Payment disputes in banking: A pathway to deeper customer relationships

| Article

Card-related payment disputes between customer and merchant occur in fewer than 1 percent of transactions, but these moments can have an outsized impact on bank-customer relationships. The cornerstone of these relationships is customer trust—that their deposits are safe, and that they can dispute a charge or transaction and get a quick, factual answer.

For customers, disputes can be a source of frustration and inconvenience. For banks, disputes are expensive to resolve. If banks treat disputes simply as issues to be minimized, they can miss the silver lining: the opportunity to strengthen their relationships with customers.

This point is particularly important today. Globally, customers continue to move away from cash and checks toward electronic payments (Exhibit 1). Overall, this trend is a positive development for banks and card issuers, but as card transactions grow, so does the number of disputable transactions (and the incidence of fraud), putting pressure on dispute processes that often are already overextended, and leading to increasesi  operating costs of hundreds of millions of dollars.

Electronic payments continue to gain share, increasing the pool of payments that can be disputed.

As they benefit from growth in credit- and debit-card use, in other words, banks and card providers must ensure that the customer experience is not degraded. A number of banks are therefore rethinking the way they manage disputes with an eye toward maintaining and even strengthening their trust-based customer relationships—all while processing disputes more cost efficiently and at lower regulatory risk.

Complex operating models and other obstacles

Today, most banks face a number of costly obstacles as they seek to address a rising number of card disputes:

  • Complex operating models. Most banks’ traditional, budget-controlled functional structures lack a clear, end-to-end chain of responsibility for the disputes experience. The operating model’s complexity obscures distinct ownership of the client experience, and in some cases there is a definitive separation among the functions required to take in and resolve disputes (the call center, dispute research, and the back office). This disjointed model leads to more customer pain points and delays in time-to-credit and overall dispute resolution. To exacerbate matters, most dispute organizations are scattered geographically—some US-based banks have as many as 15 dispute-resolution locations—which intensifies operational complexity.
  • Overprocessing of disputes. Banks and issuers usually process all disputes above $25, regardless of dollar amount, customer, or the merchant’s dispute history, and apply the same process across the board. This lack of triage allows dispute volume to rise and puts pressure on dispute teams to process disputes quickly to save costs and avoid regulatory infractions.
  • Long, complex research process. In most banks, the dispute process involves multiple IT systems and tends to be driven by the technology the bank has rather than the technology it needs. In some cases, banks must turn to third parties to conduct additional research on fraud.
  • Ineffective quality assurance. At some banks, the process of reviewing dispute decisions is overly focused on checking regulatory boxes rather than on the overall accuracy of the decision. Banks and issuers thus miss the opportunity to catch and fix incorrect decisions. This problem is especially prevalent at larger institutions, where intense regulatory scrutiny often inadvertently makes compliance a higher priority than quality control. Instead, leading banks are now embedding regulatory processes into more effective dispute processes.
  • Inadequate performance management. Process complexity of the kind that characterizes dispute resolution at most banks results in highly variable individual performance. As mentioned, banks are often overly focused on preventing regulatory violations at the expense of performance. McKinsey has seen differences in productivity of more than two and a half times between the top and bottom quartiles of dispute-research analysts who process disputes. Limited use of metrics, a lack of individual and team performance targets, and a dearth of fact-based coaching to improve individual performance and identify problems all exacerbate the problem.
  • Over-reliance on the case-management system. Years ago, many institutions implemented case-management systems for dispute resolution with the expectation that the technology would drive down costs. However, most realized only marginal improvements in efficiency and speed of resolution. In McKinsey’s experience, case-management systems deliver full value only when they can integrate effectively with the various systems used to investigate disputes—fraud prevention, ATM, and transaction history systems, for example—and can follow a streamlined process to resolve them. Typically, the handle time for resolving a case does not go beyond an hour, but it can take up to five days to provide an answer to the customer.
  • Increasing regulatory focus. Customer-protection regulations—such as Regulations E and Z in the US—tighten the timelines for banks to resolve disputes and raise the pressure on teams to process disputes and provide credit faster. However, while providing credit faster alleviates one pain point, many firms see it as a simply a Band-Aid, leaving the underlying issue unresolved.

Would you like to learn more about our Financial Services Practice?

Simpler, smarter processing

New insights, tools, and capabilities are emerging that enable banks to address many of the obstacles they currently face in the dispute-resolution arena. An integrated, next-generation operating model based on these new capabilities improves the customer experience (providing 100 percent of customers with real-time, provisional credit decisions), reduces costs (with efficiency gains of 30 percent or more) and financial losses (by up to 5 to 7 percent), and lowers regulatory risk (reducing customer-impacting errors by 80 percent). To achieve these results, banks and card issuers can focus on five imperatives (Exhibit 2):

At most banks, silos result in limited end-to-end ownership of the disputes experience.

