Balancing High Tech & High Touch in B2B Credit Decisioning

Trade credit is a popular financial tool in any environment. In the U.S. alone companies owe $3.1 trillion in account receivables on any given day. While that’s a massive number, as the B2B ecosystem experiences friction in external markets and lending tightens as a result, those numbers may soar even higher as buyers turn to suppliers to free up cash flow and sustain operations.

And with supply chains experiencing major disruptions and consumer demand falling fast, increasing activity in trade credit could prove to be the ultimate stress test. The results of that test may not be far off, as many expect that we could soon see an unprecedented rise in trade credit insurance claims as buyers become unable to meet their financial obligations.

For some time, finance teams within the B2B world have been taking a nod from big banks and other lenders that are increasing their reliance on artificial intelligence to not only help automate the end-to-end credit process, but to actually make smarter lending decisions. With uncertainty abound, the predictive ability of artificial intelligence and machine learning is indeed a welcomed sight, but does this mean that humans should be cut from the credit equation?

Mitigating Risk With Machine Learning

At a high level, machine learning has an unparalleled ability to digest large sets of complex data to pinpoint the factors that denote risk. By applying those factors to equations and models, machine learning is able to objectively predict the level of risk involved in a given lending situation in ways that human-based models and processing power cannot.

Aside from its predictive ability, and knack for saving human teams from crunching massive amounts of data, machine learning has also been lauded for its ability to extract human bias from the credit equation. Historically, human discretion and the bias that comes with it has permeated lending. Lenders often use human discretion to gain a competitive advantage by taking on more risk that leads to greater rewards and strengthening customer relationships. But of course, risk appetite is not in large supply for most of today’s B2Bs.

However, before we write off the credit analyst or the paradigm of relationship-based lending, we must remind ourselves that machine learning isn’t infallible. Credit models are not 100% accurate in the best of times, and when models are created under far better economic conditions than the unprecedented times we find ourselves in, robot analysts are surely in for a test of their own.

And while more sophisticated machine learning models can account for evolving circumstances and an increasing array of often overlooked factors, such as a company’s social media activity, there are some things that only humans can account for.

Striking a Super-Human Balance

Nothing is more important to buyers and suppliers than their relationships themselves. And as supply chains dry up and consumer demand continues to wane, those relationships become even more crucial. This is where the B2B credit analyst and A/R teams, and the human discretion they bring to the equation, become more important than ever before.

While credit scores and financial statements can be extremely reliable signals, things like buyer history, and a deep understanding of a buyer’s business, are factors that models can’t always compute. Similarly, when models and formulas may otherwise preclude certain lending decisions, human knowledge and subjectivity can lead to very favorable outcomes.

Aside from predictive ability, humans also enhance the overall customer experience (CX) when it comes to credit. Similarly to how technology has allowed the B2B collections process to transform from a friction-filled process to a strategic relationship builder and CX enhancer, B2Bs have the opportunity to leverage the best of high-tech and high-touch to take credit to the level.

While economic uncertainty is indeed a challenging time for business leaders and their teams to weather, for many, it can also be a time of great opportunity. And one of the greatest opportunities that exists is forging bonds now that last far into the future. Companies are reliant on their partners now more so than ever before, which means customer experience is more critical than it’s ever been.

AI and ML are smart, but they aren’t a silver bullet. They can’t completely remove risk from the equation, can’t always identify opportunities that humans can and can’t create a remarkable CX on their own. When it comes to credit, businesses should look to this type of technology as a supplement to human teams, not a replacement. This superhuman approach will allow machines to do what they do best and empower humans to navigate the challenges of today in ways that pay dividends in the future.

Derek Bluestone is EVP of Product for Billtrust, a provider of B2B payments services based in Lawrenceville, N.J.