Updated: Sep 6, 2021
Welcome to "AigoraCast", conversations with industry experts on how new technologies are transforming sensory and consumer science!
AigoraCast is available on Apple Podcasts, Stitcher, Google Podcasts, Spotify, PodCast Republic, Pandora, and Amazon Music. Remember to subscribe, and please leave a positive review if you like what you hear!
John Fuisz is the CEO of Veriphix. John has a diverse background, with an undergraduate degree in Physics from Georgetown, a JD from Catholic University, a Legal Masters from George Washington. John was previously a board member at Fuisz Technologies, a drug delivery company that helped originate soft chew vitamins. During that time John was also an IP litigator and consulted on IP strategy. Now, at Veriphix, John studies belief change in population with applications to market research.
Transcript (Semi-automated, forgive typos!)
John Ennis: John, welcome to the show.
John Fuisz: Thank you very much for having me, John.
John Ennis: My pleasure. Yeah, very interesting person and I think we have a lot of common interests that I look forward to talking about. So, for people who maybe aren't familiar with you, your work, and your company. I think it'd be good if we could just start with the story of your journey on how you started off. I mean, I guess like me, it kind of quantitative background, but then you went through this whole legal path. If you could take us through your journey, I think it will be a good place to start.
John Fuisz: Sure. Happy to because it is a slightly funny journey, but it was all-important, I think, to get me where I am today. And that is basically with my starts with the physics degree on some level. And that gets me used to quantum mechanics and dealing with noise and probabilities and then the first practical application really was Fuisz Technologies and Fuisz was a drug delivery company. Delivering drugs is no different than delivering food which is where we started which was essentially putting products into cotton candy machines started with oils. So oil that got encapsulated into the sugars incorporated that into meat to provide mouthfeel and to provide flavoring, et cetera to foods. We had a chewing gum division. We then moved ultimately into then drugs, fastest is drugs, et cetera. But what I took away, I think most what most impressed me from Fuisz was the placebo effect in how belief has a physical impact on people, right? And actually changes how they experience things and what they believe experiences, impacts their selection experience, value perception. So I took that knowledge that over to the legal career. So for twenty-five plus years, I was an intellectual property attorney dealing in large litigations, and convincing someone that something tastes good or worked is no different than convincing a jury. So it's the same type of process in terms of people's initial belief about something about a person impacted their experience. So apply that to juries, strategy and licensing, etc. Ultimately got tied up in the misinformation, disinformation debates and found myself in Veriphix understanding how to play with noise. I think over that time period I was fortunate enough to meet a lot of famous people who really understood noise and how to use it.
John Ennis: I see. And so, okay, well, we kind of went fast forward through your career there. I just want to make sure I understand all the different pieces of it. So Fuisz originally, the Fuisz technologies that was your father's company, is that correct?
John Fuisz: Yeah, correct. So that was founded back in the early 80's. He was literally playing with a cotton candy machine. So I would go out and I would purchase cotton candy machines from carnival supply stores and he would come home and start dumping things and he break them. And that all kind of becomes that entire kind of fast dissolve area that really didn't exist prior to Fuisz.
John Ennis: Yeah, that's strange. I work from my father's company also. Actually, that's where I got started out of my postdoc, so I have some good experience with him. So you went to that on your legal journey and you started to realize that a lot of the ideas from the food world, the placebo effect also especially if you're looking at the interaction between pharmaceuticals and food, which, of course, I mean, consumer health care is a big area. You were influenced by that, now, the misinformation piece you mentioned, can you tell us a little bit about how did you get caught up in that? So you're doing your IP work and you're working as a lawyer. So what kind of drew you back into looking at behavioral psychology?
