Challenge
- Traditional customer-level underwriting can now be augmented with transaction data to improve predictive accuracy of payment outcome
- Creditors with slow, batch-based decisions, platforms and strategies will be outperformed by players with more nimble decision and processing architectures with real-time access to emerging data and services
- Variable payment terms on new POS payment options / products inject new dimensions into traditional underwriting models
- Purchase content / purpose
- Term
- Down payment
- Location
- Device metadata
- Technology has now enabled options in card transaction authorization platforms – enabling real-time credit, fraud and affordability assessment which is deeper and more precise. This capability is been very sparsely understood or adopted.
Implications
- This is a further opportunity to build competitive advantage in risk-adjusted margin – both fraud and credit
- Deep expertise supported by leading data analytics is required to understand and implement these capabilities
- This capability is required to support POS finance at optimal level
- An API-centric, data driven technical architecture supporting lending decisions is required
- The best talent in data analytics and artificial intelligence require an environment with critical mass, forward-looking and bold ambitions and meaningful investment in data and training
Response
- Prepare for a technical architecture based on real-time events – without traditional batch processing. Design to lowest response time latency requirements – e.g., credit or fraud authorizations
- Centralize decision technology and event processing as the core technical asset – connected via API’s and streaming data events to any data or service, internal and external
- Build expertise in decision sciences, data analytics and technical platform as a competitive asset
