Why AI May Offer a Chargeback Solution for Financial Institutions


Chargebacks are nothing new for those professionals operating in the payments industry. However, when we discuss payment disputes, we tend to think of it as an issue that affects merchants, first and foremost.

That’s true on a certain level. According to a report published by Javelin Strategy & Research, direct chargeback costs came to roughly $31 billion globally in 2018, and merchants absorbed nearly two-thirds of those overall costs. That said, payment disputes carry consequences for every member of the value chain, including card issuers.

The report mentioned above reveals that financial institution losses in 2018 as a result of chargebacks came to $11.4 billion. Given that chargeback losses magnify by roughly 20% each year, we’re looking at $16.4 billion in projected direct losses in 2020. Dispute liabilities represented a large chunk of this total. With issuers often willing to write-off disputes on low-value transactions rather than file a chargeback, banks already see several billions of dollars in losses right there.

Card network fees, technology, and fulltime-equivalent (FTE) labor were significant generators of cost as well. That’s just considering the direct costs; the indirect costs of chargebacks, including margin compression, and the tendency of chargebacks to create more chargebacks due to positive reinforcement, could be several times higher.

There are other ambient concerns to think about as well. As a result of the COVID-19 crisis, for instance, consumers are doing more through remote channels like eCommerce and click-and-collect. These card-not-present transactions are much more susceptible to disputes than card-present sales. At the same time, businesses in key verticals like travel are seeing a surge in chargeback activity tied to COVID-19.

It can be tempting to just ignore this problem; to accept chargebacks as inevitable. However, the longer we allow this situation to progress, the more serious those long-term consequences become.

Chargebacks a Growing Problem

Failing to address this problem could translate to tighter margins for banks, as well as ongoing inefficiency, higher costs, and stifled innovation in payments and finance. Institutions in markets that can’t adapt and manage this ongoing issue will risk losing their competitive edge to more disruptive and adaptable challengers.

The problem is that we’ve conditioned buyers to view the chargeback process as a quick and painless option compared to resolving issues through merchants. This is a practice known as “friendly fraud,” and incidents are increasing by roughly 20% each year.

While some consumers knowingly abuse the chargeback process, friendly fraud isn’t typically a deliberate or malicious act. Instead, consumers simply don’t understand their expectations and responsibilities in the transaction process.

The chargeback process can be very opaque. Cardholders don’t know they’re doing anything wrong, while merchants and institutions can’t reliably separate legitimate disputes from illegitimate ones. While Visa and Mastercard introduced their own respective policy updates to try and address the issue, the number of fraudulent chargebacks filed continues to grow year after year.

AI Holds the Key

Part of the problem is a lack of consumer education. At the same time, the legacy infrastructure that upholds the payments industry is just as much of a problem. This will complicate any initiative that hopes to push for comprehensive, coordinated change.

What we need is a solution that can be adaptive and focus in on the unique vulnerabilities of each institution. Then, once we’ve engaged in diagnostic work, we can deploy a solution. I believe that artificial intelligence holds the key to this process.

AI would allow issuers to automate many key fundamental operations. Take dispute management, for instance; banks could use AI to review transaction data, analyze cases, and compile bodies of data over time. While this currently demands extensive human oversight and coordination between different parties, introducing AI tools could automate most of this work.

Using AI would speed up processes, save resources, and reduce overhead. At present, financial institutions devote as many as 39,000 fulltime-equivalent employees to the management of chargebacks. With better AI tools, though, institutions could reallocate staff. This would make internal operations much more efficient.

The key advantage of artificial intelligence is to amass more reliable data. This will pay dividends over time; we could see more collaboration between issuers, acquirers, and merchants.

At present, all the stakeholders in the chargeback process are trying to operate with a limited capacity for collaboration. AI could essentially fill the gaps between parties to facilitate smoother interactions, easier data sharing, and faster dispute resolutions. It would be easier to filter out invalid disputes, creating a more reliable pool of chargeback data. This would ultimately lead to a reduction in chargeback issuances in the mid- to long-term.

Just the First Step

Adopting artificial intelligence would have additional knock-on effects. For instance, it could make the implementation of new technologies and sales channels a smooth and simple process.

AI could also make it easier for stakeholders to adapt to new consumer preferences.For example, the COVID-19 crisis has led to a substantial growth in consumers’ use of click-and-collect sales channels. With more consumers buying through digital channels, we could see a sudden surge in chargeback activity. AI could remove some of those road blocks, though, facilitating faster adaptation and a dynamic response to changing consumer preferences.

Artificial intelligence is a promising tool, but it can’t deliver the kind of change we need to see on its own. Ultimately, we’re going need a permanent fix for the broken and outdated chargeback process in the long run. This demands further collaboration between all stakeholders, including financial institutions, merchants’ groups, card networks, and national governments. There’s no sign yet that such a coordinated response is forthcoming, though.

In the meantime, artificial intelligence is the most promising solution currently available for chargeback management on the financial institution side. Implementing this tool would allow for streamlined processes, reduced costs, and an overall better standard of service for custom

Monica Eaton-Cardone is global Fintech executive with extensive experience developing agile technologies and products, optimizing ecommerce profitability, analyzing risk relativity, and creating payment processing solutions. Shedeveloped Chargeback911‘s robust chargeback-mitigation solution, which combines human insights with agile technology, that is now used by thousands of companies worldwide. Eaton-Cardone was instrumental in co-founding a number of successful companies, including Chargebacks911, eConsumerServices, and Global Risk Technologies.

This content is for CU BUSINESS eMagazine , Special Deal: 2 websites , and NEW! The Leadership Team Builder Group Subscription members only.
Log In Register

Share post:



More like this

Rebecca White Joins Texas Trust as Vice President of HR

ARLINGTON, Texas – June 6, 2022 – Strengthening and...

There’s more to overdraft reform than just fees

The great overdraft debate has focused on fees. And...

Rising Inflation Impacting Non-Prime Borrowers Most, Though Many Consumers Proving to be Resilient under Challenging Circumstances

New TransUnion study explores consumer resiliency amid rising inflation Chicago,...