C (3:22)

I'm doing great, Ellen. Great to see you again. Thanks for having me.

B (3:25)

It's great to see you too. I. I think it's been 12 or 13 years at least since we've last talked.

C (3:30)

You actually look better than you did 12 or 13 years ago.

B (3:34)

You're being very kind. You look great as well.

C (3:37)

I was waiting for that. That was a long pause, Alan.

B (3:40)

That was. Yeah, that was a really.

C (3:41)

We're.

B (3:41)

We're going to edit that out, so it's going to be a lot shorter. You are an accomplished author. You've written two books of fiction which were really well received, and one of them, I think, was even optioned to a movie studio. So what made you decide to go the perilous track of writing a business book?

C (4:00)

I just got to that point. I was about 20 years into my data career and a couple things happened. One is I sort of looked around and realized that I was the expert in the room on data. And that was a nice moment, I suppose. And then I began to reflect on the classic of what I've learned and what I would tell my younger self. At the same time, I was prompted to do so even more urgently because reflecting on my time at Comcast and on my friendships in New York City and beyond, I was realizing that there were a lot of people I knew who were fantastic at what they do and especially in the advertising and marketing space. But when they had kind of decided that they didn't need to understand data, they severely limited their career path. And I would see, especially at Comcast, I would see some sort of elder statesmen which, you know, my youth in my 40s at that time, you know, people in their mid to late 50s who were incredibly accomplished, who are great leaders, who were managing and inspiring thousands of employees and managing billions of dollars in revenue, would just kind of step out of their careers because they just felt like the situation in the industry, which is now focused on data, had passed them by. So my Notes to myself, my younger self became that much more urgent and I wanted to sort of see can I articulate what are the things that really express the essence of working with data? And can I articulate that to another person? Especially to a non expert because it's such a classically intimidating topic. It's like technology or medicine. And that same fear comes over you like when you're talking to your doctor about something sophisticated and they kind of use all these sorts of jargon and you feel like dumb and small. Like that's what happens to us when we're talking about a topic, when someone else has expertise and we don't. And so the, the hill I wanted to climb was can I write about data, express it in a way people can understand, but also make it fun and engaging through storytelling.

B (5:58)

So what comes to mind is, you know, the image of that person in, you know, their mid to late 50s who sort of, you know, closing out their career. And then you're in your maybe you know, late 30s, early 40s. And it's interesting because I remember being one of those people to look at like all of the, you know, quote, I'm using air quotes here, envelope salespeople from the Direct Marketing association and thinking all those old dinosaurs, they don't get it. I'm in the Internet, I'm in new media and I share that because there's a little like self schadenfreude here because as I look into what's going on with AI right now, and now we're probably right off the bat getting way off topic here. But what the heck, that's great. How does one then try to equip yourself for stuff that isn't going to be second nature because you are now in the getting closer to the back nine of your career for sure.

C (6:48)

