My colleague Amit Parekh recently hosted a webinar about decision optimisation, which got me thinking about best practices for financial service organisations.
To recap, decision optimisation helps risk managers answer complex questions like:
- How do I run a profitable line increase campaign whilst reducing losses and exposure?
- How do I choose the best collections treatment for delinquent customers, to maximise repayment whilst minimising attrition and costs?
- How do I make sure I am setting the most appropriate initial credit line for my credit card applications?
Decision optimisation differs from traditional predictive analytics (e.g., behaviour scores) in that it seeks to make the best or optimal decision for a particular problem given a set of circumstances and constraints. For example, what is the best initial credit line for a particular customer given the information we know about him/her?
Custom decision optimisation has been successfully implemented for more than 10 years - FICO has worked on over 100 projects across multiple decision types - and is often chosen by large organisations with complex requirements and goals. However, it can also be leveraged quite effectively by organisations that are smaller, have less complex requirements or simply wish to make a quicker impact on their portfolios through more advanced strategy design.
In fact, we have worked to implement optimisation solutions with a number of clients who have witnessed positive result indicators after an 8-10 week project. The FICO solution—which we call Quick Start—simplifies the underlying decision model whilst still addressing the company's key drivers of customer profitability, such as revenue, losses, costs and attrition.
This, in turn, ensures that a project captures the majority of the benefits from a full custom optimisation project in much less time. It can be implemented for a range of decision types:
- Credit Line Increase (CLI)
- Initial Credit Line (ICL)
- Instalment Loan Price and Amount
- Early-Stage Collections
- Accept/Reject for originations
FICO clients are clearly seeing value when undertaking such projects. For ICL projects, clients have seen increases of up to 9% in revenue and 7-40% in profit. For CLI projects, we’ve see gains of up to $12.36 in profit per account for a UK bank.
For organisations looking to make best use of limited resources, maximise profits and stay ahead of the competition, decision optimisation is definitely worth considering.
