Speech Recognition Technology Offers Customers Self-Service
Picture an insurance agent riding in a car on the way from the office to a social engagement. The agents cellphone is on “hands-free” and a call has been placed to a carrier for a policy the agent is attempting to sell to a prospective client.
System: “Hello. Im Julie, the automated assistant for XYZ New Business Processing. How can I help you today?”
Agent: “Hi Julie. Id like to change the name of the beneficiary and the face amount on the 10-year term life policy I just submitted to underwriting for a potential customer.”
System: “Great. Lets continue. Please say your agent ID.”
System: “Thank you. Please say the name of the proposed insured.”
Agent: “The proposed insureds name is John Doe.”
System: “Thank you. Please say the name of the new beneficiary.”
Agent: “The new beneficiarys name is Jane Doe.”
System: “Thank you. Please state the new Face Amount.”
Agent: “The new Face Amount is 500,000.”
System: “Thank you. Can I help you with anything else today?”
Agent: “No. Thank you. Goodbye.”
System: “Goodbye. And thank you for calling XYZ New Business Processing.”
The agent ends the call by simply hanging up and the carrier system and database are updated with the new information without a single interaction with an expensive live customer service representative, a mouse or a keyboard.
This is just one example of the convenient and cost-efficient self-service that speech recognition technology, if properly deployed, can bring to the insurance industry.
Although speech technology interfaces have been in use for quite some time now, there are currently a number of different technologies being utilized for a variety of applications. The conversation above is a sample of speech recognitionwhich may be the most potent technology for cost-efficient customer self-service in the insurance marketplace–but there are other technologies currently in use. Following is a brief overview of the three most widely deployed speech technologies:
Text-to-Speech: Text on a computer screen is converted to sound by the computer. The customer cannot “speak back” to the computer and have that speech recognized; the communication via sound only goes from the computer to the caller.
Interactive Voice Response (IVR) Systems: Callers are given the ability to navigate through menus of information and utilize services by pressing buttons on a touchtone telephone keypad in response to computer-generated instructions.
Speech Recognition: This is the ability of the computer to recognize general and naturally flowing speech. The caller can talk to the computer and the computer understands what the caller is saying. The computer also has the capability to respond with “speech” and a dialogue may be conducted between the computer and the caller.
Speech recognition generally is used to allow callers to speak to computers in an everyday conversational manner in order to accomplish self-service.
How does it work? A caller speaks through the phone or the Internet to the computer. The computer then sends it to a “recognizer,” which breaks the “utterance” into sound called phonemes. An attempt is then made to match each individual sound to a database of phrases the caller might say. What the caller originally said is considered recognized by the system when a match is found. The system may then respond back vocally to the caller as well as perform internal system changes based on the match, such as an update to a database-driven Web site.
IVRs are the form of speech technology the majority of callers have encountered before. Some IVRs provide a form of very basic speech recognition; for example, the user can say “one” or press 1 on the telephone keypad. However, for anything more than simple menus and simple business commands, this type of simple speech recognition becomes more difficult and less efficient as commands and menus get more complex.
There are many industries, in addition to insurance, where speech technology is widely deployed. Speech technology is widely used for customer self-service in financial services (mortgage banking, commercial and investment banking, credit card processing), telecommunications, travel and hospitality, and retail.
There are a number of benefits from using such technology. Businesses today are faced with an increasing volume of interaction with customers that need information and services. Traditionally customer service representatives (CSRs) were used to handle all customer phone interaction followed by, in some cases, a combination of IVR and CSR support.
Although IVRs were supposed to solve cost and efficiency problems, callers now are often frustrated by the ineffective and difficult-to-use push-button interface. Web-based self-service may eventually prove effective, but most customer interactions still occur over the phone. The best way to cut down on the cost of servicing an individual caller (be it an agent or a customer) is through either the partial or total automation of customer service functions through speech.
Deploying speech recognition technology in the enterprise is relatively easy and convenient, and security can be as strong as is requiredincluding the use of biometric access controls that are recognized by the U.S. government. Speech Recognition technologies can be easily integrated with current carrier architectures and meet CIO system and security requirements.
Following are some of the technical advantages that speech recognition provides:
Security and Voiceprints: Every human voice has unique characteristics that translate into an individual “voiceprint.” A voiceprint is a digital representation of the unique features of an individuals voice. Voiceprints are an approved biometric technology and are being used currently by the U.S. government to ensure the confidentiality of sensitive data and information. As with fingerprinting, a voiceprint contains the unique characteristics of an individuals “vocal tract” and cannot be altered, forged or stolen.
System Architecture: Most speech recognition technologies are Java-based and can be implemented as a part of Java 2 Enterprise Edition architectures including pre-existing user directory services and databases. The industry standard Voice XML component can be easily integrated with any ACORD-based XML environment. Speech recognition technology is compatible with any type of enterprise security infrastructure, including those that meet HIPAA and GLB requirements.
So what does all of this have to do with new business processing? There are a number of areas where speech recognition can both lower costs and speed up the process of bringing in new business. Key areas that can be addressed include: digital signature via voiceprint for policies; voice forms data collection for application information, application name and address changes, and change of beneficiary, rate quotes, claims, and commission and pending case status for agents.
In summary, voice can give agents and customers alike the ability for full self-service during the new business process (as well as ongoing and in-force updates and changes) while at the same time both lowering the cost of ownership for carriers and speeding up the overall process from first contact to closing.
Jonathan Roth is director of business development for WorkForce Technologies, based in Herndon, Va. He can be reached at firstname.lastname@example.org.
Reproduced from National Underwriter Life & Health/Financial Services Edition, October 31, 2003. Copyright 2003 by The National Underwriter Company in the serial publication. All rights reserved.Copyright in this article as an independent work may be held by the author.