And thanks to that last comment again to echo your earlier pause, the moment I had about AI a few years ago was this is so new that the. There's a starting point about what's being published and if I can, for, for myself, if I can tap into that as soon as possible, then basically I can be as current as anyone. Assuming that I'm not going to take a coding class or computer science or theory class about it. But as a business person, I could be on as on top of this as anyone. And there are two things that really set me on the path. One was reading a book by Melanie Mitchell, who is a professor at the Santa Fe Institute. If I'm getting it right, her book is called Artificial Intelligence A Guide for thinking humans. And then I read the original paper for attention is all you need, which is about the famous creation of the large language model while reading the original attention is all you need paper about LLMs. It's just saying, okay, I'm going to read, I'm going to read original scientific publications, I'm going to read research papers, and If I understand 40%, 60%, 80% is better than zero. And that's sort of how I got on the path. And so doing the research for the book was also a challenge about what kind of balance to strike. By the way, this is how slow book publishing is that you write a book proposal, you sell it, and then the process takes so long that, like, the whole AI revolution happened between when I submitted my book proposal and the book came out. And so I had to make a decision with my agent and my editor. How hard do I pivot towards AI? And I decided, I made the decision to stay on the same level with AI that I was about data, which is I'm going to explain it, I'm going to frame it, I'm going to get you engaged with it as a, as a business thinker about AI, But I'm not going to, nor would I be equipped to do so. We're not going to get into, like the Greek language lettering and talk about, you know, small language model versus large language model versus rag versus whatever. And I think it worked out really well. And I mean, my curiosity did take me back to the early days of Charles Babbage and Ada Lovelace coming up with their algorithmic machine and Alan Turing writing about intelligent machines. And it was really fascinating that especially in the case of Alan Turing, none of this stuff is new in his paper. I think I'm getting the title right. Intelligent machines had the concept that if we had an algorithm with sufficient computational power and sufficient time to do its calculations, we could in effect, create a machine that could adjust itself and it could answer any question. And this is what Charles Babbage and Ada Lovelace were thinking as well, that you could feed a machine music, you could feed a machine science and art. And it could provide responses and answers to things that are not just proofs. They would not just be mathematical questions that you would asking the computer, which people used to think of as being something that it would actually compute, like 10 +10. It could provide more substantial answers. And so all of that was thought of in those previous centuries. But really what happened was Alan Turing's, you know, in 19, I really should have these dates. I think it was 1925 when he wrote that original paper. These seemingly impossible bars were cleared by progress. So when he said with, you know, with sufficient time to do the calculations that can answer anything, it's like, yeah, right, it's going to take infinity. But when Atari invented the graphics processing unit in the 70s for video games and when big data storage became almost infinite for sort of pennies, then you had a combination of computational power that was so great that you in effect had all the time in the world and you had all in effects all the data you could possibly want to throw any question at it. And so here we have machines that are composing music just like Ada Lovelace predicted.

B (10:59)

Okay, so what I'm hearing here, and I think that going back to the story of Turing, absolutely fascinating stuff. But what I'm hear hearing here is like the key to long term success in your career and probably also life happiness is stay curious. Because I don't know how many of the people that we were railing about in the DM and the envelope salesman were necessarily remaining curious about the things that maybe aren't entirely in their core competency, but they're adjacently have some expertise. And so you went in, you dug in and learn a whole bunch of stuff and then try to bring some of those insights into the book, which is kind of cool.

C (11:40)

I appreciate that. I think when I reflect on myself as a human being, I definitely think that curiosity is one of my virtues and I give my parents a lot of credit for that. They were college professors and just there was no book or historical period or religious phenomenon or painter that they or poet that they were interested in. And that was just kind of ingrained in me from early life.

B (12:07)

So I had a question and an observation here. So I'm a songwriter. As a songwriter I can tell you that there's a certain amount of ego involved. I mean the idea that you think anyone is going to want to listen to your creation, it takes a little bit of chutzpah. And so I'm curious, how are you approaching the act of creating where maybe not every business experience or war story is going to be interesting to others. And that's something I think you do really well within this book. And so I'd love to understand your process a bit.

C (12:40)

Yeah, well, I guess a listener who is, who doesn't know me and why should they would listen to me so far you think I would some kind of quirky writer or professor. And I, you know, I've been doing data innovation for my whole career and I Think. One of the exciting things about working with data I've worked with at Nielsen, the startup that where we got to know each other, Comcast and now Samsung, and the combination of data and a creative mindset, I think is very powerful. Data is John Graunt, who I would call the first data scientist from the 1640s. He was a haberdasher in London during the plague. And he whatever was very quantitative. And he got frustrated with the government officials that they weren't using death data with sufficient sophistication and action to really help Londoners stay alive using data like finding out where the death rates were higher by neighborhood. And he called data airy blossoms. And in the introduction to his bills of the mortality, he said, I hope these airy blossoms will bear fruit. So when you have a creative mindset and something as ethereal as data, you can really get up to a lot of fun with creativity. And I actually had this moment in Nielsen where I was just learning about this Prism segmentation system we had. It was all hooked to address. It could potentially be zip six or zip plus four. And then you could take attributes onto that, and then you could take other attributes as long as you had the hook. And then suddenly I had this kind of moment where as long as we can find this ID and put it someplace, we can attach all this attribute information to it. And once I had that moment, I mean, that was just before we got to know each other, I realized you could take all this offline data and put it online. And that was my first moment. Of course, that's been, you know, the state of the art for a long time. But there was a period where there were just a handful of people, like the guys from Datalogix and guys from a couple other places where we were really the first people doing it. There's only like 50 people who are really doing it. And that was my kind of moment where, okay, I want to be an innovator. This is what I can do, and this is what I can do with data. And I've been coming up with ways to make money from data for all those corporations I listed ever since. So now I forgot your question.

