KANA Automates Customer Response

If your call center agents are overwhelmed by having to respond to hundreds or even thousands of customer emails each day, natural language capabilities may help. KANA today announced its partnership with Banter to embed Banter Server into KANA Response email support software. Banter's customer base varies by vertical focus, but the customers all have one thing in common: large volumes of email coming into their contact centers. Some of Banter's smaller customers receive several hundred emails a day. Some, however, receive as many as 50,000 emails each day. "Call centers are getting a rising tide of customers communicating with companies over email. Our solution enables companies to respond to emails in a consistent and accurate way," says David Zinman, vice president of marketing at Banter. Using natural language capabilities, the Banter Server, which will be preintegrated with KANA Response, can understand the intent of customers' emails and provide a suggested response for agents. Essentially, the KANA system sends an email to the Banter Server, which scans the text and sends back a classification of the email, such as a "check reorder request" with a confidence level. The Banter Server can then suggest various responses for the agent. KANA can then route the check reorder request and suggested responses to the right agent. The agent can read the customer's email and quickly select the appropriate response without having to manually write one. "All of the emails will be consistently written and composed answers," Zinman says. "This takes a company's ability to manage email to a new level in terms of productivity and consistency and accuracy of emails." Banter Server offers cross-channel support and handles dynamic and informal email as opposed to static and formal types of communication. Banter Server also provides real-time feedback to ensure accuracy and promote self-learning capabilities, the ability to communicate in 12 different languages the ability to understand multiple intents in a single message, and the statistical likelihood that an answer is correct, and not a relative ranking. The big issue that has traditionally faced natural language technology is accuracy. However, Zinman says it should not be a bottleneck for customers. "If we give you the top-five suggestions for an email, almost one hundred percent of the time the correct answer is within the top five," he says. In fact, so accurate is the Banter Server that some customers have elected to use Banter's auto-response feature, which responds to customer queries without any agent intervention. Banter integrates its natural language solution with several other contact center software solutions from companies like Avaya and Apropos. However, the KANA partnership is particularly significant, Zinman says. "KANA is one of the leaders in email response-management systems. This partnership shows we have a dominant positioning the market."