3 Ways Human+Machines are the Real Future of AI in Customer Service
Don’t count the humans out yet. While some industry experts predict that artificial intelligence (AI) will replace humans in a few years, I think we’ll be sticking around, at least for a while.
It’s the human+machine equation that will be the real future of AI and the power that will drive AI innovation. I especially see this in the realm of delivering on the promise of a truly intelligent customer experience, one that combines the best of both traditional and digital worlds. Although we will see AI combine with digital capabilities to take over some of the basic customer service needs, there will be more complex questions that will still require the inimitable magic of human problem-solving.
Forrester makes a similar assessment. Daniel Hong, research director at Forrester, writes that “humans will play a critical role in the ongoing optimization of AI,” as “having a successful AI-driven customer service or sales program will depend on the processes that support a blended AI approach.”
While blended AI is taking many different forms, essentially it involves combining the distinctly human qualities of common sense, emotional intelligence, creative problem-solving and intuition with the super-human abilities of AI. In the blink of an eye, AI can sift through a gazillion data points to identify complicated patterns and automate repeatable processes without ever getting tired or discouraged. Just imagine what we can accomplish when we put these capabilities all together.
Digitize. Automate. Enhance.
One of the keys to implementing blended AI in customer service involves digitizing the massive amounts of data that are still in voice form. While technology is getting better at converting voice to digital formats, AI still requires the assistance of humans to provide context by tagging and labeling the data. Once the data is tagged, then AI can be off to the races, identifying patterns and opportunities for increased automation and improvement. First you digitize, then you automate the analysis, then you enhance. Here are just three ways that blending humans and AI will help deliver a more intelligent customer experience by improving the journey every step of the way:
1. Drive continuous improvement.
AI pattern recognition will identify exponential opportunities for enhancement in both digital and voice-based customer interactions. For example, using AI you can get the insight you need to create smarter digital content so you can aggressively pivot from voice to self-help by identifying the conditions when customers called support after first seeking answers online. With blended AI, analytics and customer teams can look at where the breakdown in communication occurred, figure out what’s missing and develop more effective content that answers the customer needs. At the same time, you’re also finding opportunities to deepen customer personalization, which can drive digital engagement, sales and customer satisfaction.
2. Deploy cobots, collaborative AI and human augmentation techniques.
Blended AI can be used to enhance human performance by giving human agents a “cobot” via an AI-powered engine that helps every agent emulate your best performers. Derived from the phrase “collaborative robot,” cobots got their start in the manufacturing environment as computer-controlled robotic devices designed to work alongside people and assist them physically. Now, we see the cobot concept extending to the customer service world in the form of a brainy AI assistant. This is beginning in online chat, essentially like a chatbot that works in combination with a person, so human and machine are working together to have a conversation with the customer.
The next step will be deploying cobots to help agents during voice interactions, which my SYKES colleagues and I see as the ultimate goal. The AI cobot will listen to the conversation and make recommendations to the agent on the fly—kind of like what we’re experiencing with Siri and Alexa—so the agent and “digital coach” are working together on behalf of the customer. But the call cobot will need to be more nuanced than the chatbot version, to avoid distracting agents as they’re speaking to customers. Think how annoying your car’s mapping system can be when it’s insisting, “Turn left. Turn left. Turn left,” when you’re simply taking a different route. Instead, we envision the call cobot as providing recommendations to the agent for next steps in the conversation, likely provided in a popup window the agent can see at a glance.
3. Improve agent training, onboarding and in-the-moment coaching.
At SYKES, we’re already using AI to help hone agent training. In his article, Thinking Outside the Class: How AI Can Transform Customer Care Training, my colleague John Kruper, SYKES Chief Learning Officer, describes how we’re combining AI and data science with adaptive microlearning techniques to give agents on-the-job coaching and training on demand when the need is identified. The next step is to use the power and speed of AI to identify improvement opportunities, experiment and test new approaches to agent training during the onboarding process. This way, we’ll be changing the experience for both the customer and the agent.
4. Make sure AI is performance-ready.
While Forrester predicts that blended AI will help reduce customer servicing costs and improve sales outcomes, the industry analyst also includes a caveat. Forester research director Hong warns that customer satisfaction levels could actually drop as more companies drive customers to chatbots, self-service and chat before those solutions are really ready for primetime.
At SYKES, we don’t lose sight of the fact that there will still be times when two people really need to talk to each other. It’s necessary, it’s meaningful, it’s emotional, and it’s complicated. There are always going to be customer challenges when it takes problem-solving, empathy and situational awareness—all functions that are beyond a robot’s capabilities. What’s important is to remember that in today’s digital world, customers usually have already tried several avenues to solve their problem before they pick up the phone to call customer service. When they get to that point, they are at their most frustrated, so you want that interaction to be the very best. By adding AI to the human equation, you can help make sure that’s exactly what happens.