Voice Verification and Fraud Detection – Interchangeable or Complementary Solutions?
May 27, 2014 Leave a comment
- Voice verification technology has advanced substantially and vendors are introducing reliable products capable of positively identifying legitimate customers, by matching their voices with previously-recorded voice prints, with a very high level of confidence.
- Fraud detection solutions have entered the marketplace and are being used successfully in many security-sensitive markets to identify known fraudsters when they call into a contact center by comparing their voice to previously recorded voice prints stored in a fraudster data base.
Voice recognition technology has improved enormously over the past five years which has resulted in several effective tools to assist contact center operators in verifying the legitimacy of virtually every caller entering a queue in a customer care environment. This biometric approach to customer identification provides better accuracy and a level of security beyond that of the previously-used security question approach to customer identification (e.g., What is your mother’s maiden name? and, What are the last four digits of your social security number?). Over the past few years, fraud detection systems that match a caller’s voice with previously-recorded and stored voice prints of known fraudsters, have been installed by many companies wishing to deter fraud in their customer care environments. These systems have been successful in mitigating fraudulent transactions in the financial services market segment, if not stopping them altogether. More recently, voice recognition solutions have taken a more positive turn and are designed to compare the voice of legitimate customers to their previously-recorded, passively collected voice prints before allowing a risky transaction, such as an address change or money transfer, to be completed. These systems are also showing positive results in many banks, credit card and retail customer care environments.
That being said, the next major question that must be discussed is whether a company that has implemented a customer voice verification system can logically assume a previously-used fraud detection solution is no longer required. In other words, does positively verifying legitimate customers negate the need to attempt to match a caller’s voice to lists of known fraudsters before authorizing a transaction? In a perfect world where we are able to collect voice prints of all our customers and use them to verify that the caller is who he/she says they are, this might be a logical assumption. However, in the world in which we live, an enterprise should never assume it can successfully have every customer’s voice print prerecorded and ready to match to a caller’s voice for verification. In any case, multiple levels of security are still recommended because the costs of letting even one fraudster through the screening process undetected can be extremely high in terms of loss of dollars, customer trust, and corporate reputation. Therefore, at least for the time being, we strongly recommend that layered levels of security combining customer voice verification, fraud detection and possibly even a spoken identification or account number be used in tandem to minimize the potential for asset loss.
Today customer voice verification and fraud detection systems are being sold as separate and distinct applications requiring enterprises to invest in both to maximize security in their customer care and support departments. Since these offerings are not interchangeable solutions, but very complementary pieces of the security puzzle, it would be logical for vendors of these systems to bundle them together in the future in order to offer a consolidated security package at a bundled price. If you are looking to better secure your contact center transactions and considering either solution, it may be wise to request a bundled offering from your vendor prior to making a product selection.