A new white paper from Frost & Sullivan details why for many businesses, customer care is something of an oxymoron. Call centers are often viewed as a necessary expense: one that must be accepted to stay in business, but which generates very little value.
Lately, this philosophy has led businesses to employ chatbot technology in their customer care channels, where many routine interactions are handled by automation. However, while chatbots can reduce call center costs by reducing the number of customer care specialists on staff, they often do considerable damage to the business by virtue of their inability to handle complex interactions and by forcing customers to click through endless menu options, only to find that they must talk to a human being anyway. Put simply: Chatbots are great, except when they are not.
The solution is to employ true interactive voice recognition (IVR) technology, operating at a high CISR, that is enabled by natural language processing and machine-learning-enabled comprehension of the customer’s needs. Combined with interaction analytics on the IVR side, rather than the agent side, it is now possible for companies to achieve real automated customer care that does not aggravate customers and which enables each customer interaction to drive up sell opportunities, based on a positive contact experience.
Frost & Sullivan research has disclosed that interactive voice response (IVR) technology is the key to unlocking the value of chatbots while addressing the customer need for a more personalized experience. In such an arrangement the chatbot is providing a necessary interface, but the IVR technology is delivering the value: as much as 95% of the value by some market estimates.
Yet, the IVR must come first: by delivering customer sensitive responses across the various customer contact channels (the omni-channel), gains in customer satisfaction and retention provide the means to fund and implement more effective chatbot technology. In fact, it is often forgotten that chat language is an application specific subset of speech. This means that first you have to automate and build up the system to understand speech, and only then you can go ahead and do chat. Given also that the ROI primarily comes from IVR, this means that all technologies that do not present a unified system to treat both IVR and chat can never deliver chat with a measurable ROI.