Converus Inc.

CEOCFO-Members Login

August 7, 2017 Issue



Q&A with Todd Mickelsen, President & CEO of Converus Inc.



Todd Mickelsen

President & CEO


Converus Inc.


Interview conducted by:

Lynn Fosse, Senior Editor, CEOCFO Magazine, Published – August 7, 2017


CEOCFO: Mr. Mickelsen, what is the idea behind Converus® Inc?

Mr. Mickelsen: Our vision is to create new ways of assessing the credibility of individuals. Credibility assessment is the generic term. The term that people would probably be more familiar with is lie detection. We were seeking an alternative to what most people recognize today as the polygraph device, which measures physiological changes in the body when you’re asked questions. It’s a process that’s very intrusive, expensive and subjective. It has been around for almost 100 years, but it has many challenges with it — because of all the reasons I just mentioned.


CEOCFO: Would you tell us about EyeDetect?   

Mr. Mickelsen: EyeDetect is poised to disrupt the lie detection market — which has been dominated by polygraph for almost 100 years — because it’s a new, completely different approach to detecting deception. It’s the world’s first nonintrusive lie detection technology that accurately detects deception in 30 minutes by analyzing eye and other behaviors. In addition, an EyeDetect test is automated and 100 percent unbiased. EyeDetect’s technology is based on the premise that lying requires more mental effort. If you think about it, when you tell a lie, you’re making something up. You’re balancing in your mind whether or not you’re appearing truthful. Those activities require more mental effort than to just tell the truth. That increase in mental effort is exhibited in an involuntary way through the eyes, the most diagnostic of which is pupil dilation. Basically, when we lie there is an increased mental effort to generate that lie. That causes the pupils to dilate. Although you cannot see the dilations with the human eye, EyeDetect can pick it up with a high-speed, infrared eye-tracking camera. EyeDetect has custom software running on a computer with a high-speed, infrared eye-tracking camera attached to the bottom of the monitor that the individual is looking at. As that person is presented with statements to which they are simply answering true or false, we’re tracking the changes that occur in the eyes, including increases in pupil diameter. That increase in pupil diameter gives us a high degree of accuracy of whether or not the person is telling the truth regarding the true/false questions presented to the person.


CEOCFO: Are they answering verbally?

Mr. Mickelsen: They are simply using a keyboard or a mouse to click “true” or “false” to the statements. These are simple statements like, “I am guilty of stealing from my employer during the last year,” “I am innocent of using illegal drugs during the last year,” or “I have no ties to terrorist organizations.”


CEOCFO: If someone does not think taking pens is stealing from your employer would that show? It just seems like a scenario that many people do not even think of it as stealing.

Mr. Mickelsen: Obviously, to create a lie you have to believe that you’re lying. You have to know that you’re lying and there needs to be a consequence. There has to be jeopardy involved for there to be a physiological change. We launched EyeDetect in 2014. We launched it outside the United States to begin because there are some areas where there is a higher degree of corruption than the U.S. For example, in some countries, to be a grocery cashier, you will be tested for whether or not you’re stealing from your current employer or a previous employer. In the test, we would define what stealing means at the beginning of the EyeDetect exam. Therefore, each exam has a preamble. It uses an audio/visual-based presentation using what are called route maps. This is a proven technique used to present concepts to an individual to improve comprehension of the test topics. In that preamble, we describe what you’re going to be tested on and how it’s defined. This eliminates any confusion. If I’m asking about stealing from your employer, then it might say, “Stealing is defined as taking product or money from your employer in excess of $25 worth of value. It also includes sharing confidential information with someone else that enables them to take product or money from the company, for which you get a kick back or some kind of benefit.” Therefore, you define what it means to steal from the company up front and then the statements are simply based on that.


CEOCFO: What might or might not affect the results?

Mr. Mickelsen: That’s a great question! We’ve tested this in the field. We currently have 350 customers in 34 countries taking tests in 15 or more different languages. What we’ve found is it doesn’t matter whether you’re speaking Spanish in El Salvador or whether you’re speaking Arabic in the Middle East; people react the same. Their eyes react the same. Culture, language, gender, location or upbringing does not affect the accuracy of an EyeDetect test.


