Why Customer Service Staff Should Never Make a CEO’s Endangered Species List

The evolution of chatbots, propelled by advances in cognitive computing, has promised to upend the realm of customer service. But any business leader already imaging mass redundancy plans for their frontline staff needs to take a breather.

In fact, at least one artificial intelligence expert—and several CEOs—are confident that the human touch will still be a crucial element of client communications long into the future.

“There are many jobs that are inherently resistant to automation,” California-based author and futurist Jerry Kaplan told Chief Executive in an interview. “You’re much more likely to get a positive action on something if you feel that you’ve made a personal connection to a human being that has some empathy, understanding or care with your particular needs.”

To be sure, there’s no doubting that the rise of AI has become a major concern for customer service workers as pioneering companies such as Uber allow riders to connect immediately with cab drivers, eliminating the need for middle men. Traditional customer-facing businesses, such as airlines and utilities, meanwhile, are increasingly turning to interactive voice response, or IVR, technology to handle customer phone calls.

“We typically overestimate the short-term impact and under-estimate the long-term impact of technology, and I think bots and AI as it relates to communications is probably falling squarely into that camp.

But the idea that artificial intelligence will soon become smart enough to have more nuanced conversations with customers is over-hyped, according to Jeff Mason, the CEO of cloud communications company Twilio.

“A lot of those bots are really just about simple questions and answers back and forth. It’s not artificial intelligence,” he recently told a conference in San Francisco. “We typically overestimate the short-term impact and under-estimate the long-term impact of technology, and I think bots and AI as it relates to communications is probably falling squarely into that camp.”

Don’t let cross-selling opportunities go begging
Kaplan, meanwhile, said there is indeed scope for automation to help companies cut customer service costs. But he agrees that present-day technology only tends to be capable of addressing relatively common, yet simple, problems—say checking if a flight is on time, for example.

“To some degree technology probably expands the class of problems that you can solve using customer service automation, but, frankly, I don’t think that’s necessarily that big a deal,” said Kaplan, who is also the founder of four Silicon Valley startups and a teacher at Stanford University.

His reasoning is that a personal touch between a human representative and a customer will always be a great way for companies to build relationships, and possibly cross-sell or increase product sales. “I think it’s going to be a very long time before computers will be very effective at doing what I call consultative selling,” he said.

Twillio’s Mason said CEOs would be better off concentrating on allowing customers to communicate with human representatives the way they want, such as via instant messaging rather than calling. “Companies that don’t figure that out—who make you call them, who make you wait on hold, who make you walk into a branch to do something—they will be the companies that slowly will have their customers trade away,” he said.

Distinguishing science fact from science fiction
Admittedly, though, chatbots have come a long way of late. Some companies are even developing virtual assistants that can respond to users’ emotions. SRI International, the outfit that invented the Siri personal assistant on Apple iPhone, is developing technology called SenSay Analyitics that can identify emotional state based on cues including typing patterns, speech tone and facial expressions.

Kaplan, who says he knows the people that run Siri, suspects advances in the system are more likely to be incremental rather revolutionary, gradually allowing a greater number of user questions to be answered via automation. He warns, however, that the system will still have its limitations. “The problem is there’s no way to easily communicate to the user of the system what it’s capable of and what it’s not capable of,” he said. “That’s the big drawback of these sort of general purpose natural language-type interfaces.”

AI has also evolved to the point where people can tell their washing machine when to start working. But it could be at least a decade before robots will be able to perform more sophisticated household functions, according to Stacy-Brown Philbot, CEO of the micro-gig website TaskRabbit. “I think that people are always going to be seeking a reason to relate and a reason to get to know each other. That emotional side of human interacting is what makes the TaskRabbit experience what it is,” she said.

Kaplan says that the idea of a Jestons-like robot capable of carrying out multiple household tasks remains, at least for now, more science fiction than reality. “One of the biggest mistakes people make when they think about AI is they’ll see a bunch of different advanced and remarkable applications and they will over-generalize,” he said. “Forget it. We’re nowhere near having some general purpose home robot.”


  • Get the CEO Briefing

    Sign up today to get weekly access to the latest issues affecting CEOs in every industry
  • upcoming events