E-Handbook: Virtual agents not yet ready for primetime customer service Article 2 of 3

Graeme Dawes - Fotolia

AI chatbots augment humans working in contact centers

It might seem like contact centers are all being automated by AI chatbots, but for now, the jobs of human agents are still safe while all the kinks are worked out.

As artificial intelligence gets more sophisticated, companies are beginning to integrate the technology into contact centers in the form of AI chatbots. AI has evolved to the point where it can handle simple customer requests, and some companies have already deployed them. But experts caution that if you're wondering whether AI replaces humans in contact centers, the answer is no -- or at least, not yet. Also, before deploying any chatbots, experts recommend thorough testing to ensure that the technology will work as expected.

The current state of AI technology allows enterprises to use AI chatbots for lower-level tasks, such as changing an address on file or getting an answer to a frequently asked question, to free up humans for more complex requests, according to Shep Hyken, a customer service and experience expert and chief amazement officer at Shepard Presentations. Introducing the technology into the contact center is the right word to describe the scenario, he added, because the technology hasn't been perfected for most companies and will not replace humans.

For that to work, a good system will have a seamless connection between the AI chatbots and human agents and recognize when a customer is struggling to understand the answer, according to Hyken. Then the computer will transfer the customer to a human automatically, resulting in a positive experience.

Some companies are using AI chatbots to support their own employees while the employees are helping customers, Hyken said. The computer listens to the conversation that the customer service representative is having and feeds information to the representative, but ultimately, the human makes the final decision as to what to send to the customer. This approach, however, does allow the customer service rep to match customer profiles and buying patterns and make appropriate suggestions for upselling or anticipating support needs.

Chatbots aren't a panacea

Like other technologies, chatbots influence every aspect of delivering customer service, from the types of interactions to routing methodology, according to Brad Cleveland, senior adviser and former CEO of the International Customer Management Institute. "A common mistake we see is assessing these new capabilities in a vacuum, not considering how they shift workloads, impact agent skill requirements or influence customer behavior," he noted.

There's no technology that can be 'the answer' to a better customer experience in the absence of effective strategy and .
Brad Clevelandsenior adviser and former CEO, International Customer Management Institute

While the goal of chatbots is to make customer service easy on both sides, the biggest con of AI chatbots isn't actually because of the chatbots themselves, Cleveland said. "It's expecting them to be a silver bullet for staff reductions or service improvements. There's no technology that can be 'the answer' to a better customer experience in the absence of effective strategy and ," he added, noting that chatbots can anger customers if they aren't offering the help needed.

This communication gap can stem from chatbots not being programmed properly. Hyken cited an example of a customer service chatbot that he used to find a docking station for a computer. Instead of returning information about the docking station, it responded by asking which computer he wanted to purchase.

"That chatbot did not give me an answer remotely close to what I was looking for," he said. "It was programmed to sell a computer, not answer questions."

Additionally, there are just some things chatbots can't do because of regulatory requirements, according to Matt Grofsky, CTO of communications software maker Ytel. "A good example of this is quoting a mortgage rate or handling certain types of monetary transactions," he said.

A test being administered between a human and computer.
A Turing test helps demonstrate whether a chatbot is human enough.

Ensuring AI chatbots deliver customer service

Like the computer manufacturer's chatbot that returned irrelevant answers, the biggest mistake companies are making right now is releasing chatbots to the public too soon, according to Hyken. Beta testing is critical to ensure readiness.

Additionally, automating the contact center with AI and chatbots requires constant maintenance, learning and adjustments to keep up with customer expectations, Grofsky said. "The knowledge of how intents and dialogues work in a chatbot environment [is] required," he added. This may mean steep development and maintenance costs that come with someone with AI knowledge as well as a developer or development team.

While the technology is still evolving, experts do expect AI chatbot capabilities to grow by leaps and bounds. But for now, they're best used to supplement what human contact center agents do, enabling humans to handle more complex customer service needs, such as specific tech support needs, and fulfilling routine requests like address changes or quoting shipping costs. Enterprises just need to ensure their AI chatbots are tested and ready to use with the public.

Next Steps

How customer service chatbots improve efficiency 

Dig Deeper on Customer service and contact center

Content Management
Unified Communications
Data Management
Enterprise AI