Customer experience (CX) agents are the backbone of any company’s interaction with its customers. It can be a very tough job, however. They’re juggling multiple applications, information screens, and clunky software integrations all while trying to solve a customer’s problem. This is the daily reality for many CX agents, leading to frustration and a revolving door of employees.
A major hurdle for agents is the sheer volume of data and applications they must navigate. They often need to sift through countless systems to find the information they need to assist customers. Juan Jaysingh, CEO of Zingtree, comments,” Agents are drowning in data and apps. They must sort through huge amounts of data and resources in countless different systems. That means toggling between apps and spending significant time searching. It’s frustrating for both customers and agents, not to mention mentally exhausting”.
This can lead to generic responses that fail to address a customer’s specific needs, ultimately leaving them feeling dissatisfied. Customers are frustrated with having to repeat information across interactions. This is typically due to siloed systems, lack of context sharing between departments, or simply inefficient workflows. For agents, this translates into wasted time and a sense of powerlessness in resolving issues quickly.
This frustration turns into high staff turnover, where the average agent tenure is less than two years. According to the 2023 Australian Contact Centre Industry Best Practice report the average annual agent attrition rate in Australia is surveyed at 32%, which is significantly higher than the global average of 22%. Jaysingh points out, “Industry turnover is high. Training new employees on the many policies, processes and platforms requires significant time. New agents get overwhelmed while existing agents must assume additional workloads until the new hires get up to speed. This situation creates a challenging work environment”.
To tackle the high staff turnover, some companies are looking to replace human agents with AI chatbots. Jaysingh explains, “The AI chatbot doesn’t need breaks or vacations, and training requires a fraction of the time. When implemented effectively, chatbots are a powerful tool for enabling a superior customer experience. A chatbot integrated with all of a company’s systems can serve as a central knowledge resource for customers”.
Ideally, chatbots can answer routine questions, freeing up agents to handle complex issues and provide a more empathetic touch. However, AI is not a perfect substitute for human interaction. Customers often find themselves trapped in frustrating loops of automated prompts when dealing with complex problems.
Replacing agents with chatbots is a race to the bottom
“AI is not a replacement for humans”, advises Jaysingh, “it’s a tool to improve customer service. Chatbots are great for simple tasks like checking order status or canceling a subscription, but companies should not channel every inquiry through bots. Some issues require human decision-making and empathy”.
“Have you ever gone through two dozen chatbot prompts trying to get a complex problem resolved? It’s very frustrating. Companies should always provide an easy path to talk to an agent. Ultimately, customer service is about people and relationships, and AI is not a substitute”.
There are significant risks associated with using AI in regulated industries, such as healthcare and finance. A major drawback is the lack of transparency in AI decision-making. “One significant drawback of current AI is its opacity. In many cases, it’s impossible to understand how an algorithm reached its decision, which could allow bias or flaws to persist. Without proper training, AI may operate outside of approved policies, possibly leading to regulatory violations and negative consequences for customers or patients”.
Humans and AI working together
The ideal customer service experience leverages the strengths of both humans and AI. Jaysing advises, “Before adopting any form of AI, organisations must build a system that interprets data, applies rules, and takes actions guided by those rules. This framework enables automation even without AI. When layered on top of the decision tree, AI operates within established guardrails. You can easily audit the technology because you have the roadmap it used to make each decision”.
Companies looking to improve customer service should focus on building a strong foundation before deploying AI. “Many organisations implement the chatbot and then try to make it learn the rules. In these instances, the bot encounters scenarios you never thought of, resulting in a car sold for $1 or an airline on the hook for a discount that doesn’t exist. Those mistakes are costly to the company and negatively impact customer service. If you implement AI on top of a well-designed system of action and automation, you can trust that your chatbot will follow your rules and policies”, says Jaysingh.
He adds,” You should also automate as many processes as is reasonable. The more tasks customers can accomplish on their own, the happier they will be. Additionally, the less tedious work agents must do, the better service they can provide. Integrating systems and creating clear decision trees empowers your teams to meet customers’ needs with much less stress”.
By integrating systems and creating a well-defined framework, companies can empower their agents to deliver exceptional customer service with less stress. Ultimately, AI should be seen as a tool to enhance human capabilities, not replace them entirely. “Additionally, AI without integration is useless. You must connect your systems so the AI platform can access all necessary information to answer a customer question, support an agent or execute a requested task”.