* Editor's note: This article is co-produced by Sukhpreet Kaur
How do we communicate today?
“According to research, face-to-face communication is conveyed with 55% non-verbal, 38% vocal, and only 7% words in the words spoken.” – Albert Mehrabian, Professor of Psychology and a researcher of body language.
In the American TV series, Lie to Me, which ran from 2009 to 2011, the premise of the show is that our facial and behaviour cues reveal more than the words we say or the body language we show.
In the paper, Stress and Deception in Speech: Evaluating Layered Voice Analysis, the authors posit that human oral communication contains features, which can be used to provide useful information about a speaker apart from the linguistic content, or meaning, of what was said, and that this indexical information can prove exceedingly helpful in forensic work.
Enough research let’s get down to the real world. The application of voice or speech analysis has many interesting applications and use cases, from talent acquisition and management for HR managers to criminal investigations by enforcement officers, and even in the quest to better customer service in call centre operations.
Layered Voice Analysis (LVA), as developed by Israel company Nemesysco, uses a “proprietary set of vocal parameters found through research to correlate with key human emotions, and in various combinations to be able to identify deceptive intentions in “real life” scenarios.”
According to James Ellender, CEO of Behavioural Cues, the easiest way to explain LVA is to look at behavioural detection and behavioural cues as part of behavioural science.
“As humans, we will display certain behavioural indicators in very different ways when we communicate. The layered voice analysis and technology detect 151 biomarkers within the voice to identify genuine emotional responses."
James Ellender
Behavioural Science is the study of when and why individuals engage in specific behaviours by experimentally examining the impact of factors such as conscious thoughts, motivation, social influences, contextual effects, and habits. Several disciplines fall under the broad label of behavioural science, including Anthropology, behavioural economics, cognitive psychology, consumer behaviour, social psychology, and sociology.
Use cases
Setting aside the academics and focusing on commercial and commercial applications, Ellender explains that the application of LVA technology is mostly around risk mitigation and protection.
“A 15 to 20-minute interview over the internet or in person where a series of questions are asked for employment assessment, the technology picks up if the responses are genuine. Insurance companies globally use voice analysis to help detect fraud and credit risk in the underwriting process. This assists the banks during finance and loan applications. It is also widely used in call centres,” he elaborated.
Required expertise for LVA adoption
Ellender stressed that the hardware technology needed for LVA is minimal as is the software integration. The LVA technology, as developed by Nemesysco and marketed by Behavioural Cues, can be installed onto a laptop.
“On a larger scale, it needs to be integrated into the current operation. In the majority of cases in the commercial sector, it only takes an hour or two to deliver the training to the persons that are going to be operating it,” he added.
Prerequisites for use of LVA technology
Ellender explains that as part of the work at Behavioural Cues, the company assists customers, for example in employee screening perspective, to conduct anonymous interviews on behalf of the recruiter or the HR department. We then analyse the results and provide the report to the client.
He stressed, however, that LVA is just an additional tool. “It is not there to replace what the companies are currently doing. It is to give them further insight,” he insisted.
LVA integration challenges
New technologies almost always present challenges for organisations not only to understand the value of the innovation but its application and benefit to the company.
For Ellender one of the biggest challenges is having trust in the process.
He cites the case of a typical financial institution that has many established processes in place to run different aspects of the business. These processes have been established and accepted over long periods, and to introduce a new way of doing something, for instance, LVA will almost always introduce scepticism, as is natural.
Ellender explains that “you are looking to integrate, in this case, LVA, into the current systems so you just have to project manage it through that process.”
The good news, when it comes to the LVA, is that it is language-independent, it analyses the responses in any language. “It is not analysing the words that are said, it is looking at the biomarkers within those responses,” he continued.
Privacy considerations
Ellender is confident that the application of LVA itself will not require changes in an organisation’s existing privacy policies and practices.
“We are not trying to push the boundaries in respect of personal data. All the policies and procedures exist within organisations now, and the integration is quite seamless. For instance, call centres have disclaimers that calls are being recorded and will be monitored for operational and training purposes,” he elaborated.
Click on the PodChat player for a deeper understanding of LVA and its applications in your industry.
- Define layered voice analysis (LVA)? How does it work?
- In what areas of business is LVA being applied today?
- If an organisation intends to apply layered voice analysis, what technologies or processes need to be introduced? What expertise is needed to make LVA investment effective?
- What are the integration challenges that come with LVA?
- Any potential privacy issues?
- How does an organisation determine if LVA is what they need?