Clarifying the current capabilities of speech leads to the conclusion that speech is ready.
Posted Jun 1, 2005
The use of speech recognition in customer service operations is gaining momentum. Industry trends are driving technology innovation that makes the move to speech recognition faster to implement and easier to support. As the speech technology adoption rate increases, business management is turning its attention to this efficient capability to represent their company to its customers and complete tasks only live customer service agents (CSA's) could accomplish in the past. As businesses begin to appreciate the power of speech recognition, a realization dawns: "Speech can revolutionize the way we conduct business with our customers."
Why is speech recognition not ubiquitous in all self-service operations? The answer: three myths and one changing reality have up until recently served as impediments to the rapid deployment of speech for customer self-service.
Myth 1: To Reduce Cost of Operations, Quality of Service Must Be Sacrificed
Customer service operations across industries have undergone many consecutive years of service degradation, driven by the need for severe reduction of operating costs. A recent consumer survey found that consumers identify call center agents and self-service systems as the face of a company and base their purchasing and loyalty decisions on the quality of service they receive more than on any other company or product attribute. More than 50 percent of those surveyed had stopped doing business with a company because of a negative call center experience.
Speech naturally addresses both customer satisfaction and operations efficiency. Without speech, customers who zero-out of the IVR to transfer to a live agent nearly double call time, tie up facilities, and use live agent resources for tasks that can easily be automated. When speech recognition is added to self-service operations, the number of zero-out transfers to live agents are reduced, there is an increase in calls completed in self-service, and a decrease in overall call handling time by the system in well-executed speech applications.
Myth 2: Speech is Not Yet Robust Enough for Utilization in Enterprise Mission-Critical Applications
The leading companies of every services industry have already implemented enterprise-level speech recognition applications, or are in the process of doing so. Bank of America, Sprint, and Sears are just a few of the companies that have received industry and press attention for their implementations of major speech recognition applications as significant parts of their self-service operations. These implementations prove that speech is robust enough to meet the exacting standards of some of the largest IT communications operations in the world.
Myth 3: One Part of the Solution is More Important Than the Rest
For a speech solution to be successful, it must have a good balance of all the components and forces creating and managing the blend of automated and agent-driven customer interactions. The components include the platform (IVR or voice browser); Voice Application Server, and tools to create, test, and maintain applications; Speech Recognition and Text-to-Speech Engines; Voice User Interface (VUI) and application. There are a handful of vendors today that can deliver all these components and offerings quickly and integrate the solution with existing operations and infrastructure.
In the last few years improved technology has greatly reduced the level of effort it takes to deliver speech solutions. This has come in the form of modules and fully pre-packaged speech applications. When done well, these packaged offerings combine all the elements guided by experience, best practices, pre-written integrations with common data sources, and best process guidelines into a quickly deliver, quickly customized offering. In these offerings, the myth of one element being more important than another simply recedes in the reality of quick and effective self service.
The Changing Reality: The Speech Apps Being Delivered Today Are Far Superior to Those Implemented Just a Few Years Ago
Technology has vastly improved accuracy of recognition, even in tough situations like mobile phone usage. Tools and techniques have greatly reduced the time it takes to create and implement custom applications. Packaged offerings now bring 60 percent to 80 percent of desired functionality and integration out-of-the-box reducing implementation, maintenance and upgrade costs for low overall TCO. Speech applications built on old technology are years behind the abilities of today's applications to deliver value to customers and businesses.
Dispelling three myths and clarifying the current capabilities of speech leads us all to the single final conclusion: Speech is ready.
About the Author
Chris Nichols is senior director of product marketing and management for Edify Corporation. He spent 11 years with Nortel in increasingly senior marketing positions, and later worked with other software companies focused on contact center and CRM solutions. Before coming to Edify, Nichols worked at Ernst & Young, Aspect, and performed independent consulting services for other companies in the industry. Visit www.edify.com
Sponsored By: Genesys, Avaya, Verint, and Aspect
Sponsored By: Informatica