SaaS, Big Data and More: Speech Analytics in the Contact Center
Speech analytics has caught on in the contact center. It's likely to become even more popular, thanks to Big Data, SaaS and other trends.
It was front page news in 2005 when it was revealed that the U.S. National Security Agency had secretly captured emails and recorded phone conversations, then used sophisticated software to conduct key word searches. That’s nothing compared to what many contact centers are doing, thanks to increasingly sophisticated speech analytics applications. At some centers, software is used to record and analyze every phone call, email, Web form, survey and social media contact.
This Call May be Monitored, Recorded and Analyzed
DMG Consulting reports there were only 25 traceable speech analytics implementations in 2004. Two years later there were 176,825 seats, and by last August the number had increased 11-fold to more than 1.9 million. The benefits of such solutions are not confined to the call center. DMG attributes the products as delivering "quantifiable … benefits to the entire enterprise, not just the contact center."
"Customer care executives realize that their value to the organization grows exponentially if they are able to organize and analyze the information stored within the customer interactions coming into their department and share it throughout the company," said Nexidia COO Jon Ezrine. “Tracking the customer journey and understanding their experience has an effect on everyone from senior level executives to marketing to product development and being able to harness the ‘voice of the customer’ [VoC] has impact far beyond just the call center.”
In addition to adding seats, speech analytics vendors continue to add functionality to their products. Here are some of the top speech analytics trends.
Big Data Driving Voice of the Customer (VoC) Initiatives
As companies start incorporating Big Data, they are increasingly turning to speech analytics to get a better understanding of customers. Most speech data is unstructured, and speech analytics can mine that data to help determine the sentiment of customers. Executives can then use the data to make changes to improve quality and boost customer loyalty, sales and revenue.
"For organizations focused on understanding customer behaviors, in addition to wants and needs, speech analytics helps identify the root cause of phone conversations – as well as deliver actionable results to the enterprise," says Daniel Ziv, vice president, Voice of the Customer Analytics at Verint. "For example, speech analytics can help identify the type of interactions that lead to customer frustrations, what makes them happy and the trigger that often transitions between the two states."
Switching to SaaS
Like other parts of IT, speech analytics is moving to a software-as-a-service (SaaS) platform. Aurix, CallMiner, Nexidia, Utopy, Verint and others all offer hosted services.
"Customer care executives are increasingly demanding hosted solutions because they want to see rapid return on investment by getting started quickly and avoiding the costs and time required to deploy the infrastructure of an onsite system," says Ezrine.
Due to technical restrictions, companies used to be limited to analyzing a random sample of customer calls. This approach would miss infrequent, but potentially disastrous, issues. Speech analytics software can now analyze data from hundreds of thousands or millions of calls, allowing contact centers to mine all of the calls they receive.
"In order for contact centers to understand the impact of a problem on their customers, it’s important to understand the frequency and severity with which it’s occurring," says Ezrine. "In addition, other common metrics such as first call resolution and agent performance are best measured and improved when comparing trends using 100 percent of audio versus a random sample."
Creating a dictionary of search terms or a predetermined set of reports is great when the company knows exactly what information it needs. However, it won’t expose topics or problems that aren’t known.
"In order to ensure that audio can be searched at any time, for any word, speech analytics solutions need to be dictionary independent,” says Ezrine. "This trend is being driven by the desire to explore and test new hypotheses, as they arise, about issues uncovered in phone calls, or to compare results found in other channels with those found in the audio without the need to re-index or update a dictionary."
Analyzing Multi-Channel Interactions
Customers no longer just call in; they also text, tweet, email, Skype, chat and post on FaceBook and Yelp. Speech analytics software is expanding to take these other channels into account. It is also important to track a single customer’s interactions as they move between these channels in communicating with a company.
"VoC has become a key stakeholder across global enterprises, especially as social media is used more often as a platform by customers to elevate their thoughts, opinions and feedback on service experiences," says Ziv. "As a result, managing social media and the VoC plays a strategic role within these organizations."
Moving Outside the Contact Center
As the ability to provide real-time information on customers improves, this information becomes more valuable to those outside the contact center.
"Analytics ownership will not only reside in the contact center – it will be a key tool across the enterprise, breaking down silos and playing a role in sharing information across departments and channels," says Ziv. "Further, eventually, speech recognition and analytics will be used real-time during the actual customer interactions to guide customer service representatives to deliver high-value service."
In Part Two of this series, we will examine some of the main call center speech analytics products in more detail.
Drew Robb is a freelance writer specializing in technology and engineering. Currently living in California, he is originally from Scotland, where he received a degree in geology and geography from the University of Strathclyde. He is the author of Server Disk Management in a Windows Environment (CRC Press).