Small Data in a Big Data World with Jeff LoCastro

Jeff Kavanaugh: Welcome to the knowledge Institute podcast, where we talk with experts on business trends, deconstruct main ideas and share their insights. Today we're discussing human data analytics. I'm Jeff Kavanaugh, head of the Infosys Knowledge Institute. And today we're here with Jeff LoCastro, Founder and CEO of Neener Analytics. Thank you so much for joining us.

Jeff LoCastro: Well, thank you, Jeff. It's a pleasure to be here. And I'm certain I won't forget your name.

Jeff Kavanaugh: Awesome. Jeff L and Jeff K.

Jeff LoCastro: There you go.

Jeff Kavanaugh: Mirror Review judged you as one of the “10 most disruptive entrepreneurs to look for,” which is an amazing statement. Being a pioneer in this field, what is it about human data analytics and small data that excites you?

Jeff LoCastro: I'm innately attracted to things that, and narratives, that no one else is telling, stories that no one else is telling. When people tend to go right, I tend to go left. When they go left, I tend to go right. I'll just be contrary because it's just more interesting. So when we talk about how we're solving this problem, that big data versus a small data approach, no one has that conversation. And, hear it every day, literally. A call yesterday with a very large credit card issuer who we're also partnered with and they were like, "I’ve never heard this before." Being able to get on a call, hopefully at the end of that conversation, say, "Yeah, you're right. No one's doing this. This is pretty cool stuff." That excites me.

Jeff LoCastro: I often say that we're not a new flavor of ice cream. We're entirely different food group. That's pretty cool. It also takes some explaining for people to understand I have this, what I call it, the 20/40 rule. It generally takes someone about 20 minutes before they say, "You know, I think I get what you're doing." In about 40, they say, "Oh, my good... I get it now." And then they have questions because we tend to have to undo a lot of what they think we might be doing or what they... preconceived notions. And that's just... that's just fun.

Jeff Kavanaugh: Your company, Neener Analytics, talks of small data results in a big data world. What does this mean in layman's terms?

Jeff LoCastro: The differences between small data and big data, the example I like to use is this: Back in 1940, when my grandfather bought his first house, he walked into the bank and the banker said, "Hey, I know you." His name was Bert. "Hey Bert, I know you." And that's what he meant. I know you. My grandfather walked into that bank as small data. He was small data. My other grandfather, in this example, if you were to walk into that same bank, the banker would say, "Oh, you know, I don't know you." Didn't live in the same city or state. "I don't know you, but I know people like you." Big data.

Jeff LoCastro: And that is the definition of big data. I don't know you, but I know people like you. Big data's nothing but bins and aggregations based on behaviors or affinities, clumped together to create correlations that masquerade as something specific. But it's not. It's an aggregation. It's a pool of humans in a bowl that you're pulling them out of and giving conditions on them as to some kind of behavior that might result in something specific. Small data is an individual matrix on each and every human that manifests in binary outcomes. So the difference between a big data solution and a small data solution, big data produces scores and behaviors. It's another aggregation. Scores and behaviors. And behaviors, you always need a behavior to make sense of the previous behavior. An outcome, this is what human beings, and when we're talking about small data in terms of AI, exactly the way human beings make decisions about other human beings based on an outcome, a binary outcome. So that's the main difference. I explained that in what, maybe 20 seconds, but it's enormous in terms of what it reveals and how you get there.

Jeff LoCastro: For example, the AI that we've developed makes decisions exactly the way human beings make decisions about other human beings, not based on what they talk about, but rather how they talk about those things, which is exactly what you're doing with me right now. You're not listening to what I'm talking about. You listen to how I'm talking about it, because the what is big data. The what is the aggregation. The how is the small data. In the example of my grandfather, that banker didn't make a decision, say, "I know you," based on that he was... He lived in Chicago... That he was a White Sox fan, that he liked the color red, that he liked Italian food or whatever. That wasn't what he meant. He didn't say, "You know, I know aggregations about you." He said, "I know you. I know you."

Jeff LoCastro: And the, "I know you," comes from the conversations and communications he had with him over time, based on the how my grandfather communicated with him and those produced binary outcomes, exactly what we're doing right now. I'm doing it with you. This is how human beings stay alive on the planet. We make binary decisions based on how people communicate with us, not what they talk about. When I'm talking to someone, I say, "This is the code we've cracked, is being able to intuit what human beings do naturally to survive on the planet, moving that to be able to do it technologically, to be able to say to every consumer, "I know you."

Jeff Kavanaugh: You've been treating humans as either, “look at this third party thing where we're looking at,” or maybe even these wonderful, innocent creatures. A lot of people, especially if they're working with your clients or with someone who's going to sell a product or assess some kind of risk, they have an incentive to game the system. Other words, lower the interest rate they're being charged or lower some other conditions that are being placed on them. So is what you're doing trying to overcome their ability, not to lie to you, but to give a false narrative, or when they do that, do you have ways of looking at it, cutting through to the authentic?

