Speech analytics technology is helping companies gain a better understanding of their customers.
When it comes to customers' calls to a contact center, there's an upside to anger. And embarrassment, excitement, and frustration. Thanks to speech analytics technology, companies are going beyond quality assurance when it comes to monitoring their customers' calls and are measuring underlying information -- like emotion -- to improve everything from customer service and agent performance to sales and marketing functions.
Speech analytics searches unstructured audio data (e.g., customers' conversations with agents) to find metadata to analyze. Emotion detection is just one (admittedly sexy) component; the technology can detect customers' stress levels from not only word choices, but also from such elements as call tempo. It mines data for such information as the reason for the call, agent politeness, competitors' names or products mentioned, and call length.
It sounds like the stuff of science fiction, or at least a James Bond movie, but it's a reality: Contact center managers can now detect and analyze, in near real time, hidden data within a call -- data that can be used to help boost customer satisfaction, increase cross- and upsells, improve service processes, and cut defection rates. They can measure whether a customer is angry, frustrated, or satisfied with her experience, and then can react accordingly.
"The technology is hands down the most exciting thing I've seen in the last 24 years in this market," said Donna Fluss, president of DMG Consulting LLC, which recently published the 2007 Speech Analytics Market Report.
Speech analytics technology, including emotion detection, has its roots in government, and has been in use there for a long time, Fluss said. But in the commercial market, speech analytics is in its infancy. Gartner research director Jim Davies, a coauthor of the report "Unravel the Complexities of Call Center Speech Analytics," said fewer than 1% of call centers are using the technology right now.
Still, it's a fast-growing market: According to DMG, speech analytics implementations grew from 25 in 2004 to 603 by the end of 2006 -- representing a compounded annual growth rate of 391% -- and are expected to grow 100% in both 2007 and 2008. The ability to quickly gain insight into customers' needs and wants is driving implementations, according to Fluss, but the benefits go beyond customer service. Sales and marketing can gain value, as well as the call center itself.
"Call centers are getting hundreds of thousands of calls from customers every day," said Yoel Goldenberg, vice president, contact center and enterprise solutions, of Nice Systems. "They're calling to get information, or complain, or ask questions, but in many cases they're also providing very important information as part of the conversation -- maybe a new offer they got from a competitor, or a new service they like or don't like."
By analyzing calls using speech analytics, companies can discover rich data that helps improve everything from a call center agent's individual performance and script adherence to how a customer responds to an advertising campaign or an upsell opportunity. It can flag customers at risk of closing their accounts, and track how many mentions a specific marketing campaign or a competitor's product gets.
Emotion detection technology like Nemesysco's QA5 solution even allows contact center agents to identify and respond to caller emotions (such as frustration) in real time, and alerts supervisors to problematic calls while enabling them to intervene or guide the agent in a "whisper" mode.
For Lending.com, which implemented Nice Systems' speech analytics tools at the beginning of the year, the technology is a way to help the company provide what it calls "over the top" customer service. Lending.com uses speech analytics and emotion detection to take the guesswork out of observation. "We can objectively listen to customer interactions to pinpoint the ebbs and flows of the conversations," said Carrie Kelleher, COO of Lending.com. "This enables our managers to use evidence-based coaching to help employees make every interaction a great one." Kelleher declined to give results, but the company has the capability to flag customer product requests automatically, which helps Lending.com focus on finding specific loans for customers.
Beyond the initial benefit of training, the technology has yielded measurable results. "We knew that the emotion detection aspect of this technology would help us create a better experience," Kelleher said, "but we didn't expect the immediate impact it had on helping us close more sales. Gaining insight into the emotional component of a sales interaction has allowed our loan officers to improve at aligning the right selling tools at the right time."
Another benefit to both companies and customers is a better understanding of upselling and cross-selling opportunities. When speech analytics provider CallMiner analyzed the emotional context of calls relating to upselling and cross-selling offers for one prospective large financial services client, it found that the offers were negatively affecting customers—even though they were increasing sales.
"Sometimes customers are so stressed by an upsell offer that you're actually decreasing the customer satisfaction experience and taking away from the value of the upsell," said Jeff Gallino, CEO of CallMiner.
Having that information allows a company to delve deeper into the data and discover how different customer demographics respond to different upsell and cross-sell opportunities, and then determine how to optimally offer upsells without upsetting its customers.
Yet for all its benefits, speech analytics will provide the most value for companies when paired with other, more operational, customer analytics, Gartner's Davies said. The audio data "is only one aspect of what you've got, and it's quite dangerous to make decisions based on one area."
Reprinted with permission from 1to1 Media. (c) 2006 Carlson MarketingWorldwide.