With online payment fraud now entering its second decade, recent fraud studies are somewhat encouraging and show that the percentage of revenue lost by eCommerce merchants to online fraud is declining. But these same studies also underscore that with the steady growth of online commerce and digital goods and services, the actual sales revenue lost to payment fraud continues to increase year over year. The eCommerce paradox is that while online payment fraud is dropping in relative terms, it is growing in absolute terms. And with eCommerce merchants expected to lose as much as $2.8 billion to payment fraud in 2005, according to the 2006 CyberSource Annual Fraud Report, the problem is demanding greater management attention, the development of industry best practices, and the pursuit of new solutions.
Understanding Payment Fraud in an Online World
How do online merchants know if a purchase is fraudulent? They actually don’t know at the point the customer hits the "Buy Now" button. Through the use of sophisticated fraud screening techniques, they might have a high-degree of confidence that a transaction could be good. But ultimately it all comes down to whether or not they get paid for the goods—and if they do get paid, whether or not they get to keep the money.
For merchants, payment fraud appears on their doorstep in one of two forms. A customer approaches them with a problem—such as the goods never arrived—and asks for a refund. The merchant might suspect that the request isn’t legitimate, but often provides the customer with a refund in order to maintain a good relationship. Either way, the merchant is out the sale—and in many cases they have lost their inventory as well.
In a more costly form, payment fraud arrives via a "chargeback", a term that eCommerce merchants have learned to hate. In the world of payment cards a chargeback refers to the process by which a cardholder disputes a payment made to a merchant. Cardholders generally initiate chargebacks based on claims such as the transaction was not authorized by them, the products or services were not as promised, the goods were not delivered, or it was a duplicate charge. After filing their claim with their issuing bank, the bank will either issue a request for more information about the purchase or initiate a chargeback—a reversal of the transaction that results in the money being taken back from the merchant. Getting the merchant’s side of the story generally takes the form of a retrieval request (also known as an RFCO or ‘ticket request’) that asks for more information about the transaction. Retrieval requests only occur about 10% to 15% of the time. Usually, the bank will immediately issue a chargeback and reverse the funds flow.
eCommerce merchants can dispute (or "represent") the chargeback back through their acquirer with evidence of purchase, delivery, or service consumption. If the evidence is strong, they might get to keep the money earned from the sale. If the evidence is weak, they are out the sale, out the inventory, are usually out the interchange fee, and are accessed a chargeback processing fee by their acquirer. If the representment leads into arbitration (1) between the issuer and the acquirer, the merchant might also be assessed with the arbitration costs if they lose their case.
Online merchants define payment fraud as either "real" or "friendly". "Real fraud" is usually committed by a third party masquerading as the cardholder, and is increasingly being conducted by organized crime. “Friendly fraud” refers to purchases disputed by the cardholder as "not authorized" even though they really did perform the transaction and receive the goods. Since they obviously don’t have the cardholder’s signature as proof, eCommerce merchants are often unsure about how to respond to these types of chargebacks.
Chargebacks in either form—real or friendly—are a problem for almost all online merchants, and a serious problem for some. In addition to the lost revenue, lost merchandise, processing fees, and staff cost, the card associations will often classify merchants with high chargeback rates as "high risk" and start assessing additional transaction fees and other fees on a monthly basis. If high chargeback rates are not reduced, these merchants risk being assessed hefty financial penalties and ultimately losing their merchant accounts.
Learning from the Leaders
To better understand how eCommerce merchants are responding to the chargeback challenge, Glenbrook Partners recently undertook a comprehensive study of its eCommerce clients to determine the cost of payment card chargebacks and best practices for mitigating losses. The study focused on eCommerce merchants’ operational costs, the write-offs associated with chargebacks, and the representment practices used to fight chargebacks.
