'Can I Help? You Seem Stressed.’ Now Chatbots Are Getting Emotional
WHY YOU SHOULD CARE
A growing number of firms are using a new wave of artificial intelligence apps to improve customer service.
By Molly Fosco
You’re just settling in for a night of Netflix and chill, glass of wine in hand, when the Wi-Fi goes out. You reset the router. Still nothing. Before throwing your modem out the window, you look for the provider’s customer service number, anticipating a full-on shouting match. But on its website, a chat bubble pops up asking if you need help turning your internet back on. The automated message seems to know what you want before you ask for it. You follow the prompts, and your Wi-Fi is back on within minutes.
Just the thought of calling a customer service line makes most of us want to tear out our hair. The prerecorded menu — the type of artificial intelligence (AI) in customer service we’ve heard for years — seems to treat someone at their wits’ end with an intractable problem the exact same way it treats someone with a basic query. But now, a new wave of AI applications is changing that.
Founded in 2015, New York-based Pypestream is using AI and automation to create emotionally intelligent, helpful customer service chatbots. The messaging platform launched with just two customers, and today has more than 30 brands using its product, most of which are Fortune 500 companies. San Francisco-based Intercom has been doing the same since 2011, and has 30,000 companies on its rolls. Mattersight, headquartered in Chicago, is using AI to connect callers directly to human representatives in a more efficient way than conventional methods. In a world where Alexa and Siri can answer millions of questions in a matter of seconds, these companies want to make interacting with businesses just as accessible. Customer service is finally — albeit slowly — becoming more friend than enemy.
When we want something, we should be able to ask a simple question and get a simple answer.
Saied Seghatoleslami, COO of Pypestream
The rapid advancement of AI in the past five years has dramatically improved health care technology, cloud computing and the ability of robots to understand human intent. Earlier this year, researchers in Portugal published a study in which they were able to successfully teach robots to anticipate select human movements and intentions. While AI chatbots are far from being mind readers, they’re emerging as a go-to customer service solution for a growing number of companies.
“When we want something, we should be able to ask a simple question and get a simple answer,” says Saied Seghatoleslami, the chief operating officer of Pypestream.
The AI chatbots created by Pypestream are specifically designed to understand natural written language. Their messaging widget and SMS integration use AI-based machine learning to identify the customer’s intent. Once trained, the bot won’t ask questions that have already been addressed — a pain point of many conventional automated service agents — and will either answer the user’s question directly or connect the individual with a human representative, depending on the difficulty of the inquiry.
Lincoln Center for the Performing Arts uses Pypestream on its website to help visitors quickly find events and answers to their questions. Lincoln Center has seen “an overall increase in conversion rates and reduction in bounce rates,” says Peter Duffin, senior vice president of brand management at Lincoln Center. “Customers are more satisfied overall.”
Seghatoleslami points out that when you’re interacting with customer service, something is already broken in the normal process. “These are moments of truth and moments that test a company’s commitment,” Seghatoleslami says. If someone has a positive experience interacting with customer service, they remember it. Pypestream aims to help the companies they work with stand out in their customers’ minds.
Intercom is also helping organizations improve the customer experience with AI and automation. Through its online messaging platform, companies can chat directly with customers to help solve issues. But before a question reaches an actual person, Intercom’s AI bot, “Operator,” can respond by suggesting helpful articles from the company’s website. When human assistance is necessary, the bot quickly identifies customers’ needs and connects them to the right person.
But Intercom prioritizes the use of chatbots only in situations that AI can handle. It’s important for the interaction to “still feel personal,” says David Katz, director of Americas and Asia Pacific sales at Intercom. “We don’t force it,” he says. Intercom sees the improvement of customer interaction as a key to customer retention. There are so many options to choose from, it’s become necessary “to go above and beyond to retain your customers,” says Katz. He believes automation combined with the ease of customer interaction on the platform provides that level of service. It’s about “finding that balance of personalization,” Katz says.
Mattersight is skipping the chatbot entirely and using AI to connect customers with agents right away. The company’s algorithm determines which human representative is best suited to solve the caller’s problem based on the individual’s personality. “A connection is more important to some customers than the actual answer,” says Mattersight chief technical officer Chris Danson. The company’s clients appear to agree. In a client case study conducted by the company in 2018, a Medicare provider that uses its Predictive Behavioral Routing technology wrote that because calls are directed so quickly, it saw increased efficiency and customer satisfaction right away. Danson acknowledges there is a growing interest in the complete automation of call centers, and he’s helping to steer the company in that direction. “We’re looking to optimize connecting customers to bots in the same AI-assisted manner that we do to agents,” Danson says.
But AI designed to interact with humans is far from perfect. Although Pypestream, Intercom and Mattersight are highly invested in AI’s ability to improve customer service, they’re also critical of its current capabilities. It’s still a challenge for chatbots to answer some questions appropriately, and “there’s no empathy or understanding of tone,” says Katz. Seghatoleslami agrees there are roadblocks. “Technology is never 100 percent,” he says. “There’s a good chunk of stuff that will be automated, but it won’t be everything.”
Human beings will likely remain critical for the customer relations industry, and AI won’t replace them fully. But even if automation only helps reduce the likelihood of you wanting to scream “REPRESENTATIVE!” into your phone when calling a customer service line, that’s still progress. At the very least, AI appears set to make one of modern society’s most painful tasks less frustrating.