Speech-recognition technology may have now been under development for over four decades but, for customer service purposes, accuracy remains a persistent problem. This leads to a major question for the CIM industry; will automated systems ever become an effective replacement for the call centre?
Alongside providing a more convenient, accurate service for consumers, one of the major advantages of a dedicated CRM system is the ability to help businesses save valuable time and resources. Readily available data on all previous interactions between a customer and an organisation enable customer service operatives to resolve any issues with greater speed.
Indeed, collecting information on common issues can allow a brand to create an automated online chat help system, returning relevant troubleshooting advice for consumers based on keywords in a query. Despite these advances, speech-based services remain the customer service channel of choice for many consumers.
Seemingly, speech-recognition technology holds the obvious solution to allow businesses to benefit from the time-saving elements of automated service while satisfying this preference. Unfortunately though, this technology has lacked the sophistication to represent a viable alternative to human involvement, often misinterpreting speech patterns.
Chris Schmandt, a founding force in MIT’s media lab with a history of designing speech interfaces, underlines three key focuses he believes must be addressed in order to change this.
Primarily, two-way communication, allowing users to feedback on how well the system’s response has answered their query, is necessary to improve accuracy. Secondly, Schmandt feels combining visual and audio advice on smartphone calls can make guidance clearer, particularly in situations where complicated audio instructions may be difficult to remember.
Perhaps most importantly, voice-recognition systems should integrate other information sources to provide greater context to queries. In terms of CIM, this means using everything from scripts of previous help requests, to browsing and purchase history in order to tailor responses to the individual; a common practice for chat services but not for voice, according to Schmandt.
Combining these factors, we may be on the way to finally seeing a scenario where customer service assistants can take a back seat to CIM voice technology.