John Fuisz: Because it's the same process. Actually, it's no different. Influencing a jury is no different from influencing the population, no different than selling a product. You have people sitting in the jury box that you know nothing about your base demographics. You have to decide what trial team you need. You need to decide what they believe, what facts you need to emphasize, what information you need to communicate to them. And it's an identical process. And so you have two hundred plus years of law, of federal rules of evidence, of everything that actually quantify some of the basic elements of human behavior. What should we trust that really gets to what's admissible, what's not admissible? And so by having that foundation originally Fuisz and asking why and trying to connect the dots. In a deposition, why do certain questioning work and elicit a confession? Why does a certain argument work and get a judge on your side? And I skip over a fact. If you believe strongly in my client, I used to call it John's five free facts. I find that we basically could ignore five logical inconsistencies in a movie, perhaps, and people never notice because they would believe the original brand. So it was all tied together. I did some work on disinformation. When the Libyan government film represented one of the groups there on disinformation and so when the US government had its issue, it was honestly it was a somewhat simple process, somewhat simple problem to solve because it was simply a question of not treating them as an isolated problem, but looking at it from a much broader standpoint. It's no different than changing, getting someone to believe something tastes good. Same process. Feeding disinformation, misinformation to get the brain to taste something in a different way.
John Ennis: Right. Okay, I don't know how much you can talk about that word and I'm sure some of that is sensitive, but it is quite interesting. So your transition then through looking at misinformation and more of a political context to looking at it in a marketing context, is that correct?
John Fuisz: Because the best, we quickly learned was that the market for the ability to kind of detect and understand what was happening in the misinformation, disinformation space was a defensive work, really wasn't there? And to really accelerate technology development, you had to work on the commercial side and turn it into a commercial tool because what commercial advertising marketing is, it's no different than the flip side of misinformation-disinformation. Is Coke way better than Pepsi? I mean, it's a matter of preference on some level, but creating that preference. And so we created the essentially a tool that brands and agencies and researchers can use to discreetly and privately measure believe, change, understand that we can automatically extract the nudges that change belief and then test that. And it's a new level of understanding of how you can influence groups.
John Ennis: Right, and I think influence is the keyword, right? Because the disinformation campaign, that's an attempt to influence advertising campaigns that influence. So can you talk a little bit then about these sorts of measurements you're making, because I think there is a temporal aspect to your measurements that are quite interesting? And then how do you leverage that information to provide recommendations to your clients for how to influence their populations of interest?
John Fuisz: So our system is definitely unique and that we have to collect data in a very specific way and what we are focusing on is in terms of we'll take the same panel and have them interact with our system, answer the same questions. Each week on the same day, over 4 to 5-6 weeks in a row, that gives us then a forward-looking time sequence that actually is very important, because what we're looking at is we're basically calculating the occurrence error that's occurring from week to week which is where then we start to extract the belief that nudges and we're able to do some testing. So it's slightly different from a single survey which gives you a single point in time. Or someone gives you their aspirational answers. We have to break through that. Trying to get an unfiltered response and then look at on time sequence to extract, started to extract meaning from the noise.
John Ennis: Okay, and so then put that to be the kind of real for our listeners here, you might have a client that you're doing this kind of work for. So is your questionnaire then geared to the product category or you try it, this is general attitudes and beliefs? What sort of questions are you tracking over time?
John Fuisz: So we're able to use up to 15 questions at a time. At that point, we start to lose the panel member and we need a sufficient number of questions to essentially get them induced into the system. So we have a free flow of information out of their brain. So we use 15 questions, all related to the underlying topic. We always baseline. So for Nike, for instance, I think I can talk about we did looked at women in sports, right? So there would be 15 questions associated with women's exercise sports and people's general perceptions. Testing when you have the 15 questions and you're looking at that data, then over a series of weeks you want to start to pick related issues that you think either should track together or track an opposition to determine if your underlying suppositions are actually correct. If the economy is getting better, will people buy more shoes? Or is that not actually tracking together when you start to look at overtime series. And you start to tease out some of those relationships and understand really what's driving people and why their behavior is changing.
John Ennis: And so this information, though, I suppose is very rarely time-sensitive, is it? Once you get some information, then you've got to act on it fairly quickly. Is that the case?
John Fuisz: Yeah. It comes out in two weeks so when it comes out of our system, it's all processed by software so that our clients have it roughly within an hour or two after the collection window ends each week. There are certain issues that will change dramatically week to week. There are other issues that we have a slower trend. So we've gone back. We were looking at misogyny, especially as it related to women, and we've gone back two to three times and run those same questions a year later. There's not a huge amount of movement, unfortunately, in the population which you would hope that people would be more open to change. But on some of these issues, they simply don't change that much. When we were looking at travel and covid this last month, there are some issues that are changing dramatically week to week which you can imagine with travel. No one wants to travel this week and the next week, to buy their ticket.
