FEATURE: Using AI to improve consumer lending outcomes and drive financial inclusion 

7th May 2024

With the rise of artificial intelligence (AI) and machine learning, the way consumer loans have been offered and priced has been quietly revolutionised, opening up opportunities for more consumer-centric, personalised products.

As a data scientist and advanced analytics consultant to the banking industry, I’m fascinated by modern AI tools and their ability to give pricing committees a much deeper view of their lending portfolios and empower them to influence strategic outcomes.

For example, AI-driven pricing analytics and price optimisation tools can enhance loan personalisation, create greater transparency, and widen financial inclusion. This, in turn, helps address the need for value, a key outcome of the Consumer Duty from the Financial Conduct Authority (FCA).

There are several distinct ways in which AI-powered pricing analytic tools can fulfil growth, profitability, regulatory compliance, customer inclusion, and other strategic objectives. Let’s explore some of these in more detail.

Price optimisation – getting the best outcomes for consumers 

Customer expectations in terms of credit provisioning have evolved significantly, leading to considerable disruption and innovation in the lending sector. Higher expectations and greater competition have made it essential for banks to keep up with the latest technologies.

At the same time, consumers are dealing with increased living costs and interest rates that have remained at a 15-year-high since the start of 2024 in a continued effort to curb inflation. When it comes to seeking credit – be that unsecured loans, credit cards, or car finance – whilst consumers have more choice, they also have more nuanced needs.

Personalising your offers and optimising pricing can give you, the lender, a competitive advantage whilst finding the right financial fit for each customer. To do that, your approval processes must be as streamlined and transparent as possible, so you can present customers with pre-approved choices at the point of decision and tailor your offerings to their preferences and circumstances.

It’s crucial to develop AI-driven strategies based on real market, competitive, and customer data, allowing you to test your models and adjust the results accordingly. Different customers will have different priorities and behaviours. The advantage of AI-powered risk-based pricing is that you can fine-tune your products in line with a particular customer profile or the current market conditions. By offering a more tailored product, you’re producing a better customer outcome, increasing the likelihood of loan acceptance and approval, as well as setting the foundations for a long-term relationship with individual consumers.

Personalising loans to widen access to finance and help vulnerable customers  

Analytical modeling is also a powerful instrument in identifying and servicing marginalised customers where traditional underwriting processes might not have been so responsive. For instance, in consumer lending, banks could deliver personal unsecured loans for small amounts to customers who may have a thin credit profile or lower credit score, and normally may not qualify for a loan.

There could be a number of reasons for this, such as a change in circumstances or because they’re self-employed with a variable income. It might be that they’re new to the country and have a strong credit profile in their native land that isn’t recognised within the UK system. When fixing the interest rates (and fees) for such loans, several operational obstacles are evident in traditional decision-making processes.

First, an accurate evaluation of the risk of default for these customers might not be available to the bank or finance provider if they’ve been traditionally bundled together with other customers requesting slightly larger loans and/or with better credit scores. Credit risk officers and pricing managers have a difficult time evaluating the true costs of such loans to the lender, leading to increased rejection rates or lengthy and manual underwriting processes.

Second, loan quoting engines for different channels might not support the level of granularity needed to enable differential pricing and underwriting for such specific segments, especially if the criteria by which segments are defined will vary over time (for example, the level of average disposable income over the last three months).

And lastly, understanding the true repayment ability of the customer might be limited by past exclusions of such customers from the base for which loans were offered.

Loan personalisation is about delivering the right amount with the right terms to the right customer at the right moment. AI has a transformative impact by considering these variables and setting the parameters of the loan.

AI-driven analytics enables lenders to optimise their pricing strategies and fine-tune their products in line with a particular customer profile or the current market conditions, so they can put the most relatable and personalised offer to each applicant. It’s a win-win approach, as by meeting customers’ needs, the provider will deepen banking relationships and build loyalty.

A number of banks have already reported impressive results using such an approach. At one UK high-street bank, real-time personalisation contributed to a 300% increase in loan sales among mobile users, whilst a leading European bank reported a 9% volume increase in unsecured loans in the first year alone.

Using technology to support compliance and enhance transparency   

The use of such tools can also support compliance commitments. Credit providers in the UK are dealing with intense regulatory pressures, with recent research among C-level executives in financial services showing regulatory compliance is their greatest reputational concern. Technology, however, can and should be supporting compliance, if fully operationalised across all areas of banking activity.

With the implementation last year of the Consumer Duty, which is described as possibly the most extensive piece of regulation in defining – or redefining – the relationship between customers and financial services providers, there’s a renewed focus on how pricing is calculated.

Lenders must ensure products and services meet the customer’s needs and offer fair value. This brings with it a need for greater transparency in terms of how prices are set so that customers can understand the products and rates offered.

Although still in the early stages of implementation, the effects of Consumer Duty are already evident. Working with many UK banks, I’ve seen great examples of major investments to improve customer service and promote greater consumer understanding within lending practices.

Improving pricing processes with deeper insights 

To enable such improvements, an unseen transformation continues to take place in the back office, where advanced analytical tools and automated models are being put directly in the hands of pricing managers and decision-makers. This represents a significant change from the standards of the past where all analytical tools required heavy involvement from the IT and data science teams.

Today, AI-driven technology has matured to a point where pricing managers are able to implement quick adjustments and better personalisation of the products and services such as interest rates and fees. These modern AI tools and models directly impact the product catalogues, and their pricing, and hence are integrated with the main systems within the bank. This transformation started almost a decade ago.

Today, it can empower product owners and pricing managers to focus on customer inclusion, developing more personalised products and services for their customers, as well as offering better alternatives in line with economic conditions.

In my role as a data scientist, I keep witnessing modern analytics offering better visibility and data-driven insights to the decision-makers within banks. This leads to increased transparency in customer communications and greater perceived value in the interaction between the bank and the customer.

How banks benefit from harnessing new technology 

The key to success isn’t only in deploying advanced technology, but also in operationalising the analytical tools and processes available to banks today, keeping in mind the actual users of such tools. AI-driven models and automated processes can be total game-changers only if they empower business users instead of making them rely on IT teams as they have done in the past.

Such models and processes must be flexible enough to be changed at high frequency with minimal back-office and IT requirements. When deployed correctly, the results include fascinating improvements in speed-to-market, customer segmentation, and increased pricing granularity, helping credit providers decrease risk whilst driving growth.

Future-proof your processes: the consumer-first mindset 

In terms of the Consumer Duty, the FCA is yet to provide its view on how successful the initial implementation has been, but it’s an ongoing process, not a “once-and-done exercise”.

Providers will need to demonstrate they have thoroughly embedded the consumer-first culture and put continuous monitoring in place.

Price and value are core components here. Lenders need to be able to demonstrate that customers are paying a reasonable price for the product they’re taking, and show that the product is aligned with the market and fits the customer’s specific financial profile.

Pricing analytics and optimisation can help credit providers do just that, consistently and systematically, whilst leading to overall greater customer satisfaction, building trust, and supporting long-term relationships.

Giovanni Oppenheim, Director of Banking Solutions, Earnix   

To find out more about the real-life applications of AI-based analytics in consumer lending, visit https://earnix.com

Giovanni Oppenheim is a Director, Banking Solutions at Earnix – a global provider of AI-driven dynamic pricing, product personalisation, and digital decisioning solutions. For almost a decade, Oppenheim has headed up the delivery of analytics and implementations for global financial institutions at Earnix, including consumer lending, car finance, cards, mortgages, and more.