Jeff LoCastro: It's one of my favorite questions. Absolutely. This one of my favorite questions. It's interesting because we tend not to get asked this anymore, but early on, absolutely. This was something, well, I'll just be a different person. And then how do you do that? So this is not something that we focused on or a problem that we tried to solve, but it's absolutely a derivative of using human data. Not aggregations of what someone might think is human data. Authentic conditions and domains that exist within that individual human being. It's really hard to change those things in a way that is authentic.

Jeff LoCastro: How do you know… to your question, in an application process, consumers know what the right answers are, right? They don't have the right... They know lots of money's good, not money is bad. They know lots of debt is bad, no debt is good. For talking even insurance, they know when the question comes to smoking and your health, they know they know which box to check, right? They know.

Jeff Kavanaugh: Exactly.

Jeff LoCastro: They're not the algorithm, but they know, right? Now, applications are, by their nature, a system gaming device, because the consumer goes into it already knowing what the answer is. It's beyond cheating on the test. When we're talking about human data and these conditions and domains that every human being possesses, that they innately communicate that we're hardwired individually to channel to. How do you do that? Right? Our AI, for example, could be listening to this conversation within the first three or four minutes, ARIA would have been able to decision both you and I, and whether I will pay back the lender or whether I will not, and whether you will, but we haven't talked at all about getting a loan because it's irrelevant.

Jeff Kavanaugh: All right. What is relevant in the world of lawyers, is that…

Jeff LoCastro: The lawyer question's my second favorite question.

Jeff Kavanaugh: Well, you're going to be asked. Well, great. If you're so good at this, then the next thing is, put your hand over your wallet because wait a minute, did I give you permission for all this really interesting information that you're getting from me somehow? Did I give my explicit approval or where's it coming from? Could you talk about that for a moment?

Jeff LoCastro: Yeah, absolutely. We were actually in Paris May 30th of 2018 when the GDPR went into effect, as an example, and everyone was racing around trying to figure out, oh my God, are we compliant? We were like, "Yeah, we're compliant." We're compliant by just by virtue of how we access the data. We're a hundred percent opt-in. We're 100% opt-in and this is self-created data. We fall right off the table of any GDPR or any privacy concerns. This is self-created data. And 100% opt-in. The consumer knows what they're doing, why they're doing it and the benefit that they should receive from it. Which is why we have, for example, in Latin America, we have mid 90, 96, 97% opt-in rates, globally 87% opt-in rates.

Jeff Kavanaugh: Is it because people understand, someone's either mentioned to them or they infer that by doing this, they're going to get a better deal? Or is it just natural?

Jeff LoCastro: That's a cool question because, and it actually leads me to some of the social science implications of what we've discovered about the algorithm of the AI. That a part of the engagement is some kind of an incentive, right? They're being told, "Hey, click here and something..." For example, if our solution is being delivered to a consumer at rejection, "Hey, click here," or "Talk to ARIA," whatever it might be, depending on what channel the customer's choosing, and "We don't want to lose you. We want to find a way to say yes to you. Sorry. We had to reject you," kind of a thing. "Click here” and three in five rejections, or whatever the ratio is for that to lender, become approvals. That's a very powerful incentive. That's a built-in incentive.

Jeff LoCastro: In the application process for example, there might be other incentives, "Hey, click here and you have a shorter application process." Whatever it might be, they're incentivized in some way. But the interesting thing is that we do tend to see, we don't guarantee this, but it is a social science derivative of what we have created, about a 44% increase in completed applications.

Jeff Kavanaugh: Well, that helps.

Jeff LoCastro: That's huge. And we would see these things... For example, in a customer engagement, we'd say, "Hey, look, how many applications do you typically run?" "Oh, we run 20,000." "Okay, cool." And we were seeing 25, 24, 23, 26, 27. And we'd say, "Are you advertising this?” “No, we haven't changed anything.” And it took us about eight months to figure this out. What happened is that consumers were figuring it out. They were figuring out that they were suddenly, by this click, that they were small data. Now, they weren't thinking small data. And I'll go back to the example of my grandfather. Why did that first grandfather go into that bank and not the other one? Because he knew that that bank would say, “ I know you.”

Jeff Kavanaugh: He knew they knew.

Jeff LoCastro: Exactly. We're hard wired for this, Jeff. We're hard wired for this. Now that bank could have been 150 miles away walking distance, and he's still going to go there. He is not going to opt, no one would opt for, "I don't know you, but I know people like you." No one would choose that. Every human being is hardwired for a small data engagement. And I know this to be true. Every time you walk into a new event, a party, a room where you may not know somebody, are people that you look for the people you don't know, or the people that you know? You always look for the small data. It may be a nanosecond. It may be longer. We're hard wired for this. So it's that really cool social science derivative of what we've created. Consumers have just figured it out. Which is wild.

