Learn more about Nobl at: https://www.wearenobl.com/
Find Cedar Milazzo on LinkedIn here: https://www.linkedin.com/in/cedarmilazzo/
JC: Well, hello everybody. And welcome to this day’s episode of the Future of BizTech. I have with me, the founder of Nobl, Cedar Milazzo of WeAreNobl.com. Cedar, thank you so much for coming on the show and tell us a little bit about yourself and a little bit about what your company does.
Cedar: Absolutely. Thanks for having me. So I’m Cedar Milazzo, as you mentioned, the founder. So Nobl is what we call a brand responsibility company. We use artificial intelligence and machine learning to identify things online like misinformation and hate speech and, and other toxic kinds of content. And then we provide services to brands to allow them to avoid sponsoring that content. So there’s been many cases. In fact, you know, you hear about them in the news, you almost every day, where advertisers are accidentally running ads in front of an ISIS recruiting video, or, you know, in front of the Christ church manifesto, you know, things like that, things that are obviously not something that they really support.
Cedar: But they are sending money to the people who are actually publishing that sort of content. And so what we’re designing our software platform to do is essentially allow those advertisers to say, I want to be responsible with how I spend my money and I want to avoid misinformation. I want to avoid toxic content. And I want to make sure that my ad dollars are spent on legitimate content, on things that actually help our society instead of hurting it. So, you know, supporting the real legitimate news and, you know, things like that, as opposed to just random websites out there that just have things that are divisive or, or incorrect, or, you know, just hate-filled
JC: That’s incredible. You know, if I have an obvious question, but I am curious, you know, what motivated you to start this company? I mean, what was that? What was that day? I felt like every owner of a CEO of a company has that day where they’re like, just something clicks and they’re like, screw it. You know, I’m doing this, you know? This is the company I want to build. And here’s why, so what was your motivation to start Nobl?
Cedar: It’s actually a pretty interesting story, at least I think so. It was right after the 2016 elections and there had been a huge amount, as you probably know, as everybody knows of discussion in the news about, you know, what they were calling fake news and other types of misinformation. And right about this time my wife was actually working or volunteering at an animal rescue center. And my father in law lives up by Yosemite and he was down visiting and this was actually for Thanksgiving. He came down to visit and he was telling us these stories about how everyone in that area, where he lived was going out and killing possums. And the reason was because there was a bunch of information online that says that they were spreading rabies to everyone’s dogs. And so everybody was going out and getting rid of them as quickly as possible.
Cedar: Well, because my wife actually worked at an animal rescue center. She knew that she didn’t have to have rabies shots to handle the possums. And therefore it was very obvious to her that they actually don’t carry rabies and they can’t spread it to anybody else. And so it was an obvious bit of misinformation out there on the, on the internet. And they got into this huge discussion about it. And it ended up being well, you know, I saw it online. I know it must be right kind of thing. Right. And it, at that point, a kind of a light bulb went off in my head saying, you know, there’s two kinds of things happening right now. One is little types of misinformation. Like this, people are doing silly things because of information that they find online, where on the other extreme, you have all this misinformation in the political world that’s causing major problems, not just in, in how we vote and who we believe in where we’re thinking about our politics.
Cedar: But also right after that, there was an incident in India where people were sharing stories about strangers coming through small villages. And, and they were molesting children. And this is what the story went well in India, in the small villages, they started actually lynching, every stranger who came by because of this Instagram post. And so about 12 people actually were killed because of this type of misinformation. So it was a big problem from, from something silly, like possums all the way up to people dying, right. And everything in between. And at that point I was like, okay, there’s gotta be something I can do. I have a technology background I’ve been in software development for 30 plus years. I know there’s gotta be some way to cut back on this. And, and I decided at that point, something had to be done.
Cedar: I had the skills to actually do something around it. And so I quit my job and started on this full time. And haven’t really stopped since basically started writing up some machine learning models that help to identify some of the basic stuff right off the bat. And within a couple of weeks was able to put something together to have like an 85% accuracy rate, obviously not good enough for production, but it can show that you can actually do something about the issue. Right. And so over the past couple of years, we’ve just been kind of refining those and making things better. But the idea essentially was let’s identify this stuff out there. And at the very least we can tell people to be careful before they actually go off and believe it.