CEOCFO: What about medications?

Mr. Mickelsen: Medication would, obviously, have an effect, depending on the medication. In the industry, it’s referred to as a counter measure. Let’s look at polygraph. In the movies, you may have seen that some people will bite their tongue while taking a polygraph. Or they will put a tack in their shoe. Or maybe they will practice trying to maintain their blood pressure, their heart rate, whether or not they sweat. Basically, the test is comparing those physiological changes on one set of questions verses a base line or a control set of questions. In the case of EyeDetect, it’s not based on those physiological changes. Therefore, if you’re nervous in a polygraph, you could appear to be deceptive. In EyeDetect, it doesn’t matter how nervous you are because it’s simply measuring an increase in cognitive load that occurs when you create the lie and when you tell the lie. That’s what is exhibited in the involuntary changes in your eyes. Your pupils dilate ever so slightly. We’re talking less than a tenth of a millimeter. Our high-speed, infrared eye-tracking camera is measuring changes up to 60 times per second in each eye — at up to one hundredth of a millimeter change. Your blink rate changes. Your reading behavior, in terms of how you read the items that are presented to you on the screen, changes. What medications could affect that? If you come in intoxicated, if you’re on other drugs, then that would affect the result. We’ve built into the software measures that can pick that up. They have some people that will go to the eye doctor, for example, and the ophthalmologist will dilate their eyes before they come in for their exam. We pick that up because the pupils don’t dilate as they normally should in the standard times presented.


CEOCFO: Is the result from the machine, does the program gives the results? How clear cut are the results? Are there borderline areas?

Mr. Mickelsen: Most people do not understand this, but in the case of a polygraph, a properly administered polygraph exam is two to three hours long. The questions are generated by the examiner. The examiner interprets the data that is captured by the sensors, and then he or she makes a decision. In the case of a polygraph, on average worldwide, 14 percent of exams are called inconclusive. That means the examiner cannot determine whether the person is guilty or innocent, lying or truthful, given the data collected. Then, outside of that, the accuracy for a generic screening type of exam (which is different than an investigation), if I ask you, “Did you touch the murder weapon,” then that’s a different scenario than me asking you, “Have you stolen from a previous employer during your lifetime.” The accuracy on polygraph is somewhere between 71 and 82 percent for generic screening exams. That is outside of the 14 percent that are inconclusive. In the case of EyeDetect, we allow the computer to present the items and interpret the results, so that the test is a preconfigured, standardized test. If I’m testing a group of people, they get the same questions because the computer presents the items. While the items are presented and the user is simply clicking true or false, we’re capturing all the data. Then that data is uploaded to our secure servers in real time in the cloud. In less than five minutes an algorithm is run against the data and a score is generated. The score is what we call a credibility score. It’s a number on a scale of 1-99, where the closer you are to 99, the more credible you are, the stronger the data that we have saying that you’re telling the truth. The closer you are to 1, the more deceptive you are. We automatically indicate that if you get a 50 or higher, then you’re considered credible. The customer decides which number is credible or deceptive. If I’m hiring two people and one scores 50 and the other person scores a 99, and they’re equal in all other areas, then I would probably hire the person that got the higher score (99). The computer algorithm scores tests consistently for everyone because it’s automated and unbiased. And there are never any inconclusive results. It makes a definitive decision; is this person lying or not? The algorithm assigns the number. How well does it do? Several scientifically validated studies show EyeDetect has a mean accuracy of 86 percent. That means that out of 100 people, statistically it may incorrectly score 14 tests.


CEOCFO: When you are presenting to a potential customer, is there skepticism? Is there an aha moment when they understand and their eyes light up? What do you find?

