There’s no denying that interest in chat bots and AI have soared over the past few months. Chat bot technology like ChatGPT has brought AI conversations front and center into the mainstream culture.
However, the concept of AI and chat bots isn’t new to one area of business… the contact center. Referred to as “one of the most fertile testing grounds” for this technology by the Wall Street Journal, chat bots are further along in this area of a company than many others.
But what’s the greater impact of a business for testing AI and bot technology? It’s created a work environment in contact centers which carefully toes the line of using technology to drive efficiency and lower call volume while still providing the excellent level of service customers expect. To help elaborate further, below are a few ways to use chat bots in a contact center without negatively impacting the customer experience (CX).
Routing basic customer inquiries automatically – One of the primary and first use cases of a chat bot is to route issues. A customer can tell the bot “cancel a phone plan” and it will automatically move them to a conversation with the Retentions team to work on the customer relationship. However, chat bots can have issues with routing complex customer issues, and some contact centers still have a human component in this process. As a simple example, if a customer says “I want to cancel a phone plan because a family member passed away” you wouldn’t want to send them to Retentions who is going to try to keep their business. A human can still pick up on nuances like this that a bot sometimes struggles with.
Gauge the sentiment of a customer before they reach an agent – One area of nuance that contact center chat bots have improved in significantly over the years is accurately gauging the sentiment of a customer. Emphatic statements like “Transfer me now!” can be analyzed by a chat bot to indicate the customer has low sentiment and the technology will send them immediately to an experienced agent. Phrases such as “Thank you” or “That works” can indicate positive sentiment and an escalation process is likely not necessary for this specific customer.
Prompt suggested actions for agents based on the issue – The two examples listed above highlighted a customer interacting with a chat bot. But there are also components within bots that can help agents who provide chat support to customers. For example, a bot can prompt an agent during a chat conversation with recommended next actions based on what the customer has typed. If a customer asks “I will stop by your Savannah store but what are the hours?” the bot can quickly display a prompt for an agent to click on with the appropriate hours. This eliminates the need for the agent to look up the hours manually, creating a faster conversation and a better customer experience. With all of this said, relying too much on bot assistance can be troublesome. While it’s a good source for factual responses like hours of operations, suggestions for open-ended and judgmental responses aren’t nearly as accurate. This is why an agent must use common sense and their own experience to determine whether or not the help of a chat bot is valid.
Evaluate bot and agent success to make adjustments – Even though the contact center is a testing ground, AI and bot technology is still new and is constantly changing. It’s important to continually measure the success of the technology and how it’s impacting the conversations between agents and customers. Some areas of a bot may be ignored because they are not useful, while others can result in confusion and lead to an increase of customers abandoning the conversation. Continually monitor metrics around a chat bot and assess how it factors into your contact center operations to understand what’s working and what isn’t so you can adjust in the future.
In short, a contact center can use chat bots without negatively impacting the customer experience by maximizing what works and minimizing what doesn’t. A bot should never replace an agent or drive all customer conversations. Instead, it should help to make the conversation better and improve the experience for both the agent and the customer.
