Clients
Product News
Contact Sales
Whitepapers

“Chargebacks and Consumer Behavior”

by Mark Betz

In the age-old interchange of buyer and seller, the development of the modern payment card network is perhaps as significant an event as the invention of portable coinage. The ubiquity of consumer credit, and the near-universal acceptance of the payment card as a means of exchange, has revolutionized the practice of retailing goods and services. Payment card use at retail points of sale has been growing an average of twenty percent annually over the last four years, according to consultants Edgar, Dunn, and Co. The industry is now seeing card acceptance take root in retail environments, such as fast food, that were traditionally not thought to offer profitable applications. This year, the credit card networks operated by Visa and Mastercard will process more than 15 billion payment transactions.

All of the parties in the card value chain have benefited tremendously from the growth of this universal and convenient method of payment: consumers have more purchasing power at their fingertips, and more convenience of use, than ever before; retailers have access to more qualified buyers, and enjoy credibility by association with the banks that control admission to the networks; the banks themselves have established huge and profitable businesses in issuing cards to consumers, and acquiring purchase receipts from merchants.

Over the forty-plus years of growth in the card business, the demographics of the participants have changed in contrasting ways. The number of cardholders and merchants has grown: there are now approximately 500 million active card accounts in the U.S., being used to make purchases from more than 4 million points of sale. Meanwhile the number of banks involved has dwindled, with consolidation putting the bulk of the market share in the hands of the top ten on both the issuing and acquiring sides of the market. While vertically-integrated players like American Express and Discover have profited and grown, the volume in the business remains heavily concentrated in the Visa and Mastercard association networks.

Perhaps in response to the increasing power of the banks, and their proxies the card associations, consumers have benefited from more rights attached to their participation in the system. In 1968 the U.S. Congress passed the Truth in Lending Act (TILA), which was later implemented in the infamous Regulation Z of the U.S. Code of Federal Regulations (12 CFR, Part 226 (2002)). Reg Z, as it is known in banking circles, established guidelines to regulate key aspects of the practice of providing open-ended credit to consumers. Reg Z made banking compliance into an industry, and put a tremendous amount of power into the hands of the consumers whose complaints about early abusive lending practices spawned the law.

It is worth emphasizing that because of Reg Z consumers enjoy nearly-absolute protection when using a credit card to make purchases. A consumer can purchase a piano, over the Internet, from a vendor in another country, and should it not arrive suffer no consequences whatsoever. The good news is that the nature of the credit card system, and the checks and balances inherent in it, make it extremely likely that the merchant, by virtue of having been approved by a bank to accept credit cards, will come through with the instrument. For the merchant, it is also very likely that the purchasing consumer is a real person with a real ability to pay, for the same reasons. The two parties don’t have to know each other to have a pretty good idea that they will each be true to their commitments in the transaction.

How, then, can we explain chargebacks? A chargeback occurs when, for reasons discussed below, consumers call their card-issuing bank, refer to an item on their bill, and typically state that they did not make the purchase. A smaller percentage of the time the consumer will acknowledge making the purchase, but will state that the merchant did not fulfill the transaction, that the goods were not satisfactory, or that the billed amount is incorrect. Because of Reg Z, the issuing bank must resolve the matter for the cardholder, and there the weeping and gnashing of teeth commences. Not on the consumer side, of course! It’s all smiles in buyer land when it comes to the effects of Reg Z on transaction disputes. A grimmer reaction prevails throughout the rest of the industry, but primarily in the accounting departments of issuing banks and retailers, where the burdens of this problem primarily fall.

And the burdens of the problem are significant. Chargebacks currently average about .20 percent of transaction volume. That means something in the neighborhood of 30 million chargebacks per year, and the number is growing. The costs of processing a cardholder inquiry and the associated chargeback exceed $50 end-to-end. Some estimates place total annual costs to the domestic card industry at more than $5 billion. The costs of processing chargebacks accounts for approximately 15 percent of the cost of issuing credit cards, and 25 – 30 percent of the cost of acquiring merchant transactions .

Where are all those chargebacks coming from? To fully understand the sources of the problem, it is necessary to start measuring at the front of the pipeline: the card issuer’s call center. Not all issuers track the reasons for incoming calls meticulously, but some do. These issuers uniformly report that 60 to 75 percent of incoming transaction inquiries (what we call the first report of a consumer complaint that may eventually lead to a chargeback) are presented to them as unauthorized use. In other words, the consumer is alleging that fraudulent use of the card has occurred. Is it possible that 25 million transactions per year are fraudulent? In short: no. The overall incidence of actual fraud is somewhere in the neighborhood of 8-10 basis points of total volume. That is at most .01 percent of volume. This number is higher in some sectors, such as Card Not Present, than in others, but no matter how you slice it you don’t get to .20 percent of volume.