B (15:03)

Why are you so awesome? That was the question.

C (15:06)

No. Well, so I'm trying to establish my credibility for answering whatever that question was.

B (15:10)

Well, you know, I think that there's a little bit of audacity in anybody thinking that their war stories are super interesting. There was a Seinfeld bit back in the day where, like, one of his uncles had just proudly announced that he was sharing a memoir of all of his experiences. Like, you could see everybody in the, in the, in the, in the scene kind of rolling their eyes. And so, you know, what made you think that you are so awesome?

C (15:36)

Well, that wasn't your question, so I'm not going to answer that version. But yeah. So what more stories to include? I mean, I just, when I reflected on some of these scenes that I had in my career and they were important for me. Yeah, it's like to tell a story. And one of the ones that people have responded to is when I was at Nielsen and I was in the corporate strategy group, we were taken over by private equity. And I think everyone who works at Nielsen has kind of a love, hate relationship with it. And on the love side, I had fallen in love with data and I really loved the business. On the hate side, you know, there were these people who would kind of swagger around and, you know, I have a couple scenes where one time I'm, I'm sitting on the sort of the back bench as a group of McKinsey consultants and a group of Nielsen executives sort of duke it out over who's going to be responsible for transforming the business. And another scene where, and this is probably the one that affected me the most, where some of these big dogs at Nielsen were sitting across the table from the private equity guys. And the private equity guys were like, well, explain your business. And the data guy starts by basically saying, you couldn't possibly understand my business, even even though, you know, you went to Oxford and Harvard, and even though you have 20 lawyers and 15 bankers sitting behind you, you could not possibly understand my business because the data product I make has a hundred inputs. And then we go this process and we have a 64 cell grid and we do all these things. And the bankers were the hero of that story because they very patiently, yes, with all their brains and all their money, sat there across the table and said, okay, a hundred data sets. What's the first one? And the Nielsen executive who thought he was going to be able to kind of dance and BS his way out of this, had the list, all the data sets that were involved, all the steps in the machine learning, all the clients and all the use cases. And the banker, yes, they went to Harvard and Oxford, didn't have any expertise, but they were able to pull it apart in very simple language by asking very simple, maybe stupid questions. And that was such a learning for me in my career that when you're talking to a vendor or you're talking to a colleague and they get arrogant with you that you can just slow it down and say, I don't understand, I need to pick this apart. Let's take it piece by piece. And as I say in the book, if they will do that with you, then you know you've got a, a collaborator. If they, if they won't, then you've learned something else, which is they, you know, they don't respect you enough. To tell you the truth.

B (18:15)

I love the concept of a data bully. That person who's like, oh, you'll never understand my business. What's interesting is that when you, when you translate that into the ad space, you certainly have your share of data bullies, but you also have a share of folks who only understand a percentage of it and think they understand the entirety of the business. And so that adds an entirely different complexity. I don't know how you, you know, how one would categorize that flavor of people, but I feel like there's a sort of a special set of opaqueness in the ad space, often in, but, but, but companies and, you know, tend to use that to their advantage, at least smart ones.

C (18:52)

Yeah, I, I think that's, it's a shame when, when people do that to say, use sort of their information advantage, especially with a client, to their own advantage on the inside. When companies and executives are trying to innovate with data, I think one of the, one of the trip ups they have is they actually don't take themselves seriously enough. Because really, data and AI are servants to your business goal. And what a business owner of any flavor must remember is that they are the expert in their business. Whether it's a geography or a type of customer, or it's an industry or it's a process, they are the expert. And that also means that they have to have the courage to, to appear to be stupid by not talking data talk, but by talking business talk. And I'll interject here that one thing we all learned at Nielsen after the private equity takeover was being simple is very hard work. And you can take paragraphs and paragraphs of methodology, but you really distill it down to the handful of sentences that a client can understand is a ton of really worthwhile work. So in the same way that the general manager or the executive who's trying to launch something innovative has to do that same process with themselves and say, what am I really trying to do? Well, I'm trying to make a decision between 10 different things and understand what's the best price for them or something. And you really have to understand at that level. And then you have to pivot and you have to be humble, and you have to understand that. Okay, now to answer that question, I'm up to bring in people with other expertise. And the way I'm describing it maybe makes it sound trite and simple, but it's that focus on the business problem that, weirdly, people lose track of, especially the business people, because they think they're trying to fake being jargon speakers and they really shouldn't.

