Why E-Commerce Needs a New Way to Assess Creditworthiness

By Christian Mangold, CEO Fraugster

Key Takeaways:

Credit checks are designed to answer a simple question, “Can I trust this person to repay the amount they are borrowing?”. Traditional bureaus such as Equifax, Experian, Transunion, and major regional players like SCHUFA perform regular credit checks on customers to assess their total debt, repayment history, previous types of loans, and length of credit history.

Such measures of creditworthiness work well when used in the case of high-value personal loans or mortgages with elongated interest-based repayment cycles. However, they are not fit to assess e-commerce risk against low-value purchases. The use of such traditional credit checks adversely affects BNPL services by driving false positives, service denials, and increased costs.

A customer’s ability to repay $80 of credit against an online sneaker purchase is bound to be different from their ability to repay an interest-based home loan spread over ten years. Shorter, interest-free repayment cycles further differentiate the risk of default faced by BNPL providers compared to traditional forms of credit. Given the rapid adoption of BNPL as a form of payment, providers need new ways of assessing creditworthiness that are better suited for their use case.

The fact that somebody failed to pay parking fines 3 years ago says nothing about the readiness to shop regularly for sneakers at the favourite e-store

The problem of false positives and potential loss in revenue:

The reliance of credit scores on historical data discriminates against people who have availed little or no credit in the past. This is especially prominent amongst younger cohorts and marginalised communities. Over 100 million adults in the EU are unbanked and thus credit invisible. This can have a significant impact on BNPL service providers where younger cohorts form a major consumer group.

Additionally, traditional credit scores don’t travel well, which is to say they are not always portable from one country to another. An immigrant with an exceptional credit score in their home country may arrive in a new country and be credit invisible. This magnifies the potential of false positives and a loss in potential customers for BNPL solutions.

Discrepancy in data:

Credit reports and scores often differ between credit bureaus. This is because a credit report is based on the information supplied by various lenders, and they may not report every piece of information to all the credit bureaus. Further, soft credit pulls performed by BNPL providers currently don’t report the most recent activity of the borrower. These gaps in data make decisions riskier and less accurate.

Recency and reporting issues

Missed payments currently impact a borrower’s credit score for as long as seven years, and therefore could be classified as a lagging indicator of a consumer’s risk. Due to this a perfectly qualified borrower would be rejected from accessing credit for a causal-impulse purchase online, thus presenting a loss in revenue for BNPL service providers.

Risk of fraud

BNPL solutions have been prone to ATO and synthetic identity attacks by fraudsters. This can lead to a significant fall in consumers’ trust in the service. To stay a step ahead, BNPL providers should not only assess the buyer’s risk of default but also verify their identity to prevent losses.

Additionally, performing credit checks also proves to be an expensive exercise for BNPL providers, often eating up a significant share of the trade margin. This adversely impacts the profits of BNPL providers, especially given the expected growth in processing volumes in the years ahead.

Need for alternative data sources to decide on offering BNPL:

Compared to traditional forms of credit, BNPL offers a unique proposition for buyers to avail quick and easy credit. Given this differentiation, solution providers need to look at additional data sources that better suit their use case and assess a consumer’s true financial position more accurately. Some of these alternative sources and entities include:

• Digital Footprint (Device Attributes), email address which provides insight into historical purchase behaviour
• IP address and GPS coordinates to better track and identify customers and the location of where the transaction is being initiated from.
• Account History data points such as customers’ total spending, unpaid amount, etc

In addition to these, artificial and network intelligence can be leveraged to develop key insights from networks of returning buyers, allowing for more accurate credit decisions even for guest users. These technologies not only analyse more data points compared to traditional scoring but also enable BNPL providers to focus on positive transaction history, allowing them to approve more good customers than ever before.

The World Bank projects that alternative data could help provide formal financial services to up to a hundred million more adults. This presents a key opportunity for BNPL providers to benefit from more up-to-date, real-time risk assessments which aid in enhancing their approval rates and avoiding false positives.