All articles

Debt collection gets a vital assist from AI

Interested in how AI can streamline and transform the process of debt recovery? Here’s some insight into the ways Applica Robotic Text Automation (RTA) and robotic process automation (RPA) are raising the bar for profit and performance in this sector.

Automation of document processing is big news these days, and companies in any industry can benefit from taking paperwork and workflows into the future. After all, innovative and targeted AI solutions take human error out of the equation. And they replace it with superhuman – and scalable – accuracy and speed.

One branch of business that is especially primed for robotic-based solutions is debt collection. The high volume of incoming documents means that text processing and extraction of information are among the most important operations for a collector or collections department. Names, aliases, locations, addresses, dates, debt types, notification histories, docket numbers, verdicts, dismissals, court fees, credit ratings, bankruptcy histories, collateral details, bank account numbers, and so on. All of this is important and much of it is easy to miss, mistype, or mislabel. And because so many of the documents in question are court-issued or otherwise legally binding, any mistakes in processing are not only potentially costly, but also fraught with risk of non-compliance or outright illegality. Luckily, implementing AI eliminates the problem of human error, and thus also much of the risk.

But there is more. In the world of debt collection, a significant number of documents require action on a tight deadline, sometimes within a day or days of filing. However, the influx of these documents is not constant. Peaks occur. Thus far, collectors had to allow for fast turnaround even during such peak times, either by maintaining maximum-capacity workforces full-time, or by outsourcing express work in special circumstances. And when such peak times result from events that affect the whole industry all at once, in-demand outsourced help can be scarce and expensive – or outright unavailable, with so many companies struggling with pandemic-related staff shortages or lockdown bureaucracy. This problem of “enough hands-on deck” may well be the number one challenge faced by every company and department in the sector.

The work performed by people is not scalable, period. More documents, more scrutiny, more data points and more decision-tree sorting means more time. The minutes or hours it takes for a person to process a document can turn into days and weeks when a backlog occurs, even with many people working overtime. A big collection agency typically receives upwards of a million documents every year. Often, it’s several million. The amazing advantage of AI consists in its astonishing speed and scalability. Documents get classified, analyzed, and sorted in seconds. And this is as true for a dozen documents as it is for thousands of them at a time. In fact, this is another way that the right machine learning-based solution eliminates risk. In this case it’s the risk of crucial deadlines being missed, a vital thing to watch out for in this industry.

Once the decision is made to consider automation, some challenges to implementation are organizational and cultural. Will the CEO sign off on such a departure from the status quo? Will employees thrive in new roles that challenge and stimulate them? Will the numbers add up to boost profit enough to sway the more conservative members of the board?

Other challenges are functional in nature, and can come as a surprise to unsuspecting clients. These include the burden of preparing the data sets required to train machines. What can seem like a mere technicality typically winds up revealing the inconsistency with which documents were processed in the past. It turns out that every person has a unique fingerprint when it comes to classifying, analyzing, sequencing, labeling, tagging, and sorting information. Getting the data points to line up in any one case is sometimes hard enough. Getting a thousand similar cases to line up in the way they were processed is actually why we need AI in the first place. But it can be done and it’s what lets the machine learn what to do down the line. Invariably this challenge is worth it. In fact, clients report that, while trickier than expected, this step reveals the relative inefficiency and imperfectness of the formerly manual workflows. It’s as if the challenge involved at this stage is the best proof that automation is exactly what was missing.

So far, debt collection has been the domain of banks and specialized collection agencies. Companies with customers in the red have generally outsourced the collecting, or they have sold off their debt at some loss to their bottom line (seventy cents on the dollar, say). It is an exciting question in these dynamic times whether the ability to deploy intelligent automation to expedite the collecting of debts will make some companies hold on to their non-paying customers’ accounts rather than make them someone else’s problem. After all, there is tremendous profit lying dormant in every debt portfolio. And it’s profit you’re owed. You just need the right tools.

To discover how Applica can help you maximize profit from debt collecting while minimizing risk connect with an Applica expert today.