The study was based on six months of actual chargeback data from eight large U.S. eCommerce merchants, ranging in size from $100 million in annual online sales to several billion in sales. Most were in the $500 million to $1 billion annual sales range. Four of the merchants sell "tangible goods" such as general merchandise, apparel, and electronics, while four sell "digital goods" such as travel reservations, downloadable music, electronic books, information services, and event tickets
In order to quantify the comprehensive costs of chargebacks to large eCommerce merchants, we asked each client to provide cost and related data that included:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
||
|
From their responses we saw that the merchants exhibited a range of understanding and expertise in chargebacks management. Because the merchants were able to readily identify and manage non-fraudulent chargebacks (i.e., the merchant dropped the ball in some fashion), we asked for and received just fraud-related chargeback data. Importantly, because all of the merchants had a difficult time distinguishing between “friendly” and “real” chargebacks, we focused on fraud-related chargebacks in aggregate and did not try to drill down into the two different forms.
Key Lessons
Based on our analysis of the 350,000 plus chargebacks that were received by the eight merchants in the second half of 2004, we drew the following conclusions:
Merchants Can Successfully Fight Chargebacks for Digital Goods
Quite a few merchants believe they can’t win the chargeback fight for digital goods and services—and are incurring significant losses as a result. This belief is generally based on a number of factors including feedback from their acquirers, perceptions of the associations’ rules and how they are enforced, as well as their own inability to provide compelling evidence that cardholder did in fact consummate the transaction.
Our analysis showed that best-in-class digital goods merchants are choosing to fight 35% to 40% of their chargebacks (most likely friendly fraud) and are achieving 70% to 80% win rates.
While not quite as effective, the digital goods merchants that are “throwing everything against the wall” (i.e., representing everything with varying levels of research and support) are still seeing 30% to 40% win rates.
Respectable Win Rates are Achievable for Tangible Goods Merchants
As might be expected, the percentage of chargebacks fought and won did vary between merchants selling tangible goods as opposed to intangible goods. Based on the hypothesis that tangible goods (e.g., consumer electronics, jewelry, etc.) are more prone to real fraud, there isn’t much the e-tailer can do to fight the chargeback, resulting in fewer representments and a lower win rate on those that are represented.
The data showed that best-in-class tangible goods merchants are achieving 30% to 40% win rates on the chargebacks they choose to fight.
Given that and the fact that tangible merchants tend to have a much higher cost-of-goods sold relative to digital content merchants, it is a given that they are much more reliant on their up-front fraud screening tools.
Chargeback Processing Strategies can be Successfully Optimized
It is clear from the data that those merchants with a comprehensive understanding of their chargeback processing costs, the nature of their chargebacks, and their ability to win a given chargeback, can optimize the ROI on chargeback related labor and fees. For example, the data indicated that merchants with the lowest chargeback rates spend the most time per representment, which is more than offset by their higher win rates. The data also showed that merchants with lower representment rates had higher win rates, also demonstrating the benefits of focusing resources on those items with the highest probability of return.
Everything is Negotiable
Respondents reported a wide range of chargeback-related fees and fee policies. Chargeback fees varied widely but were generally in the $0 to $5 range, reflecting the strong bargaining power of large merchants (smaller merchants generally pay between $10 and $25 per returned item). (2)
One of the more valuable findings for study participants was learning about the wide range of acquirer policies regarding fees on chargebacks. Some merchants were refunded interchange fees, acquirer processing fees, or association assessments. (3) Specifically, about half of the merchants had the interchange fees on the transactions reversed, while a relative few had the association assessments and their acquirer’s processing fees refunded. In addition, American Express and Discover are refunding the merchant discount on returned items for approximately half of the participants (but not necessarily for the same merchants that are receiving refunds from their bank card acquirers).
Extrapolating from the study results, merchants with less sophisticated tools can expect to spend between 15 and 20 minutes on chargebacks deemed to be winnable (i.e., the ones they really want to fight). As an example, those with more sophisticated chargeback support tools have the relevant data available to them on a single system (e.g., from CRM tools, customer databases, web logs/server data that show what was used, what was downloaded, etc.). Study data shows more sophisticated merchants spending approximately 10 minutes to fight winnable chargebacks.
Other Key Findings
Getting new headcount tends to be so hard in large organizations that merchants are considering outsourcing all or parts of the chargeback function in order to avoid adding headcount to meeting growing volume, or to redeploy their FTEs in higher priority areas.