John Ennis: Right. There's some announcement or something that happens.
John Fuisz: And it could run and it can be an announcement and must just be a background issue which is what most of the things could change on that background noise that impacts people. Right. So we're looking at the travel data. We had a downward different people's view of travel, but it didn't occur to us at first. It took 24 hours, unfortunately, realized, but it was because it was tax day and Tax Day wasn't in April was in May, and then also that was tax day it made sense that everyone just paid their taxes. And so there was a negative movement that week. Our perception of spending rebounded the following week. But if you were going to send out your ads or are trying to get people to travel, that was not the week to do it because everyone was pessimistic.
John Ennis: Right. And Bitcoin took a dip around that time, too.
John Fuisz: Right. Bitcoin is exact same thing. It's just a question of, bitcoin is what fear hedge plus playing off some of those other occurrence noise type biases that status quo that if you can track you kind of get a sense of what's happening.
John Ennis: That's very interesting. Okay, so let's talk a couple of things I'd like to kind of dig deeper into noise and nudges. Right. So let's start with nudges because this is, I think a very prevalent idea in behavioral science. Can you talk a little bit about what is a nudge and then are you providing recommendations to your clients about nudges they can make? Are you looking more at the reaction to the population to being nudged like how do nudges factor into your research?
John Fuisz: So we identify issues that if amplified will actually move the population.
John Ennis: I see.
John Fuisz: And what we have uncovered is basically once we can identify belief change week to week. Place our panel members than on that spectrum panel where they sit relative to each other and we can then identify the periphery of the group. The periphery of the herd, right? Because real change occurs at the periphery. Someone who is in the very center of a group is not the one who is going to change the group. It's going to be the periphery. Things migrate. New ideas migrate around the periphery where people are more risk-averse and then they actually migrate to the center. So we look at the periphery of the groups and so we kind of apply the information theory. We're looking at resonant issues that keep appearing across those 10-15 questions on the outskirts of the periphery. Issues that those people are seeing, justifications that those people have that are different from the center. So those are those that become our nudges. We get both positive nudges and negative nudges to move people. Negative nudges would be things you want to avoid, right? Positive nudges are things you want to include either in your messaging or have your message next to it. Those we can then test we can do an ad test to actually quantify the impact of a nudge. And when we do a raw nudge extraction that first week, they tend to test, well, 60 to 70 percent of the time. They're not one hundred percent. I wish they were, they're not yet, but we can do that subsequent testing and quantify. So then people know, okay, I have issues A, B, C, D. And then there was going to be time to use. Nudges then do become time-sensitive because facts change.
John Ennis: Right. So these are basically potential points of influence, ways that people may be open to being influenced starting on the periphery, and then if you can move those people, then the movement might carry through the whole population.
John Fuisz: Yeah. When we were able to move a population with a nudge. I can't tell you who in the population will move. I can't target an individual. I simply can tell you what level of increase we will see because that issue resonates with a large enough group to actually effectively move the population.
John Ennis: Right. And you shared with me some of the travel results. Is it okay if we talk about this?
John Fuisz: Sure, absolutely.
John Ennis: I thought it was quite interesting, some of the stuff you were sharing with me, for example, some of the information you're extracting around the use of masks and pictures. I thought that was very interesting. One of the things I like about the research is that it was didn't seem to be normative in the sense it wasn't prescriptive about how the way the world should be was more about the way the world is and that people don't want to see masks and pictures when they're thinking about traveling. That's not, that it wasn't positive for people to see those pictures with masks.
John Fuisz: But it's somewhat, it's counterintuitive up into a point until then you realize and it's like that makes total sense. And essentially, people are scared.
John Ennis: Right.
John Fuisz: So telling them you have a clean hotel does not help. I mean, that's cleanliness was a negative notch. So you say that we are doing great cleaning our planes. You're actually subconsciously reinforcing the fact that there is danger if it's not clean, oh, my goodness, am I going to die?
John Ennis: Yes.
John Fuisz: Right. And so masks, et cetera. I mean, things like that actually are counterproductive. What we were then based on the positive and negative nudges out of travel. What we were recommending to some groups is get covid off your web page.
John Ennis: Right.
