What are the most common applications for: speech recognition and speech verification?
Traditionally, speech recognition and verification have been used by contact centers to automate and improve self service applications and to authenticate customers automatically in an attempt to reduce transaction time. Today, these technologies are being applied to customer relationship management, customer feedback, risk management and compliance, quality monitoring, and performance management as well as agent coaching and development.
When deployed in an integrated environment as part of call recording and quality management systems, speech analytics can provide insight from customer interactions that is valuable to the entire organization, not just the contact center.
For example, by using a list of key words (such as competitor names), the sales department of an organization can understand what competitors are doing regarding product and pricing. Sales and marketing executives can get this critical competitive information efficiently and quickly, and use it to help make key product, pricing and process decisions. Customer feelings and intentions can also be discerned by analyzing interactions for words such as "cancel, or "refund." Alerted to such instances and combined with emotion detection to validate customer intent, proactive actions can be taken to address these customer issues before they become customer defections.
Speech verification can help an organization detect and reduce fraud. Savings from fraud reduction can be re-deployed into customer programs, helping to increase customer satisfaction or growing the business. Speech verification can also help to streamline customer interactions, allowing agents to spend more time in productive interactions with callers.
By using advanced speech technologies, organizations can extract insights from customer interactions not only to improve agent quality, customer satisfaction and retention, but also to become a more efficient and, most importantly, a more effective operation
In your opinion, where and how will we be using speech technology in say 5 years time and what challenges will we need to overcome to get there?
I believe that over the next 5 years, and some of these will evolve sooner, speech technologies will become ingrained into our daily lives and into the tools we use to conduct our business.
In contact centers agents will receive authentication of user identity by means of speaker voiceprints. This analysis will take place as part of the conversation itself and not as a separate "authentication event".
We will interact with "machines" (PCs, cellular appliances) using natural language just like we do with our mobile phones today. We'll even dictate our email.
PDAs will translate languages as we speak
Contact centers will widely adopt systems that analyze our voices and our emotions and will use this information to help provide better service. If we are upset, they won't try to up sell us, etc.
These advances, powered principally by stronger, faster CPUs and denser, cheaper storage mean that speech analytics will be widely used in fully integrated end-to-end implementations that will enable the contact center and the entire enterprise to extract insight from interactions. By end-to-end I mean solutions that integrate speech with a whole range of other data sources, such as CTI, business data, call-flow analysis, screen activity, and customer feedback.
Speech analytics will be used in the contact center for analyzing what customers say. Every call will be automatically transcribed, classified and later mined for customer intelligence. The contact center will become a primary source of information on customer habits, wish lists, the competition, market trends and more, and by turning into the eyes and ears of the enterprise will offer the enterprise tremendous benefits.
Many of these "futures" are very exciting, but I don't want anyone to get the impression that they should wait five years to deploy speech technologies. While we will see improvements and new applications, these technologies are now sufficiently mature to help companies achieve real business improvements.
What are the common misconceptions of speech recognition?
As speech technologies become more widely deployed in service automation solutions (such as speech-enabled IVR), there is a tendency to underestimate the challenge involved in moving from this to more complex speech recognition solutions capable of analyzing freely spoken agent/customer interactions in a contact center. There is also confusion between using speaker-dependent speech dictation or voice commands solutions that are trained to recognize specific speakers, and even specific words versus trying to analyze conversations by speakers unknown to the system.
A third common misconception is that a discrete, stand-alone application can perform the kind of analysis that is required to truly gain insights from customer interactions. Only an end-to-end, fully integrated speech solution that correlates as many data sources as possible can provide the full picture, place the customer's words in context and deliver uncompromising accuracy.
Technically, how advanced are we today with speaker recognition?
Today we are able to analyze human voice and achieve a very high level of authentication. Current technologies can identify a person with an accuracy that approaches 99.9%. While some challenges still exist when moving to an unsupervised environment, technology keeps improving and has already reached the stage of practical implementation. Speaker recognition can be used to provide secured access to systems such as banks accounts, and implement password resets without reaching an agent.
Speaker recognition can now also be used to fight identity theft and give consumers better protection, prevent fraudsters from opening new accounts, avert the issuing of false credit cards, etc.
How will these technologies change the way contact centers function?
Customer inputs are critical to understanding what an organization needs to do to stay ahead of the curve, and the contact center is a strategic focal point and hub for gathering these inputs. Every day, customer service representatives speak with customers about the issues that matter most: product or service feedback, spending patterns and competitive offerings and more importantly their level of overall satisfaction
Speech technologies are adding an additional dimension by helping to identify and classify customers who should be given additional, even special attention. This allows contact centers to deliver what customers are asking for, provide better feedback to the rest of the enterprise and increase their overall worth to the corporation. This results in higher profiles for contact center management as well as the overall function and helps to shift the contact center into a profit center rather than a cost center .
Will the technology replace the role of the agent?
Technology can not and will not replace the role of the agent. Rather it provides the agents, their supervisors, management and indeed the contact center overall with tools to deliver better service. Regardless of how much technology we implement, "at the end of the day", an agent speaks to a customer. This interaction can make or break the customer relationship. There is a wealth of information "inside" this interaction and speech tools help us to gain insight from these interactions. Assuming these insights are properly used, the agents will better serve their customers, be more successful selling them additional products and services and as a result feel better about their jobs. This will have the add-on benefit of increase agent satisfaction which results in lower agent attrition, cost saving for the organization and most importantly, higher levels of customer service. There is endless research that tells us that happier employees provide better service to our customers.
Contact center agents are critically important to an organization's ultimate business goals of increased customer satisfaction and retention, profitability, and market share growth. Technology in general and speech technology in particular, helps to emphasize and empower agents to drive the success of the enterprise.
Executive Interviews
