Problems with online approvals? See what the creditors are looking to match............ http://www.experian.com/newsletter/vision_fall_authentication.html Verify customers online with Authentication Solutions Internet traffic is on an explosive growth curve with no end in sight. More than 1.6 million new users start surfing the Net every month-that's 1.7 users every second.* Moreover, e-commerce is creating unrivaled opportunities expected to total $250 billion by 2002, a 525% increase over 1999.** At the same time, the Internet is spawning a whole new breed of fraudsters. According to Louis Harris & Associates Research (1999): E-commerce has contributed significantly to the rise in identity theft Internet fraud ranges from 1-5% of total sales, compared with less than 1% for traditional retailing Online credit card fraud has affected up to 6 million consumers Stolen cards and account numbers account for up to 50% of Visa's discovered Internet fraud The costs of fraudulent transactions, which can mean the difference between profit or loss on the Internet, are typically assumed by the merchants or service providers. Fighting online fraud As part of our commitment to reduce fraud, Experian initiated a pilot program to verify consumer identities prior to approval of online transactions. Called Authentication Solutions, this breakthrough product provides a confidence index for accept/refer decisions based on two levels of data: In-wallet information: name, current address, Social Security number, DOB, home phone number, listed vs. unlisted phone number, driver's license number and state Out-of-wallet information from consumers' credit reports: mortgage lender and payment ranges, auto lender and payment ranges, previous address, retail card and balance ranges, student loan creditor and balance ranges, etc. New authentication model Initially, the pilot program is using a scorecard for the in-wallet data, combined with the percentage of questions answered correctly for the out-of-wallet data. This in-wallet score originates from a new generic authentication model based on the algorithms being used in other Experian fraud products. Experian's Fraud Shield, for example, already looks at factors such as how well the name and address a consumer gives online match the name and address in our credit database. It also compares Social Security numbers for matches with our social tables, which also indicate if a particular number has never been issued or belongs to a deceased person. Similar kinds of in-wallet information are available from Find Consumer, Experian's proprietary search-and-match algorithm. For the out-of-wallet data, the consumer is asked to answer up to four questions on a single screen. Randomly selected questions from Experian's Authentication Solutions typically ask about mortgage and auto loans first, if that information is present in the consumer's credit file. To verify their identity, consumers choose the correct answer from a drop-down list that's randomized to reduce the potential for someone else to guess correctly. In a small developmental sample conducted recently, 88% of those surveyed correctly identified their mortgage lender. When asked to guess at the right answer, psuedo-fraudsters were able to answer only 18% of the questions correctly. Instantaneous decisioning In just seconds, the two scores for in-wallet and out-of-wallet data are matrixed together, providing an almost instantaneous decision whether to accept or refer for a manual review (see Figure 1). If a consumer is accepted for a loan, he or she would then be allowed to fill out the remaining online application. For online purchases, the consumer is directed to complete the transaction. Protecting e-commerce profits High volume e-tailers and financial institutions granting credit over the Internet will be able to use Experian's authentication solution to: Protect their e-commerce profits Reduce fraud losses Maximize application processing turnaround Save time and money reviewing applications Adjust for shifts in their population bases Streamline decision strategies *Geoffrey Ramsey, Statsmaster, eMarketer **Boston Consulting Group Figure 1 - Confidence index matrix The accept/refer decision is derived by defining a threshold based on a matrix Out-of-wallet In-wallet 1-20 21-240 241-460 461-640 641-660 661-999 0% R R R R R R 25% R R R R R A 50% R R A A A A 75% R A A A A A 100% R A A A A A The scores across the top (1-999) are based on verifying in-wallet information. The higher the score, the less likely it represents a fraudulent application. The percentages listed vertically represent the number of out-of-wallet questions the consumer answered correctly. The resulting matrix indicates whether the transaction should be accepted or referred for manual follow-up. For example, if a consumer scores in the medium range of 241-460 and answers 50% of the questions correctly, the transaction would be accepted. For more information on how Experian's Authentication Solutions can protect your e-commerce investment, contact your local Experian sales representative or call 800 333 4930.
