Customer Service’s Future: Humans And Machines In Cahoots

The robots are coming, and they are going to take our jobs. A simplification, perhaps — but if one were looking for the simplest statement to explain the narrative following the era of automation, that would probably be it.

In a world where customers can order on mobile, use a touchscreen or talk to an artificial intelligence (AI)-based bot, the service industry will no longer need human workers, because machines will be able to do all of those jobs. While there are grains of truth mixed into that rather apocalyptic employment narrative, it gets much more wrong than it gets right about the automated future.

As Peter Rowan, Yapstone’s EVP of international and global customer support, recently noted in a conversation with PYMNTS, the questions about human workers and automation aren’t really either/or conversations. APIs, AI and technology are important advances, but the company has invested heavily in a human customer service team to work alongside them. The point isn’t to replace the people, he noted, but to help them be more efficient and to delight customers.

“We are anticipating over a million inbound contact calls per year over the phone in 2019,” Rowan said. “It is clear they want to talk to us, and we think the ability to have a human-to-human interaction is a major differentiator from our competitors. When it comes to money, people usually want to speak to a human.”

The goal for customer service, he continued, is to pinpoint the customer’s issue, then put them on the best, most comfortable and most efficient path to resolving it. However, that path will vary widely by customer and situation.

The Demographic Differences In Serving Consumers

Yapstone — which specializes in payment technologies for platforms and marketplaces — deals with a wide demographic variety of consumers, and those differences strongly influence consumers’ preferences for the service they receive.

On the oldest end of the spectrum are baby boomer customers, Rowan noted, who have a strong preference for direct human interaction and almost no desire to use any kind of automated system. The goal there is to create the shortest, most efficient journey to that representative over the phone, which means they should never be more than two key presses away from abandoning the automated system for a real person. On the other end of the spectrum are the Generation Z consumers, who don’t want to talk to a human representative if it can be avoided — and are looking for fast, automated ways to jump in, get the information or service they need and jump out.

Somewhere in the middle are Generation X and millennials, he added, with the older Gen Xers showing a greater preference for the phone, and millennials outright avoiding it unless there is an issue that absolutely demands it.

“For us, we give ownership to the customer of how they want to contact us,” Rowan said, noting that, from Yapstone’s point of view, the goal is to keep as many options on the table as possible so consumers can choose their ideal combinations.

Those preferences can quickly change, though, depending on context and specific services. In rental situations, when tenants are looking at their monthly rent and trying to remember whether it was paid on time, that is a fairly quick, direct interaction that many consumers will handle via an automated system. On the other hand, when the purchase in question is a $15,000 vacation rental payment, the customer is likely to have more questions or concerns, and may want a direct human interaction.

There are also cases when there is a possible security breach or fraud event, when the customer service team must really “rise up and directly attend to the consumer,” he said. Regardless of demographic, consumers almost universally want those types of issues to be dealt with directly, personally and immediately.

Technology In Tandem With Human Workers

In general, security is an area that points to how humans and AI automation work in tandem for an overall better result. On the technological side, Yapstone employs what Rowan referred to as “world-class” risk management, buttressed by machine learning and AI, to snuff out any security problems. Considering the large payments the company deals with regularly, fraudsters are simply a cost of doing business.

However, in the era of social engineering attacks that look to get around the tech by exploiting human weaknesses, that means human workers must have extensive training, lots of review, many feedback sessions and a strong adherence to internal procedures.

“I won’t say we haven’t experienced these kinds of attacks on our workers — people have certainly tried,” said Rowan. Moreover, he continued, many of Yapstone’s investments in machine learning and AI aren’t focused on replacing workers, but on supplementing or enriching their interactions.

For example, a customer sends a simple email query to Yapstone. Instead of sending that directly to a human agent, bots scan the email, find an answer and create an email to answer the question. A human agent then checks the email to make sure it is relevant to the original question, then adds any personal information that might be necessary. The whole process takes about 45 seconds, as opposed to the five or six minutes it once required. The human worker is still a central player, but they can do the part they are best at, and leave the fetching and gathering to the bots.

“The future really is not just humans; it’s not just machines,” Rowan said. “It is not a man or woman being replaced by machines, but, instead, agents being supported and powered by machines.”