John Fuisz: Put it into a concierge's section. Make it as a concierge that's normal. This is just another health issue. Otherwise, your masks, you have your covid warnings. If someone wasn't sure they wanted to travel before or after you just explained to them how great you're going to do it, keep them alive in this incredibly dangerous time. I mean, there's no way they're traveling.
John Ennis: Right. Yes, I really thought that was very interesting because it really seemed like it was really data-driven. I like that it wasn't about how or might think the world should be. It was like, well, this is the way it is. People don't want to see these images. And so, yes, maybe things have to be clean. You don't want there to be something bad that happens. But at the same time, talking about it all the time is not actually helping people come to your hotel. So I thought that was very interesting.
John Fuisz: I thought one of the other interesting, one of the positive nudges we had were certainly the outdoor mountain breezes, right? And for a hotel or someplace to then show pictures of their outdoor patio. That's a way then to subconsciously address that fear if there's going to be some great fun places to hang out. And if you're scared about being inside, don't worry. There's some great places to be outside. But we don't have to talk about how scared you are.
John Ennis: Right.
John Fuisz: You know, just you have a name-brand hotel. You can trust that they're clean, beautiful outdoor spaces. If you're scared about being inside and never, you can deal with that whole safety issue. Covid fear with never actually talking about it which is what the nudges were for we're coming out with at the time.
John Ennis: Yeah, now, I thought that was really insightful and I'm sure very helpful for your clients and for anybody that had access for anyone for whom that information was relevant. Okay, so let's talk a little about noise because it's something else that we're both interested in. I mean, we've talked a little bit about my work in signal detection theory and how I mean, for us in sensory like pure sensory, you have product testing, you're trying to see change integrating. Maybe you actually did this when you were at Fuisz where an ingredient is changed and there's a question of whether or not the product has changed in some meaningful way. And signal detection theory can help you with that where you can see what's the difference relative to the noise and the responses. So can you talk a little bit about noise in your research and how you're leveraging it?
John Fuisz: Yeah, because, look it's the same issue when you're measuring believe change week to week. What we do not have control over is what everyone is exposed to. There's going to be that natural background noise.
John Ennis: Right.
John Fuisz: And so recognizing there's the natural background noise and then trying to extract the signal that's sitting on top of that. That's the challenge. And I know we both are fans of Claude Shannon. One of my early cases I got to work with. It was on the DeRosa Rogoff and I got to work with Mr. Rogoff, who was the inventor of spread spectrum technology in CDMA where he was modulating the signal on top of noise and then applying a counter version of the noise to then extract the signal which is the basis for how our phones work today. And it's the concept that even if you don't have the exact copy of the noise if you have a rough approximation of the noise you can still extract enough data on the backside which was the reason we went to this time sequence approach. Once I have a time sequence approach, I can access the noise. If it's a single point, a single survey that's done on a single day, you have a much harder time extracting the noise that exists, right?
John Ennis: That's really interesting. So in your research, then, are you tracking individuals through time, or is it just the population to the time? Each individual through time, but then the result is that the population, is that correct?
John Fuisz: Correct. It's both in some respects. Our panel members stay the same for the entire process and we're tracking each of those individuals through time. But then we're using the wisdom of the crowd approach to them to approximate what's happening in the population off of that group.
John Ennis: Right.
John Fuisz: Right. So, some of the crowd approach, you want a group of 25 plus on some level, I can have a rough sense of what's happening. If I have a sufficiently diverse panel and they're not in communication with each other, I can have a rough sense of what's happening in the population that lets me understand individuals over time, population over time, individuals relative to population over time. That's then where we start to find our outliers and that's where we start to then find our vendors.
John Ennis: Right. Yeah, it's fascinating, now, I think it's really good work you're doing, and I think that it would be good for any of our listeners to check out your website. And so Veriphix.com, is that correct?
John Fuisz: Veriphix.com. The website hasn't been updated recently, so it is always the life of a startup. Some things always fall to the back of the pile, but Veriphix, we are always looking for interesting projects or pilot tests. I think there's a lot from a research standpoint, there's a lot that this tool can do. I think to help researchers on some level. I think that's really we have a base application where certain clients are using it for certain marketing, R&D, PR purposes and there's some government applications. But I think from a research standpoint of if you add belief and believe change over traditional research gives you another data set and actually may help extract additional insights from some of the based tools that otherwise wouldn't be obvious.