JC: Yeah, that’s incredible. You know, obviously it’s obvious the kind of application that would have, you know, with a, with a new election cycle coming up as well. What do you find? I mean, how, how ramped up is this now? Because I mean, it’s, you know, at the time of this recording anyway, you know we are in what are we in June, June of 2020? So we’re about, you know, five, five months out, or so until the next presidential election, you know, how do you see your, your software being used right now for gearing up for this kind of political climate coming up with all this information and misinformation coming out?
Cedar: Yeah, it’s an interesting time. That’s for sure. It’s not just the election, it’s also the coronavirus and even the black lives matter movements that are happening right now. There’s just so much going on and there’s so much misinformation out there, that it’s very hard to tell. What’s real. So in terms of where we’re at, like I said, we spent the last couple of years basically working on the technology we’ve got about six people or so on the technology side, just building things and making things better continuously about three months ago or so, right about the time COVID-19 hit. We actually started going out to customers and actually deploying it. So it’s very early stage still. But we’re seeing a great uptick in interest, you know, especially with actually the COVID-19 stuff. There’s so much health misinformation out there that, you know, people are, people are desperate, especially anybody who’s related to any companies that are related to the health market.
Cedar: We have one company who came in, we had a conversation with them on Monday, Tuesday, they were informed by their, by the press that they were supporting COVID-19 conspiracy theories. And then on Wednesday, they came back and signed a contract with us. Yeah, I imagine so, I mean, good, good timing. But, you know, because there’s so many problems out there, it’s, it’s a, it’s a solution really that’s that people are desperate for. And so we’re, we’re definitely seeing a huge amount of traction happening right now. And I think that we will be able to, to get out there in enough force and enough scale before the elections to actually be able to help people understand what’s real and to cut back more importantly on the amount of misinformation that’s actually being out there and it being spread around. And so I think the biggest thing that we’re looking for is the ability to cut the funding to these sites of misinformation.
Cedar: Right. If we can identify this thing and tell the advertisers, don’t place your ads here, then though the profit motivation essentially goes away for those sites. Right. And that’s, that’s not all of the problem of course, but it certainly is a big part of the problem is a lot of people out there who are just publishing whatever they want to say, just so they can make a little bit of money or a lot. So we’re trying to just cut that, that financial incentive out for them.
JC: So what is it, I mean, walk the audience through kind of, what is it like to implement this system? You know, I mean, I think some people might have a hard time wrapping their head around, like, okay, we get what your system does, but how does that integrate, you know, what’s it actually doing? Is it the AI reading the ads and, you know, I mean, like walk us through exactly maybe how it’s implemented and how that helps the clients.
Cedar: Yeah and it’s a really complex environment that we’re working in. Our stuff isn’t actually that complex, but the advertising market itself, especially when you look at programmatic advertising is extremely complex. There are a lot of different players, and there’s a lot of different ways that you can attack the problem, but the way we’re coming at it essentially is in programmatic advertising, they have what’s called real time bidding or auction, essentially an advertiser sets up a program that will essentially bid on every impression, every viewable ad that happens. Right? So if you pull up a web browser and there’s three ads in the site, he goes back to the backend server and says, I have three ads that are going to be displayed to JC Granger.
Cedar: And here’s his background information. He’s this many years old, he lives in this area, you know, whatever information they have on you that they’ve been collecting in violating your privacy about. So he sends all that information to the backend and the auction is then run and all the advertisers have the information and they can make a decision on whether to show the ad team or not. What we’re doing is we’re inserting another piece of information at that point, which is, this is the page that ad ad is actually on. And we’re reading that page, the content on the page, and we’re analyzing it and coming back with a score, essentially, a credibility or toxicity score, basically from zero to a hundred that says, if it’s a low score, it’s toxic. If it’s a high score, it’s high quality. So we simplify it down to that level.
Cedar: And then the advertiser basically makes a bid based not just on who you are and your background, but also on the content that you’re looking at. So decide if we want to be associated with that content.
