Early adopters in this space are already at the point where their projects are bearing fruit. This de-risks the prospect of an open banking project. So without further ado - here’s the latest data we have.
The short answer is between 12% and 78% depending on a function of the value-add of the proposition and the UX of the process. That’s not a particularly helpful answer so check out the graph above for uptake on our live affordability projects since the beginning of 2021. There’s a clear bell curve around the middle two quartiles with a median of around 62%. For more on maximising your uptake, check out the article we wrote on this here.
This has shifted significantly over the last few months. As things stand, we have about 15% of organisations delivering API - only integrations, 30% using dashboards only, and around 55% using a combination of both in some way. Most of the organisations on dashboard-only projects have deployed these as a precursor to some form of combined solution and around half of the organisations using the combined solutions are in the process of moving to an API-only integration. We’d expect this trend to continue for at least the next year as clients become more familiar with open banking data.
Time: The most common key metric we see is around the reduction in time spent manually processing applications. Clients with at-scale use-cases are reporting time savings around the 40% point for applicants that use the Open Banking channel. Interestingly, this uniformity is present across the lending spectrum from unsecured to mortgages.
Acceptance rate: This shows an increase across the board though the extent of that increase varies widely based on the customer profile of the organisations concerned. Where a lender has a traditionally prime customer base, we’re seeing net improvements of around 10%, normally with regards to customers with good disposable income from non-traditional sources. At the other end of the spectrum, we’re seeing net acceptance rates for organisations with a skew towards thin-file users increase by around 60%.
Risk reduction: The data here is less clear as we only have access to a couple of controlled studies and both are still relatively early stage. However, at the low end we’re seeing a reduction in default rates of around 40% vs the predicted default rates of those organisations when modelled on CRA data only for the same cohort. In another study with a thinner-file user base, this same reduction was at 75%
Most organisations start out with relatively simple data points outlining really fundamental affordability criteria like total income, net disposable income and basic risk factors (Gambling / Cash withdrawals etc…). As organisations become more comfortable with the accuracy of AI classified data, their scope tends to broaden with common factors including loan management and priority spending data.
We hope this post is useful if you’re looking at how to orient yourself in those early stages but if you’d like more details, don’t hesitate to get in touch. We’d love to talk.