Will AI Kill the Customer Service Star?

Ready to build better conversations?

Simple to set up. Easy to use. Powerful integrations.

Get free access

For this article, we sat down with customer service expert Shep Hyken and discussed the growing prevalence of AI in customer service and its influence on the industry.

Due to the expansion and popularisation of customer service AI, consumer expectations have shifted. Moreover, AI-assisted customer service is often coolly received by consumers and service workers alike. “Some people see machines as having the potential to replace people,” explains Shep. And that’s the crux of the issue.

The future’s already here

Customer service AI, though still developing, is already well-established in some industries. Some “personal assistant” AI programs, such as Apple’s Siri or Amazon’s Alexa, are used without thinking twice by millions of consumers. However, the customer service applications of this technology tend to make those same consumers nervous.

Making customer service smoother and easier

Customer service AI is a more marvellous technology than one might suspect. Going beyond the data-retrieval abilities of, say Siri, some customer service AI programs are able to understand language in real-time, exercise predictive reasoning, and learn as they go.

Most of us don’t grasp the subtle power of customer service AI. In fact, that’s the mark of truly effective artificial intelligence: being smooth and efficient enough to not feel inorganic. This is the case for Netflix or Pandora’s predictive recommendations, for instance.

New breakthroughs and experiments in customer service AI are booming. Just in time for Mother’s Day, 1800flowers.com announced the beta release of Gwyn, its AI concierge based on IBM’s revolutionary Watson. Gwyn lets users give “her” hints about the recipient and directs users to the perfect gift.

The need for speed

The primary reasons for the expansion of the customer service AI industry are speed and convenience. Most customers consider speed of execution to be the most important facet of customer experience.

However, Shep explains, fast service doesn’t mean cursory service. “The goal isn’t to get you out as fast as possible, but rather to make your experience more efficient,” he goes on to say. AI can access, interpret, and articulate inhuman volumes of data, thus delivering a speedier customer experience than a human could. The sheer convenience of using an AI program to perform a mundane task, such as changing a billing address or checking the status of a shipment, far outstrips traditional means of service.

Nevertheless, many balk at the suggestion of AI encroaching upon what we consider to be human territory, such as replicating a personal customer service touch. Some worry that machines could replace human customer service representatives and that customer service could be devoid of a human presence.

Customer service AI is a tool, not a threat

Many service agents, managers, or specialists, and indeed many customers are uneasy with the growing presence of customer service AI.

Will AI make customer service reps obsolete?

The major worry seems to be that AI is poised to overtake human representatives and make them redundant. Another frequent concern is that leaving service wholly in the figurative hands of machines would cause the quality of service to decline and become less personal.

This is due in part to the trope of shaky virtual assistance and artificial intelligence programs. This motif is very present for consumers: the thought of getting trapped in a loop of “Sorry, I don’t understand. Can you repeat the question?” springs to mind.

Therefore, there is a certain measure of reluctance felt towards customer service AI by representatives and customers alike. However, this fails to consider the following possibility: what if AI enhanced customer service not as a standalone replacement, but rather as a helper to human representatives?

To clash or to collude?

The question of possible conflict between human customer service agents and AI fails to take into account the possibilities afforded by their cooperation. Indeed, instead on focusing on AI outstripping and replacing human agents, consider instead our respective complementary strengths.

The evolution and development of artificial intelligence happen through enriching the technology by teaching it human reflexes. Google’s DeepMind learns to apply its previously acquired knowledge to the resolution of new problems, which is a quintessentially human trait.

This serves to illustrate how well human and artificial intelligences can enhance and harmonize with one another. Though these examples are rather daring and, for now, removed from the sphere of customer service, it still stands to reason that the customer experience can be enriched by the collusion of humans and machines.

An enhancement, not a replacement

The potential for enhanced customer service lies at the intersection of complementary abilities. Shep gives the example of AI customer service applied to banking.

When modern ATMs became widespread in the 1970s, many heralded them as the end of human tellers. Yet here we are, forty years later, and yet tellers continue to be a fixture. However, their role has changed: by letting machines take over mundane tasks of cash-counting and distribution, tellers are now advisors to customers. They handle the counselling aspect of banking while the machines complete repetitive tasks for which empathy isn’t necessary.

“Now, you can go online and be your own banker, and make your own decisions. So why still go to a real bank teller? For advising. Now the advisor is responsible for building the strongest possible relationship with a client,” explains Shep. In fact, since the advent of ATMs, the overall number of bank branches increased, in part due to being more efficient. And so has the overall number of employed tellers.

Delegating to IAs rather than bowing to AI

Shep adds, “We’re starting to move into a machine world, where AI can successfully take over some of the functions. However, the machine can’t take over every function. Don’t think of machines as a means to take over a person’s role. Think of IA rather than AI, intelligent assistant instead of artificial intelligence.”

Shep’s vision of an “advanced call center” involves an IA listening in on the conversation between the customer and the representative, and offering solutions to the former’s problem in real time, so that that latter can focus their undivided attention on the customer.

The result of the vision, which isn’t so far off as to be unrealistic, isn’t a disincarnate and disconnected version of customer service. Rather, it makes a more empathetic quality of service possible through the union of both human and machine skill sets.

Building engagement thanks to customer service AI

The science of IAs hasn’t progressed enough yet to be without hitches. Nevertheless, we are learning to stitch together technological advances and the personal aspect inherent to memorable customer service.

Powerful AI can’t save poor service

Artificial Intelligence programs, much like any other business tool, can’t salvage abysmal human customer service. A CRM program, though absolutely necessary to growing your business and shaping your sales funnel, won’t make up for a poor product. Likewise, a slick AI program, from a humble chatbot to a powerhouse like Gwyn, can’t remedy poor customer service.

Shep puts it this way: “At the end of the day, the company still has to deliver. If they don’t, it won’t matter how cool or how easy of an experience the customer had.”

An example of powerful AI coupled with impeccable service is Amazon, which keeps raising the bar for exemplary customer service. Not only does Amazon’s recommendation algorithm present us with uncannily appropriate suggestions, but those suggestions make the customer feel a personal touch. They achieve this even while the user browses the world’s largest virtual retailer. “When I log on, I’m greeted with ‘Welcome back, Shep! Are you still interested in these items?’ It’s like visiting my favorite store,” says Shep. Amazon has mastered faultless service by having its service execution live up to the might of its AI.

Know your audience

“In order to be relevant, a retailer has the sell the way a customer wants to buy,” warns Shep. If your customer service AI doesn’t simplify the customer journey, it will make it more frustrating. When developing an AI of any type, businesses need to make sure that it moves the customer journey along, which is why journey mapping is so important. Digital customer engagement of any kind means making hard choices. Shep adds, “you can’t be all things to all people,” and that successful customer service AI ties in with brand identity.

Therein lie the real difficulties facing effective customer service AI. That’s why the adaptive abilities of AI are so crucial: “AI should be programmed to learn — about you, your business, and your customers.” It’s not about the conflict between IAs and human agents, but rather in developing AI tailored to your business, your identity, and your customers. The issue shouldn’t be a strength contest, but rather a balancing act.

Published on April 7, 2017.