What is embedded finance? Four key challenges

In its most simplistic form, the term ’embedded finance’ is the integration of financial service products (insurance, lending, payments, upsell packages etc.) within a non-financial service customer journey. The most prevalent example and the primary focus of this discussion being the online buying journey and the provision of financing options at checkout.

“Point of sale” finance has existed for a number of years, for example with Auto Finance loans being sold through car showrooms. Many of us may have even funded fashion, furniture or white goods through in store finance. However, the digitisation of the buying journey, alongside a change in consumer preferences is driving the increased adoption of financed services. Consumers are beginning to move away from traditional payment methods, with many now expecting to be offered a range of personalised financing options within their buying journey. It is this change in distribution channel that makes the journey different, as while the core principles remain consistent, there are a number of factors that are unique to this flow, and these have to be understood to ensure success.

Embedded finance in numbers

A study conducted by Juniper Research[1] in November ’22 found that revenue from embedded financial services will exceed $183 billion globally in 2027, increasing from $65 billion in 2022, that’s an anticipated growth of 182%. Fuelled by the rapid growth of Buy Now Pay Later (BNPL) in the e-commerce sector, and the development of innovative technology solutions that allow financing options at checkout, the dynamics of the personal loan market are changing.

Our data tells us that 4.2 million unique customers have funded at least one purchase through a BNPL provider in the four months between December ’22 and March ’23. The ‘always on’ API driven network, which joins the checkout journey to the payments network and platform providers, has meant that more agile FinTechs have been able to capitalise on the opportunity presented by Embedded Financing. Given the size of the opportunity it is no surprise that we are now seeing merchants, tier one banks, and mono-line card issuers looking to build market share. But by the very nature of the journey there are a number of different challenges that have to be addressed.

Embedded finance – a different set of challenges

Eligibility

The need to deliver certainty in outcome without the introduction of unnecessary friction is an ongoing challenge, and one particularly prevalent in embedded finance. Consumers are demanding instant gratification, and while the provision of financing options at checkout has been seen to increase advocacy, increase basket size and drive conversion, any offer of credit must be fair and appropriate. The buying journey can also be highly emotive, and it is therefore important that where an offer or option for credit is presented, consumers understand the terms associated with it and implications of non-payment. Unlike traditional aggregator/marketplace flows where consumers can ‘shop’ the market to find the right offer for them, the dynamics of an embedded finance journey mean financing options are only shown where the lender has an agreement with the merchant (noting there are journeys where the finance is provided by the merchant, such as Telco and Mail Order, however this model is different again).

Where a panel of lenders, often managed by a 3rd party are at play, eligibility is key. Some merchants can provide up to 10 different payment or financing vehicles and those that get it right only show options where the customer will be accepted, and the associated terms meet their requirements. The presentation of several ‘unqualified’ offers can often be confusing and cluttered.

Consumers don’t want to go through numerous application flows before they find a provider that is willing to fund the purchase. This causes huge frustration, delivers poor customer outcomes and drives ‘cart abandonment’ which loses the merchant sales and negatively impacts consumer advocacy.

Over 50% of all new credit card agreements and personal loans are originated through an aggregator flow. This change in consumer behaviour is being driven by certainty in outcome, and we are seeing increased adoption of this capability within the checkout journey that is unique to the embedded finance flow.

Credit Risk

Where merchants work with finance providers on an exclusive basis, it is important they offer a range of products and services that are positioned to maximise conversion but also manage risk. When interacting with merchants in this way, lenders can’t control the “through the door” population in the same way they would if distributing via an aggregator, or for the larger tier one banks, on a cross-sell basis.

As such, existing scorecards may need to be recalibrated and, in some instances, we are seeing bespoke scorecards (which deliver uplift over our next gen Retail Finance scorecards) being built at both the sector (e.g. fashion) and merchant level which are reflective of their customer base. Lenders wanting to offer embedded finance options may not have performance data that reflects the potential customer base, and their previous use of retail finance. As such, anonymised data samples are often required which can then be aligned to the merchant’s customer demographic. Once the customer profile has been identified, these can then be matched to consumers who’ve taken similar products previously so the lender can use the data to build predictive models that drive the required level of offers while remaining within risk appetite.

Affordability

Similar challenges also exist with the development of an affordability strategy that is proportionate to the level of risk associated with the product being taken. To deliver the ‘straight through’ journey that consumers and merchants demand, it is imperative that providers make the very best use of the data-sets available to them. To achieve this, we are seeing providers vary their approach by basket size, and the product to be offered (both in terms of monthly contribution if revolving credit, alongside payment and term for fixed).

What is common across both is the compliant use of Current Account Turnover data (CATO) and the insights this can drive. Whether this be through the verification of a minimum income, or the testing of disposable income against residual Current Account balance once expenditure and commitments have been considered. Even within the most optimised strategy there will always be customers who have insufficient information on the credit bureau to perform a robust affordability test, and there will be others where additional information is required. Unlike traditional lending journeys, referrals aren’t an option, and this is where customer consented data through Open Banking or income verification at source (payroll) come to the fore. Although both of these options introduce a degree of friction for the consumers, it’s a necessary step to validate affordability for bigger ticket items.

Using the right API integrations, the assessment can take place in session, the consumer does not have to leave their device, and results are immediate.

Fraud

The quick expansion of embedded finance could offer new opportunities to fraudsters, looking to purchase goods in the hopes of less sophisticated fraud detection systems. In fact, the fraud profile of each transaction or buying journey is governed by basket content. For example, solar panels that are bolted to the side of a building carry a very different risk profile to high end electronics that can be moved on with ease. Consideration should also be given to the collection or delivery mechanism alongside the strength of relationship the customer has with the merchant. Alongside existing customer data, many online merchants have fraud profiling solutions of their own, this information when combined with behavioural or physical biometrics, and appropriate datasets layered with machine learning, can be used to develop powerful scores which drive a binary accept/decline decision.

We’ve seen the deployment of a number of ‘step-up’ challenges, whether this be through knowledge-based authentication or document verification. Although effective, both introduce friction, they take the customer away from their device which in turn drives higher abandonment rates and where the journey is completed, we observe a degree of negative selection. Understanding when it is right to introduce additional verification is an important factor in completion rates.