Voice automation has traveled a long way since its introduction in 1961 with the IBM Shoebox, which converted the spoken word into electrical impulses. Considered the first voice recognition system, it was able to recognize and respond to 16 words spoken into a microphone, including the numbers 0-9. Commands were simple. For example, “Add 4 plus nine.” Shoebox worked by prompting an adding machine to calculate and print the answers to these basic math problems. By the end of the 20th century, subsequent breakthroughs began to occur more frequently and with greater commercial impacts. New algorithms, like the Hidden Markov Model that assessed the probability of utterances being actual words, became available and could run on the faster and faster processors being developed in accordance to Moore’s Law. Expanding beyond single user solutions like Dragon Dictate software, BellSouth introduced VAL, the dial-in interactive voice recognition system that is widely recognized as the forerunner to today’s sophisticated Communications Platform as a Service (CPaaS) solutions and interactive voice response (IVR) trees. By 2011, voice automation grew so sophisticated, IBM’s Watson computer competed with human Jeopardy contestants and won.
With non-stop advancements in voice technologies—and consumers increasingly growing accustom to quicker, more personalized service—voice has taken front and center stage. Home devices, cars and even wrist watches are now able to recognize human speech, some in multiple languages. While this is exciting on many levels, there is an underlying notion that interaction with digital agents, no matter in what context, should be more exact than ever. The reality, however, is that their ability to recognize and address every problem lacks human powers of intuition. As a result, customers still may want the option to speak with a human agent who can work in tandem with a digital bot to guide the customer in their journey, creating a customer experience combining the best of both worlds.
Considering the power of voice available today, implementing natural language processing (NLP), AI and machine learning (ML) has the potential to boost a number of sectors, particularly hospitality and travel-related companies, which are now experiencing a resurgence as economies reopen and people start traveling again. While airlines have done a good job with enabling online bookings and sending automated travel updates, call-in customer service remains stuck in the past. News reports are filled with stories of airline customers waiting hours to speak with an agent. The same holds true for car rental companies. In addition to a severe shortage of vehicles, many car rental companies (and hotels) that furloughed employees are suddenly understaffed and unable to respond to high call volumes.
For organizations that want to improve CX yet still provide human interaction, CPaaS solutions are ideal, as they can blend not only conversational assistants with live agents but also self-service options, which a growing number of customers are preferring. Healthcare and insurance offer excellent use case. An independent third-party administrator (TPA) of self-funded employer medical plans, for example, wanted to add more conveniences to its high-touch support services for customers, including self-service automation capabilities. Though it was recognized for superior customer service—where every caller is supported by a live representative—75 percent of incoming calls were transactional in nature. In other words, only about 25 percent required a live agent. To augment its person-to-person customer service, the organization implemented a customized CPaaS solution to offer clients, brokers, providers and plan members the ability to access information and complete common processes in every area of health benefits—around the clock. In addition to voice-enabled call flows for inbound and outbound interactions and dashboards for actionable on-demand visibility into all customer interactions across channels, the platform included a wide range of integration and configuration support services as well as a drag-and-drop workflow builder that integrated with its existing system. This allowed customers to solve their own issues and escalate them to the best agent based on their location, the product they use or a specific issue.
Well before COVID, forward-thinking companies in the travel and hospitality sector were implementing modern communications technologies. In fact, a 2018 report from Oracle found that 89% of hoteliers felt AI could significantly reduce operating costs and minimize human error. Using AI, businesses that collect guest information can gain valuable insights from the enormous amount of data that their customers create throughout their lifetime. From search and booking to checking-in and out, advanced analytics can provide the ability to automate processes and bolster decision-making power that ultimately improves CX.
Mordor Intelligence expects the global CPaaS market, valued at USD 4.54 billion in 2020, to reach USD $26.03 billion by 2026. It’s not surprising. The addition of AI and NLP that now supports many languages enables a better customer interaction, irrespective of whether through a friendly bot or the voice of a well-informed contact center agent. With personalized communication foundational to businesses, the ability to provide targeted, relevant information via the channels customers most frequently use (whether phone, text or email) will likely be essential for any consumer-facing industry, especially those like hospitality and travel in recovery mode.
ABOUT THE AUTHOR
John serves as Chief Technology Officer at IntelePeer and leads his team in delivering advanced and innovative solutions for IntelePeer’s infrastructure and its partners and customers. He has authored many of IntelePeer’s patents. John brings 25 years of telecommunications and systems experience holding architecture and engineering positions at Level 3, Hewlett Packard, and MCI. John holds a Bachelor of Science in Computer Science from Colorado State University.