B (20:41)

Yeah. And look, there's certainly a component of the legal profession who operate in a similar way, which is just like, you know, let me come down from the mountaintop and explain this stuff to you. And, like, I don't personally find that to be a super effective way of doing things. I think that if you can. That. I built my entire career on the idea that, like, no, no, no, no, man, I'm one of you. I'm a. You know, I am a business person. I'm going to speak to you and hopefully get you to speak to me as a peer. It doesn't work either way if either side is sort of up on their. Up on their high horse. It just. It doesn't facilitate communication. And without communication and trust, there's just almost no point.

C (21:20)

Yeah, yeah, exactly.

B (21:22)

So one of the reasons I wanted to have you on the podcast is that you were so optimistic about data, and not just in terms of how value it can be. I think we agree there. But in terms of how data can be such a force for positive change in the world. I would love it if you shared a couple of examples either from the book or just from your other experiences.

C (21:44)

Yeah. There are two characters in the book that I really just fell in love with, and one's become a friend where they use data completely outside of big tech, completely outside of marketing and advertising, to solve really important problems for human health and survival. One was my friend Adam Green, who had been a banker and decided to give that up when his father died of dementia. And the father's dementia really kicked up after my friend's mother died. And the father was alone, and he was in New York City, and he could join temples and museums and groups and. And he was very gregarious, but he. He kind of went into this hole of loneliness, and my friend watched him get lonelier and lonelier, and then the adventure kick in, and he was convinced that loneliness had killed his father. And, you know, the surgeon general kind of now corroborates that, and he set out to do a startup, became Obsessed with it to help seniors in senior facilities get connected and defect, cure loneliness. And the challenge he had was, you know, they would come up with these programs and they wouldn't know who to encourage the senior home to kind of shunt into the programs. So they came up with a loneliness score and they had all sorts of, you know, methodological challenges to apply this. And they ended up going with a way where they would actually do a, a three minute phone call with the senior. But the words they use, the vocabulary they use, their tone would all inform an algorithm that they could say, okay, Justin's lonely, but Alan is not. And then therefore, Justin needs to get the lonely program. And they were also able to measure the impact of the programs, you know, to do the before and after once they had the accurate test. And he's, you know, now on a mission to make this AI driven so it can be much more scalable and operate much more quickly. And if his program works, you'll have all these, you know, senior, thousands of seniors in American senior homes who are getting connected and getting programs that are helping them be more engaged with their peers and happier in their lives and not going through this internal, you know, what loneliness does is it kind of sets a stress response in the body and which accelerates many maladies, including dementia. So, you know, if he can make that work, if he can use that data based approach to scoring someone's loneliness, then he can make all these people that much happier. The second person I met with was the head of epidemiology at the Bureau of Communicable Diseases, which I say slowly because there's a lot of syllables in New York City. It's a, it's a big sort of green anonymous building in Queens. And actually this is another person named Green, Sharon K. Green. And she was the person who was in charge of finding a way to, to attack Covid hotspots.

B (24:50)

Yeah, I remember this during COVID Yeah.

C (24:53)

