The State of Speech Analytics
QY Research valued the global speech analytics market at $950 million last year and expects it to grow at a compound annual rate of 13.8 percent through 2025. Speech analytics, QY Research says, has become increasingly popular among businesses because of the huge amount of data produced as a result of rising call volume.
In fact, Jeff Gallino, founder and chief technology officer at CallMiner, says basic speech analytics will be table stakes for companies in 2020.
The Year in Review
The past year saw speech analytics adoption move up significantly, with the solution being adopted much more than text analytics, according to Donna Fluss, founder and president of DMG Consulting. “More and more companies are using both digital and speech analytics as well as text analytics. It’s a highly valuable resource.”
Additionally, there was a noticeable increase in speech analytic applications, such as “next best action,” according to Fluss. Next best action can alert contact center agents when to escalate calls to managers, before callers get frustrated enough to ask for the transfer. By escalating calls more quickly, call times can be shorter and customer dissatisfaction can be kept at bay.
In the past 12 to 18 months, buyers have become more strategic in their speech analytics purchases, says Carmit DiAndrea, vice president of portfolio market strategy for voice-of-the-customer analytics at Verint Systems. “Companies are starting to see speech analytics as a longer-term initiative to start building better customer relationships and to reduce customer churn. They’re using speech analytics to improve customer engagement.”
In 2019, speech analytics was also used increasingly in financial services and to aid predictive analytics teams across various industries, according to DiAndrea.
Another advancement was that speech analytics became more automated, making it much easier for companies to work with ever larger amounts of data, DiAndrea says.
The rise of the platform was one of the major changes in 2019, according to Gallino. As such, speech analytics becomes part of companies’ digital transformation strategies rather than simply point solutions that need to be cobbled together with other technologies.
“Vendors are providing more functionality with fewer technologies,” Gallino says, adding that artificial intelligence (AI) became more incorporated into speech analytics in 2019 and will become more ingrained in the technology in 2020.
QY Research’s data supports this, finding that a key to the industry growth is the increasing use of artificial intelligence integrated with voice and speech analytics. Voice and speech data categorization based on AI can analyze words, acoustics, and feelings automatically and derive hidden opinions and emotions, it finds.
Modern speech analytics enables contact centers to combine speech recognition software with text analysis and pattern spotting to categorize interactions according to a set of custom rules. As the technology improves in speed and word identification as well as from quickly evolving sentiment recognition, speech analytics is becoming more valuable for enterprises looking for ways to differentiate themselves from competitors through providing better customer experience.
Basic speech analytics software reviews every customer conversation after recording, then translating it into machine-readable text. With the influx of AI, speech analytics has started moving increasingly into the realm of analyzing the actual sound (words spoken, tonality, etc.) without the interim step of converting the conversation into machine-readable text. Analyzing speech in sound format means companies can capture sentiment analysis, which translation to text often misses.
A Look Ahead
In the new year, experts see speech analytics continuing to evolve as artificial intelligence becomes more ingrained in the technology. Will Hall, chief creative officer for RAIN, expects speech analytics to increasingly move from a technology that analyzes transcribed text of conversations to one that increasingly analyzes speech as it is spoken, which does a much better job of capturing sentiment analysis—an important factor that speech analytics can measure much better than other analytics. Is the person at the other end of the line mildly annoyed or extremely annoyed? Did the emotion continue to build during the entire interaction, or was there some element that sharply changed the customer’s rate of speech, tone, and wording?
“A lot of firms tend to measure voice the same way they do other [key performance indicators], like total views, interactions, time spent onsite, etc., but voice is different,” Hall says, pointing to sentiment as the prime differentiator and prime reason he expects to see a growing number of companies adding speech analytics to their customer experience measurement capabilities during the next 12 months.
DiAndrea also expects an increased focus on the customer sentiment feature of speech analytics in 2020. By combining sentiments and other elements of speech analytics with other real-time data, contact centers can make more informed decisions about handling customer interactions, offering higher-value customers a higher level of service.
Employing speech analytics can result in a 5 percent improvement in customer churn rates, according to DiAndrea.
Fluss adds that companies are starting to see the value of feeding speech analytics into their customer journey analytics and contact center tracking.
Hall points out that speech analytics are more of a middle- or bottom-of-the-sales-funnel measurement tool than something that provides value at the top of the sales funnel.
Fluss expects 2020 to be a very good year for speech analytics providers nonetheless, with a serious increase in speech-analytics-enabled quality management offerings.
“Analytics-enabled quality management (AQM), capturing the voice of the customer, compliance management, and sharing/leveraging results from interaction (speech and text) analytics throughout the organization all have a quantifiable payback and provide significant customer, agent, and enterprise benefits,” Fluss says.
Gallino also sees the competitive landscape having an impact on analytics adoption. As companies find value or see their competitors getting better customer retention as a result of speech analytics, they are more likely to add it.
As an added benefit, speech analytics can help companies determine which agents are adhering to requirements in collections, sales, and service calls and demonstrate compliance with company, industry, and government regulations, Fluss adds. “Speech analytics solutions can identify calls where required language is or is not spoken, enabling contact center supervisors to provide targeted coaching to employees who are not meeting expectations in this area.”
DiAndrea also expects AQM to be a major driver for speech analytics in 2020.
“Vendors can motivate their customers by having the right speech analytics offerings,” Fluss adds.
However, despite improvements in the offerings over the past few years, speech analytics technology is still very complex to implement, Fluss cautions.
Speech technology providers and users could struggle in 2020 with determining whether voice data is considered personal data, according to Gallino. If it is considered personal data, there will be various privacy rules that will have to be considered.
Phillip Britt is a freelance writer based in the Chicago area. He can be reached at email@example.com.