Question

You operate a deployed churn-prediction model for a subscription business. Input feature distributions monitored by Vertex AI Model Monitoring have remained stable for months, and there are no alerts on feature skew or feature attribution drift. However, the business team reports that model accuracy, measured against actual churn outcomes as labels arrive, has steadily declined over the last quarter. A recent pricing change altered which customers tend to churn. What is the most accurate characterization of what is happening, and what should you monitor going forward?

Accepted Answer

This is concept drift — the relationship between features and the target has changed; monitor model performance metrics against ground-truth labels and trigger retraining when accuracy degrades.

Answer

This is training-serving skew caused by inconsistent preprocessing; add a Dataflow transform step and compare training versus serving feature statistics.

Answer

This is data drift in the input features; increase the feature-drift alert sensitivity thresholds in Model Monitoring to catch the shift earlier.

Answer

This is prediction drift; monitor the distribution of the model's output scores and alert when it deviates from the training baseline.