Speech Technology’s Decade of Evolution
It has been 10 years since Information Today Inc. (ITI) acquired Speech Technology magazine and its associated products. Back then, the speech technology industry was still wiping a lot of egg off its face from the many failed enterprise deployments. Has the technology improved during the past 10 years? Absolutely. Has it evolved enough? That’s debatable.
There certainly hasn’t been a shortage of innovation. During the past decade, Apple’s Siri, Microsoft’s Cortana, and Amazon’s Alexa skyrocketed in popularity, breathing much-needed life into the industry. Thanks to these virtual assistants, users are much more aware and appreciative of speech technology. Other promising innovations have surfaced in that time as well. In fact, this issue of Speech Technology magazine reveals the tremendous opportunities that deep neural networks and the Internet of Things bring to the speech technology industry and the organizations that are willing to invest in the pairing of these technologies.
But what about the speech-enabled IVRs that exist today? Unfortunately, for various reasons, many of them are still not consistently delivering positive customer experiences. Part of the problem is that expectations were set extremely high right from the start. Unlike other technologies, speech systems are often compared to human-to-human interactions and, as a result, are expected to perform as well as, or better than, interpersonal communications.
It’s not unlike the challenges that vegan and vegetarian food manufacturers have endured over the years. Today, there are some really delicious mock meats out there. (What companies like Beyond Meat and Gardein are doing with their vegan products is really incredible.) But, if I were to say to you, “Try this vegan hamburger,” unless you’re already vegan, you’ll likely compare it to a regular hamburger. And if it doesn’t taste exactly like a meat-based burger, you might be disappointed. But if I could prevent you from establishing a preconceived notion of what the product should taste like and, instead, say, “Try this new lentil patty,” you’d be able to appreciate it for what it is, and not what you think it should be. As a result, your satisfaction with the product would likely be higher.
Similarly, people’s opinions of a speech-enabled IVR are largely influenced by their expectations of it. If organizations compared a new speech-enabled IVR to their antiquated one, they’d be less likely to compare the new system to a live agent. As a result, they’d establish a more realistic benchmark for success and appreciate the newer system more. Jenni McKienzie proves this point in her column “With Voice Design, Why Don’t People Expect More?”
So to answer the question posed at the beginning of this column, if you were to compare today’s speech technology innovations to previous speech technologies—and not human-to-human interactions—you might come to the conclusion that the industry is evolving quite nicely.
It is my hope that the industry continues on its evolutionary path for many more decades.
****
If you haven’t done so already, mark your calendars for our upcoming SpeechTEK conference at the Washington Marriott Wardman Park in Washington, D.C. (April 24–26, 2017). As in past years, this event will feature industry experts covering a wide range of speech technology topics. For more information, visit www.SpeechTEK.com.
David Myron is the editorial director of Speech Technology. He can be reached at dmyron@infotoday.com.