Consumers have more ways than ever to communicate with the brands they buy — be it through private chat or in public on social media sites such as Twitter. If a conversation conveys a negative sentiment, it can be detrimental if it’s not addressed quickly. Many companies are leaning on early stage AI tools for help.
Companies can use artificial intelligence in customer service to build a brand that’s associated with excellent customer experience (CX). This is critically important in an era in which consumers can easily compare product prices on the web, said Gene Alvarez, a Gartner managing VP, during a September 2018 webinar in which analysts discussed ways artificial intelligence in customer service can drive business growth.
“When your price is equal, what’s left? Your customer experience,” Alvarez said. “If you deliver a poor customer experience, they’ll go with the company that delivers a good one.
“This has created a challenge for organizations trying to take on the behemoths who are doing well with customer experience, with the challenge being scale.”
AI in customer service enables companies to understand what their customers are doing today and to quickly scale CX strategies in response. Chatbots can be deployed relatively quickly to handle customer requests around the clock, while social listening tools can track customer sentiment online to gain insight, identify potential new customers, and take proactive action to protect and grow brands.
With that, AI technologies including text analytics, sentiment analysis, speech analytics and natural language processing all play an increasingly important role in customer experience management. By 2021, 15% of all customer service interactions will be handled by AI — that’s 400% higher than in 2017, according to Gartner.
Where AI for customer service makes sense
With the current hype around AI, companies may rush into projects without thinking about how artificial intelligence can help execute their vision for customer experience — if it’s appropriate at all, Alvarez said.
“Organizations have to ask the question, ‘How will I use AI to build the next component of my vision in terms of execution from a strategy perspective?’ [and] not just try AI at scattershot approaches,” he said. “Look for moments of truth in the customer experience and say, ‘This is a good place to try [AI] because it aligns with our vision and strategy and the type of customer experience we want to deliver.'”
Gene Alvarezmanaging VP, Gartner
For example, an extraordinary number of companies have deployed chatbots or virtual assistants or are in the process of deploying them. Twenty-five percent of customer service and support operations will integrate bot technology across their engagement channels by 2020, up from less than 2% in 2017, Gartner reported.
But chatbots certainly aren’t the right choice for all companies. Customers who shop a luxury brand may expect a higher level of personalized customer service; self-service models and chatbots aren’t appropriate for customers who expect their calls to be answered by a person, Alvarez said.
And it’s no secret that virtual agents haven’t delivered the success companies hoped for with AI in customer service, said Brian Manusama, Gartner research director, in the webinar. All the experimentation with chatbots and virtual agents has, in some cases, hurt the customer experience instead of contributing to it. Companies have a long way to go to learn which technologies to use for the right use cases, he said.
“Companies really getting into [AI for CX] are disproportionally getting rewarded for it while companies that don’t do well with it are getting disproportionally punished for it,” Manusama said.
Match the product to the CX
The first step in choosing software for artificial intelligence in customer service is to understand that there is no single tool that works for every customer in every scenario, said Whit Andrews, an analyst at Gartner. For example, a customer who buys an inexpensive product may be fine interacting with a chatbot about that purchase, but not other types of purchases, he said.
“You have to identify the people who want to work with a chatbot and be realistic about the fact that if someone says they’d rather work with a chatbot, they might mean that for one situation but not another,” Andrews said.
To put a finer point on it, Jessica Ekholm, a Gartner research VP, advised companies to “pick the right battles” with AI tools by examining where the customer pain points are and developing a CX strategy that uses artificial intelligence in customer service strategically.