Interactive voice response (IVR) systems are—how do I put this nicely?—not one of the most beloved technological advances of the last century. They can be cumbersome, confusing to use, and flat-out frustrating sometimes. We’ve all had that IVR experience where, after navigating through a maze of choices, we find ourselves suddenly disconnected.
It’s no wonder then that some pundits like to point to IVR as on its way out. Press “1” is the past, they say; services like Siri (Apple’s voice recognition software) are the future. But IVR does have a strong future, in large part because of technology like voice recognition. And companies like Molo Innovation (a Dialogic partner) are helping to make it happen.
Voice recognition is a complex science. It’s also an imperfect one. Today’s IVR systems often employ some type of rudimentary voice recognition, usually as a side-by-side alternative to touchtone prompts. The reason for the dual touchtone/voice approach is simple: voice recognition isn’t 100% reliable on most IVR systems, and so enterprises and service providers need touchtone as a fallback. As voice recognition solutions evolve, however, they’ll help to redefine the IVR experience.
There are several ways that new voice-based technology will make IVR systems more interactive in the future:
You’ll have more choices (and get to them faster).
Most IVR applications are heavily scripted and inflexible. Navigation is limited to a handful of fixed options or yes/no responses, and users often have to progress in a linear fashion through these until they find what they want. More recently, IVR applications are using keyword detection to let users drive the conversation (e.g., “Tell me what you’re looking for?”). As voice recognition software gets better at parsing conversations into keywords, IVR systems will be able to direct users almost as intelligently as human agents. For example, Molo created an IVR application for one of the largest wireless service providers in South Africa that allows subscribers to look up share prices using their voice rather than their fingers.
Your voice will be your password.
The human voice, like the human fingerprint, is unique from person to person. Voice biometrics is the mapping of the human voice into its “digital DNA” as a means of personal identification and authentication. Voice biometrics could become the personal identification number (PIN) and password of the future. Particularly in industries where personal identification is critical (e.g., government, financial, healthcare), IVR systems that employ voice ID would have significant benefits.
Machines will replace live agents.
Speech synthesis can make a machine sound human. Artificial intelligence is now demonstrating that machines can carry on basic conversations and even detect changes in tone. It’s a bit far off in the future, but it’s not unthinkable for IVR systems to be completely automated yet still offer “live” customer service through artificially intelligent agents. Expect more Siri 3.0 than HAL 2000, but it would still be a big step for mankind considering where IVR systems are today.