So it was, it was a hell of a story. So for them, it all started when they had this, this meeting called the Doc of the Week, where someone comes in as a doctor and they're facing something weird because the Bureau of Communicable Diseases tracks leprosy, attracts gonorrhea, attracts Zika, it attracts every anthrax. Anything that's, that's communicable, it's in New York City, can kill us. And this time the Doc of the Week came in and said, there's something happening in China. There's a disease, there's no test. And there's no cure and it's fatal. And within a few weeks, they, you know, it had reached New York City. And we had one of the highest death rates in the world. New York City had a death rate higher than Delhi, higher than Hong Kong, higher than London. And we had full hospitals, thrumming ambulances in the back of parking lots, and not enough ventilators full of patients and victims. And what Sharon's and her team's approach was, if we can find the hotspots where the COVID is spiking and we can send testing and medical attention to those areas, then we can take the death rate down in those areas, prevent the spread, and we'll save lives. But of course, because Covid was new and there was no test and there were no trends, they didn't really have a way of what's a spike when you don't have a baseline? And so they had to use all their science and techniques to come up with a new way of doing it. And they decided to do it where they would look for an average positive testing level and then look for spikes in the positive testing level, even though it wasn't necessarily a spike in the population infection rate. And that seemed to work. And so they launched a program to identify hotspots, flood the zone, so to speak, with testing and equipment, and they brought the death rate down. And, you know, I did some back of the envelope calculation. If, if the New York City death rate had continued as it had, without this type of technique, there would have been an additional 300,000 people dead in New York City from COVID So Sharon K. Green, you know, was the hero of the story. And she, you know, I'm just like sitting across from her at a lunch table and like in Queens, eating a salad. And she's just like, you know, she's a Norse, a nice, obviously very high iq, but normal person. And she said, you know, this was my wartime service and I'd do it again. And I get emotional just thinking about it. I mean, that was my privilege to talk to these people who were, I mean, I wouldn't say laboring in obscurity, but they're just doing normal jobs for them and using data and using those skills to do wonderful things for the world. And, you know, they're not, they're not, they're not working at Facebook, getting tens of millions of dollars in shares. They're normal people using data the right way. And I'll actually pivot if I, if it's not going too boring and I'm talking too Long? No, not at all.

B (27:49)

Keep going.

C (27:50)

Thank you. I would pivot that. Like one of the things that's really powerful, that's happening now in a negative way is that as the current administration is yanking data from public access, you now see how important it is. So there was a time where actually, Alan, I was even very self conscious about the question you asked me. I was self conscious that, hey, man, Shoshana Zuboff, who is a genius, writes this incredible book about the perils of data. And Kathy o', Neill, who's wrote a wonderful book about the perils of data. And I'm like, those guys are cool. I'm this apologist for data. I'm looking at the positive side. That's going to seem silly or something, but now, now we see that when you take these databases about climate change and the way that temperatures are changing in the ocean and how the coastlines are changing, and you take those away and you take away that access, now there are thousands of scientists that are starved for data to do proper research to try and save the planet, and now they're denied. And so you think about Sharon K. Green, and this is what made me think of it. She was using a database that would take all the testing results from all the hospitals in New York City and put it one place. And it was very secure. And you have to have a login and all that stuff, but. And a limited number of people have access to it, but it's a public database and it saved, you know, potentially 300,000 people in 2020 and 2021. And if we didn't have that, then, you know, all those people would have been at risk. And it just, it's a, it's a negative reminder about how important data is to how we function as a society today and how important these true believer practitioners are to data's proper use.

B (29:37)

So you raise a really interesting point. Big data. And I don't mean that as the term everybody was invoked a decade ago, but large swaths of data have gotten a very bad name. And the problem is not big data in and of itself. The problem is that incentives are misaligned. And a lot of that has to do with the fact that for better or worse, we've allowed seven or eight companies to own most of the data. I mean, so your problem here. And by the way, this is why I've sort of pivoted where I don't even see myself as a privacy lawyer anymore. I'm like. Because I don't think you can have a discussion about Privacy without talking about competition and antitrust. And so it's sort of interesting that like everybody, not everybody, but there is a narrative that like, well, it's, what we need to do is get rid of the, you know, these large swaths to data. And this goes beyond what, like what the current administration is doing. I'm just saying, like, you know, philosophically, I've been battling with the privacy advocates for, you know, a decade and a half, two decades, you know. Well, we need to, we need to limit data. Well, the problem isn't data. The problem is, is that a data is in too few hands. So there isn't really a question there. I guess I got up on my soapbox, but like, but I, I feel like I'm saying a very similar thing you're saying, although you're saying it a little bit more from the business perspective and, but there's also kind of a regulatory perspective there as well.

