Blurred Lines: Merging Customer Service with AI
“Bot” is the word on everybody’s lips. Ever since Mark Zuckerberg took to the F8 stage in April, the tech community worldwide has been bombarded with articles and opinion pieces on the subject. Love ’em or hate ’em, the influx of chatbots reflects the evolution of customer service and the need to fix the disconnect between brands and consumers — to turn each interaction into a meaningful relationship.
Bots themselves are nothing new. In the 1990s, they were called “virtual assistants,” and they were a massive failure. And the Twitterverse first experienced with AI 10 years ago. It’s only been in recent history that the titans of Silicon Valley like Facebook and Google have started investing in chatbot technology. The public’s renewed interest comes at the heels of advanced NLP algorithms, faster machine learning, and more sophisticated AI software, which make them better at anticipating needs and responding appropriately.
>> Related read: The Chatbot Roundup: Here’s What You Need to Know
Interacting with chatbots is more effective and less frustrating. For brands, bots present the opportunity to reduce costs in customer service and streamline every interaction. However, brands must remember that bots as they currently stand are a means to an end — not an end in themselves. Carefully implementing bots into your customer care strategy is an intelligent way to boost customer connection at a reduced cost, but leaving it completely in the hands of AI is dangerous. Bots can’t do it all.
Bots: two sides of the story.
Several queries fielded by customer service reps can be answered through an automated bot. (Think: tracking an order or resetting a password.) But not all requests are clear cut. Bots definitely have their time and place, but sometimes you need to escalate the conversation to a human being. Remember these for things when adding AI to your CX.
1. We’re open to bots...but they aren’t for everyone. Now is the right time to introduce bots into your CX strategy. The public is more receptive than ever to dealing with AI — many people are even excited about its future implications. Would you be interested in interacting with a White House bot eight years ago? Not likely. Although many of us have opened our minds to bots, others are unconvinced. Someone looking to engage directly with a customer care professional (CCP) won’t be happy if they get a robot on the other end. (And, yes, they can tell.) Wise brands have a human being on deck to seamlessly transfer any conversation that starts going south.
2. A poorly scripted bot is bound to frustrate customers. Bots today run largely on scripts — that is, they aren’t the sophisticated AI-driven machines we imagine. Instead, they’re tools for automating specific tasks and not much better than traditional IVR systems. “Crappy bots are almost the same as crappy calls,” maintains Rob LoCascio, LivePerson founder and CEO. “Automated bots cannot solve user-specific problems.” If a bot’s script takes a customer down the wrong path, the end result will be even greater frustration. And this is bad for business: Seventy-six percent of users would stop doing business with a brand following a negative experience.
3. Bots: an agent’s best friend. Bots and humans are good at different things. Though bots are seeing vast improvements, they simply enhance — not replace — an agent’s ability to work by taking care of repetitive or automated tasks. While a bot is capable of resetting a password or confirming an order, human inquiries are often nuanced and complex. There are thousands of ways to ask a question, and bot technology can’t anticipate how two individuals may inquire about the same thing. It many cases, human discretion is far more accurate than an algorithm. In short, bots should work; humans should think.
>> Related read: Lasting Digital Transformation Must Be Tech-Powered but Human-Led
4. The missing ingredient is empathy. “Care is not about reach,” says LoCascio. Despite what science fiction imagines, you can’t form a real relationship with a bot. Canned responses lack an emotional element that connects two people. There’s a delicate balance between AI and HI (human intelligence), and, when in doubt, it’s best to fall on the human side. Only a live person can understand true emotion. Only a live person can say “I’m sorry” or “I feel your pain” — and mean it.
Technology is supposed to empower and enable people. For brands, it’s supposed to make their lives easier as well as the lives of their customers. So we remain optimistic about the future of bots. Technology is making great strides, and there will soon be a day when a bot interaction comes as naturally as an exchange with a friend.
However, brands that over-promise on bots too early fail us in the long term. People understand and appreciate the power of the human touch. That means customer service is no place for brands to cut corners. There is a time and a place for bots. But that can never take away from the value of a live person.