From 15 June 2026, Google Analytics data collection is governed by the Google Signals consent setting, decoupled from the ad_storage consent signal that Google Ads now relies on exclusively. Until this change, most teams treated consent mode as a single lever: a user accepts or declines tracking, and both analytics and ads data move together. That assumption is now wrong. A user can decline ad personalisation while still allowing analytics, or the reverse, and GA4 and Google Ads will report different populations for what looks like the same traffic.
The practical effect shows up as a reporting mismatch nobody planned for. Conversion counts in Google Ads and session counts in GA4 start drifting apart for reasons that have nothing to do with attribution windows or click-through delays — they’re measuring different consented populations. Teams that don’t track consent state as a first-class dimension end up chasing a discrepancy that isn’t a bug in either tool, it’s two different consent signals doing exactly what they’re now designed to do separately.
Data Points to Track
- Consent state per session, split by signal: Google Signals consent and ad_storage consent captured as two independent values, not one combined “tracking allowed” flag
- Modelled vs observed event volume: the share of GA4 events that are consent-mode-modelled (statistically estimated) versus directly observed, since a growing modelled share means the visible numbers are increasingly an estimate
- Cross-tool population overlap: what proportion of users consenting to ad_storage also consent to Google Signals, to quantify how much the two datasets have actually diverged
- Geography and consent-banner variant: consent rates broken down by region and by which consent banner copy or design a user saw, since consent rate is now a UX metric as much as a legal one
- Reporting discrepancy size over time: the gap between GA4 session/conversion counts and Google Ads conversion counts for the same campaigns, tracked as a trend rather than investigated fresh each time it’s noticed
Setup Steps
- Confirm your consent management platform sends both signals independently to Google tag, rather than a single merged consent value, since the June 2026 split makes this a hard requirement, not a nice-to-have.
- Add a consent-state dimension to your GA4 reporting so every downstream report can be filtered by which signals a session actually consented to.
- Build a standing comparison view between GA4 and Google Ads conversion counts for the same campaigns, so drift is visible as a trend rather than discovered ad hoc during a monthly review.
- Track the modelled-event share over time and flag any sharp increase, since it usually means consent rates are dropping, not that GA4 is broken.
- Test each consent-banner variant’s effect on the Google Signals consent rate specifically, not just overall accept rate, since that’s now the number that determines analytics data completeness.
Actionable Insights
A widening gap between GA4 sessions and Google Ads conversions for the same campaign, with a stable ad_storage consent rate, points to Google Signals consent falling — the fix is in the consent banner and its Google Signals wiring, not in campaign targeting. A rising modelled-event share is a warning that an increasing share of your GA4 numbers are statistical estimates rather than observed events, worth flagging before anyone makes a budget decision off a headline number that’s partly modelled. And if consent rates vary sharply by banner variant, that’s a low-effort lever — a clearer consent prompt can directly increase the volume of real, unmodelled data flowing into every report downstream.
Related Resources
Need help tracking this in your app?
Our team sets up analytics pipelines for mobile and web teams every day. Talk to us and get your first events flowing in under an hour.
Talk to an expert