Google’s AI-powered modelled data refers to the machine learning techniques Google uses to estimate conversions and user behaviour when direct measurement is not possible due to cookie consent restrictions, browser privacy features, or cross-device gaps.
When users decline cookies, Google Consent Mode v2 sends cookieless pings that provide limited signals to Google’s servers. Google’s AI then uses these signals, combined with patterns from consented users, to model the conversions that likely occurred among non-consented users.
This approach recovers significant measurement data that would otherwise be lost, without processing personal data from users who declined consent.
Google’s modelled data becomes more accurate as the volume of consented users increases, as the models are trained on observed patterns from opted-in users and extrapolated to opted-out users. Accuracy improves with higher consent rates and larger traffic volumes. However, modelled data is an estimate, not a precise measurement.
Businesses should understand that modelled conversions carry inherent uncertainty and should be interpreted alongside direct measurements. Google reports modelled data separately in its analytics and advertising platforms, allowing businesses to distinguish between observed and estimated conversions.
Seers’ integration with Google Consent Mode v2 is the key enabler for Google’s AI-powered modelled data. By properly implementing Consent Mode through Seers’ CMP, businesses ensure that the necessary cookieless pings are sent to Google when users decline consent, providing the raw signals Google’s AI needs to generate modelled conversions.
Seers.ai makes this complex technical integration straightforward, helping businesses maintain measurement accuracy while fully respecting user privacy.
Turn consent gaps into actionable insights with privacy-first modelled data from Seers AI
START FREE TODAY