Home

What Is Credit Card Fraud?

Cost Of Fraud

Products

Order

Q and A

Contact

About Us

Introductory FraudScreening.com Questions

What is FraudScreening.com?

FraudScreening.com is a service powered by a powerful intelligent database, and is provided to merchants, commerce service providers, independent sales organizations, gateways, aggregators, and similar e-commerce enabling companies. FraudScreening.com provides complete Internet Credit Card Fraud Detection that features fraud-risk scores that measure the likelihood of fraud for each online order.

What is a fraud-risk score?

FraudScreening.com uses predictive analysis tools based on its neural network technology. As each transaction is analyzed against cardholder and merchant data to detect patterns of fraud, a fraud-risk score is computed. In contrast, rules-based systems simply raise alerts or restrict classes of transactions, such as those from certain IP address ranges.

What kind of fraud reduction should we expect?

Your fraud reduction will be dependent upon many factors including but not limited to: your current rate of fraud, your product mix, your current rule settings, your eFalcon fraud policies and the efficiency of your staff in identifying frauds and rescuing sales.

System Requirements & Implementation of FraudScreening.com

How does my system communicate with FraudScreening?

When merchants first use FraudScreening, WE incorporate the application programming interface (API) library into the payment gateway code, allowing merchants to send transaction data to and receive scores from our fraud system. Merchants associated with Authorize.Net then send order requests directly to Authorize.Net using the standard Authorize.Net linking methods. Authorize.Net sends the scoring requests to both Service Bureaus, and the resulting scores are returned to the merchants.

What is the minimum transaction information that needs to be submitted for scoring?

The more information, the more robust the FraudScreening score will be. We expect to get all the order data that is available. The FraudScreening team is not insisting on any level of detail, since all merchants differ in the amount and quality of data that they collect.

How does your system handle a credit card number that it does not recognize?
As long as the credit card number conforms to the association mandated formats, FraudScreening will recognize and process the order.
If you are referring to accounts where we have no current profile, we have a methodology of establishing a default Internet shopping profile.

We return an exception code for credit card numbers that do not pass Mod 10 checks. Our models are trained to deliver scores that are robust on the first transaction that we see on a card, based upon other information in the order, such as dollar amount. Additionally, other information is used, such as purchase behavior at individual merchants, to contribute to the performance of the model when a new cardholder is detected.

If you have negative files, how do you maintain the quality?

FraudScreening provides the facility for use of merchant or CSP negative files for the execution of rules in the Policy rule engine. The FraudScreening system enables clients to maintain their own negative files at their site, however it is the client's obligation to update and maintain these files. It is extremely difficult to maintain negative files centrally because for what one merchant considers 'negative' &endash; may not be negative for another merchant.

Negative files are too static to be of use in a real time scoring neural network. Negative files get old and deteriorate quickly. If a credit card account is deemed to be a fraudulent, generally the credit card issuer will already have this on record.

How FraudScreening Technology Works

What are FraudScreening scoring models? How do they work?

The model consists of two elements, the neural network and identity profiles that capture customer purchasing behavior. The neural network is trained offline to recognize behavior patterns that are fraudulent based upon real historical examples of fraud. While in production , FraudScreening learns the purchasing behavior of customers from their transactions. As each new transaction is seen by FraudScreening in real time a profile belonging to the customer is updated to reflect the new purchase.

Why should we score every transaction?

To determine good profiles, the FraudScreening neural networks must score every transaction that comes into a merchant site. Profiling needs to see individual behavior in order to detect a pattern.

How would you know what criteria you would use to determine what transactions not to score? For example, the riskiest credit card transaction in the world is a $1 charge at a gas station. A fraudster often 'checks' a card before using it by making a test or a conservative purchase, i.e. buying a bible at a bookstore to test to see if the card actually works.

Profiling has to get into individual behavior patterns of customers at merchant sites overall to detect a pattern. If you only decide to score transactions above $50 &endash; you wouldn't be able to know the entire profile of that shopper if you are missing information about transactions made by the same person that are even a few dollars less, ie. a $49 dollar transaction.

FraudScreening vs. Other Fraud Detection Methods

The Address Verification Service (AVS) is an important security device designed for Mail Order and Telephone Order merchants. What are the top four reasons Address Verification Service (AVS) has limited applicability in detecting fraud on the Internet?

• AVS is a 'US only' product and does not provide fraud detection for international accounts.
  
• AVS only checks the first four numbers of the street address as well as the zip code. It cannot verify street names, city names, and so on.
  
• Fraudsters can easily call an issuer to change a cardholder's address to match where they want the goods shipped.
  
• Fraudsters can easily use reverse directories to locate the true addresses of cardholders and make sure they match on the orders. Fraudsters can then call the merchants or shipping companies and change the delivery addresses or place a change of addresses at the post office.
  

The Card Verification Value 2 (CVV2) is a three-digit value indent-printed on the back of a credit card that can be used to verify a card's authenticity. What are the top four reasons CVV2 has limited applicability for protection against fraud on the Internet?

The CVV2 is printed on the signature panel on the back of the card and as a result, it is frequently hidden from view by the customer signature or obscured so it is easily misread.
Number generation programs can create potential CVV2 numbers in little time.
Since the number is on the card, fraudsters at convenience stores and restaurants have learned to turn the card over and copy the CVV2 number as they copy the card number and expiration date.
Since not all cards have a CVV2, you can't assume that one is present.

Internet Fraud Control Consortium

What is the Internet Fraud Control Consortium (IFCC)?

The Internet Fraud Control Consortium is a database of credit card transactions that is used to build and rebuild fraud detection scoring models, The information in the IFCC is supplied by users of FraudScreening. The power of the consortium is its ability to provide all members an equal opportunity to detect and combat fraud. The data in the consortium is used to create neural network-based predictive scoring models.

FraudScreening.com always treats this information as private and confidential. It is never transmitted to any merchants, but rather if suspicion is aroused about a particular field of data, the Consortium will transmit an alert update to the eFalcon scoring service bureaus.

FraudScreening.com does not sell, rent, lease, disclose or otherwise provide your personally identifiable information to third parties for marketing lists nor does it use the FraudScreening.com Service to track your shopping habits.

Who are the Consortium members?

Any company that utilizes FraudScreening is required to be a consortium member. We have announced publicly that Circuit City, CyberCash, Flooz, Vitessa, ShopNow.com, Signio and ebit.net are installing FraudScreening. Since potential customers can join the consortium prior to purchasing the FraudScreening service there may be from time to time, consortium members who have not been publicly announced. It is not our policy to share information about consortium membership to other consortium until such time as their membership becomes public knowledge.

How can I get access to the Consortium Database?

You can't. The Fraud Control Consortium database is maintained anonymously and FraudScreening.com never shares the contents of the Consortium database. In fact even if a consortium member wants FraudScreening.com to retrieve and copy some of the data they previously contributed to the consortium there is a strict process that has to be followed which includes a written request from the contributing member.


Internet Fraud Prevention Advisory Council

What is the Internet Fraud Prevention Advisory Council?

The Internet Fraud Prevention Advisory Council (IFPAC) is a consortium of businesses including online merchants, merchant acquirers (credit card processors), card associations and card issuers. FraudScreening.com is a founding member of IFPAC.

The mission of IFPAC is to form a working industry group to reduce the risk and monetary cost of Internet fraud incurred by those involved in e-commerce. To this aim, members of IFPAC will:

• Share business expertise and knowledge
  
• Synchronize existing fraud prevention efforts
  
• Identify future opportunities
  
• Participate in research and industry data analyses
  
• Participate in education and the promotion of new technology