Everything's Just Speechy: Exploring the Evolution of Voice and Text Analytics

Customer Experience
Customer Experience
As analytical technology becomes more affordable, companies now have the opportunity to listen to untapped customer sentiment and recognize the underlying factors that drive satisfaction.

If consumers aren't talking, they're typing or texting. But just as quickly as they speak or text their words, companies capture that data to gather insight from every channel at any time. But if organizations truly want to make the most of today's socially connected landscape, they must hone their analytical capabilities, thereby enabling themselves to assess and respond accordingly to the mounting sentiment. Speech analytics offers many unique insights into real-time brand interactions, but organizations must integrate this voice and text data to bring actionable results to the enterprise.

In the past, speech analytics required contact center agents to record calls for both quality assurance and future training purposes. However, because "speech" has come to encompass both voice and text, now analytics are in place not only to observe and classify voice inflections, but also to explore and understand text sentiment across a variety of platforms, including email and social media. As Shep Hyken, customer service expert and New York Times bestselling author, highlights, today's approach to speech analytics typically observes two areas: technology, and the human element. Brands can easily track the first person to answer, determine the sentiment behind the feedback, and assist when the consumer is in need of help or support, but the human element comes into play when companies begin to recognize patterns and potential areas for improvement, thereby teaching representatives how to better solve certain issues.

Daniel Ziv, vice president of voice of the customer analytics at Verint, notes that the primary driver behind most speech analytics initiatives today comes from companies looking to address specific business pain points or opportunities, such as increasing customer satisfaction, while others are motivated by the desire to reduce the company's overall cost to serve. Speech analytics allows brands to interpret voice and text sentiment in order to identify the root causes of customer dissatisfaction as expressed through complaints and emotional calls. Then they can begin implementing strategies and processes to reduce and eliminate any customer issues. By learning and understanding customer sentiment, companies are also able to utilize analytical tools to eliminate repeat and unnecessary calls, increase self-service effectiveness, and reduce the average handle time of most cases.

"Speech is the richest source of customer interactions," Ziv says. "A five-minute phone call is typically more than 1,000 words, while a typical tweet is about 15 words. The interest in text analytics has been fueled by the rise of social media being an important new channel. However, many organizations find social media to be very noisy and difficult to take direct action from, thus many businesses continue to leverage other channels, such as phone calls, email, and Web chat, to prioritize and take effective action based on insights gained."

Because phone and email often serve as early warning signals before complaints go viral on social media, companies continue to focus their analytical efforts on all modes of communication, as being where the customers are remains a primary driver behind most business strategies designed to gain the competitive advantage. Yet, while most organizations have come to accept the need for speech analytics, and are beginning to employ the necessary technologies within their own enterprise, the analytical capabilities of speech recognition tools not only curb complaints, but also facilitate customer interactions.

Today, the Complete Semantic Index (CSI) allows brands to move beyond basic word capture and speech-to-text capabilities, to actually capture the sentiment behind the feedback coming into the organization. This approach moves beyond translating audio into a string of phonemes, or syllables, to then match this group of words and phrases into meaningful categories that express a much richer, smarter index of sentiment and customer insight, providing content and context that the organization can use to take action and improve the customer experience accordingly. Such capabilities then allow companies to create automated response systems that make consumers feel as if they're talking directly to another human because of the sophistication of the company's analytical tools.

"In airports, you can now walk up to a hologram of a person and ask him or her where to get pizza," Hyken notes. "With voice recognition, the artificial intelligence can quickly pick up what you said and use keywords to figure out the proper response to be made." Though it may seem like science fiction, the future of speech recognition and analytics has arrived.

Security First Insurance Integrates Sentiment and Service

For Security First Insurance, a Florida-based homeowners insurance firm, its nearly 180,000 customers offer the company an unyielding supply of sentiment, particularly during hurricane season. Because the people of Florida are exposed to more hurricanes that any other state in America, the company must be prepared to handle an exorbitant influx of claims and questions when necessary. During an average month, Security First Insurance processes about 700 claims regarding damaged property or possessions. However, when a storm strikes, claims can jump into the tens of thousands within a few short days.

"Our goal is to be there for our customers, storm after storm, year after year-just as our slogan promises," says Werner Kruck, chief operating offer at Security First Insurance. "When a claim comes in, we want to begin the recovery process as quickly as possible." By partnering with IBM, Security First Insurance now has the ability to collect and analyze customer data flowing in from social media and email alongside more traditional methods, such as phone, to get a jumpstart on the claims process no matter the circumstance. The insurance firm can easily analyze incoming messages using text mining, text analytics, and natural language processing to assess whether the given message expresses urgent distress and the immediacy of the damage. Security First will then deploy first responders in times of crisis to ensure clients are cared for and problems are resolved in a timely fashion.

Because storms have become increasingly more frequent over the past few years, Security First Insurance recognizes that, in times of need, many clients take to social media to connect with first responders and obtain assistance when the phone lines and electricity are down. Through text analytics, the insurance firm can quickly assess comments and tweets, allowing them to route the claim to the appropriate department for a prompt resolution. Text mining also allows Security First Insurance to detect profanity and filter those messages from social media so the company may handle the issues accordingly behind the scenes.

Siri Sizes Up the State of Speech Analytics

Karen Jacobsen, the GPS Girl and Australian speaking voice of Siri, created her livelihood as a voiceover artist over the past 20 years by becoming part of the ever-evolving world of IVR. However, since the start of her career, much has changed from the days of "press one" systems that offered strict limitations to the way companies interacted with customers over the phone. Now, with intelligence driving many of the systems we've come to know, customers need only ask a question, and tools, such as Siri, will quickly find the answer for them. But, no matter how advanced this technology becomes, Jacobsen believes that speech analytics and IVR tools will never replace the feeling of having a human being on the other end.

"[Technology] replicates the human connection, but it won't ever replace the human connection," Jacobsen emphasizes. "It's a reminder of each person's individual strengths in business that no other person really has. We need to celebrate and embrace that, and really allow ourselves to be the full expression of ourselves in whatever we are doing. There's more room now than ever before in the corporate world to bring creativity and flexible thinking into roles and areas that previously may not have allowed for that. Value can be seen at the bottom line, and that makes it a truly exciting time in business."

No matter how advanced and integral analytical technologies may become, people still need people to truly drive and understand the meaning behind customer feedback and the sentiment flowing into organizations around the world.