Predictive Analytics (Marketing)
Predictive analytics in marketing uses statistical models and machine learning to derive probabilities for future behaviour from historical user data, such as the likelihood that an individual user will buy or churn. GA4 provides ready-made predictive metrics for this: purchase probability, churn probability and predicted revenue.
In practice
In practice GA4 requires a minimum data volume for predictive metrics: according to Google, within a seven-day window in the past 28 days at least 1,000 returning users must have met the relevant condition and at least 1,000 must not have met it before a model is activated. If a property clears that threshold, you can build audiences on top of it – for remarketing to users with a high purchase probability, or win-back campaigns for users at high risk of churn. Smaller websites with little traffic frequently fail to reach these thresholds.