Mr. Mickelsen: We find that people are skeptical up front. Most people do not think that a polygraph, for example, works either — because of movies or whatever else. That’s clear because they’ve seen that, in reality, a polygraph is measuring physiological changes that you can potentially learn to control. Once you understand that EyeDetect is measuring involuntary responses, that it is a consistent presentation of the test topics, and that a computer algorithm does the scoring, you become a believer. We have what is called an acquaintance test. It’s a test that takes about 8 minutes and allows anyone to experience this technology for themselves. If we were physically together, I could give you this test. Here’s how it works: You choose a number between 1 and 10. You don’t tell me what it is. I never see the number you selected. You write it on a piece of paper so you cannot change it, and then you fold it up and put it in your pocket. This is a test where we basically present a set of statements such as, ”Did you write down the number one? “Did you write down the number two?” “Did you write down the number three?” We go all the way up to ten and back down again. We ask you several times about each of those numbers — one through ten — and we ask you to purposely lie about the number that you wrote down. You basically say, “I did not write down a number between one and ten.” Then, 60 seconds later, the algorithm shows which number you wrote down. It also shows a graph of your pupil dilation as you answered the questions. The test is 96 percent accurate in determining the number written down. Some people make mistakes here and there as they’re answering those questions, so that can affect accuracy. However, generally speaking, that is the aha moment. I’ve tested 20 or more senators, congressional representatives, and senior executives at some of the biggest companies in the world. When they see their number come up, they are blown away. That’s because you don’t sense that anything is happening during the test. You do not sense that your eyes are changing. You cannot feel it, yet you know that you lied about the number. And when the number you chose appears on the screen, that is the aha moment.


CEOCFO: Where does cost come in to play, if at all? 

Mr. Mickelsen: Cost is one of the key advantages to choosing EyeDetect. An average polygraph exam here in the United States at the non-federal level is $400. At the federal level, it’s as high as $1,700.


CEOCFO: Anything the government does seems to cost more!

Mr. Mickelsen: With polygraph, you have an examiner who must get recertified every year and attend significant training, and so forth. They charge the government to come in and perform the exam on the applicant. In the case of EyeDetect, we charge for each test score via our online service. The pre-negotiated government price via the GSA Schedule for a large quantity of EyeDetect tests, is only $99 per test. We charge for scoring an exam. You buy a block of licenses. If you want to score 100 tests, then you buy 100 licenses. Those licenses get consumed when you score an exam. They just go on your account. It’s like a postage meter. Therefore, EyeDetect is significantly more economical than using a polygraph. Plus, it can easily scale. For example, if you had 10,000 people to test — such as visa applicants or refugees — it can be done because the automated test only takes 30 minutes or less. It’s just a matter of determining how many EyeDetect computer stations you need.


CEOCFO: What is the plan to replace lie detectors? What are the next steps for Converus?

Mr. Mickelsen: We believe that we’re certainly disrupting the credibility assessment market with EyeDetect. It’s completely new and unique. Detecting deception in individuals by monitoring involuntary eye and other behaviors has never been successfully done before. EyeDetect is faster, cheaper, better, more accurate and less invasive. However, we also recognize that EyeDetect could be used in conjunction with what is already out there. The benefit is that you could use both polygraph and EyeDetect together to increase your confidence that the test outcome is correct. That is because now there are two completely independent ways of measuring if someone is lying. Our strategy, where polygraph is used, is to show there’s an alternative that gives you an additional way of confirming whether or not you’re getting the right answers. Therefore, we think there is opportunity for EyeDetect to coexist with polygraph. At the same time, we have many customers that fall outside of where polygraph is traditionally used. They are banks, insurance companies and others who would never subject their employees or their customers to a polygraph exam. It’s too intrusive. You get hooked up to many sensors and for two to three hours, you may feel like you’re being treated as a criminal. Whereas in the case of EyeDetect, you sit down in front of a computer and answer some questions for about 30 minutes. Nothing is attached to you. Our plan is to grow the overall market and introduce it in areas where it’s needed to help and ultimately create a society where you have honest, safe environments to work in. Therefore, we’re focusing on this untapped market where polygraph is not being used today.


“EyeDetect is poised to disrupt the lie detection market — which has been dominated by polygraph for almost 100 years — because it’s a new, completely different approach to detecting deception.”- Todd Mickelsen


Converus Inc.



Todd Mickelsen








Any reproduction or further distribution of this article without the express written consent of is prohibited.

 does not purchase or make
recommendation on stocks based on the interviews published.