1. Digitize the dispute-resolution process

Digitizing the dispute process dramatically reduces the time and effort required for a customer to file a dispute and for the bank to take in the dispute. Best-in-class institutions are simplifying their intake process by taking a customer-centric view, to the point where customers can file disputes in a couple of minutes on their mobile devices, compared with the ten minutes it takes to do so speaking on the phone. Banks are also streamlining the research process by providing a set of questions that customers must answer up front. Call-center agents and customers no longer have to plow through a questionnaire to understand or explain the dispute.

2. Redesign the process with lean principles

Traditionally, banks have sought to improve the dispute process through technology investments and process tweaks, leading to only marginal gains. Only banks that have conducted a “zero-based” design of the entire dispute process have achieved dramatic change. Designing a new dispute process from scratch is the only way to truly simplify dispute research policies and requirements. Some banks have increased productivity quickly by categorizing disputes according to their complexity (number of transactions, number of different systems required to decision, for example).

3. Apply advanced analytics

Some banks and issuers are leveraging advanced analytics and machine learning (in which computers “learn” patterns from examples and apply the insights to new data) to predict disputes and fraud (for example, by letting customers know when a suspicious transaction has occurred and giving them the option to decline it). While fraud detection has leveraged neural networks for the last two decades, the next generation of models is substantially more powerful, leveraging more data within shorter time spans. These processes work by using dozens to hundreds of variables to determine whether provisional or final immediate credit is the optimal strategy for resolving a particular dispute. Banks are also using analytics to estimate the amount of effort required to research a dispute, based on the customer’s history and the merchant profile. The value of such an approach is significant; consider that one regional US bank was conducting a complete research process for all disputes before providing provisional credit, despite the fact that it approved final credit in 98 percent of cases. Advanced analytics can also dramatically improve the customer experience by expediting time-to-funds, and it frees up research capacity to focus on more complex and higher-impact disputes.


Putting customer experience at the heart of next-generation operating models

4. Employ intelligent process automation

Intelligent process automation (IPA), or robotics, automates aspects of the end-to-end dispute-processing value chain. IPA can reduce dispute-resolution cycle times, boost operational efficiency, and improve overall customer experience. The most common use cases include:

  • Fraud prevention. Advanced analytics can flag potentially fraudulent transactions and address them before they become disputes. For example, McKinsey has worked with QuantumBlack to develop a model that uses machine learning and a random forest model to flag potential fraud cases based on 13 months of transaction history.
  • Intake review. Natural-language processing and machine learning can identify the type of dispute (fraudulent vs. nonfraudulent, PIN-based vs. non-PIN-based).
  • Dispute management. Advanced analytics can separate complex disputes from simple ones and assign them to the appropriate analyst. For example, a US bank used an algorithm to identify complex disputes and matched disputes to analysts based on analysts’ skill sets and the other disputes they were working on that day. The result was less “context changing” and significantly increased productivity.
  • Dispute processing. Banks use robotics to collect information automatically and prepare cases for human processing.
  • Postdispute processing. Robotics can be used to process credits and to send notifications and status alerts to customers.
  • Postdispute QA on fraud. Machine learning can be used to identify paid dispute cases that need to be audited.
  • Dispute support. Cognitive agents and robo-chat functionality can answer customer queries during the dispute-research process.

5. Strengthen management systems

While new technologies offer banks the opportunity to re-envision and redesign dispute processes, there is no substitute for best-in-class management systems. Banks and card issuers that excel at dispute management usually take an end-to-end approach to managing the dispute process. They assign a product owner to the entire dispute process for each product or claim, and that owner is responsible for—and compensated for—delivering better customer experience, quality, and efficiency across intake, research, payout, and communication. Clarity in expected customer outcomes and strong key performance indicators (KPIs)—such as those that measure the process from end to end, or those that assess performance from the customer’s perspective—help staff to focus on the right goals.

Making the most of disputes

Transforming the dispute-resolution process can provide banks with an opportunity to solidify bonds with their customers while driving down costs. The benefits can be grouped into three broad categories:

  • Strengthening trust. A quick resolution—and provisional funds when appropriate—can turn a negative customer experience into a loyalty-building moment. Loyal customers tend to feel that their bank “has their back”—but it often takes a dispute for them to reach that level of comfort.
  • Cutting operating costs. McKinsey estimates that the top 15 US banks spend approximately $3 billion each year, combined, on disputes processing. (About 50 million to 100 million disputes occur annually in the US, with a cost per dispute ranging from $10 to $50.) Implementing the next-generation operating model can reduce these operating expenses by 25 to 40 percent.
  • Improving quality. Given the complexities in the dispute-research process and the pressure to resolve disputes quickly, quality can suffer. McKinsey has seen banks in which 10 percent of dispute outcomes were incorrect (both in favor of and against customers). Simplifying and automating the process leads to better decisions.

The disputes process is often overlooked as a chance to build trust with customers. This is a missed opportunity, as the tools and capabilities that will enhance the customer experience can also improve bank outcomes and reduce costs. As banks seek to improve the dispute process, they should look beyond traditional incremental changes and reimagine the process from start to finish. This will not be a simple endeavor, but the rewards can be significant.

Explore a career with us