← Back to all Predictive Models
PREDICTIVEThis model is designed to predict the probability that a customer will become a top spender - defined as customers ranking in the top percentile for lifetime purchase/deposit amount over the next 3 months.
VIP Program Development
Identify potential high-value customers for exclusive tier programs
Strategic Customer Investment
Enable proactive targeting to:
- Nurture promising customers before they reach top-spender status
- Allocate premium support and personalized experiences
- Invest in relationship-building with future high-value prospects
- Population Lifecycle Stages: Customers in all lifecycle stages besides "Dormant"
- Training: Runs every iteration (unless data drift/changes require retraining)
- Inference: Runs daily in the daily process
The following outputs are stored in the customer profiles (as internal fields):
- Becoming Top Spender Score: Probability of this customer to become a top spender within the next 3 months
-
Percentile rank in Becoming Top Spender:
Percentile rank among all customers by Becoming Top Spender Score.
On a scale of 1 (lowest) to 100 (highest) -
Permille rank in Becoming Top Spender:
Permille rank among all customers by Becoming Top Spender Score.
On a scale of 1 (lowest) to 1000 (highest), providing higher-resolution segmentation than the percentile ranking -
Percentile rank by LCS Becoming Top Spender:
Percentile rank among same lifecycle stage by Becoming Top Spender Score.
On a scale of 1 (lowest) to 100 (highest) -
Permille rank by LCS Becoming Top Spender:
Permille rank among same lifecycle stage by Becoming Top Spender Score.
On a scale of 1 (lowest) to 1000 (highest), providing higher-resolution segmentation than the percentile ranking
- Training Window: default = 20 iterations
- Prediction Horizon: default = 6 iterations
- Catboost Classifier, which is the state-of-the-art model for tabular data
- Individual models per lifecycle stage for optimal accuracy
- Hyperparameters tuning in the first model training for optimal model parameters
Data Cleaning & Imputation: Handles missing values, removes constant features, and groups rare categorical values
Feature Engineering & Encoding: Converts categorical variables to numerical format using target-based encoding and applies scaling transformations
Feature Selection & Optimization: Removes low-variance and highly correlated features, selecting the most predictive variables
- Historical customer profiles (customer activity data, demographic data, campaign history, etc.)
- Minimum 7 iterations of historical data required for training (1 iteration for feature generation and 6 iterations for target definition, given the 6-iteration prediction horizon)
- Minimum number of 500 customers per lifecycle stage required for training
- Minimum number of 50 customers per lifecycle stage that became top spenders