JC: Yeah, that’s brilliant. I know that’s something that I think the real, especially now, you know, black lives matter crew and all this what not, and even the election with just how polarizing, you know politically it is you know, a lot of big brands, I think what would find this to be huge because they have the most to lose. If, you know, if a Nike ad shows up, you know, on some coronavirus is a hoax website, not a good look for Nike, you know what I mean? So, and, and you know, they, and they’re a public company, you know, they, take a half a percent a hit and that’s hundreds of millions of dollars, you know, like that that’s not small, you know, when it comes to market cap. So I, when I, when I was reading a, your website will say, that’s what I found the most fascinating. But so where do you think..
Cedar: It’s really interesting as well because consumers are becoming more and more involved in this, right? As a consumer, people are, people are saying, you know, what, if this company supports this type of content, I’m not going to buy that product. I’m not going to buy their products. Right. And that’s happening more and more, you know, I, I kind of credit it to a group called sleeping giants. They actually happen. They kind of came together right after the 2016 election as well. And sleeping giants is kind of an activist group. And they basically point at advertisers and say, Hey, your ads are running next to this. You know, this really bad content. And they send it out on Twitter and Facebook and everywhere else and get people basically to, to put pressure on the advertisers to change. But the problem with that is they average Kaiser’s, didn’t have a tool to actually do anything about after the fact that they were spending the money.
Cedar: So they couldn’t really proactively keep their ads off. They just had to wait until they had an incident in a, in a boycott and all that before they could actually do anything about it. So we’re, we’re kind of attacking the problem from the other perspective. But the other interesting piece about that is, and the reason I wanted to bring up sleeping giants is that the founder of sleeping giants is a man named Matt Rivitz. He’s actually joined us, as an employee, as our chief marketing officer. And we’re kind of joining forces to kind of attack this problem, both from the front and the back ends and make sure that, you know, people who or the companies who are not being responsible and letting their ads flow through are still gonna be, you know, caught up in this public sentiment kind of thing. Whereas those people in those companies who are taking responsibility upfront and preventing their ads from going to that sort of content are going to be in a stronger position to never have to worry about that issue.
JC: That’s brilliant. That’s a hell of a, of a get, as I say for employment, then you got a great guy that came on there. Let me ask you a personal question here. You know, what did you want to do when you were, when you were a kid? Like, you know, what did you want to be when you grew up, so to speak,
Cedar: Oh, I have so many ambitions, you know everything from the usual astronauts to, to firefighter and all that sort of thing. But, you know, when I was in high school, I really did get enamored with computers. And I went into software engineering. I got my degree in computer science and stuck with that ever since. So that really is kind of what I was super interested in longterm and what I’ve, what I’ve made a career out of. So I actually did get to do my dream job.
JC: That’s awesome. And then who were your mentors kind of coming up? You know, what was the best advice that you ever got? You know, something..
Cedar: Oh, geez. There’s so many people out there. I think my, my probably most, most valuable mentor at this time is a man named rich Wycoff. He was a tech CEO for a long time here in the Silicon Valley area. And he’s just given me really great advice, especially when it comes to starting this business. You know, this is the first company that I’ve started. I’ve been in a lot of startups, but this is the first time I’ve actually been the founder and the CEO. And, you know, there’s so many pitfalls along the way when you’re starting a new company. And he was really able to guide, you know, my decision making and guide me to avoid a lot of those pitfalls.
Cedar: You know, there’s, I think probably the most beneficial, single piece of advice for this company at least was from, was actually from Rich and he basically told me to reach out to people like Matt Rivitz of sleeping giants and, and see what we could do to partnership, right. It’s really hard to do a tech startup, especially in the ad tech world without having some strong, strong partnerships. And so when I started reaching out to organizations, sleeping giants is one of them, credibility coalition, miss Infocon and places like that, that’s where things really started to take off. It was, it was one of those things where, you know, you can build something and build something and you have a solution, but if nobody knows about it, what’s the point, right? So if you were able to start finding people in that industry and, and really make those connections and get them to be supporters and champions and, you know, and partners made a huge difference for Nobl. Absolutely.