Jeff Kavanaugh: Because you mentioned social science and this very crisp, definitive outcome answer, it conjures up, is there bias? Is it bad? Identifiers. You start talking about gender, race, and on. How are you dealing with those issues? Are they an issue for you? And if have figured it out, share your secret because a lot of folks are struggling with that.

Jeff LoCastro: Yeah. It's pretty straight forward. Our results were pretty balanced by nature. What that means is we're looking at conditions and domains that every human being on the planet possesses individually, just in different ratios that no one... I'll use the example of maybe you, your listeners have took Psych 1A and there's always a conversation in a psych class about conscientiousness. We've got over 600 conditions and domains that we look at, most of which we've discovered, but conscientious as certainly in that pool and the one that most people understand.

Jeff LoCastro: Conscientiousness. No protected group, no race, gender, creed, religion, height, weight, orientation, pick a protected group, pick a straw to pick a channel, has the franchise on conscientiousness. Nor are those same groups disenfranchised by nature of their group, disenfranchised from conscientiousness. Now, conscientiousness certainly might have implications how you were raised, who your parents were, how they brought you up, the values that they instilled, but it has nothing to do with your skin color, has nothing to do with your gender, your height, your weight. Pick a protected group. These are conditions and domains that manifest themselves individually in human beings. And we have gone through our fair lending analysis in the U.S. and we knew going into it we were probably going to see a pretty flat line. And that's exactly what we saw. Completely flat. Zero. Zero bias. These human conditions, by nature, don't discriminate based on your religion or your skin color or whatever it is. So we say, "Look, our results are pretty balanced by nature.” They are.

Jeff LoCastro: And we've been looking for someone to come out and say, "No, they're..." Because, my goodness, we'd be on the page of every paper on the planet. Like, who would say that? If you said it you're insane and it's simply not true. Which is why when we go from... We say, "Look, American humans aren't different than Brazilian humans. Brazilian humans aren't different than Colombian humans or Mexican humans." Their culture is different. Absolutely. Big data. The food they eat's different. Yeah. The music, they like... all big data aggregations, but the human conditions that we all share, which may have been influenced by the culture and these kind of.... But we don't care about that. All we care about is the you. How did it end up in you? And none of that correlates to your skin color.

Jeff Kavanaugh: It's is a small world after all.

Jeff LoCastro: Absolutely.

Jeff Kavanaugh: To leave the folks listening with some very actionable things they can do, what are the three things that someone listening can take from this discussion and go make something happen in their own workplace?

Jeff LoCastro: Yeah. I always say, "Look, we can't help you unless you are willing to admit that there's a problem." And I chuckle when I say that because it's one of those things, the blinding flash of obvious, but we do find it. Sit down and understand, and be honest, do you have a problem? Is this a problem? Do you want to say yes to more people? And I don't want to sound salesy here when I say that, but that has, as we've discussed, global implications. Do you want to say yes to more people? Is it a problem? The second thing is, is understanding are you going to approach this problem as a risk mitigator or a risk minimizer? Are you going to understand the cost benefit to this, or is this only about trying to figure out a way to maintain the status quo? If that's what it is we can't help you there either.

Jeff LoCastro: And the third is, is in that balance, as I mentioned I think earlier, is the problem we're solving bigger than any risk that you would take in understanding how this solution will affect you? If you can do those three things, we can absolutely help. We have never... and Jeff, this is not entrepreneurial hyperbole and I sometimes hesitate saying this. We've never had a situation where it didn't work. Not once. We're 11 countries. It works every time. We've cracked this. Those three things, and it really does start with admitting that there's a problem. And in any engagement is deciding who that champion in the organization is going to be. It has to start there. Who's going to champion this? Who's going to own this and bring it through? Because you will come out the other end saying, oh my gosh, it worked.

Jeff Kavanaugh: Three strategic questions, whether or not we see the wisdom in approaching you folks directly, or initially trying to work through some of those questions themselves. Great to think about, and thank you. Thank you for sharing those with us.

Jeff LoCastro: My pleasure.

Jeff Kavanaugh: What resources do you recommend so people can learn more?

Jeff LoCastro: You can always reach me directly at Jeff at Neener, N-E-E-N-E-R dot net or our website, neeneranalytics.com. We're happy to engage quickly and perform an impact audit for you if you're willing to share a little bit of information on your transaction role, and we can show you in about five minutes what the expected outcomes should be, and to see if you want to move forward with us. The two best ways to reach out and show you that it does work.

Jeff Kavanaugh: Everyone, you can find details about everything we've covered, including contact information on our show notes and transcripts at Infosys.com/IKI in our podcast section. Jeff LoCastro, thank you so much for your time and a very interesting discussion.

Jeff LoCastro: My pleasure, Jeff. It was a joy.

Jeff Kavanaugh: Everyone, you've been listening to the Knowledge Institute where we talk with experts on business trends, deconstruct main ideas, and share their insights. Thanks to our producer, Catherine Burdette, Christine Calhoun, Dylan Cosper, and the entire Knowledge Institute team. Until next time, keep learning and keep sharing.