John Ennis: Yeah, and I think it's really interesting. And I like the combination of kind of behavioral science, cognitive psychology, information theory, the data-driven approach I think is very good. Okay, so we've got a little bit of time. I do want to ask you about a few topics because I know we have a shared interest. And in fact, you and I think maybe your father has been involved in cryptography and been involved in some of maybe the kind of pre Bitcoin work on encryption, can we just talk briefly about it because it's interesting to me. So what is your background in the kind of cryptography space and the contributions that you've been involved in?
John Fuisz: So my father has been outed as being part of the CIA and over the years he'd ask for various technologies, one of which was an anonymous way to transfer currency over the Internet. So that was 98 I believe that I created the E-currency, it was dual-key encryption, a public log-based transaction system that actually we filed that the week before the Bitcoin website gets reserved and Satoshi starts and then Satoshi is in Japanese at Central Intelligence for that application actually gives rise to the pop culture rumor that Bitcoin was a CIA operation.
John Ennis: Right. I've heard that quite a bit. I'm not clear why the CIA would want to do it. But it's interesting.
John Fuisz: Bitcoin was, the concept of an E-currency was interesting because it was you have the Internet now, and if you want to buy something that's risque or something else, you really can't without leaving a trail. Right? But if you want to buy a Playboy magazine, you can go down to the 7-Eleven and buy one given cash and no one knows you bought it. But you couldn't do that over the Internet. And so it was a way to then create something that would allow you an anonymous way that allow you to essentially do transactions over the Internet. The problem was always what level of security do you need that little file on your computer is not going to get stolen or is not duplicative, not spending the same money two, three, five times which was a transaction-based thing. It was never meant to be one hundred percent secure. So once again falling back into this concept of a certain level of noise is okay. You don't need a system that's one hundred percent secure. A dollar bill certainly is one hundred percent secure because dollar bill could be thrown on the color printer and you could actually make one and just someone may realize it eventually, but didn't need to have one hundred percent security which is what that allowed. So it was interesting because it really then becomes a fear-based hedge. People don't trust the government. They don't trust their country. So living in China or someplace and I don't want to have my currency not controlled by the government
John Ennis: Right.
John Fuisz: It then bridge was crossed borders, but then it also played on even though it has no intrinsic value, necessarily plays on some other heuristics. Well, someone will suffer two dollar loss for every one-dollar game. Someone will buy it and hold it to status quo. So it had a way to create value and create the perception of value. That really have any value, it has a cost to produce.
John Ennis: Well, I think the service of transferring money through, internationally, that's something valuable to be able to transfer something through.
John Fuisz: Yes, a hundred percent but the person who is going to use Bitcoin is the person who doesn't want to use the bank.
John Ennis: Yeah. Well, it's interesting, you know we pay quite a bit in wire transfer fees when our clients pass, right. And the idea of being able to have that, and that's the idea of decentralized finance. We're getting a little bit off-topic. I am curious. Do you think there's any application of block change? This is something I've been thinking about quite a bit. Is there any reasonable application of blockchain to market research? Maybe verifying that panelists are who they say they are? I mean, have you thought very much about this in terms of the use cases for blockchain, maybe the supply chain? Obviously, it has some value. Do you come across any reasons why you'd want to supply blockchain in our line of work?
John Fuisz: So the end of the day, the short answer is yes. At the end of the day, blockchain is nothing more than a verification process.
John Ennis: Right.
John Fuisz: Right. And so the good old days before it was so easy to verify that our credit card was good, you would wait in line and someone without one of those huge books with 600 pages and they'd look at your credit card number and they'd see if it was stolen or not. Right? And even then, they weren't totally sure and they would get credit cards first came off. So blockchain is a way to quickly authenticate someone. So from a research standpoint, especially now that we're moving this work from home and remote, verifying that people are real, verifying who they say they are, or making sure I don't have the same person over and over again. There's so much gaming on especially on the survey side. Bots that are just blasting off random survey responses that's poisoning the data. And you get into misinformation, disinformation and you get into an algorithm and the move for the government to artificial intelligence and the ability to poison data.
