on your site. Meta measures conversions it can tie to an ad interaction inside its own window. Those definitions do not overlap.
Five mechanisms produce nearly every discrepancy:
- Counting unit. Google Ads counts clicks. GA4 counts sessions. One person clicking an ad twice in ten minutes is 2 clicks and 1 session.
- Attribution rules. Each platform applies its own model and lookback window before crediting anything to a campaign.
- Timing. Ad platforms credit a conversion to the day of the ad interaction. GA4 credits it to the day it happened. A Tuesday purchase from Friday’s click lands in two different columns.
- Collection loss. Ad blockers, rejected consent, cookie lifetime caps, and tags that fail to fire remove events from one system but not the other.
- Processing. Google Ads filters invalid clicks before reporting them. GA4 does not. GA4 applies sampling and thresholding. Ad platforms do not.
Attribution Windows by Platform
Window and model settings cause the largest conversion gaps, and they are the easiest to check.
| Platform | Default model | Default window | Counts views |
|---|---|---|---|
| GA4 | Data-driven | 90 days, 30 days for acquisition events | No |
| Google Ads | Data-driven | 30-day conversion window | Engaged-view video only |
| Meta | Data-driven across Meta surfaces | 7-day click, 1-day view | Yes, 1-day view |
Two 2026 changes matter here. Google moved GA4’s default lookback window for acquisition key events from 90 days to 30 days in April 2026, so first-visit credit now expires sooner. Meta removed its 7-day view and 28-day view options on 12 January 2026, leaving four combinations: 7-day click with 1-day view, 7-day click only, 1-day click only, and 1-day view only.
One trap sits inside GA4’s default. Data-driven attribution needs roughly 400 conversions for that specific key event and 20,000 conversions across the property within the lookback window. Below those volumes, GA4 falls back to last-click silently. Two properties on “the same” default model can therefore use different models.
Meta’s 1-day view-through credit alone explains much of the Meta-versus-GA4 gap. GA4 has no concept of an impression, so a conversion from someone who saw an ad and never clicked cannot appear in its campaign reports.
What Counts as a Normal Discrepancy
Compare against the expected range before treating a gap as a bug.
- Google Ads clicks vs GA4 sessions: under 10% is normal. 10% to 20% is common and usually explained by back-button clicks, repeat clicks inside one session, and ad blockers. Above 30% points to a specific technical fault.
- Meta conversions vs GA4 key events: 20% to 40% is routine, driven by view-through credit and the 7-day click window.
Google’s Ads documentation lists the expected causes of the clicks-versus-sessions gap directly: invalid click filtering, GCLID stripping by redirects, tagging errors, users leaving before the page finishes loading, and users who block or opt out of tracking.
How to Diagnose a Data Discrepancy
- Align the date range and time zone. A GA4 property set to UTC and an ad account set to US Pacific will disagree on every daily figure. Check the property time zone under Admin, then Property details.
- Align the counting unit. Compare clicks to sessions, not clicks to users, and one key event to one conversion action.
- Match the windows. Set Meta to 7-day click with no view, then compare. If the gap collapses, it was attribution, and nothing needs fixing.
- Confirm the parameters survive the journey. Redirects, link shorteners, and app in-app browsers strip query strings. A click that arrives without its
utm_sourceorgclidbecomes direct or unassigned traffic in GA4 while the ad platform still counts it. - Check the collection layer. Safari caps JavaScript-set cookies at 7 days, and at 24 hours when the landing URL carries tracking parameters. Rejected consent and ad blockers remove sessions entirely.
- Count clicks independently. A redirect-layer counter registers the click before the browser loads a page or a tag fires, which isolates whether traffic is missing or only mislabelled. Link analytics gives you that count alongside the GA4 and platform numbers.
- Fix the ratio, not the row. Once the gap is stable and explained, record it as a reconciliation ratio and track deviations. Reconciling individual conversions is not achievable.
Common Data Discrepancy Mistakes
- Assuming one number is correct. Both are correct under their own definitions. The question is which definition answers yours.
- Treating every gap as broken tracking. A 25% Meta-to-GA4 conversion gap is expected. Chasing it wastes time a 40% clicks-to-sessions gap deserves.
- Layering manual UTMs onto auto-tagging. Manual UTMs override GCLID data in GA4, changing the source attribution and creating a new discrepancy.
- Comparing sampled data. GA4 Explorations sample above 10 million events on standard properties, and thresholding withholds rows when Google Signals is active. Neither is applied in ad platform reports.
- Summing platform-reported conversions across channels. Meta and Google Ads both claim orders they each touched, so the total exceeds real revenue.
Frequently Asked Questions
Why doesn’t GA4 match Facebook Ads?
Meta counts view-through conversions and GA4 does not, so Meta reports more. Meta also uses a 7-day click window by default and credits conversions to the day of the ad interaction, while GA4 credits them to the day they occurred. Meta measures across its own logged-in users on any device, whereas GA4 relies on cookies unless User-ID is configured. A 20% to 40% gap is normal.
What causes tracking discrepancies?
Five things: different counting units (clicks versus sessions), different attribution models and windows, different timing and time zones, data lost in collection through ad blockers and cookie limits, and different processing such as invalid click filtering or sampling. Most gaps come from the first two.
What is an acceptable data discrepancy percentage?
Under 10% between Google Ads clicks and GA4 sessions is normal, and 10% to 20% is common. Above 30% signals a real technical problem such as a missing tag, a redirect stripping parameters, or a slow landing page. Conversion comparisons between an ad platform and GA4 tolerate wider gaps, usually 20% to 40%.
Why is my conversion count different in every tool?
Each tool defines a conversion differently and applies its own window. Google Ads counts conversions credited to an ad click within 30 days. GA4 counts every key event from every channel. Your CRM counts closed orders. The three will never converge, so pick one source of truth per decision.
Can data discrepancies be eliminated?
No. They can be explained, reduced, and made predictable. Consistent UTM tagging, server-side tracking, and matched attribution windows shrink the gap, but the platforms measure different things by design.
To rule out tagging as a cause, validate your campaign links with the free UTM checker at linkutm before you compare reports.