Already, a survey found that 45% of people “prefer chatbots as the primary mode of communication for customer service activities.” AndGartner predicts that by 2020, 25% of customer service interactions will involve either a virtual assistant or a chatbot—a huge jump from only 2% in 2017.

Are customers giving up on the “human touch?” No. But it turns out that they don’t need that “human touch” to always come from a human. IBMsays there’s general agreement that “advancing the ability of computers to interact with us in a more natural way is critical for the AI-human relationship to reach its fullest potential.” And to that end, within 3 to 5 years, “advances in AI will make the conversational capabilities of computers vastly more sophisticated.”

This may sound all well and good, but truly excellent customer service representatives don’t just speak in a natural way. They also tailor their messages and their styles to each customer. Can AI really compete in this way?

Increasingly, yes. A team of Scottish researchers reported that, just as people align their communication when they interact, “there is strong evidence” human-computer interaction can achieve the same.

But what about unspoken cues? Customer service personnel need to take note of changes in a buyer’s tone of voice, facial expressions and body language. This presents AI with perhaps its steepest challenge. “Analyzing emotions in real timeis a mathematical challenge of astronomical proportions,” according to Nova. Learning to decipher such things means wading through a “tsunami of data.” But it’s happening. Steadily, AI is being taught to analyze and even predict emotional responses.

In fact, AI is already doing a much better job than most people realize. A survey found that only 37% of respondents say they’ve used AI. But among those who thought they had not, 63% actually had.

Before joining my current company, Nextiva, I worked in retail. I believe it will be one of the first industries to see seismic impacts. These will take place not just when buying things online, but also in brick-and-mortar stores. Expect to see more interactive kiosks to recommend and help you find items, and new technologies at checkout that suggest other purchases you might want to make, either right then or the next time you visit. (These technologies will then send you reminders.)

Already, we’ve seen humanoid robots taken out for a spin at Asian retail stores. Amazon , which controls nearly half the e-commerce market and 5% of all retail spending, is already implementing AI across its operations. And fierce competition will give more and more brands the incentive to follow suit.

In any industry, the businesses that do the best job of harnessing AI for customer service will be those that know their customers’ journey and their personas. If business leaders don’t understand their customers and their needs to begin with, AI can’t help them. (In the case of B2B organizations, this means knowing not only the buyers’ needs, but also the needs of their customers.)

While these changes will eventually mean less need for hiring people to answer phone calls and handle live chats, there will be an even greater need for employees to help build and improve these new AI technologies and to make them work for your company. So it doesn’t mean jobs will disappear altogether, but it will likely put more pressure on low-skilled workers and open up more opportunities for data scientists, for example.

Of course, the crystal ball is always a bit fuzzy. Just how soon all this will happen we can’t be sure. But it’s likely that a recession could move up the time frame, since “interest in automation accelerates during economic downturns.” Already, the algorithms that power AI are getting more powerful every day, and businesses of all sizes are starting to experiment.

So it’s time to set aside proclamations that the AI revolution in customer service is far off. Based on what we know today, it may be just beyond the horizon.

source : https://www.barrons.com/articles/holiday-toy-sales-could-be-light-but-2019-could-improve-51545929589