Increasing profitable marketplace lending with Affordability Engine

The lender tested how Infact's Affordability Engine – a frictionless, responsive, performant API built to cater for marketplace volumes – would impact their lending performance relative to their incumbent solution.

The data showed a transformed pre-approval process, achieving a potential 4.23% increase in lending.

Client: A major UK personal lender

Objective: To achieve at least a 1% uplift in the affordability and income verification pass rates

Background: The lender’s existing pre-approval process was multi-stage. At each stage, they anticipated drop-off as their hard-fail rules, then credit checks declined applicants.

The set up

The lender did not have confidence in the ONS-data used in their expenditure models and their existing model was causing more affordability-based declines than they felt was reasonable based on economic research.  

This was a frustration as it meant the lender was declining applicants who had good creditworthiness and excluding people based on poor availability of affordability data to drive a responsible lending strategy.  

In addition to this, they had a large percentage of applicants they could not verify income for using their CATO-based models, resulting in larger manual referrals.

What did Infact do?

Infact's Affordability Engine was tested against the incumbent solution to assess the potential impact on the lender's pre-approval decisioning process, and the net reduction in affordability-based declines by:

1. Replacing the in-house ONS-based expenditure model with expenditure components.  
2. Supplementing their existing CATO strategy with Infact’s income components, giving 100% coverage of applicants.
3. Creating a swap set inference to assess the positive swap-in vs. negative swap-out impact.

The solution proposed offered frictionless delivery of personalised income and expenditure metrics on 100% of applicants, giving the lender more nuanced affordability data and a higher degree of configurability and control over their model inputs.  

The lender also benefits from continuous model refinement ensuring output accurately reflects the current economic environment, while offering full explainability that improves transparency for customers, auditors, and regulators. Additionally, the solution reduces the cost per acquisition of new customers by driving down OPEX costs.