JC: Well, you know, I own a marketing agency. So when I see something like this, this makes my, my brain speed up and think, you know, is this a platform that we could use to help with our clients, for example, to make sure that they don’t end up on ads, you know for the marketing agencies out there that might be listening is your, is your system something that’ll be available if not already in more of a, kind of a sass, easy sign up monthly subscription style? Like, what is it like to actually even purchase your service here?
Cedar: It is a SAS model. Absolutely. So you can kinda actually we’re, we’re taking early early applications now basically for free trials, essentially to get people to come in and great year.
JC: Me, I definitely want to play with it. I’m a complete nerd when it comes to this stuff and this could, I could actually see this being something that we could actually really benefit from. So I want to check out.
Cedar: It’s interesting because you know, we have the service on the backend that allows people to kind of protect themselves. And when we present that to customers and to agencies and people like that are they’re interested, but they don’t quite a hundred percent get it, but we have also a dashboard and we show them the dashboard and they can actually see what their what’s really happening, how many dollars they’re spending on hate speech, how many dollars they’re spending on like an audit.
JC: Wow. Could you do an audit of their current campaign and just show them, wow, that’s gotta be impactful. Like seeing that for the first time, seeing that audit run, I mean, Holy crap, we got like 10% of our ads are on ISIS pages or something stupid like that.
Cedar: Exactly. That’s exactly what, what is actually getting our business at this point is actually demonstrate the product and show them that this is actually going to be beneficial and valuable to you. And so when you, when you sign up, you get access to this, this dashboard and it shows you not just all that sort of information, but also every single place that your ad is being run. So really nowhere else in the ad tech world, do you have the ability to get that sort of transparency? I mean, if you use Google advertising, for example, if you pay for, or get the very detailed log information, they might get you down to the domain level, but we have every single page in the context of that page. So you can really understand exactly where you’re at.
JC: And it’s learning all the time. It’s just going to keep getting better and better. That’s amazing.
Cedar: Yeah that’s the other piece of it, it’s always learning. Exactly
JC: So in the spirit of the title of this podcast, the Future of BizTech, tell the audience here, two things, one, why you think your Saas is the future of the industry that you’re helping out. So, which I know market inches are one of them and also just big brands in general. Right? Right. And then, but then also, if you can, if you, if you’re able to give us a preview of anything coming down the pipeline, that’d be fantastic too. So why, why would it be the future of the industry and then, and then what’s coming down the pipeline?
Cedar: Sure. So, I mean, I think, as I mentioned before, consumers are getting more and more aware and interested in the corporate behavior of the brands that they support, in the products they buy, right. They’re becoming more and more interested in where you spend your money. What are you doing? That’s good for the environment or good for the society in general, right. Whatever it happens to be. And so it’s going to become more and more important that brands become not just pay lip service to being responsible, but actually really proactively become responsible with how their advertising dollars are spent.
Cedar: And so just from a public perspective, I think that’s going to become more and more important to these brands. From a more insider view, there’s some big changes coming to the ad tech world right now. One of them is the death of third party cookies. And third party cookies are essentially how Google and these other big advertising companies track everything about you as you go around the web, right? So you might have a cookie on website A, and a cookie on website B, and they both sell that information about when you came to their website to Google. And now Google knows more about you and every advertising company is doing that, buying all this information well, that’s becoming illegal very soon. And once those cookies are no longer allowed, there’s going to be, it’s going to be very difficult to target you specifically based on all your, your past browsing habits.
Cedar: And so it’s going to be much more important that you understand the context of where your ad is being shown, because you don’t know who you’re showing it to necessarily. You have to at least know about the context that you’re being shown in.
JC: That actually has a lot of ramifications. You know, when you think about it that way, it could actually change the entirety of how advertisers target people in general targeting the context of what they’re looking at. Not necessarily, you know different pieces of information about that, you know, very more psychological. So that’s my background is psychology. So when I, I sort of think about this stuff, I think about the, the, the psychological side of it, more consumer psychology, for example..