John Ennis: Right.
John Fuisz: I mean, that's real. And so the way that's the government's data being poisoned, whether or not that's market research being poisoned, I will say there's not a bright line between what's government and what's commercial anymore for two or three of these projects. There are active campaigns between countries on these same topics. Travel effects. US business travel alone is one percent of the worldwide GDP. It can affect a local country's GDP by four to 10 percent. Yeah, scaring people not to travel can destabilize the country.
John Ennis: Right. Well, I've seen that countries in the South Pacific are really getting hurt like some of the poor, more poor countries that depend on tourism are getting totally hammered right now and could be destabilized like you said.
John Fuisz: Yeah. So even US companies operating overseas, we've seen one or two situations where there was an active campaign by another government against a company simply because they were competing then likewise with the company. But that all then get back to the research and what we're doing for research and moving these companies forward in such which is the importance of the blockchain and making sure that things the data you are using is decent data.
John Ennis: Right now. I think that's a huge issue with it. Like you said, with people at home filling out surveys. You know, my wife's a professor and she ran a survey not too long ago and had to deal with the kind of scourge of all these spammers coming to, yeah, I mean, it does poison the data. She learned a lot about data-clean like there's a whole field devoted to, well, I mean, it's probably related to misinformation but there is a whole field devoted to verifying survey data. And it was quite interesting for her to learn about that. But it would be better if everybody that took the survey was, in fact, verified in some way and there was no chance that you were being deceived when you're putting a survey out there.
John Fuisz: Yeah, I was just watching a video earlier today. It was the woman's name is escaping and she does a Google analytics team, but talking about data and bad data and outliers in your data and then trying to figure out what to do with them. So once you have all the spammers in and my artificially cutting the data out, do I really know that it's wrong? Even in our data, you know, we will see people that give us answers on the extremes and we tell people not to and you really should never or will find some of the types in the same thing in terms of their free-text answers as it's hot. We don't remove them because on some level there should be a counterbalancing effect of other people that have maybe not a single person, but maybe five people have a negative bias and it should work its way out. But we will test our data removing people and saying does it really something?
John Ennis: Yeah, that's right.
John Fuisz: That's a huge issue.
John Ennis: Well, John, we actually are out of time. I can talk to you for quite a bit longer with many common interests. We can talk about Bitcoin for an hour.
John Fuisz: Maybe we should some time. Get a look at Bitcoin pros.
John Ennis: You know, I went to the Bitcoin conference. It's quite interesting.
John Fuisz: It looked like a scene.
John Ennis: It was very fun. It was mainly outdoors, like a state fair. I had a great time. It was really enjoyable. Yeah, it was pretty wild, but it was fun. Okay, but let's just wrap up here. So how can people find you? We mentioned Veriphix.com.
John Fuisz: Correct.
John Ennis: Okay, and what about LinkedIn, Twitter?
John Fuisz: LinkedIn is good. So it's John Fuisz. I do have a cousin named Jonathan, but you will see that only one of us has a Veriphix graphics. But I'm always on LinkedIn.
John Ennis: We will put your LinkedIn on the show notes. Okay, and then last bit of advice, we always like to get advice for young scientists. What would be your advice that you would have?
John Fuisz: I would say for young scientists, stay general. I mean, I was just talking to someone about the beauty of physics and that there are all the physicists think they're geniuses and they can jump in anyone's field. And I said because we're masters of nothing. Right. We understand a little bit of everything, but we really don't understand anything which so much of what I've learned over my career. It's the bits and pieces of food, the drug, the law to juries. So much of it is related. And having that breadth of experience and not shutting doors, especially for younger people, some of those doors open or even if you love your science classes, maybe take an occasional poetry class just simply because you never know what that inspiration is going to come from.
John Ennis: And I do agree with that. And there's always, you know, the 80-20 where like just a small amount of learning in any field can usually pay a big benefit if you don't know anything.
John Fuisz: Absolutely. 20 percent of the effort can get your 80 percent of the output. And that's so huge.
John Ennis: Okay, John, it's been a pleasure. Thank you very much for being on the show.
John Fuisz: Thank you.
John Ennis is a leading expert in sensory science and consumer research, with extensive experience in statistical analysis and product development methodologies.