The cross-merchant data from our benchmarking initiative required somewhat customized, merchant-level analysis to be most useful and actionable since each merchant’s chargeback performance is a function of:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Ironically, chargeback rates can be too low. While a detailed analysis of this hypothesis was outside of the scope of this study, there were sufficient data to at least highlight this as a potential issue. In fact, several merchants are collecting and analyzing data on an ongoing basis in an attempt to lower their “false positives” (i.e., declining a good, profitable order because of suspected fraud). One potential indicator of overly restrictive fraud screens is a chargeback rate far below norms for the merchant’s vertical segment. To properly perform the associated cost/benefit analysis, merchants need to have a reasonable understanding of their chargeback processing costs.
Recommendations
Work the Chargebacks and Pay Attention to Reason Codes
Successful strategies for both digital goods and tangible goods merchants include analysis of the chargeback reason code. The chargeback reason code often does not accurately reflect the actual reason the transaction is being charged back, since consumers and issuers often insert arbitrary reasons to get the chargeback process going. A merchant cannot accurately analyze its chargeback situation and subsequent chargeback dispute decision-making and actions based on simply relying on the chargeback reason codes associated with the chargebacks it receives. Successfully fighting these chargebacks requires analyzing the situation and correctly recategorizing the chargeback reason code—in other words, looking beyond the stated reason code to make a compelling assertion on whether a given chargeback is a case of true fraud or friendly fraud.
While some large merchants may have the systems resources to facilitate cost-effective data collection, those that don’t, such as most mid-sized and smaller merchants should consider viable third party tools and outsourcing opportunities.
Focus on Winnable Chargebacks and Present Compelling Evidence
Sophisticated merchants also recognize that fighting chargebacks can have important, albeit somewhat less quantifiable benefits as well. In addition to optimizing operational processes and customer experience via root cause analysis, there is anecdotal evidence that supports the hypothesis that merchants develop both good and bad chargeback processing reputations with issuers—to the point that it affects the number of chargebacks generated to a given merchant and the intensity with which that issuer works the merchant’s representments.
If a merchant has a reputation, for example, of representing all chargebacks—independent of supporting data—issuers will tend to more liberally charge items back to the merchant knowing that the merchant isn’t really “working” the items, and will correspondingly and effectively challenge the representment.
Understand the ROI and Use It To Make Better Decisions
While merchants recognize the associated cost of goods sold and acquirer fees, they do not generally have a comprehensive strategy for understanding the true costs of chargebacks to their organizations, taking full advantage of the valuable learning available through chargeback analyses, nor for truly optimizing the appropriate quantity and quality of resources devoted to working chargebacks.
For example, this survey highlighted the fact that many merchants do not have an adequate understanding of the ROI on an additional chargeback FTE. Merchants should have the data available to assess on a periodic basis how cost effective it is for them to work chargebacks at various dollar thresholds. This analysis is a function of staff costs, chargeback and related fees, and win rates.
Conclusion
While many myths and misinformation exist about chargebacks and what merchants can and should do with them, chargeback analysis and chargeback management can and should be viewed as a strategic opportunity for merchants. Rather than viewing chargeback processing as a tactical nuisance, thoughtful analysis and tight management can help improve merchants’ revenues and profits, as well as serve as a leading indicator of the health of the business and customers’ buying experience. Best in class eCommerce merchants use chargeback analysis as a feedback mechanism to understanding how to improve business operations and online processes, and of course, to reduce their chargeback rates.
References
- Inside the chargeback process, issuers and acquirers have a well-defined procedure for settling disputes. When both sides feel strongly about the case, it can move into pre-arbitration, and eventually to full arbitration with the losing side usually paying significant fees.
- A word of caution may be in order: a number of acquirers (but not all) develop merchant-level P&L’s when preparing to negotiate contacts/renewals, implying that they may just try and recoup lower negotiated fees somewhere else. Having said that, in general, chargeback processing fees are a large and profitable source of revenues for acquirers.
- MasterCard and Visa acquirer assessments are 9.50 and 9.25 basis points (.0925%) on volume, respectively
Acknowledgments
Several of my partners at Glenbrook (Jay DeWitt, Russ Jones, and Dennis Moser) helped shape this report.
Publication History
Initial Publication Date: March 24, 2006
