CallMiner Adds Semantic Building Blocks to Eureka Speech Analytics

CallMiner, a provider of speech and voice-of-the-customer analytics for contact centers, has updated its flagship Eureka speech analytics product with a Semantic Building Blocks functionality that allows contact centers to more rapidly identify calls and customer contacts that contain specific behaviors or conversational flows by reusing previous language searches or analytical queries.

With Semantic Building Blocks, CallMiner enables analysts to leverage categories (previously defined queries that automatically tag calls containing certain topics) as building blocks for subsequent searches or categories. The new functionality lets analysts use the same search logic currently supported in Eureka to stitch together sophisticated conversational flow analysis without having to code lengthy queries.

"We've had for years a library of categories that allow people to find calls that match certain criteria," explains Scott Kendrick, vice president of marketing at CallMiner. What's different about Semantic Building Blocks, he says, "is that you can take any of these categories and sequence them together.

"You can stitch things together more easily," he adds.

Mining speech, Kendrick points out, "is still not an easy thing to do," but with Semantic Building Blocks, "the work you've done before is so much more available."

Semantic Building Blocks enables the following types of analyses:

• Identifying timely or undesired agent behavior within the call—agents who are putting customers on hold within the first 30 seconds to finish their notes from the previous call;
• Trending sentiment—calls that begin negatively but end on a positive note;
• Correlating topics—customer sentiment or competitor mentions near product, service, or promotion references;
• Measuring sequence of events, such as compliance statements that must occur within a specified order, sales procedures that drive revenue (such as offering a gold plan before silver), and service best practices that improve the customer experience (such as demonstrating empathy after customers' expressions of dissatisfaction).

"You don't want to just find calls with agent empathy and customer frustration in them, but [also] when the agent showed empathy after a customer expressed frustration," Kendrick says.

Though the Eureka product does not have a specific emotion-detection engine, it can measure emotional states "through expressions of language and acoustics that suggest that a conversation or part of a conversation is emotionally charged," Kendrick points out.

Semantic Building Blocks are not just limited to speech anayltics during recorded phone conversations. The same technologies and algorithms can be applied to text-based interactions, such as email, chat sessions, text messaging, or social media.

"The building blocks can be applied across the board," Kendrick says. "You can look at all different types of content to find specific scenarios."

And, "they can be used for service, sales, and support applications," he adds. "It's about helping people understand the entire customer journey."

---

Related Articles

CallMiner Updates Eureka Product

New analytics platform offers real-time guidance for agents based on the current conversation.