Speech Technology Magazine

 

New Uses for Speech Analytics

Artificial intelligence and machine learning are transforming many industries and technologies. Speech analytics is no exception--take a deep dive into the new uses and applications for speech analytics.
By Phillip Britt - Posted Mar 6, 2019
Page1 of 1
Bookmark and Share

Speech analytics can provide users with valuable contact center insights that they didn’t have before, according to experts that discussed the subject during a recent Speech Technology webinar.

A New Paradigm

Speech analytics is much more than transcription, said Steve Chirokas, director of product and channel marketing for CallMiner. Speech analytics identifies intent, effort, sentiment, emotion, using contextual visibility.

It covers all of the communications in the contact center, so it’s a massive amount of data, Chirokas added. Automated scoring enables the user to focus on certain areas.

To handle the massive amount of communications, speech analytics is scalable, running in real-time through a secure cloud application.

“We’re able to use automation and things behind the scenes to bubble up topics that you may not have thought of,” Chirokas said, pointing to a provider of telephone accessories that found that speech analytics provided by CallMiner not only helped identify the caller’s issue for the agent more quickly, providing significant savings in terms of call handling times, but also helped the company recognize opportunities for product innovation.

Several Opportunities from AI

AI-driven speech analytics can help with sentiment analysis, IVR, and digital containment analysis and customer journey analysis, said Abby Monaco, senior product marketing manager for NICE nexidia.

Sentiment uses machine learning to help determine if a customer interaction is positive, negative, or neutral by analyzing positive and negative words and phrases, pitch and tone, as well as other “tells,” such as cross-talk (agent and customer talking at the same time) and laughter detection.

Companies use sentiment analysis to help with agent performance measurement, compensation packages, and enablement and quality programs.

Customer journey analytics uses machine learning to connect seemingly disparate customer interaction sources into a single, consolidated journey, providing the user with critical business insights.

Companies can grade the customer journey by evaluating customer feedback (based on churn, renewals, complaints, upsell/cross sell success and sentiment); explicit and implicit experience indicators such as surveys, complaints, journey duration, channels used and customer intent. Then those factors need to be combined with historical data, such as past interactions, sentiment as scored throughout the customer journeys and customer satisfaction over time.

IVR optimization visualizes the IVR flow for all customers, identifying choke points or dropped calls, with an analysis for improved self-service.

Digital containment is an analysis solution that identifies customer journeys that dropped out of web or mobile apps, while identifying bottlenecks or problems to be eliminated in order to improve the online customer experience.

Four Examples

Carmit DiAndrea, vice president, portfolio marketing strategy for Verint discussed four separate case studies in which an organization benefited from employing speech analytics.

One company reduced its “super detractors” by 16.4%, converting most of them into promoters. Another organization used speech analytics to successfully improve the company’s ability to predict customers likely to churn. When a churn candidate was detected using the speech analytics, the organization made a personalized offer—based on the customer’s lifetime value—in real-time to urge the customer to stay.

Speech analytics improved the organization’s churn prediction rate from 60% to 75%, while also decreasing the false positives from 40% to 25%.

Speech analytics helped yet another company pinpoint the best- and worst-performing contact center agents based on sales conversion percentages. The analytics helped identify the best practices for language usage and avoidance as well as approach training and other factors. Once identified, the langue and other factors were used as a basis for training and monitoring other agents.

The fourth company used speech analytics to help a health insurance company recognize certain key words related to certain medical condition in order to provide the agent with a link to a relevant article in the knowledge base in real time, improving CSAT scores by 15% in the first to months.

For more complete details, see the complete webinar at http://www.speechtechmag.com/Webinars/New-Uses-for-Speech-Analytics-1230.htm

Page1 of 1
Learn more about the companies mentioned in this article in the Speech Technology Buyer's Guide:
Learn more about the companies mentioned in this article in the Vertical Markets Guide: