Yoav Degani, CEO of VoiceSense, has explained how the company’s predictive analytics solutions can assist enterprises in analyzing their voice-based customer interactions and improve risk assessment, sales and customer retention.
Enterprises use predictive analytics for assessing the probabilities for behaviors and future outcomes based on analysis of trends, patterns and characteristics of current and past data.
VoiceSense has an innovative approach in predictive analytics that leverages speech and voice-based customer interactions.
Almost all enterprises can use predictive analytics to forecast the likelihood of future events and improve their decision making processes in areas such as marketing, sales, service, production, operations and human resources,
“Predictive speech analytics can help organizations monitor and anticipate customer behavior, enabling them to identify opportunities and provide personalized service,” explained Raul Castanon-Martinez, senior analyst at 451 Research. “The range of use cases that can be addressed make it particularly relevant for high-touch customer interactions, which can have a direct impact on a company’s bottom line.”
Speech-based predictive analytics
Using speech-based predictive analytics can further shorten and improve an enterprise’s decision making processes. Examples of voice-based customer interactions for which speech-based predictive analytics can be applied include scenarios to improve risk assessment, expand sales and increase customer retention. There are also applications for human resources in terms creating personal profiles for candidates and staff.
VoiceSense offers an innovative biometric concept based on personality profiling behavior of consumers. VoiceSense focuses on an analysis of the prosodic speech parameters of a customer. This refers to the non-content parameters of speech, such as intonation, pace and stress. It provides an automated process that can be used to analyze either recorded voice files from a big data source or live audio streams.
VoiceSense said its predictive models and signal processing techniques are independent of language and content and assess over 200 prosodic speech parameters. This analysis builds a personal profile of the speaker with evaluation of the individual’s characteristics, such as levels of risk affinity or aversion, tendencies for impulsive behavior and rule abidance, personal integrity, conscientiousness, sociability, wellbeing and so on.
Enterprises can integrate a prediction score provided by VoiceSense’s analysis for a specific consumer behavior into its decision-making processes and CRM data. Enterprise can use this score in real-time to make go/no-go decisions or determine approaches to address a customer.
Demand for predictive analytics
Enterprises are still building the infrastructure and environments to facilitate such predictive analytics – automated data collection, data storage, analytics and statistical platforms, data scientists training, connecting the enterprise systems into their data centers. Most predictive analytics efforts are focused on the data exists in the organization’s data systems. At the same time, the use of speech in predictive analytics is still in its early stages.
“We are involved in a number of projects, including a number of banks and insurance companies where we have demonstrated improvements in the loan default predictions and insurance claim probabilities. We recently officially launched our solution – we call it SEAL – and we are very pleased with the demand we are experiencing for it,” Degani said.
Many larger enterprises maintain an executive position of Chief Analytics Officer (CAO), responsible for predictive analytics. For enterprises that do not have a dedicated CAO, research and information departments of the organization takes care of such predictive analytics work.
Enterprises conduct the predictive analytics activities in cooperation with the strategic business and marketing decision makers due to the significant contribution that predictive analytics offer to the core business activities.
Speech-based predictive analytics vs speech analytics
Typically, speech analytics is used for speech recognition in order to spot keywords during customer-agent interactions at call centers. Such analysis is focused on sentiment analysis and aims to detect mainly customer dissatisfaction and other call related events. This type of speech analysis can only resolve immediate customer service issues taking place within the call center.
Speech-based predictive analytics has a much wider scope and focuses on customer tendencies and behaviors. By focusing on analyzing voice-based customer interactions, technology companies such as VoiceSense can make a highly accurate prediction in real-time during the course of a phone conversation with a customer.
Recommendations to CIOs
CIOs and IT heads should select a platform that offers actual, clear and well defined predictions. Many of the platforms offer mainly the infrastructure to collect data and perform the analytics, but may not offer capabilities for predictive analytics.
IT decision makers should select a system that offers a range of prediction types that are relevant to business objectives. CIOs and IT heads should make sure that the system supports the type of predictions and analysis methods that match their operational needs.
CIOs and IT heads should aslo select a platform that demonstrates proven, accurate validated predictions. The predictive analytics promise is very big, but actual proven results that improve the business bottom line can be elusive.
Image source: ABB