C (31:02)

One of the things that I've been really encouraged about in the last five or 10 years is that there's really no problem that you can't point data at and improve it. And I think that's true. I mean, obviously of scientific inquiry and academic inquiry and nonprofits and non advertising and marketing companies, there is a broad set of data now being employed. I mean, I could just think of a couple of individuals and organizations I've sort of brushed up against that. You know, museums are using data more now to try and pursue their mission to have a more equitable distribution of art in the world. I worked with as a volunteer for a nonprofit that was trying to simply find out that there was a community garden nonprofit. And they just wanted to know what's the impact of starting a community garden? You know, what's the impact on the economy? What's the impact on community, what's the impact on food supply? And because they felt, they felt if they could quantify that somehow they would be able to better raise funds and open more community gardens, which is almost undeniably a good thing. So, you know, if you can be a museum or a community garden, you know, you don't have to be in big tech to put data to use to solve a problem and make your environment more efficient or more productive or further your organization's goals. So I think in the spirit of your comment about, you know, you didn't use the word, but democratizing data and making it more available to more people, I mean, that's really the goal of my book. I want, I want folks to reach for the book and read it and feel empowered to apply data to any substantial problem they're trying to solve.

B (32:45)

And that's a laudable goal. One of the other. And I'll get off the regulatory stuff in a minute, but. No, but there's sort of an important thing that what a lot of the state privacy laws have done over the last five years is they have significantly broadened the type of data that covered by their laws. Okay, fair enough. But they've also taken the definition of de identification and made it such that you actually can't really deidentify a data set and move it outside of the rules of the law. What that does is that that means that like there isn't a safe way to manipulate data or there isn't one to remove it from the rule set. And that's rather unfortunate. I'm not expecting you to comment on that. I'm just sort of an observation as I'm, as we're kind of riffing back.

C (33:30)

And forth, you know, we need some Alan moments here. That was, that was at least two. That's good.

B (33:37)

Fantastic. So I want to pull one concept from your book, the dark room problem. And with your permission, would you explain what you mean by that? And then is there some way to tie that back to the ad space in some capacity?

C (33:54)

Yeah. So the darkroom problem is a concept with some pride made up. It's a riff, I would say on the, the lemons problem from George Akerloff's Nobel winning paper about the used cars and the information asymmetry challenge. But the darkroom problem is you don't want to enter a dark room. So my metaphor is it's a market space where there's not enough information that you feel comfortable investing either time or effort. So the metaphor is you don't, you don't know what's inside of a dark room. It could, it could have no floor and you could fall, it could have a killer clown and you don't, you don't want to go inside. And so the, the two kind of specifics I use, you know, it's, it's a little bit like the, the lemons problem because you, you don't know what you're buying when you buy a used car. Only the seller knows and the buyer doesn't know. And that creates an inefficiency. In this case, the lack of information creates a lack of investment and the hold back growth in a sector that we could be investing in if only the information. So the story I tell in the book actually relates to my work at Samsung. It was like right after the COVID broke in 2020 in March, and suddenly all of our clients, people knew that we had ACR automatic content recognition data, which tells you about how opted in users are using new televisions and the phone was ringing off the hook and all the clients are saying, well, my gosh, now everyone's staying at home. What are they doing? Like, how do I reach them? What are they watching? I don't, I don't know how they're, how to reach them right now because there's been such a disruptive change. And a billion people, by the way, started streaming overnight globally. And what we did was we put together data, we ginned up a storyboard, like, what do we think people? The answers people want. We created a dashboard, we started circulating that information to clients as insights. And we really put a lot of advertisers at ease about investing into the streaming ecosystem, which before they didn't really understand at all. I mean, which is crazy to think about now because we were doing it for so long. But in 2020 it was that much newer. So they would say, oh, I'm interested in men 18 to 34. And we would say, oh, this is how men 18 to 34 do streaming. They watch this many apps, here are their favorite apps, here's how they spend time with day parts, all that as an aggregate. And so having lit up the dark room with this information about how many 34 stream, then the advertisers are willing to put money in and in the business career room. So it's, I think it's a decent illustration about how data and information at an aggregate level, per my story, allows people to have comfort in a new space that before was just a space of ignorance.

