Conversational Voice AI: The Agent Staffing Solution

We've all had this experience: you ring up a call center and are put on hold. A few minutes pass, then a half hour, then an hour. While it might seem like expected behavior from our overburdened service sector, what you're experiencing is caused by a perfect storm of pandemic-related adverse events, including the Great Resignation and the economic downturn.

Customer service centers are suffering from high levels of attrition. The Bureau of Labor Statistics reported that agents in the 20-34-year-old age group generally stay in a job for only about one year, a metric that has declined over the past few years. Additionally, budget cuts mean managers are dealing with higher demands and fewer resources to tackle them, and are sometimes even forced to let go of agents.

These factors contribute to unprecedentedly high wait times in customer service. Call center analytics firm CallMiner recently reported that the length of service calls has risen by several minutes since COVID. That's a huge customer experience problem.

For too long, we've thrown the wrong technology at the problem, and the results have been mediocre at best. Humans have been doing the heavy lifting when it comes to customer service, and the automated tools, seen as a cheaper alternative, haven't eased the problems enough to impact the industry in positive ways.

Why? Those robots, also known as interactive voice response (IVR) systems, are limited in the queries they can solve (think: pay a bill, change an address on file). So humans are left holding the bag.

The solution, then, is to create a hybrid system. This is where voice AI comes into play. And it could be the customer service staffing solution you need.

Voice AI uses machine learning models to converse with customers in a way that supplements, and in the case of mundane, repetitive tasks, replaces, human customer service agents. It's unlike Alexa and other voice control systems in that the responses are never canned. You can have an entire conversation with a digital voice agent, which can sometimes be nearly indistinguishable from a human agent, provided the AI has been trained with the correct data.

The data itself is collected from various sources, including your databases of business and product information, email, and customer histories. Sometimes, it's even trained by actual human conversation data, with an aim to enrich domain-specific knowledge. In other words, you train a robot and the robot answers customer questions just as a human would.

Voice AI Can Solve Staffing Issues

Voice AI can have any number of applications, ranging from pre-call assistance (gathering information from the caller and forwarding it to the assisting agent) to in-call assistance. In these instances, the AI listens in and provides real-time, supplemental information to the agent to better help assist the caller.

We can also train AI to manage post-call assistance, including automatically creating call summaries for the human agent to review and accept.

Voice AI can handle these necessary-yet-tedious tasks quickly and error-free. It's like a seamless scaling of your support team. The better we train it, the more it will improve your customer experience— an experience that doesn't need to abide by 9-5 limitations. And your human agents don't need to spend valuable time performing tasks better suited for robots.

How many times can someone ask a customer to verify their name in a day? A lot, it turns out. And repetitive tasks like this tend to kill motivation and job satisfaction. Current data shows that contact centers have at least a 35 percent to 40 percent attrition rate due to factors like low employee morale.

These quick, uninteresting calls can be handled by voice AI, while more complex issues can be forwarded to human agents. As a result, human jobs become less boring. A study by Salesforce shows that 77 percent of customer service agents believe their job is more strategic than it was two years ago. This is in large part due to new automation technologies like voice AI.

AI systems also can be indiscriminately sympathetic and kind, allowing them to bear the brunt of whatever verbal abuse might be on the other side of the line and defuse anger and frustration before human agents can step in. More pleasant phone experiences inevitably mean employees become less prone to burnout.

Voice AI ushers in a modern approach to customer service, consisting of a collaboration between voice bots and humans. By employing voice AI to perform monotonous tasks, companies need fewer human agents. Additionally, the improved quality of life and job satisfaction of the team leads to lower attrition rates. Staffing shortages will be revealed for what they truly are: inefficiencies in an outdated system. Voice AI will become the norm before we know it and introduce us to a better way of working.


Sourabh Gupta is co-founder and CEO of Skit.ai.