Leveraging scientific Speech Analytics in business
Voice technology and speech analytics is a trending topic for many businesses today. Companies are beginning to realise that the only way to drive real results in their organisation is to gather as much data as possible – including the information usually locked within the phone conversation.
Voicesense is an innovative speech and behavioural analytics company capable of analysing various markers within a person’s voice to create customer profiles and even forecast future behaviour. The applications of Voicesense’s technology can extend across multiple industries, including finance, healthcare, and even HR.
We spoke to the CEO and Founder of Voicesense, Yoav Degani, to find out more about how their pioneering behavioural solutions can change the future of the voice conversation.
Voicesense is a company dedicated to the concept of behavioural predictive speech analytics. According to Yoav, while most speech analytics focuses on speech recognition and content, Voicesense focuses on the acoustics of the speech, including factors like pace and emphasis.
“We offer emotional analytics, but we don’t just focus on the current state of mind for the speaker. Instead, we develop insights into the personality of speakers based on their speaking acoustics. We can use the information we gather to provide real personalisation for customers, predict future behaviours and develop in-depth user personas.”
The technology available from Voicesense gives companies an automated framework that they can use to predict the actions of customers, improve upselling opportunities and even enhance satisfaction. The AI-driven personality profiles built by Voicesense create personality profiles and predictive scores via the analysis of over 200 prosodic parameters
Degani told me that no other company he knows of provides the same level of behavioural analytics that Voicesense can offer. The technology creates a personality profile and predictive score through an analysis of over 200 prosodic parameters in speech, including things like intonation, emphasis, and pace.
“It took us years to find parameters that remain stable across geography, language, and culture. The number of combinations we can generate is endless”
With machine learning and AI, Voicesense can create customer profiles without knowing the content of conversations, or anything about their customers. “After years of tracking emotions in speech, we’ve noticed that some people have repeating, typical speech patterns. We’ve used this to build an objective insight into the people most likely to adhere to certain kinds of behaviour.”
Yoav told me that the technology removes the prejudice and stereotypes from customer profiling and looks at hundreds of highly scientific measurements instead. Their automated solution is available for enterprises to access securely on-premise or in the cloud.
Voicesense sees opportunities for growth in a range of markets, starting with call centres in the enterprise, and moving onto healthcare, personal smart speakers, finance and more. “Even car manufacturers with connected vehicles want to be able to monitor drivers and passengers to prevent accidents and deliver better experiences.”
Yoav told me that he has a strong background in the science and technology industry, but he was also a clinical psychologist too. “At a certain point, the two worlds collided into what we offer here at Voicesense.”
Voicesense can even offer opportunities for employers and HR. “We can track employee attrition and the risk of burnout for instance. Several rail companies we’ve worked with have noticed a high degree of churn in their workers. Using voice technology, we can determine which employees are likely to churn.”
When it comes to hiring new employees, the Voicesense technology can look at video interviews and screen out candidates who may not be appropriate for the job in question. There’s even the opportunity to determine which agents in a call centre may be more likely to drive profits and results than others. “In our research, we’ve found that the people we predict would be more able to drive upsells managed to successfully get an additional sale within 80 calls. The people who we determined as having less potential took 300 calls to get the same results.”
AI-based tools like voice analytics have the potential to continue growing and improving through machine learning solutions. Degani told me that AI is a statistical software, and whenever they go to an organisation to create a predictive model that works well, they’re building off the statistics within that business. This means that the AI applies specifically to the experiences that unique company has.
“However, our objective is to find voice tendencies that can predict behaviour everywhere. We want to reach generalised models that remain consistent across different models, industries, and environments. We collect a lot of data from a variety of environments, and we will continue to improve our solution based on what we learn.”
Voicesense also offers its customers an opportunity to keep upgrading their own predictive systems. “We provide a service where we go back to the company every quarter and gather information, then update the models accordingly. The learning process is essential to boosting performance.”
Traditional systems for predictive analytics work by looking at demographics and probability. By adding speech to the mixture, Voicesense gives machine learning solutions more sources of information. The addition of the behavioural concepts offered by Voicesense gives customers insights into their customers, unlike anything they could have achieved before.
“Look at banks, for instance; they have a lot of basic information about their clients, about whether they’ve defaulted on loans and what their credit history looks like. However, they don’t know what kind of person they’re serving, whether they’re impulsive or prone to risks.”
“We succeed with our clients every time by helping them to get insights into behaviour and personality”
Voicesense is also language independent and remote. They can install their system into mobiles, on-premise, or on the cloud. “Because we’re language independent, any customer around the world can be a candidate. Think of depression. This chronic disease takes up millions of psychiatrists’ hours each month. In clinical trials, we’ve proven that we can use our markers to track changes in mental health state remotely over the phone, without one minute of insight from a professional.”
The biomarkers that Voicesense can offer to the healthcare space are a medical breakthrough and a huge differentiator for the brand – and that’s only in one industry. “Almost any enterprise is a potential prospect – every company worldwide. Our business strategy is to take this technology worldwide with business vendors in every market we’re in.”