Using AI to Resolve Longstanding Chargeback Concerns

There’s a tendency within the payments space to view chargebacks as a problem that primarily impacts merchants.

It’s true that merchants absorb more than two-thirds of overall costs connected to disputes. However, every payment dispute carries consequences for each member of the value chain. Financial institutions are exceptionally vulnerable in this regard; in fact, they see billions of dollars in chargeback losses every year.

Acquirers stand to lose processing costs and card network fees in the short-term, while also facing other liabilities. For instance, let’s assume a merchant sees a sudden spike in chargeback issuances – something that’s very easy to imagine in the age of COVID-19. The merchant’s acquirer could end up footing the bill if the merchant doesn’t have sufficient cash in reserve. That’s on top of other long-term threats like margin compression resulting from lost merchant activity.

For their part, issuers could suffer lost processing costs and chargeback liability losses. Also, many issuers opt to write off low-value transactions, rather than engage in the chargeback process. This leads to further losses.

It can be tempting to just ignore chargebacks. However, we could see serious ramifications with long-term consequences for credit unions if chargeback issuances continue at their predicted pace.

Financial institutions will be constrained by tighter margins, stifling innovation in payments and finance. This would mean ongoing inefficiency and higher costs. Institutions in the U.S. would lose their competitive edge to institutions in other markets in which financial institutions are able to manage this problem.

Abuse and ­Outdated ­Technologies Causing Problems

The core of the matter lies in consumer impressions. Buyers increasingly see the chargeback process as a fast and simple alternative to requesting returns from a merchant. When this happens, it’s a clear-cut case of friendly fraud.

Chargeback issuances tied to friendly fraud are on the rise, increasing by roughly 20% each year, according to research from Chargebacks911. Further complicating the matter is the fact that, while these individuals are abusing a consumer protection mechanism, it’s typically not a deliberate, malicious act.

The chargeback process is very opaque. As a result, cardholders who commit friendly fraud typically don’t even realize they’re doing anything wrong. From their vantage point, there’s no meaningful difference between a return and a chargeback. Thus, it’s no surprise to learn that only 14% of cardholders, according to our research, contact a merchant before filing a dispute.

Despite all this, not enough is being done to really confront the matter in any meaningful way. Visa and Mastercard have introduced policy updates to try and address the problem, yet chargeback issuances continue to mount.

Consumers don’t have any incentive to change their behavior. Thus, consumer education will be key to resolving this problem. At the same time, legacy infrastructure in the payments space makes it hard to accomplish any kind of comprehensive, coordinated change and solutions for this matter are less cut-and-dry.

AI Holds the Key

Credit unions can’t sit around and wait for change to come. Instead, they need to take this issue up themselves to mitigate chargeback losses.

Conventional approaches to chargeback management have been unsuccessful at addressing the problem from the financial institution perspective. What we need is an advanced, adaptive solution that can focus in on each institution’s unique sources of loss, then deploy a precision solution. That sounds overly-idealistic; however, it’s possible if we employ artificial intelligence in the right way.

AI could allow credit unions to automate key components of the dispute process. Reviewing transaction data and analyzing cases currently demands close human oversight, as well as a drawn-out exchange between multiple parties. Better AI tools could allow institutions to free up resources and significantly reduce overhead.

Financial institutions devoted as many as 39,000 full-time equivalent employees to chargeback-related activities in 2018. Adopting AI tools would make it possible to reallocate these staff members, providing substantial savings and greater operational efficiency.

Artificial intelligence offers the prospect of better connection between parties. In effect, it would close the gap between credit unions, acquirers and merchants. AI could also eliminate potential hiccups in adapting to new consumer preferences; for example, the rapid growth of click-and-collect adoption since the onset of the COVID-19 crisis.

The result could be a much more straightforward and collaborative chargeback remediation process, reducing costs for everyone involved.

In the long run, we need a permanent fix for the broken and outdated chargeback process. Credit unions will need to work alongside other financial institutions, as well as merchants, card networks and governments to develop a solution on that front. In the meantime, though, artificial intelligence may be the most promising solution available to streamline processes, reduce costs and provide better service to credit union members.

Monica ­Eaton-Cardone COO & Founder Chargebacks911, Clearwater, Fla.