Accounting for customer transactions, discounts, and returns, predictive model attributes anticipate the potential lifetime value of a profile through their considerations of churn risks, predicted orders, and spending amount.
In this article, you will find a comprehensive list of our predictive model attributes, their definitions, and their types.
The following table explains the predictive attributes available in the hub.
The profile’s predicted spend in the next 12 months based on their historic purchases.
The profile’s predicted number of orders in the next 12 months based on their historic purchases.
Predicted Order Value
The profile’s predicted order value for next 12 months based on their historic purchases.
The profile’s likelihood of churning based on their historic purchases, represented as High, Medium, or Low. Typically used to suppress high risk customers.
Lexer's inferred gender, based on the person's first name, scored against the probability that the name would be given to a male or a female.
These attributes can be so powerful when used as part of customer growth and retention strategies, and the video below showcases this.
And that's it! Now you know which predictive model attributes are available, what data is used to create the recommendations, and how the recommender model uses this data to allow you to analyze your segments effectively. You can use these attributes to build your segments for activation and engagement. If you need any further help please reach out to Lexer Support ([email protected]).
Updated 13 days ago