Cookieless Tracking in GA4: What's Possible Without Consent

What GA4 Collects in Cookieless Mode

When analytics_storage is denied, GA4 sends a stripped event payload called a cookieless ping. This ping contains no user identifier, no session identifier, and no personally identifiable information.

It does contain: the measurement ID, the event name, a timestamp, the page URL, the user agent, and a consent state indicator (the gcd parameter).

These pings are used by Google's machine learning models to infer aggregate behaviour patterns for non consenting users, feeding into the conversion modeling that fills in the gaps in GA4's reported metrics.

Importantly, cookieless pings do not appear in GA4's standard reports as discrete events.

They are consumed by the modeling pipeline and their influence surfaces as modeled conversions and session estimates rather than as real event counts.

This means you cannot query cookieless pings directly in the GA4 interface or in BigQuery, they are an input to modeling, not raw data you can analyse.

What You Lose Without Cookie Consent

The capabilities that are entirely unavailable without cookie consent are significant. User recognition across sessions is gone, without the _ga cookie, each session from the same user is treated as a new anonymous user.

This means metrics like sessions per user, user lifetime value, and return visitor rate are unavailable or inaccurate for non consenting users.

Audience building is unavailable, users who do not consent cannot be added to GA4 audiences for remarketing or personalisation.

User explorer reports, which let you drill into individual user journeys, are unavailable for non consenting users. Path analysis in Explorations is available only for the fraction of users who consented.

For properties with high non consent rates, above 40 to 50%, these gaps mean that GA4's individual level analysis capabilities are available for less than half the user base, making behavioural insights based on Explore reports potentially unrepresentative of the full audience.

Strategies to Maximise Signal in High Non-Consent Environments

Three strategies help maximise measurement quality when a significant portion of users decline cookies.

First, improve consent rates: a well designed CMP with a clear value exchange for consent, explaining what analytics data is used for and offering users a genuine reason to consent, can meaningfully increase acceptance rates.

Second, invest in server side first party analytics that does not rely on cookies for basic session level measurement.

Server-side request logs provide counts of pages served, referrer data, and basic geographic distribution without any client side tracking.

Third, use the GA4 Data API or BigQuery to quantify the modeling contribution, comparing observed vs modeled sessions over time gives you a sense of the reliability of your reported metrics.

Properties where more than 30% of reported conversions are modeled should treat any conversion based optimisation with appropriate caution and validate against backend order data before making significant budget decisions.