We spoke with Amir Liberman, founder and CEO of Nemesysco, a leading provider of voice analysis solutions. The Israeli company’s Layered Voice Analysis (LVA) technology correlates vocal data with human emotions for use in assessments and personality tests.
Usually, assessments rely on self-reporting from the candidate, the same person you’re trying to assess. That doesn’t strike as ideal.
Well, leaving aside the concept of misleading answers, people don’t really know a lot about themselves. You need to look at yourself in the mirror in order to know how you behave in a group.
I’ll share with you a story about hiring. When we started to look into the field of personality, I requested that all my team take a personality test. I wanted to see what it told me about them. And from the lot I got three identical reports, identical personalities, the vertical personality types on three different types of people. They were all lions. They were all leaders, they were all thought leaders. They were all pushing people, and pulling people up behind them, leading people to their goals. And these are three completely different people..@NemesyscoLTD's Layered Voice Analysis technology correlates vocal data with human emotions for use in assessments and personality tests. Here's our conversation with CEO Amir Liberman. #HR #HRTech Click To Tweet
Now keep in mind, these are people I know because they’d been working for me for many years. And they’re three completely different types of people. I’m looking at this and I asked them, “Okay, so is this report reflecting you?” they all said, “Yes, all of it. Pretty much so…” “Are you a thought leader?” “Yes. I am a thought leader.” And I just didn’t see that.
So we said, “Okay, you know what? If that’s what it tells a manager, then there’s a significant link missing.” Because these people strongly believed that these were the right characteristics, because they responded to the questions and they responded with what they thought about themselves. But the results didn’t give me the profile that I needed in order to know that I’m making the right appointment—that I could assign the person to a job and know what to expect of it.
You know, there are different roles and different types of people in different areas within the same organization. I want my marketing manager to be very creative, but I want my CFO to be strict. And I want my office manager to be very precise.
There are key factors that I want to know that I can trust in someone. Not because of the way he thinks about himself, but based on his previous record, based on the things he likes to do and doesn’t like to do. Because I know people will normally tend to do things they like and try to avoid things they don’t. You know, maybe I’m the perfect dishwasher, but it hurts my back. That creates stress for me. That’s the type of thing I want to know when I hire people. I want to know their likes and dislikes, and then I can build a team around them.
How does your approach solve for that?
Very easily. We don’t deal at all with whatever’s said. The technology ignores it completely. The questions are phrased to present something as a good quality, and then we ask you to relate to that good quality. Either you’ll relate to it positively or you won’t. We don’t care. We don’t even look at the text. We only look at how this topic makes you feel.
For example, we can ask, “Tell me about a time when you had to work alone, for long hours and manage your own time. How was that? How did you manage to do it?” If that was something you liked to do, you’d respond with excitement. You’d respond “Yeah, that was fun. I was actually making my own shifts.” If it was something that wasn’t pleasant, you might actually say the same thing, but your voice would convey the difference.
Could you explain how voice analytics works?
Yes, sure. Keep in mind that the technology itself was intended to detect risk, to detect inconsistencies and detect fraud. So everything we pick up from the voice comes in very tiny elements that can’t be controlled or normally heard by others.
When I talk, I hear myself, I hear the way I speak. I can act in a certain way. I can suppress certain emotions if I suddenly get too excited. But the elements that we pick up from the voice are actually so small and so tiny, the technology is like using a microscope on the voice. We suddenly see all these micro-organisms that actually make you well or make you ill.
We have 151 biomarkers that we pick up from the voice. We found ways to combine them, then found significant correlations between these biomarkers and certain key emotions. Key emotions such as excitement, stress, confusion, uncertainty, anticipation, hesitation, concentration, mental effort, imagination and anger and happiness.
Now you might think, “Okay, I can hear that with my own ears.” As when somebody says, “I’m very angry,” and you’re like, “Okay, he’s very angry.” But what if he wasn’t angry? What if he was kidding? Or when you meet someone you don’t really remember and you go, “How nice to see you?” You sound very happy, but you don’t really remember who the person is.
Humans use this masked voice, or masked emotions in the voice, to reflect different things. Our technology looks at the underlying layer. We scrape out everything that a person can control, that they can influence and that has to do with the language and the words they choose. We look at the basic ingredients, which are far more precise in terms of understanding emotions.
Let me ask you to sum that up. What are the advantages of the technology over the human ear?
I think it’s the combination that counts. Nothing can substitute for human understanding. Not today, at least, though maybe one day. Understanding the context, understanding the uniqueness of certain events that can only come from the human ear. Understanding emotions—the human ear is very incapable of doing that. The fact that I’m smiling doesn’t mean I’m happy. The fact I’m talking with a low voice doesn’t mean I’m sad. Hey, there are actors who are trained to do this. They fool us by making believe that they feel certain emotions. We, as observers, look at them and think, “Yeah, that guy really is going to save the universe.” But he’s not, right? In that sense, I’m not sure we want to rely on the way people broadcast themselves.
Can you put your technology into context? What role does it play in the world of assessments, and getting more value out of them?
I think that every manager needs to get the best information he can in order to make an optimal decision. That’s the basic assumption, and I think any manager would agree. No manager likes to make a decision based on partial information. The more critical the question, the more critical the mission, the more information you’d like to have. What we provide with our technology is information that was never available before.
Also, we find very predictive values that help you really understand what you can and can’t expect from a future employee. The technology itself isn’t going to say hire or not hire. It’s not going to be in any way deterministic. But it will tell you, “Listen, when you hire this person, these are his real strengths, these are the things he likes to do, these are the things that you can’t expect him to complete well, regardless of the effort.”
I can tell the guys, “Cross your T’s, dot your I’s,” but it just may not be in them. And as a manager, I may have to make the hard decision to take out this person and attach another person to complete a task. But then maybe the creativity of the first person is so strong that it can surpass all of the less-strong qualities of his personality. I need a full picture.
We want to develop technology that can be integrated into any HR procedure. We want to look at multiple partners from multiple fields. I think that what we can offer is, in a way, similar to what exists in terms of its output, but it’s very different in terms of quality and very unique in terms of the information it provides. So, I think we can complement any type of system out there, in any type of recruiting process.
Nemesysco is a leading provider of voice analysis solutions. The Israeli company’s Layered Voice Analysis (LVA) technology correlates vocal data with human emotions for use in assessments and personality tests. You can learn more here.
Sign up for our newsletter here.