Technology May Save Money but Human Agents Make the Difference
By Peter Lyle DeHaan, Ph.D.
We’ve been hearing a lot about artificial intelligence (AI), and we’re going to hear a lot more about it. Some claim AI is the future of the call center industry, saving money and retaining business.
Others fear it’s the end of customer service as we know it. Neither is right, nor are they both wrong.
But AI isn’t the only technology in our call centers. We have digital assistants to help our agents and automated bots to help our customers. Before that we had interactive voice response (IVR) and auto-attendant solutions.
Regardless of the technology or the era it comes from, each innovation brought with it the inherent promise to speed resolutions and reduce labor expenses. To some degree, they accomplished this. Yet they also fell short of meeting expectations.
In most cases, however, the implementation of technology has brought with it a corresponding ire of the customers it’s supposed to help.
In some cases, technology—especially AI—can make a real mess of things. When this happens, human intervention is the only way to correct the problem. This assumes, of course, that people are available to intercede to fix technology’s error.
Here are some things human agents can do that technology can’t do or can’t do well:
Correct Miscommunication
Technology struggles to correct its mistakes. When it determines what path to take, it persists on that course even if it’s the wrong one.
Often, miscommunication devolves into such a quagmire that the simplest approach—sometimes the only one—is to terminate communication and start over later.
Yet this is an ideal time for human intervention to clarify the customer’s concern and redirect action toward the right solution. This means that human agents need to have the ability to override technology. They also need to have both the training and confidence to know when to do so.
Calm Frustrated Customers
Technology isn’t good at realizing when customers are upset and responding in a truly comforting way.
Though through algorithms, AI can detect anger or frustration, customers will likely discern any attempt to diffuse their concerns as disingenuous. This will escalate their tension, not defuse it.
A successful outcome requires a real person, someone who will listen, comprehend, and offer sympathy. Though no human agent can accomplish this all the time, their chance of success is much higher than that of a machine,
Respond to Complex Issues
Convoluted problems can escape the ability of AI to accurately comprehend and successfully navigate. This is especially true when a situation is unique, something AI has not yet encountered and never will again. Yet human ingenuity shines in these situations.
Offer Empathy
Sometimes customers feel a need to vent. Ironically, this is often over the failure of technological solutions to appropriately address their concern. Though AI can determine the need to give an apology and mimic the right words to say, can it do so with empathy?
Will the customer feel they were heard? Will the response come across as sincere?
A person has a much better chance of doing this successfully than a computer.
Conclusion
Though AI technology will continue to improve, causing fewer problems and producing more satisfyingly complete solutions, don’t plan on replacing your staff. Though you will not need as many, you will need some.
And the skill set of these super agents will carry higher requirements than current ones.
Being able to offer the human touch will distinguish contact centers from their technology-only counterparts.
In an era when technology surrounds us and threatens to overwhelm, a human customer service agent stands as a core distinction between offering solutions that are close versus ones that are comprehensive and complete.
Don’t forget to offer the personal touch of a human agent to best serve customers whenever needed.
Peter Lyle DeHaan, PhD, is the publisher and editor-in-chief of Connections Magazine, covering the call center industry.
Read his latest book, Call Center Connections.
