Chat is a favorite interaction channel for customers and is increasingly offered as an option on websites and mobile apps. In our NICE CXone CX Transformation Benchmark, chat has the highest satisfaction of all channels with 47% of consumers saying they were very satisfied with chat, as it provides convenient, quick, and helpful service. According to WhosOn, 68% of people prefer to have root canal without anesthetics, a tax audit or a dinner with mother-in-law, rather than to wait on hold.
What’s the point? Consumers have already accepted chat! It often eliminates the “on hold” frustration of voice calls and lets contact centers increase productivity by having agents work with 3 or 4 customers at the same time. Many organizations feel pressured “top down” to apply AI technology to automate chat interactions entirely. Chatbots are the most common application of AI for customer self-service and are not as successful (yet), as the hype would indicate. A recent study found that only eight percent of consumers used an AI-enabled service like chatbots, and just 27 percent were highly satisfied.
Even though the best-case scenario for chatbot containment rates will be 20-40% -- that’s okay. It’s a great way to move them to chat with a live contact center agent. Instead of looking at this as a failure to contain, lets look at this as an opportunity to get rid of the repetitive questions chat agents spend time answering. And if your bot is built correctly, you create a seamless transition to a contact center agent with full visibility into information collected, at which time the bot itself converts from virtual agent to agent assistant. Many organizations in 2019 will take a split approach: more aggressive use of AI to automate repetitive agent after-call work; while taking a more targeted approach with simple and high-volume self-service use cases. If you can deflect even 10-20% of routine asks to a chatbot and let agents spend more time with customers than on after contact work -- that is success!
