Speech Analytics and AI Is a Winning Combination

Article Featured Image

Speech analytics is getting a new lease on life courtesy of artificial intelligence (AI), machine learning, and the digital transformation. Vendors in most IT sectors claim to provide AI-enabled solutions, and the speech analytics providers are no exception. To be sure, speech analytics has leveraged core AI technologies—natural language processing (NLP), natural language understanding (NLU), neural networks—to varying degrees for years, so many speech analytics vendors are justifiably calling what they do a form of AI. 

But this is just the beginning. These applications are being pushed to the next level by more advanced AI-enabled technologies, like supervised, semi-supervised, and unsupervised machine learning and predictive analytics. 

Improvements in Speed and Accuracy

As the pace of business has accelerated, the demand for real-time speech analytics has increased. Enhancements in speed and accuracy are yielding what speech analytics vendors have claimed to provide for years: accurate and timely insight to impact an ongoing conversation. Powerful speech engines, increasingly delivered via the cloud, send reminders to agents to give required disclosures within prescribed time frames, identify potential fraud situations before protected information is released, and deliver timely guidance on the right product or service. 

QM: From Antiquated to Automated

The main obstacle to progress for traditional quality management (QM) is that few companies can afford the resources to properly staff a traditional QM function. The future of this process is analytics-enabled QM (AQM). Speech and text analytics are used to listen to/read customer interactions and provide feedback to the enterprise (general trends) and agents (what they do right and how to improve). When speech analytics first entered the commercial market in 2003, it was not ready for AQM, but it is today. 

VoC Unfiltered

Speech (and text) analytics provides the benefits of surveys with none of the work for customers. They can mine customer interactions from all channels, including social media, to capture the voice of the customer (VoC) firsthand. It is an effective approach for listening to customers and provides input from transactional and other analytics solutions to gain insight into the customer perspective. VoC results from speech analytics can be used to replace traditional enterprise feedback management (EFM) programs, or can be integrated into EFM reporting and used to augment findings. Enterprise executives who are not thinking more broadly about the uses of solutions like speech analytics for capturing and understanding the voice of the customer are leaving big cost savings and benefits on the table. 

Customer Journey Analytics: Speech Is Along for the Ride

Enterprises are starting to leverage sophisticated AI-enabled technologies—intelligent virtual agents (IVAs), robotic process automation (RPA), predictive modeling, and speech analytics—to thrive in the era of the personalized customer journey. Leading organizations are combining self-service with live agent support to pre-engineer effortless service experiences. 

Speech and text analytics provide essential input into the CJA process by capturing spoken and written conversations and converting them into structured data for analysis. Speech analytics can transcribe the self-service interaction between a customer and an IVA or chatbot and continue transcribing the conversation after the call has been transferred to a human agent. The results are available for conversational analysis—topic and emotion detection; intent and sentiment analysis. Analyzing contacts across individual customer journeys, or contacts from multiple callers with the same issues, can help determine the root causes of trending issues. This information can be used to identify the actions required by every customer-facing department or employee to expedite resolution, minimize impacts, and enhance the customer experience. 

In real time, speech analytics changes the service paradigm from reactive to proactive. Real-time speech analytics can be used with predictive analytics to determine the best sales offer or service solution to present to a prospect or customer, or identify the next best action to enhance the customer journey or reduce the risk of customer churn. 

What’s Next

As the solutions above indicate, the next big thing for the speech analytics market is AI. DMG expects a new round of investments in AI-enabled technologies to make significant contributions and improvements to speech and text analytics solutions. In the future, AI will be used to enhance the ability of speech analytics to identify issues and recommend ways to address them. Semi-supervised and unsupervised machine learning is starting to be used to enable SA solutions to quantify the impact of new trends and issues, with minimal (or no) human intervention, to improve the understanding of what people are saying and to predict human behavior.

Donna Fluss (donna.fluss@dmgconsult.com) is the president of DMG Consulting, a provider of contact center, analytics, and back-office market research and consulting.

SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues
Related Articles

IVAs Enter the Business Conversation

IVAs boast big potential in a wide variety of verticals and functions

Voice Assistants Are Changing Shopping—Are You Ready?

Voice shopping will only become more prevalent going forward, and retailers must take an interest in the technology or risk being left behind

11 Keys to Designing Effective IVAs

Businesses and customers alike can benefit from quick, personalized, intelligent self-service.

IVAs Can Remake the Self-Service Landscape

Customers prefer self-service if the solutions are effective—which is where IVAs come in.

Voice Biometrics Can Secure Hassle-Free Caller Verification

It minimizes fraud and makes ID'ing customers an easier process. The big challenge: educating the market