Start by importing your Google Analytics 360 data into a Google Cloud BigQuery dataset. A service account is used to ensure that MorphL has access to the BigQuery tables. MorphL will automatically begin ingesting the data. See documentation.
Depending on the model that has been selected, the generated predictions will be displayed for each individual client ID. Different thresholds can be defined to segment users into low, medium or high quartiles.
Every prediction can be pushed back into Google Analytics 360 and saved in a custom dimension. Custom audiences are defined using predictions thresholds and can be further used in 3rd party platforms such as Google Ads to run predictive marketing campaigns at scale. See documentation.
Customers that are at the risk of not returning for a new order in a given time interval.
Conversion rate Average order value CLTV
Automatically identifies the common characteristics of customers that completed transactions.
Conversion rate Average order value Cart abandonment
Users that are more likely to add a product to cart, go to checkout or complete a transaction.