The real answer is hinted at by a recent change that Mastercard has made to its set of “reason codes” that attach to a charged back transaction. In April of 2003 the association added a code for “Do not recall/Unrecognized.” This change aligns the Mastercard reason code set with a secret that the best issuers have been in on for years: in the vast majority of alleged fraud cases the consumer actually did make the transaction, and either doesn’t remember the activity, or doesn’t recognize the activity or merchant from the billing description. The associations recognize this problem as a data quality issue: there isn’t enough information on the bill to prevent the problem, or the information that is present is misleading; for example when a merchant corporate name is used on the bill, and it differs from the retail DBA name. One issuer that we have spoken with now initially labels all incoming fraud reports as “Unrecognized” transaction activity until proven otherwise.

What about the other 25 to 40% of inquiries? Some of these are initiated by the card issuers themselves, in cases where a card should not have been authorized for a transaction, as an example. Another portion comprises what the issuers often call “friendly fraud” or “soft fraud,” i.e. consumers taking advantage of the system. The rest typically involve some complaint by the consumer about the merchant’s fulfillment of the transaction. Some of the more common causes for this type of complaint also involve lack of item detail, as in cases where a merchant splits a shipment, and bills only for the shipped portion. The consumer, not recognizing the amount, may call the issuer to complain. In cases of consumer dissatisfaction, the association rules require the issuer to ensure that the consumer and merchant have made a good-faith effort to resolve the problem before a chargeback is initiated. In practice this is very difficult for issuers to enforce. The role for the chargeback system then becomes that of a very costly and unsatisfying dispute resolution mechanism for arbitrating between the merchant and their customer.

These problems are further aggravated by the growing number of transactions occurring in Card Not Present environments. Currently CNP transactions account for roughly 20 percent of volume, and they are growing. In a normal card-present transaction an issuer, faced with a cardholder alleging unauthorized use, will forward a documentation request to the acquiring bank, which will work with their merchant client to essentially get a signed sales receipt back to the issuing bank. Many merchants report that a signed receipt resolves approximately 50 percent of these inquiries without a chargeback occurring. However, documentation requests are costly to all parties. Even large merchants may pay as much as $8 per research request; acquirers and issuers have their own associated correspondence costs. Because issuers are well aware that in a CNP transaction there is no sales receipt, they are often unwilling to pay the costs of a documentation request, and instead immediately initiate a chargeback. Such “preemptive” chargebacks are very damaging to merchants, because the difficulties of winning an appeal, and the consequences of losing one, mean that most will simply credit the purchase. One merchant we spoke with referred sardonically to this practice as “credit without return.”

This analysis clearly indicates that there is a large body of costs that can be addressed if sufficient detail can be added to card billing statements, and if consumers and merchants can be enabled to communicate directly with one and other to resolve questions. How large? If we take the most conservative view, something more than 60 percent of the costs attributed to the chargeback problem may be addressable through these means; the number could be more than $3 billion. That is certainly enough reason to look at solutions. The question that remains is whether there are solutions that can actually be applied? Are there remedies that are feasible and practical?

The idea of adding detail to statements is not new, and in fact some steps have already been taken: acquirers are working with CNP merchants to rationalize corporate name use, and get customer service phone numbers and URLs into billing descriptions; and some acquirers have begun to offer “soft descriptors” that allow merchants to be more specific in their billing descriptions. It is difficult, however, to cram enough data into the limited record formats associated with settlement and billing, to be certain of interdicting a significant number of inquiries. Livermore Research recently performed a review of transaction descriptions. In looking at their work it is hard not to conclude that the numbers are often presented counter-intuitively, again because of record length limitations that forbid the liberal use of things like spaces and hyphens. When the number is toll-free the ‘800’ or ‘877’ tend to catch the consumer’s eye, but a non toll-free number, presented without formatting, looks a lot like some kind of reference number. For these or other reasons, consumers still seem inclined to call the issuer first.

One of the problems with adding detail to payment card bills is the cost of bill production and delivery. Some large issuers mail more than 300 million monthly statements per year, at a cost of approximately $.30 per bill. Anything that increases the size or mailing weight of these statements would generate additional costs that could very easily overwhelm any savings from a reduction in transaction inquiries and chargebacks. This situation is changing with the increasing adoption of online bill delivery. In the online environment additional information can be added to the statement at very little, if any, incremental cost. It is also the case that additional merchant information can be applied to the problem in the issuer’s call center, even if it cannot yet be printed on the bill. If the issuer can mitigate some significant percentage of inquiries by providing more information on the first contact with the cardholder, this will have the effect of lowering costs for all the parties in the chain.

