BPA Featured Article

Speech Analytics Provide Contact Centers Internal, External Benefits



By Paula Bernier, Executive Editor, TMC
April 20, 2018


Automation and understanding are two words being used with increasing frequency in communications and collaboration circles these days. And analytics help power both.

Contact centers that embrace analytics can reduce operating costs while keeping customer satisfaction levels high. They can do that by automating processes such as call monitoring and scoring. That way, they can lower human resources requirements, and people are freed up to do more value-added work.




But, as indicated above, contact center analytics is more than just a cost savings play. Contact centers can also use solutions in this category to better understand their customers.

Speech analytics does that by collecting, aggregating, and structuring data from calls, chats, emails, and social media. It then looks at that data to identify the cause of problems and identify trends.

That can help contact centers understand when and where in the customer journey problems frequently arise. That way, contact centers and other departments within enterprises can take action to improve those processes. Speech analytics can also alert contact center managers when the tone, phrases, or words callers are using signal frustration, the likelihood to purchase a certain product, or other noteworthy signposts.

Companies that implement speech analytics can benefit from more sales, happier customers, better first call resolution rates, and more effective agents, blogged Monet Software’s (News - Alert) Chuck Ciarlo. Yet just 15 percent of contact centers were using speech analytics as of July, added Mike Bourke of Aspect.

Bourke says there are two basic kinds of speech analytics solutions used in the contact center. One is the phonetics variety. It allows for fast processing but actual searches tend to be slow and often turn up false results. The other is a transcription-based approach known as Large Vocabulary Continuous Speech Recognition. This version of speech analytics, Bourke says, is more accurate but requires more time for processing since it’s working from a larger library.




Edited by Ken Briodagh

Home