Thoughts of fraud can cause call center leaders to sit bolt upright in bed at night. The damage to a company’s reputation from agents stealing credit card information is a serious threat. Further, companies are constantly trying to combat a low, but steady percentage of consumers’ attempts to not pay for goods they have received.

Despite the huge real and potential costs, the current strategies for fraud reduction do not leave contact center leaders sleeping better at night: Consumers and agents are committing fraud every day, often without the leaders even knowing it. As any Six Sigma practitioner knows, when you have outcomes with serious consequences, you have to begin thinking about error-proofing strategies.

There are two sources of fraud that plague contact centers: agent fraud and consumer fraud. The primary and most serious manifestation of agent fraud is credit card theft. Agents remember or write down the credit card information on a piece of paper or in a computer notepad. They then use that information to try to purchase goods for themselves or they can post the information on websites for a fee.

The second source of call center fraud is consumer fraud. The primary manifestation of consumer fraud is chargebacks, where consumers have purchased and received an item, but call their credit card company to report it as an errant charge.

Costs of Call Center Fraud

The costs associated with both types of fraud are huge. In the case of agent fraud, you could argue that there are no costs to the merchant from identity theft. That is true provided the theft does not get traced back to the merchant/outsourcer. If it does get traced back, the liability and, more important, the damage to the company’s reputation can be huge.

Moreover, in places like India and the Philippines where BPO outsourcing is a huge industry employing hundreds of thousands of citizens, lawmakers there are moving to pass tough legislation to deter citizens from committing a crime like this and hurting the industry and jobs in that country associated with it.

Calculating the costs of consumer fraud is not easy. One blogger cited an industry report indicating that fraud cost ecommerce merchants $2.6 billion in 2004. Certainly, all of this would not have been chargeback fraud and all of it would not have been from call center transactions. Still, the companies we have spoken with indicate the number is large and that there also is a significant cost associated with the rework process itself.

In addition to the dollar value of the disputed chargeback, here are the basic steps in the chargeback resolution process:

  1. Client files complaint with card issuing bank
  2. Bank researches the transaction
  3. Bank debits merchant via merchant provider and credits cardholder
  4. Merchant account provider sends you documentation
  5. You can then refute the chargeback by providing evidence. This includes screenshots of the customer’s account (for service usage), copies of receipts, relevant communications, etc. In the case of phone recordings, this is good evidence, but they can be hard to find, as not all recording systems allow you to search by credit card number or order number.
  6. Further evidence may be required, or a decision is made. Due to all of the to and fro, it can literally take months to resolve one of these disputes.
  7. In the event that the chargeback challenge is successful, the cardholder is debited again
  8. If the chargeback challenge ultimately fails, the merchant is not only out the goods, out the revenue and out the cost of fighting the chargeback, but to add insult to injury, they are hit with a chargeback fee. If they get enough chargeback fees, the fees they pay the credit card companies may be increased.

Further complicating the situation, the customer has six months to dispute a transaction, in many cases. If you offer subscription services, the situation is even worse. Let’s say you provide access to a subscription service and your longest subscription plan is two years – cash up front. Customers can dispute the transactions up to 2.5 years after they originally made payment.

The fraudulent chargeback process is a classic example of non value-added rework – the proverbial “hidden factory.” But companies do a lot of it every day. They just regard it as one of the costs of phone sales

Root Causes of Fraud

The root cause of fraud gets into agents’ and consumers’ multi-faceted motivations to steal. The ultimate root cause in the case of agent fraud is the agent seeing and hearing customers’ credit card information. If agents could take the order without ever seeing or hearing the credit card information, there would be nothing to steal.

In the case of consumer fraud, the root cause and what opens the door to chargeback fraud is that the “credit card is not present.” If you go into a Best Buy, for instance, and buy goods with a credit card, there are no fraudulent chargebacks. The customer was there. They swiped their credit card (and the cashier never saw the number). They signed for the goods authorizing the purchase. They walked out with the goods. Chargebacks are virtually unheard of for purchases like this.

In MOTO (mail order/telephone order) call centers, the customers are not swiping their cards. They are providing their numbers and the agents are entering them into the system, and this allows the consumer to claim they never ordered the goods or services.

Error-proofing Strategies

Both types of fraud can be error-proofed with the same simple solution. An applet is embedded into whatever CRM (Customer Relationship Management) solution is being used. When the customer is ready to place their order, the agent asks the customer to enter their credit card number and security code using their phone’s keypad. The app reads the tones and enters the number into the CRM. The tones the agent hears, however, are masked so that the agent cannot record them for later deciphering – a type of signaling called dual-tone multi-frequency. Also, what the agent sees in their system is masked so the agent cannot see the information. A feature of the app is that it checks the number as it is being entered to make sure it is a valid credit card number. If there is a problem, the agent can ask the customer to reenter the number.

This solves the two kinds of agent fraud. The agents never see or hear the number, so there is nothing to steal or misappropriate. It also greatly improves the ability to fight fraudulent consumer chargebacks, because the consumer has to enter his or her own number. It is not totally equivalent to swiping, but it proves that the agent didn’t enter the number.

One additional step also helps fight the chargeback. Before the order is put through, the agent clicks a button to play an audio message summarizing the details of the charge to a card ending in XXXX, which the customer can hear. This audio message ends by asking the customer to confirm their agreement by stating their full name. This small transaction, referred to as a “verbal signature,” is recorded and then emailed to the customer, either as a separate email or as a part of the standard emailed order confirmation.

This approach, technically, has not error-proofed consumers from submitting fraudulent chargebacks; they can still submit them. However, the given process is a deterrent to customers submitting them in the first place (they know they had to enter their number) and the confirmation recording provides significant evidence – akin to confirming the purchase on the keypad in a store – that can dramatically improve the successful resolution of fraudulent chargebacks.


The results to date have exceeded all expectations. Though it is hard to track improvements from events that you did not know about or that rarely happen, the inability of the front-line agents to see consumer credit card information has everyone breathing a much easier. Offshore centers using the technology have reported a reduction in customer escalations and requests to be transferred back to someone in the United States.

Several companies implementing this solution have reported that every fraudulent chargeback they have received has been overturned. It also has made it easier to research and respond to the chargebacks as the information is all digitized and easy to look up and submit as evidence.

Black Belts on the manufacturing side know that if something bad can happen if you don’t do X before Y, then it is prudent to make it virtually impossible for the person to do Y without doing X first. This is the power of error-proofing.

This mindset is not as strong on the service side of most businesses, usually because the consequences are not as dire as they are in manufacturing. But if a simple app can reduce the chances of reputational damage from credit card theft and reduce the loss of millions of dollars from reworking fraudulent chargebacks, it should at least allow call center leaders to cut back on their sleep aid and rest soundly at night.

About the Author