Models / Customer Analytics & CRM
ETB Conversion Propensity
Predicts which existing customers take up a further product in 90 days.
AUC (out-of-time)
0.784
Top-decile lift
3.1x
Segment forecast MAPE (1 month)
4.8%
Segment forecast MAPE (4 months)
9.6%
1 more metric
Next-product top-3 accuracy
0.72
About this model
Estimates, for each existing-to-bank (ETB) customer, the probability of taking up a further product in the next 90 days and which product is most likely. Customer-level scores roll up into conversion forecasts by segment, so Consumer Banking, which owns the model, can size campaigns and set RM targets. Built with Group Data & AI; business owner Mei Ling Tan. The model behind the ETB conversion analytics use case in the Enterprise Data Platform programme.
Intended use
Selecting and sizing ETB cross-sell campaigns, forecasting conversions by segment, and feeding the next-best-offer and lead tracking flows. Not used for credit decisions or pricing; any credit product it suggests still goes through normal underwriting.
Training data lineage
Trained on customer-360-mart monthly snapshots (Jan 2023 to Mar 2026) joined to casa-transactions behaviour features and campaign-history responses, read from the lakehouse gold layer in BigQuery. Features are served from the Vertex AI Feature Store; the model is trained in Vertex AI Pipelines, registered in the Vertex AI Model Registry and monitored for drift by Vertex AI Model Monitoring. Tier 2 model, validated by Model Risk Management.
Limitations & bias notes
Customers with under 6 months of history score close to the segment average. Salary-credit features dominate, so customers paid into another bank are under-scored. Input drift is monitored weekly; a campaign or product change can shift the feature mix quickly, and retraining is expected when PSI on key features passes 0.25.
Ownership and sensitivity
Who approves access
- Owner, Customer Analytics & CRM, Mei Ling Tan
- Data Governance Office, Ahmad Faizal
Customer-level data under banking secrecy and PDPA, and models that decide on customers.
Named-user entitlement with the data owner's consent and a stated purpose; reviewed on expiry.
- Banking secrecy
- Customer account information protected under the Financial Services Act 2013.
- Customer PII
- Contains or processes personal data about customers (PDPA 2010).
Try it
Live sandboxFeedback
2.0(1)Details
- Updated
- 2026-04-20
- Latest version
- v2.1
- Licence
- BU Restricted
- Access
- Restricted
- Framework
- XGBoost
- Language
- N/A