B (36:33)

Yeah, it's funny, that's an optimistic view and it's a really good example. I tend to take somewhat of the pessimistic view with respect to the larger ad space because the challenge is that there are still too many places where things are opaque. And then, then just wearing my privacy hat, there's too many instances where access to data is cut off in the name of privacy. And what that does is it creates a larger trust problem. Because if you just have to trust the black boxes, numbers, you're going to have to do that to a certain extent, I suppose, but it certainly doesn't facilitate trust within the marketplace. And that ends up being kind of its own set of challenges. So I want to finish on one story because I kind of feel like it comes from your book and I think There's a couple of different ways to look at it, and I would just really welcome your thoughts. So the book recalls a story of a meeting with three entrepreneurs who had gamed the system to be able to build a data business on the back of 200 million Facebook profiles. And I'm wondering if you'd share with the audience that story, because I think it actually demonstrates for me both what's bad about the data business, but also what's kind of cool about the data business. Can you kind of start with, you know, recalling that story?

C (37:51)

You know, I guess it's a. It's a disclaimer. I can only report on what I saw and surmised in the room in this one meeting that I described. I can't confirm that any of it was true, but a colleague of mine and I went to go meet with a startup who claimed that they were selling social media data, which, in whatever the hell the year it was, was a valuable thing that we all wanted. And so we go down to this startup, checkbook in hand, ready to sign a license for whatever they have, and they told us, and we didn't know you just gave away the number. But, you know, we. We go in and we say, well, you know, what do you have in terms of social media data? And they said, well, we have really rich data on 250 million Facebook users. And we said, no, you don't. That's ridiculous. Like, why would Facebook be giving you the profile data on 250 million of their users? We were like, oh, it's super easy. We just have this little game app that we get people to opt into, and then when they opt in, and I remember from this time, I don't know if you do, but, you know, when you would sign up for a new app or a game, it would say, do you want to opt in all your friends? Or something like that? And you would say, sure. And so in those days when you could opt in for yourself and everyone in your social graph, apparently, according to, you know, the Cambridge Analytica scandal, you know, that can really work, and you can end up with just mountains of data. So. But this was years before the Cambridge Analytica. I mean, I think maybe six years before the Cambridge Analytica scandal broke, and we didn't know. So we're in the room with these guys, and they're telling us that they have 250 million Facebook profiles, and we're like, no, you don't. And actually, my colleague, who is a lot spicier than I am, just actually kind of Lectured them. He sort of humiliated. It's really nice.

B (39:36)

You are horrible people.

C (39:38)

Yeah. He's like, you're. He's like, you're stupid. You're bullshitting us because my language. You're. You're lying to us. How dare you? This would never happen for these 15 reasons. And we return to Rome and walk out in the house six years later. We read the paper, the Cambridge Analytica scandal happened, and we realized that, I mean, the keywords analytics scandal was basically treated like it was this one off horror. But in fact, they were probably hundreds of companies who had little puzzle games on Facebook who were walking away without much data, I suppose, just not using them for such nefarious purposes.

B (40:12)

Well, you had like the. The flashlight company that was also collecting precise location data and just amassed like a huge database of that type of data. There was. There's definitely a lot of wild west going back then.

C (40:23)

Yeah. And by the way, I shouldn't be laughing about this, but the story is kind of comical.

B (40:27)

Okay, so I'm going to share a story that actually predates the Facebook era. But I knew a group of entrepreneurs and I have no idea how they did it, but they were able to scrape a whole bunch of MySpace profiles. And in doing so, what they discovered at an aggregate level is that consumer behavior on the open web and this is like what 2005 was morphing pretty significantly, and as a result of understanding that they were able to build a business. So there's sort of two sides to look at this. And it's odd that the, that the data guy is going, oh, this is horrible. And then the legal regulatory guy is going, yeah, but there's a. There's a silver lining to some of this where sometimes if you're doing it, I don't want to vouch for how they got this data, but I do want to say that the use case here was innovative and informative, and a number of companies were able to benefit from that because they were able to understand the larger space a little bit better.

C (41:31)

Okay, sorry, I'm not as good on the follow ups as you. I'm sorry, I'm sorry.

B (41:34)

That's okay. That's okay.

C (41:37)

That's a great story. I'm glad you shared that. Thank you.

B (41:40)

I'm also the king of the interviewers who, like, aren't asking questions. I'm just. I. This gives me another opportunity to pontificate.

