The Delicate Dance Between AI and Human Agents

Artificial intelligence will soon take center stage in your contact center — if it hasn’t already.

Artificial intelligence (AI) uptake increased dramatically over the last few years. A 2022 PwC report revealed that more than 70% of companies were already using or planning to deploy AI in some form within their business operations. Business leaders are using the technology to realize an array of benefits, from product innovation to enhancing data-driven business decisions.[1]

AI helps companies operate more efficiently and address more customer issues with less direct human involvement. This outcome is key for AI’s return on investment. It’s natural, then, that many would wonder what role humans should play in the delicate dance of customer service. As they rapidly improve, are bots pushing humans off the dance floor?

We may not yet know exactly what it will look like, but there’s good reason to believe that there’s plenty of room for robots and humans to perform in harmony to create quality customer experiences. In fact, their ability to stay in step will prove to be essential. In the CX of tomorrow, bots aren’t pushing humans off the dance floor — but humans will need to change up their footwork.

The self-service revolution

Customer service used to be, essentially, a dance between two people: the agent and the customer. Providing a quality experience primarily meant training agents to engage customers with empathy and address their concerns. But this has all changed in the age of self-service.

Thanks to AI, many customer issues can now be resolved without the involvement of a human agent. From basic banking transactions to checking on the status of an order, today’s chatbot and conversational IVR solutions can address many customer needs while reducing the costs of service. Unlike human agents, they can provide 24/7 support and handle multiple customer queries simultaneously. New forecasts predict that AI agents could save the contact center as much as $80 billion in labor costs.[2]

These benefits don’t apply only to businesses, though. Customers want self-service. According to a report from Aspect Software, 73% of customers want brands to allow them to solve issues and answer questions on their own. Sixty-five percent say this makes them “feel good” about a company.[3] In other words, they sound happy to invite AI to the dance.

Where AI still falls short

There’s a catch, though. Customers are perfectly happy to go a turn or two with AI — as long as it resolves their issues. When bots don’t work well or create barriers to customer service, they cause more problems than they solve. A recent UJET survey even reported that 80% of customers found that interacting with bots only increased their frustration with a company.[4]

To some degree, the reasons for this are obvious. We’ve all experienced what happens when AI interactions go awry: the doom loop of incomprehensible or just plain wrong responses with a chatbot or the bot trap that won’t let you escape to the safety of a human agent. These are familiar scenarios in today’s service economy.

These AI limitations are still real, even as the technology improves and aims to move past them. Today, though, the limitations of AI can be more subtle — and, perhaps, more dangerous. Consider the recent hype around ChatGPT, the latest and most ambitious chatbot released by OpenAI in late November. This bot achieves a new level of conversational fluidity with its ability to respond to open-ended questions, even writing convincing humanesque essays in response.

ChatGPT seems so authentic that it could easily deceive anyone who is unaware.[5] It’s clear that, despite AI’s significant advancements, it still can’t quite meet the challenges of complex human tasks and interactions in the way that real humans can.

Setting the stage

Ultimately, these limitations point to what the new rhythm of customer service can feel like. AI hasn’t supplanted human agents — and likely won’t anytime soon. But it can support human agents in crafting a more complete customer experience.

AI can enhance customer service beyond simply replacing human-driven interactions with bot-driven ones. AI can augment agents’ capabilities in several ways. AI can easily collect vast amounts of customer data and turn it into personalized recommendations that human agents can use to enhance CX. It can seamlessly serve up customer information directly to agents so that they can solve customer problems faster. And it can accurately route issues to the right support teams so that the best-equipped agents take the call. This type of give-and-take represents the future of customer service in contact centers.

Moreover, by handling routine and mundane tasks, AI can offload interactions from agents, ensuring that they are available for more complex interactions. Experts have foreseen this for a while now. In its 2017 report, “The Future Of Jobs, 2027: Working Side By Side With Robots,” Forrester predicted that robots would complement and enable humans to focus on tasks that involve more complex intelligence, aesthetic judgments, and unique skill sets.[6] This, in turn, helps human agents improve to meet these new challenges — 79% of support agents report that handling more complex issues enhances their skills.[7]

To truly set the stage for a smooth performance, however, contact centers must ensure that their AI technology is dialed in and ready to perform. Training agents to do their part is essential, of course. But if your chatbots and conversational IVR systems are unreliable, they will only create more problems for your human agents, no matter how prepared they are.

That’s why, in the age of AI, a comprehensive testing solution is more important than ever. Contact centers need complete confidence that the chatbots and conversational IVR systems they deploy not only stay up and running but also avoid causing problems through clunky or unintelligible customer interactions. Cyara is the only CX assurance platform that covers your customer support solutions from end to end. With Cyara, you can ensure your AI-powered CX is performing at its best — and prepare for the main event of human-to-human interaction.

The dance may be delicate, but it can be especially beautiful when all goes smoothly. Take a look at our Building Smarter Chatbots ebook to learn more about how we can help you ensure that it does.

[1] PwC. “PwC 2022 AI Business Survey.

[2] Tech Monitor. “AI in call centres could save businesses $80bn.”

[3] “Report: Poor Customer Service Pushes Consumers Away.”

[4] Forbes. “Chatbots And Automations Increase Customer Service Frustrations For Consumers At The Holidays.”

[5] Business Insider. “I asked ChatGPT to do my work and write an Insider article for me. It quickly generated an alarmingly convincing article filled with misinformation.”

[6] J.P. Gownder, “The Future Of Jobs, 2027: Working Side By Side With Robots.” April 3, 2017, Forrester

[7] HelpScout. “Will AI-Powered Customer Service Replace Your Job?

Artificial Intelligence, Machine Learning