Get ready for the next big thing in contact center workforce optimization: Speech Analytics. Even though speech analytics was introduced to the contact center industry back in 2003, Aberdeen Group reports that only about 15% of contact centers currently use speech analytics. In its research, Aberdeen found that speech analytics users enjoy about 6 times better year-over-year profit margin improvement and about 3 times better year-over-year customer effort score improvement than non-users of speech analytics. Why is the penetration rate of speech analytics so low? How can such a valuable tool be overlooked by so many contact centers?
Well, until recently, speech analytics faced two primary barriers:
- The voice recognition technology inherent in speech analytics was not advanced enough to achieve highly accurate textual transcriptions of voice calls (required to perform sophisticated cross-channel analytics)
- The upfront investment for speech analytics solutions were quite daunting, making ROI justifications highly questionable from the start
Let’s consider each of these in turn.
Just the past few years have seen remarkable improvement in the word error rate of voice recognition technology. Last month at Google’s I/O 2017 developer conference, CEO Sundar Pichai revealed that the company was achieving a 4.9% word error rate for phone-based Google searches. That’s pretty darn good when you consider that human beings have a manual transcription error rate of about 3%, and only one year ago, Google’s error rate was about 8.5% (down from the 23% word error rate in 2013). Few technologies have improved so dramatically in such a short period of time. Voice recognition technology is closing in on what a human quality analyst understands while evaluating a call. No wonder recent market research projects the growth rate of speech analytics to be about 22% CAGR through 2020. Speech analytics technology is becoming as fundamental to contact center operations as workforce management and quality monitoring.
There’s little doubt that speech analytics is a complex process that requires quite a bit of memory and CPU horsepower to convert potentially thousands of unstructured voice conversations every day into structured, categorized and searchable data. With traditional on-premise delivery models, the combined hardware and software cost was often prohibitive, especially for small and mid-sized contact centers. Of course, in the past few years, we have seen a rapid race to the cloud with many workforce optimization applications now available in a subscription pricing model that’s quite attractive. Speech Analytics is one of those WFO applications that’s increasingly implemented in a private hosted or public cloud environment eliminating the large initial Capex investment and turning it into a more palatable monthly Opex expense.
As of today, practically any contact center can enjoy remarkable benefits from speech analytics technology, and do so with an excellent Return on Investment (ROI). Aspect has sponsored new research by Pelorus Associates that clearly documents some of the many ways in which speech analytics technology can improve the quality and operational efficiency of individual agents, teams and the entire business as well as provide valuable insights into the voice of the customer. We’ve compiled this research in a new eBook: 10 Reasons to Invest in Speech Analytics.
You can get a free download of this informative new eBook here. It’s packed with information to help you understand why speech analytics may be a technology you want to consider since it can bring some huge benefits to your contact center. If you like what you read in our new eBook, contact us at 888-547-2481 for more information on Aspect’s powerful speech analytics solution – Aspect Engagement Analytics.
Latest posts by Mike Bourke, SVP & GM Workforce Optimization (see all)
- Time to Talk Seriously about Speech Analytics in the Contact Center - June 16, 2017
- Optimizing the Workforce in an Era of Digital Employees - March 16, 2017
- Performance Management Bridges the Divide between Big Data to Big Knowledge - November 28, 2016