Cedar: Right, so now you’re gonna have to look at, okay, what’s the psychology of a person who would read this type of content and, you know, decide in real time you have 10 milliseconds to decide whether you want to bid on that ad impression or not. So you have to very quickly make a decision of whether that impression is something that you want to bid on or not knowing that the toxicity, the credibility of that page makes it much easier to understand and to decide whether to actually target that page or not. So contextual targeting is going to be huge shifts in the ad tech space over the next couple of years. And we’re really at the forefront of that because we have we believe the very best contextual information about the content. Our algorithms are far beyond any other sort of what they call brand safety companies out there. And the way we’re interpreting the information, the way we’re training our models is so different that our contextual awareness is just light years beyond anybody else. We believe that we will be really kind of at the forefront of who takes control when you don’t have the information about the user anymore.
Cedar: And how are ads targeted and how are ads placed? So, you know, it’s a $200 billion a year market that basically is going to be turned on its head within a couple of years within a year, really. And we believe that we have the technology that will basically enable the next stage of this. So we think we’re sitting in a pretty good spot there.
JC: This is fascinating to me. Is there anything that I didn’t ask or we didn’t go over that you think would benefit the audience to know?
Cedar: I’m not sure how important it is to go into the technical side and describe how our, our models are built or anything, but it is something I think that would be interesting to talk about just because even from a layman’s perspective, just saying an AI looks at it and interprets it, isn’t really helpful. Right? And so essentially the, well, what we do is we have we have scientists on our team who are actually rhetorical scientists. These are language scientists. So they’re not computer scientists. These are people who understand the language and how it’s used for different purposes. And they have come up with these, these indicators of what is toxic and what is high quality, and there’s about 25 or so of them that they’ve identified so far. And we’re continually looking for more. And, and it’s everything from, you know, the clickbaiting is how closely does the title of the article match the actual content down to very esoteric things like the ratio of adverbs to verbs and nouns and, and all those sorts of things kind of play together. And we take all these 25 indicators out of every page and we run them through a, a machine learning model and that outputs the final school.
Cedar: So that machine learning model score and the, the indicators are actually trained based on what we call expert data. They’re not just scraped off the web, every single piece of data that we’re running into our training set is all hand reviewed by multiple people. So, and these are all experts in the field. So we get two or three reviews for every single piece of content that goes into our training step which is very different than some of the competition out there that is basically saying, “I want to script the New York times and I want to scrape natural news.com and everything from, from New York times as good. And everything from New York from natural news is bad”, but that’s not really true. I mean, there’s varying levels of goodness or badness from each of those sites, right? I mean, if you have a editorial on the New York times, it’s not the same level of credibility as a news story.
JC: I was going to say they got, they just got some heat for running an op ed, you know from, from a politician that a lot of people didn’t agree with it. Imagine if your ad ran on that, you might be thinking, Hey, I want my ad to run on New York times because it fits with my demographic, you know, in our, in our line of thinking. But that particular page would have been insanely problematic for a company to have their ad on that op ed page. So that’s why that’s brilliant. I think it’s great that you guys had the foresight to have rhetorical scientists, you know, language scientists rather, or not necessarily like, rather than, I’m sure you also have computers just did the algorithm stuff, but that you, that you employed people that specialize in language and have an actual human intuition factor that you’re importing into this AI. And I think that’s fantastic. Listen, I want to absolutely thank you for coming on here. How can people reach you if they’re interested in signing up or, you know, just people listening might be going, Holy crap, I need this. I mean, I’m one of them, right? So we’ll talk after this because I want to, I want to get signed up here and check this, but how would people reach you if they were interested in this?
Cedar: Well, the easiest thing is to go to our website. WeAreNobl.com. WeAreNobl.com. There’s a link there too, to send us an email or you can just email us at hello@WeAreNobl.com and that’s the easiest way to get in touch with us and we can go from there.
JC: Well, that’s awesome. Cedar, thanks again for coming on, you have a fascinating business. I’m so glad that this type of technology is coming out at a very much needed time. So thank you again for coming on the show.
Cedar: Absolutely. Thank you for this, man. This has been great.