In today’s post we’re going to look at the prerequisites required to apply predictive audiences.
In order to apply the predictive audiences, there are several prerequisites. If they are not met, there would be insufficient data to use predictive modeling. If the data meets the correct conditions to predict the behaviour of the audience, the predictive metrics become available, such as determining purchase probability and churn probability.
The first requirement is to configure the purchase event and send it to the GA4 property.
The second requirement is that there must be a minimum of 1000 positive and 1000 negative samples. This means there should be at least 1000 users in the website who have triggered the predictive condition to purchase, and 1000 users who triggered the predictive condition to churn.
The last requirement is model quality and regular traffic generating purchase events must be sustained over a set period of time. That means the traffic requirement as stated above must be sustained over the set period of time.
Don’t want to wait for part 3? Get all parts of this series, including how to set up predictive audiences in up in GA4, FAQ’s and even a case study? Download Napkyn’s Ultimate Guide to Predictive Audiences in GA4.