Which Google Analytics feature employs machine learning to measure conversions that aren’t directly observable?

The selected answer, conversion modeling, is correct because it leverages advanced machine learning techniques to estimate and attribute conversions that cannot be tracked directly due to various limitations, such as user privacy constraints and limitations in cookies and device tracking.

Conversion modeling predicts the likelihood of conversions by analyzing user behavior patterns and demographic data, allowing businesses to understand and optimize their marketing efforts without needing direct attribution for every single interaction. This method enhances the accuracy of conversion data, providing insights into how different channels contribute to overall conversion performance.

Event tracking, user-ID, and audience triggers are all important features in Google Analytics, but they do not have the same machine learning capabilities as conversion modeling. Event tracking focuses on monitoring specific interactions on a site, user-ID is intended for tracking individual users across devices, and audience triggers are used to activate marketing tools based on user behaviors. Although these features provide valuable data and insights, they do not estimate unobservable conversions in the same predictive way that conversion modeling does.