The automated speech community's flagship event led off with a keynote address from the self-styled "voice of the caller," Paul English. The selection was controversial, as departing conference chair John Kelly noted, primarily because English raised his profile in Web 2.0 fashion through blogging. Millions of people have visited and contributed to the list of the ways callers can game voice self-service systems and reach a human being. Much of Kelly's e-mail skewed toward "Waddya nuts!" and "Stop consorting with the enemy!"
To its credit, the general response from attendees to Speechtek was much more tame. As it's written, the first step toward self-improvement is to admit that you have a problem. In the context of GetHuman.com (which is the Web resource that English launched to absorb all the hits that the list has generated without crashing his blog), the problem arises from bad voice-user interfaces. In his keynote address, English provided a prescriptive list of improvements that could be made to the user experience which, he argued, would be more empowering to callers – more humanizing and less humiliating. All of English's recommendations can be accomplished with today's technologies. A sampling: Inform callers of the time it will take for them to reach an agent and let them decide whether they would prefer to be called back. Don't require callers to repeat personal information or the purpose of the call over-and-over each time a transfer takes place. Be more sensitive to the times when a caller indicates that he or she is exasperated, flustered or frustrated and direct them to a human. First and foremost, provide more obvious mechanisms for a caller to opt-out of the automated system and talk to a live agent.
At best, English has proven to be the irritant that will inevitably help produce a shiny pearl. As a matter of fact, his efforts have induced two leading technology providers – Microsoft and Nuance – to join GetHuman.com in an effort to produce a published "standard" of the practices that go into producing a user experience that is gratifying for the caller and, ultimately, better for the enterprise as well since it will produce customers who are happier, more loyal and therefore more profitable.
Lest we forget, however, callers can be exasperated by live agents just as easily, and more often, than with self-service systems. A recent article in Forbes.com included a top-ten list of the factors that cause caller frustration, just two had anything to do with automated speech technologies. Topping the list was "Waiting too long on hold." Number 2 was "Repeating the same information to multiple agents, followed by "Agents lacking the right answers," "Agents trying to sell other products," "Agents that are inflexible about solving problems," "Company is too slow about addressing concerns," "Agents aren't personable," "Customized solutions aren't available," "Company's computers are often down," and finally, "Company asks for too much personal information."
By my count, these problems have to do with staffing, training and the underlying business intent of the company operating a contact center. Implicit in English's recommendations is the idea that reaching a live agent is the solution to caller's lack of empowerment and overarching frustration. The Top Ten list above indicates that a well-designed interactive voice response system or a speech enabled system that is capable of reconciling a caller's core issues without resorting to the ever-present upselling or cross-selling techniques that are common to all channels solves more problems that it generates.
Thus, the recommendations generated by the troika of GetHuman.com, Microsoft and Nuance will, by definition, only address a fraction of the user experience problem. No doubt, providing callers with an estimate of the time they can expect to spend in a queue, the option to schedule a call back or explicit directions for reaching a live agent will show that the speech community is aware of some of the problems and making the effort to change.