C (41:47)

You clearly watched a sufficient number of Charlie Roast episodes.

B (41:52)

Well, let me ask a question because. Because what would happen as a result? And look for privacy, for maybe even society. The net net of the Cambridge Analytical scandal was a limitation on the amount of data. But there's a downside to that for anybody interested in using the data. I'm not saying you should get unfettered access, and it sounds like this company was really manipulating a whole bunch of loopholes in the system. But the net result since then has been for large companies to do everything they can to shut down any data access to anybody else. And that leads to the trust problem for the ecosystem as a whole. And I'm just wondering, do you have any thoughts as somebody who's, you know, who wants to be able to get access to data? I mean, does that. How are you thinking about that?

C (42:37)

Well, you've already identified that I'm an optimist and I tend to look at the positive side of these opportunities. I mean, I'm really encouraged by the opportunities posed by clean room technology where companies and entities are able to share signal without sharing data. And I think that poses a lot of opportunities for entities of any kind, commercial or nonprofit or scientific, to share data in a way that allows their partners to get answers without getting too deep into data sharing and everything that that entails. You know, Alan, you're the expert on this, so you can sober me up, but I feel like there's a lot of opportunity for collaboration between multiple entities.

B (43:20)

I agree. I agree. And I think clean rooms are certainly an intriguing tool and I hope there's more investment in that and, and in a whole bunch of other areas within the ad space. We're going to leave it there. But, Justin, first I want to plug the book. I've read it a couple of times and I've actually read the audiobook, which I kind of recommend more than the written. And the reason I do is you read it. And that's probably not always a great idea for an author, but it did feel like it was just obvious that this was you. Like it just you brought your personality into the reading, which I just thought it just was a value add. So, everybody, I would encourage you to read the Little Book of Data.

C (44:04)

Thanks, Alan. I appreciate the audiobook review and review it on Amazon.

B (44:10)

Fantastic. I will.

C (44:12)

The reader should buy it and review it as well. Well, it was a pleasure to spend time with you, Alan. It's been too long and it's great to catch up. And why don't we get a coffee where we can have you be the interviewee and I'll catch up on everything you've been working on.

B (44:24)

Fantastic. That was a great conversation. I realized that the discussion strayed a bit outside of the ad space at times, but I think the larger points Justin makes are directly relevant to anyone working in digital media. Justin is the kind of person who would make a great dinner guest, as he's fantastic at breaking down complicated topics and bringing them into the light, but also doing it in a way that is conversational. Two concepts that I want to highlight from our discussion and from Justin's book. First is the idea of a data bully. The digital media space is certainly complicated and convoluted, but in my experience, the people who do really well in this space are the ones who are able to break through the clutter of complexity. Justin does that as well as anyone. Second, the dark room problem. In his book, Justin uses the example of how the company carfax helped solve an endemic problem in the used car space, where in the absence of reliable information about a used car, most people would assume the worst. So by providing a way to shed light on the condition of most used cars, carfax was able to bolster the larger used car marketplace. But there's an analogous set of challenges in the ad space where there's still way too much darkness and opaqueness. And those challenges are are exacerbated by two privacy trends, and some of those I alluded to in the interview. But the first trend is where large platforms have taken the limiting data use to any third party. And then there's a second trend where the US State privacy law definition of personal data has effectively swallowed the definition of DE identification to the point where it's almost impossible to treat a data set as truly privacy safe. Both of those trends limit data, and both of those trends exacerbate the larger trust problems within the ad space. Finally, I wanted to give one last plug to Justin Evans and his little book of Data. It is a breezy read that uses little entertaining stories to make some really important broader points. I highly recommend it. We've got a bunch of other fantastic guests coming up on the Monopoly Report podcast over the next few weeks. Please subscribe to the show@monopolyreportpod.com or on Spotify, Apple, YouTube, or wherever you listen to your podcasts. And thanks for listening.

A (46:50)

Thank you for listening to the marketexture podcast. New episodes come out every Friday and an insightful vendor interview is published each Monday. You can subscribe to our library of hundreds of executive interviews at marketectures tv. You can also sign up for free for our weekly newsletter with my original strategic insights on the week's news at News Market tv. And if you're feeling social, we operate a vibrant slack community that you can apply to join@adtechgod.com.