By: Rajneesh De
In an exclusive interaction with Voice&Data, Yoav Degani, CEO, VoiceSense, talks about VoiceSense’s offering differ from typical speech analytics and how does a service provider or network operator benefit from your technology. Excerpts:
Q. Why do you think predictive analytics technologies are important for today’s market scenario?
In all sectors, enterprises are always looking for ways to make sounder decisions. Predictive analytics is a core tool that nearly any commercial enterprise or non-commercial organization can use for making better decisions with its data analysis budget. Predictive analytics provides the ability to forecast user behaviours and leverage insights to identify opportunities and issues. It can also help businesses to take specific actions to meet their goals.
Q. How does incorporate speech into predictive analytics improve upon existing data-driven approaches? How is this bringing innovation to this segment?
Typically, data-driven predictive analytics relies mainly on demographics and history – data on consumption history, health history and so on. For example, a 40-year-old customer with a certain socioeconomic status living in a certain area with a certain family status is more likely to purchase certain products. The enterprise can generally use this predictive analytics process for marketing segmentation purposes.
By using speech analysis within predictive analytics, organizations gain much more information, enabling more accurate predictions. For example, a prospect tells a customer representative that she has certain interests. By applying speech analysis to that conversation, you can highlight specific products that this prospect is more likely to purchase.
This approach is much more focused as well as innovative and practical as speech analysis is still mainly used for analyzing customer experience rather than predictive analytics.
Q. How does VoiceSense’s offering differ from typical speech analytics technologies used by telcos in their customer service operation?
It’s important to differentiate between speech recognition and speech patterns personalization. Speech recognition refers to what words people actually say–this is conversation content. This already has become a rather common value proposition for telcos.
However, predictive analytics powered by voice refers to how people sound. This is a highly unique approach. It points to the typical behavioural tendencies and especially to typical consumer behaviours as reflected in the customers’ speech patterns.
Our technology leverages these non-content voice patterns. These patterns are called the prosodic aspects of speech. They are physiological by their nature, which therefore means that they are language independent, highly accurate, objective and difficult to control or manipulate by the customer.
With this in mind, VoiceSense presents the telco marketing and customer service operations with real-time go/no-go upsell and retention indicators by assisting agents and representations while there are on the phone with callers.
Q. How can a service provider or network operator benefit from your technology? What kind of value is being added by your solution?
For service providers or network operators, the customer service call centre places a severe drain on expenses with little to no profit to show for it. VoiceSense transforms the customer service centre into a revenue generator: agents use VoiceSense’s predictive speech scores to personalize offers, selling customers what they want, when they want it. They use our technology to enhance the customer experience, alerting agents to caller displeasure or annoyance and suggesting approaches to alleviate issues and prevent churn.
Within the organization, we also track agent performance automatically, analyzing and scoring all calls as opposed to the existing practice for manually reviewing calls. We mark conversations with significant customer-agent events and marks significant moments within the call. The technology provides a range of scores, characterizing the conversation atmosphere and labelling the attitudes of both the customers and agents. This boosts training effectiveness, root cause understanding and operation management.
What additional applications for your technology are there in other market spaces?
The range of markets and use cases for our technology is very wide. Because we specialize in personal behavioural classification and almost any aspect of life – work, health, entertainment, social – is strongly influenced by people’s behavioural tendencies.
We offer a solution for HR, where companies can automatically screen out up to 70% of job candidates in one go. We also have a solution for finance, where institutions can determine the loan default risk and debt collection success probabilities. In the healthcare space, we allow the mental health states of patients to be monitored remotely with caregivers getting alerts to potentially dangerous situations for prompt action. We also have a use case for personal assistants with applications for smart-speakers, the automotive industry and mobile phones, where users can receive personalized content based on their tendencies and state of mind.