Assuming that more comprehensive merchant information could be attached to a bill, would it be effective? At least one pilot program about to commence will test the theory that adding information such as the merchant’s retail DBA name, a description of the business, the business’s location, and customer service contact details, will prompt cardholders to remember their transactions, and to resolve satisfaction problems with the merchant directly. Industry practice already suggests that the prognosis for using these techniques is encouraging. In fact, one sophisticated large issuer, using what information is available, such as the Merchant Category Code that describes a merchant’s class of business, as well as Internet search techniques, successfully mitigates nearly 50 percent of incoming inquiries.

It is probably safe to assume that everyone in the payment card value chain would like the consumer to remember what they bought, and hang up the phone happy, if a bit chastened. It is less clear whether everyone would like cardholders with satisfaction issues to communicate directly with the merchant. Certainly card issuers would, but what about the merchants themselves? Research in this area is ongoing, but it can already be said that attitudes differ depending upon a merchant’s business. Large, cost-driven retailers seem less inclined to assist the issuer in shedding any of these calls. A certain adversarial attitude between large players on either side of the market is partly responsible, but on the practical side these merchants often do not have the infrastructure or online information systems to add much detail to what is already on the bill. Why take the call, they reason, when we can’t do any more to address the problem?

Other merchants, including large service-oriented retailers, and most CNP operations, do have the infrastructure, and the information systems, and recognize the value of resolving the customer’s problem in a phone call, as opposed to arbitrating it through the costly chargeback process. To them, this is both a customer service as well as an expense reduction strategy. It is too early in our research to predict that this is a trend, but we can state that the trend among merchants is toward better CRM (Customer Relationship Management) systems, with better information, and thus we can expect that the number of merchants with the ability, if not the willingness, to address consumer inquiries will grow.

When discussing adding additional detail to the consumer’s credit card bill, the nirvana scenario has always been the idea of line-item detail. Presumably the availability of line-item detail on the bill would have a dramatic impact on the number of nuisance chargebacks caused by consumer misunderstanding. This expectation is validated by the experience with corporate purchasing cards. Transactions applied to such cards have long been able to include what is known as “Level 3” data, which can include line item detail as well as more information about the vendor. Most issuers report that the incidence of chargebacks on their purchasing card programs is miniscule. For these and other reasons Visa and Mastercard have active incentive programs under way to encourage merchants to pass Level 3 data for purchasing cards. Why not add Level 3 data to consumer card bills? There are a number of very significant barriers. First, level 3 compatibility would need to be implemented on all points of sale. Second, consumer card processing software used by issuers would have to be altered to store and manipulate level 3 data. Lastly, consumer card issuers would have to be willing to print the information on their mailed statements, and as discussed above the costs of that are likely prohibitive. In fact, the total costs of all of these changes have been estimated by various parties at upwards of $20 billion.

There are some encouraging facts, however. Most large and medium sized retailers do store line item detail on transactions, and at least one issuer of private label cards has closed the loop and made this information available to their online billing customers. Such line item detail, if it can be exchanged securely, and in such a way as to preserve privacy and confidentiality, can be displayed on online bills, and utilized in the card issuer’s call center, at very reasonable costs, and should have a very positive impact on the problem. Again, trials will be underway in the near-term future to examine the possibility that third-party systems operating outside the existing card settlement network can provide a facility to connect cardholders with this information, at least for larger established merchant sites.

The problem of chargebacks is second only to outright fraud in terms of the total costs to the industry. In terms of processing costs (as distinguished from the cost of goods or services stolen, for example) it is the largest single problem the industry faces, and unlike fraud the pain from it is felt more or less evenly on both ends of the network: card issuers and merchants. While objective analysis indicates that technology can provide information-based solutions to mitigate the problem of chargebacks, applying these solutions will require that merchants, acquirers, issuers, and the associations put aside their differences and work together to both provide more information to cardholders, as well as streamline communications between cardholders and merchants. Successful cooperation may ultimately reduce the incidence of chargebacks back to something nearer the incidence of outright fraud; however it is unlikely they can be made to disappear entirely. Experience suggests that where consumers are granted rights by law, they will take advantage of them.

Mark Betz is Chief Technology Officer at Incurrent; a provider of online information systems and products to credit card issuers, acquirers, and merchants. He can be reached at mbetz@incurrent.com.