Nexidia on the Advantages of Speech Analytics for the Contact Center and the Enterprise

By: Patrick Barnard
TMCnet
Oct. 19, 2007

http://callcenterinfo.tmcnet.com/Analysis/articles/12852-nexidia-the-advantages-speech-analytics-the-contact-center.htm

Speech analytics is one of the fastest growing market segments in the call center industry. Thanks to major advancements in the algorithms used to interpret speech, coupled with the increasing speed of today’s processors, these automated systems now enable organizations to “mine” their recorded calls and learn details about their customers -- and their agents -- which they were previously unable to discover using manual processes such as call recording/monitoring.

Unlike the speech analytics solutions of a few years ago, which were mainly used to perform simple “word spotting” in recorded conversations, today’s solutions can actually identify the context of what is being said, thus giving organizations the ability to more accurately categorize calls and gain new insights into customer and agent behavior. In fact, these systems have advanced to the point where they can actually detect emotions - thus giving organizations a powerful new way to measure customer satisfaction.

As such, today’s speech analytics solutions have uses beyond just improving agent performance. In fact, it can be argued that they now have greater value outside of the contact center than inside. That’s mainly because today’s solutions are capable of mining all of the call data, as opposed to just a small percentage of it, which in turn enables organizations can get a more accurate view of the “big picture” and then take appropriate action. As a result, speech analytics are increasingly being used by executives in sales, marketing and operations to increase revenue, reduce costs, increase customer retention and improve the overall performance of the call center. Interestingly, as organizations come to adopt these solutions, the role of the call center is being redefined into that of a “business intelligence center” – in other words, it has become the central “hub of intelligence” for the organization.

To learn more about how today’s speech analytics solutions are benefiting organizations, TMCnet recently interviewed Anna Convery, senior vice president of marketing and product development for Nexidia (News - Alert) Inc., a leading provider of speech analytics solutions to the call center industry. The company’s advanced speech solutions enable today's contact centers to achieve the maximum benefit of speech analytics with offerings that focus on the three areas most likely to impact the overall organization: Call Traffic, Customer Experience and Business Opportunities. Utilizing Nexidia's patent-pending phonetic search technology, contact centers can gain immediate insight into their recorded contact center interactions in days, not months. Without further ado, here are Convery’s responses to our questions:

How can speech analytics help the average call center?

There are three major benefits to organizations adopting speech analytics technology: enhanced customer experience, deeper business and market intelligence and improved operational efficiencies.

For example, upon implementation, our clients immediately deepen their understanding of caller intent. Nexidia allows them to very quickly and accurately categorize calls based on the actual conversations, not on call metadata or subjective CRM reason codes. Once calls are properly categorized, users can isolate just those interactions relating to critical business issues. The intelligence gathered from this analysis is used to improve traditional efficiency metrics such as first call resolution, script compliance, agent performance and cross-sell/up-sell rates. What is even more exciting is that many opportunities for improving the customer experience are uncovered. Finally, this ‘voice of the customer’ insight can be shared outside the call center with other areas of the business, such as product management and marketing, enabling the customer-centricity that many companies aspire to but often fail to realize.

What size call center is the technology best suited for?

The larger the volume of recorded calls in an organization, the less likely it is that they are able to effectively understand what is actually happening during customer conversations. Some of our largest customers record thousands of hours of calls each month. Speech analytics allows them to focus on calls containing issues relevant to the success of their business. That said, any size enterprise that is recording calls can benefit from speech analytics technology.

How reliable is the speech recognition needed for the speech analytics at this point? Are confidence factors high enough at this point that the results delivered by speech analytics are right all the time?

The accuracy of speech recognition has improved dramatically over the past few years. While it is unlikely that any recognition system will ever be 100% accurate, Nexidia has very high accuracy rates because of our phonetic approach to speech recognition, which focuses on matching sounds rather than words; meaning that all proper names, slang and industry jargon are immediately searchable. Other speech analytics solutions convert speech-to-text, which makes them-dictionary dependent. Text searches are made against large dictionaries looking for a word match. These dictionaries must be constantly updated to accommodate additions to the lexicon.

How hard or easy is it to get speech analytics solutions integrated with the other functions/solutions in the call center?

Very easy. Nexidia indexes the existing audio files already created by the organization’s recording system; it is not invasive. We have worked successfully with audio files from all of the major recording solution vendors, as well as many smaller and custom solutions. Several speech analytics solutions only work inside their quality monitoring environment.

Does it matter where the material to be analyzed by the speech analytics solution resides? What about distributed call centers where the data may reside in many different places?

Many of our customers, particularly large enterprises, are collecting audio data from multiple locations so it doesn’t matter where the data resides. Nexidia can retain metadata attached to the call, such as location, DNIS, agent name, etc., so that calls can be easily segregated into logical groups. For customers with complex environments who prefer a managed service, we offer Nexidia OnDemand, which provides ongoing audio-video analysis, business process recommendations, business case development for utilizing speech analytics and ongoing business intelligence support.

Are the results real-time?

We are working with customers today who analyze calls in near real-time to quickly identify and react to escalations, dissatisfied customers, etc.

Do you need to be an IT professional to interpret the results offered by the average speech analytics solution?

Not at all! Nexidia’s Forensic Search, which is a “Google (News - Alert)-like” ad hoc search function, allows a casual user to search recorded calls for specific words or phrases and requires no training. Power users, on the other hand, can build very complex queries and trend reports to support their analysis of the recorded calls, generating invaluable insight for many different users in the organization.

Speech is still generally considered by many in the call center industry, to be a luxury...and for call routing only. Can you talk a little bit about the ROI of speech analytics, which I imagine is quite higher and has faster turnaround than the speech technology implementations of just a few years ago?

Market awareness for the return on an investment in speech analytics is growing rapidly. Enterprises who have adopted the technology have discovered the depth of business intelligence once hidden in their recorded audio and cannot imagine running their call centers, and their businesses, without it. Nexidia’s phonetic-based search allows companies to index and search audio very quickly so they can generate results in days rather than months. Seven figure annual savings are a common return on some of the improvements discovered through Nexidia’s speech analytics.

Many of the features of speech analytics are already quite "sci-fi," such as emotion detection. What even better, cooler functionality lies on the horizon?

It has been our experience that emotion detection technologies alone cannot be a reliable indicator of customer satisfaction due to the cultural and personality differences among callers. For example, people who are angry may raise their voices, but many will be silent or speak very softly. Nexidia is currently working on several other technologies that dovetail into our core speech recognition engine and enhance its overall ability to interpret the intent and result of a call. Because we are the creator and owner of our technology, we are continually exploring new ways to improve and expand our capabilities. Music and silence detection, gender and speaker identification are just a few of the innovations we are currently adding to our product capabilities.