Editor’s Note: Welcome to our weekly Reality Check column. We’ve gathered a group of visionaries and veterans in the mobile industry to give their insights into the marketplace.
Smartphones are now the majority of mobile phones in countries such as the United States, and at the end of 2011, nearly 15% of U.S. consumers went online using a tablet, according to eMarketer. By 2014, more than one in three online consumers will use a tablet, the research company predicts.
Those statistics increase the proportion of inbound sales and customer support queries coming from mobile devices. In the past, the majority of those mobile-originated queries were in the form of calls, but today it’s increasingly likely they’ll be from an app or a website. When they are, there’s a good chance that users will be frustrated. According to recent market study, more than 90% of all Web-based self-service inquiries receive a wrong answer the first time. Worse, more than half receive a wrong answer even after their third attempt.
These performance issues apply to all Web-based inquiries, regardless of whether they originate from a PC, smartphone or tablet. But mobile-originated inquiries deserve particular attention because their form factors make navigating and using websites trickier and more error-prone. Every additional step on a PC due to a vague or wrong answer becomes even more cumbersome on a mobile device, and increases the likelihood the consumer will abandon the channel after a bad experience.
Enterprises, government agencies, wireless carriers and other consumer-facing organizations benefit financially when they minimize that failure rate and all of the irritation that comes with it:
–When consumers can’t get the information they need from a website, they switch to a channel that’s far more expensive for the organization. More than one out of three consumers abandoning the online experience make a call to an agent, and each conversation with a live agent costs $5 to $12, Forrester Research says. Increasing the success rate of online self-service significantly reduces the organization’s inbound sales and customer support costs by displacing more of those calls.
–Sales and customer-support teams can be more productive. When customers get most of the information they need from an app or website but still want to talk with a live agent, their interactions take much less time than if the agent has to answer all of their questions.
–Some customers and prospects will simply give up instead of calling or e-mail. Those are expensive revenue losses.
–Some existing customers will complain about their experience on social networks and blogs creating a cost burden for the organization to salvage a sullied brand. According to Forrester, the business impact of a quality customer experience for large Firms can gain $177 million to $311 million per year.
The good news? There’s a way to turn those high failure rates on their heads and answer most questions right the first time. This enhancement isn’t hypothetical; it’s running in live deployments, most recently for a major bank in the Asia-Pacific region.
Context is king
Reversing the failure-to-success ratio starts by doing a better job of understanding the questions. Suppose that a person is in a taxi on their way to the airport and wants to check the roaming rates for the countries they are traveling to. They get out their smartphone or tablet and navigates to the customer care section of her carrier’s website, where they begin by asking, “How much do roaming calls cost?”
Here’s where the problem starts because many of the deployed platforms simply carry out keyword searches in the knowledge bases. This approach breaks down very quickly for two key reasons. First, consumers ask for the same things using a wide variety of language that is difficult to predict a priority (e.g. “How much are calls abroad?” “What are the roaming fees?” “What’s the price for calls while I’m in another country?” and so on). Keyword searches are unable to interpret these different questions and understand that they are all asking for the same information. Second, a keyword search’s results have no broader context: they simply return lists of documents containing the keyword. Rather than forcing the consumer to read through lists of PDF’s and Web pages, a preferable approach is simply to answer the question – in our example above, to return the roaming rates for Mexico. Today’s platforms need to be smart enough to figure out the real question and its context, beneath the words the consumer has used.
Additionally, an intelligent question/answer platform needs to proactively guide the consumer to the correct answer. Such platforms use techniques including semantic analysis and disambiguation to ask clarifying questions. Extending our example above, to provide the correct answer, we may need to know “Where are you roaming?” or “Do you want voice roaming rates, data roaming rates or both?” The consumer’s responses enable the system to narrow down the choices and provide the right answer, right away.
Suppose that their answers were “Mexico” and “voice.” After the system gives them the voice roaming rate for Mexico, they asks, “What about Panama?” because that’s where they are going after Mexico. Most platforms don’t preserve context, so they don’t understand a follow-up question the way that humans do. But an intelligent platform does, meaning it provides the single, correct answer immediately: the voice roaming rate for Panama.
The ability to provide a single, correct answer immediately is an obvious advantage for any Web search. But it’s even more beneficial in the mobile domain, where scrolling through a list of results or pecking out a question on a virtual keyboard is about as much fun as folding a fitted sheet. There are deployed platforms today that are delivering this next generation of intelligent question-answering capability. Consumers are getting their answers the first time, and enterprises are avoiding expensive inbound calls completely. As smartphones continue their ascent to become the predominant communication devices globally, companies across all industries should take advantage of state-of-the-art question-answering technologies to get the right information to their consumers in a timely, hassle-free manner.
Dan Faulkner has led core technology and marketing organizations during his tenure in the speech industry. Today he is responsible for the development, management and marketing of Nuance’s enterprise speech solutions and strategy. Faulkner has been immersed in the communications technology market for more than 10 years, with first-hand experience in the North American and European markets. He has a Bachelor’s degree from the University of Manchester, and a Master’s degree in Speech and Language Processing from the University of Edinburgh.
