We now know the downside of AI in call center operations. As ChatGPT changes the nature of work, AI for call center is having a surprising negative impact. The results are in. Reducing complexity and increasing efficiency has surprising side effects. Erosion of agent mental health.

A call center agent’s job is tough. Today, the cost of burnout could be twenty six million or more and nearly half the average call center is at risk. It makes sense. No one calls customer service to thank them for their hard work or great product; customers call to complain. It’s taxing, so if anyone could benefit from offloading workload to AI it’s call center agents. At least in theory.

Before CX leaders started using AI for call center agents, their roles were filled with tasks beyond simply answering calls and solving customer problems. Agents were required to do paperwork on the back end, work with teams, do quick fact-finding calls, and work with the manager on customer questions.

Many of these tasks are considered routine and mundane, so it makes logical sense that the first step of burnout alleviation would be to offload the mundane using Robotic Process Automation (RPA).

Downside of AI in call center: more burnout

Today, companies are integrating RPA to offload many of the mundane admin tasks that previously consumed call center agents’ time. In doing so, they hope to empower agents to spend more time working with customers and be more engaged with customers. And that’s exactly what is happening. 

The mundane tasks are all but gone in many cases. Agents are faster and more efficient. But the result isn’t all positive: as agents spend time with more angry customers, something significant changed. The downside of AI in call center operations is this: agents spend more time under stress dealing with customer problems. The “mundane” tasks CX leaders sought to eradicate, may have been a feature of great customer experience, not a bug.

“Certain admin tasks may feel ‘effortful’ for a new employee, but for someone who is tenured, that can be completed on autopilot. It doesn’t actually strain them.” explained Tim Schocke who manages Lowes’ Contact Center Technologies. In essence, the mundane tasks provided a much-needed mental break for the agents. 

“Crunching more calls into the same amount of time becomes more efficient, but it increases the ‘burden density’ or more problems per hour.” said Schocke. More calls compressed into a smaller amount of time increases burden density. It turns out the mundane was inadvertently a chance for agents to depressurize from this burden density Schocke describes. 

Turning employees into robots, is a significant unforeseen downside of AI in call center.

---

---

Agents aren’t machines

By delegating the so-called “useless” tasks to AI, an agent’s entire day is filled with mostly strenuous tasks. One angry customer after another. It’s like asking the starting lineup to play an entire football game with no downs and no TV timeouts. In practice, the results of using AI for call center burnout alleviation lead to decreased performance, poor attitudes with customers, and increased turnover. 

As previously mentioned in our 2023 report on burnout cost, in an organization with 10,000 agents, the expense of such neglect can amount to over $26 million annually. Yes, AI and RPA are incredible tools, but it’s important to keep a close eye on your processes and routinely get feedback from all stakeholders, including your agents. Increasing burnout isn’t just a downside of AI in call center, it’s burning the candle at both ends. Adding implantation costs, and compounding turnover costs.

Icarus flew too close to the sun and melted his wax wings, plummeting to the earth below. It is important to consider whether the technology is truly taking you to new heights, or if it’s a temporary rise followed by a painful fall. “As we build towards an increasingly more digitized future,” said Schocke, “the responsibility of creating new working patterns that balance efficiency and employee wellbeing increase as well.