
Loading summary
Julia Shulman
Foreign.
Alan Chappelle
Welcome to the Monopoly Report. The Monopoly Report is dedicated to chronicling and analyzing the impact of antitrust and other regulations on the global advertising economy. In this past week's newsletter, I write about the battle over net neutrality and its impact on the digital ads business. Today.
Host
I make the point that some of.
Alan Chappelle
The early concerns around ISPs and web wide collection of data and control over the ecosystem might come off as a little bit quaint today. And I suggest that the real concern when it comes to control over the ad space might just be the edge providers. If you were new to the Monopoly Report, you could subscribe to our weekly newsletter at Monopoly Market tv. You can check out all the Monopoly report podcasts@monopolyreportpod.com I'm Alan Chappelle. This week my guest is Julia Shulman, who is a longtime ad tech thought leader, having worked in legal and senior policy roles at AppNexus and General Counsel and CPO at Triple Lift and now GC and CPO at Tele. I love speaking with Julia because I always learn something. And we may go back and forth a little bit on some of the inside baseball stuff that's gone on over the last couple of years, but I think the business community really needs to hear about some of this stuff. So I'm looking forward to. Let's go.
Host
Hey Julia, thanks for coming on the pod. How are you doing?
Julia Shulman
Thanks for having me. I'm doing well.
Host
Good, good. Hey, so I wanted to talk today about two sort of interrelated themes. So the first one is Google's policy shift when it comes to device fingerprinting. And I covered some of this in the newsletter a few weeks ago and went through some of the historical background, starting with the arms race between the browsers and ad tech and then the Google settlement with the FTC over their alleged hack of the Safari privacy settings. So my question for you what is a device fingerprint? Is it simply browser tracking without an HTTP cookie or is there like a threshold of data that you have to cross to be viewed as fingerprinting? Like walk me through that.
Julia Shulman
Yeah, you know, I did my, my Google research, pun, no pun intended, to make sure I had a crisp definition for you. But in a, I think if you take a step back, you think about it's really just a unique identifier that's created by collecting a number of details about a user's device. So that's, that's not necessarily just a browser. It could be operating system, browser version, screen resolution, installed, plugins, time zone, other hardware and software characteristics. And it's really used to create this distinctive fingerprint that can be used to identify the same device across different contexts. So it could be across different browsers, you know, different mobile oss, it could be across different connected cities and across all of those. And it can also help if someone's trying frankly to mask their identity behind various different IP addresses, et cetera.
Host
Yeah. So for me, I've always thought that the threshold was once you get into fonts or other software installed and you're recording that, then maybe you've crossed over into that threshold. But one of the points that I've sort of made is that I don't know that there's a great definition. And so we're all kind of bemoaning fingerprinting as being bad. But I don't know that there anybody really is aligned on what, what in the world it is.
Julia Shulman
Yeah, no, and I think you're right, the key is it's, it's someone trying to identify the same device across various contexts. And so it's done in very different ways. Everyone has their secret sauce and you do it differently. Again, not being an engineer, but my understanding is done differently across, you know, browsers from mobile operating systems and TV operating systems, you name it. So there's just a lot of different techniques. And then in terms of the thresholds, you know, there are vendors and folks out there that claim their fingerprints are, you know, up to 99.9% accurate. You're smirking because we are on, on video here. I will say I've never worked for a company that was only focused on doing this work. So I haven't nerded out and gotten into the weeds on this as much as I would other things, but that just seems maybe far fetched. But I think most of the industry really don't want to use fingerprints that probably go below like an 80% threshold of accuracy. Because why are you, you know, what are you getting out of that?
Host
No, fair, fair enough. So is fingerprinting always bad? Like what if anything, can make fingerprinting okay?
Julia Shulman
Yeah, and the way that I look at this is, I mean historically, you know, fingerprints really started out being used by financial institutions to prevent fraud. And then that might, you know, they used migrated to different industries, including the app industry. So like any use in our industry, whether it's bad, really comes down to the typical analysis you, me, others would do from a privacy perspective. So you kind of have to use it. Look at the use case. This is just an identifier. So what is the use case for that identifier? Is it for fraud detection? We, you know, unfortunately we have some fraud in the industry. I don't think anyone would argue that the use of a fingerprint to prevent fraud is a bad use case. What about frequency and recency? I mean, you probably watch connected TV like the rest of us and seeing the same ad over and over and over is just a terrible user experience, let alone a total waste of money, you know, for any marketer. And then you kind of go up that spectrum of different uses to analytics, deployment of first party data, third parties, building and enhancing profiles onto like, you know, some nefarious actor out there trying to collect this data to do something kind of wild and crazy. So it really comes down to the same analysis I think we all do across the board, which is, does this make sense? Does the user understand it's going on? Do they need to understand what's going on? Is this use case reasonable? And you know, we'll get to this later. But does a user need to and can it have control over in certain of those use cases?
Host
Yeah, and you bring up the fraud use case and that was sort of the first use case with a company called Blue Kava back in the day. And I remember talking to their CMO and I won't name, name them, I'm not, I'm not trying to name and shame here, but I remember telling him, I said, look, you have successfully gotten your company on Wall Street Journal by coining the term fingerprinting, but that is going to be an albatross for you from this point on. And it turned out that like that.
Julia Shulman
A little, a little like people based marketing. I think that was liveramp, like in 2016.
Host
Yep, exactly.
Julia Shulman
For a whole other podcast.
Host
So we talked a little bit about the, you know, Google has done a little bit of an about face when it comes to fingerprinting and I'm just curious, do you have a take on why Google initially came up with that policy and then maybe more to the point, why are they reversing it right now?
Julia Shulman
And this goes to the heart of why you are so good at doing these podcasts. So let's get into it because this is, I think this is the one that the business folks really need to understand. So first of all, I can't speak for Google, obviously, you know, you can't. We don't, neither of us work for Google. They are, they're rightfully very, very tight lipped around these things. But there's a lot of conspiracy theories out There and then there's a lot of us that just kind of watch what they're doing and either frankly like it's just stuff worked in their favor for the last 20 years. They made the right call 20 years ago or you know, a lot of things had to fall into place. But you know, the decisions are made for two reasons. First and foremost, privacy in a while. But I think the more important reason that they made this decision all these years ago was for commercial reasons. Like any good sophisticated company in the space that understands the value of control over and access to data. And you know, when they made this decision initially the world was a very different world. We was display, you know, first mobile certainly was on the scene. Correct me if I'm getting kind of timelines wrong. And you know, connected TV was sort of just barely out there. And so if you think about the worlds that existed when they made this initial decision, you covered off kind of all the regulatory reasons. But on the commercial side they had a great position in the space. They didn't have to do fingerprinting, number one, so they kind of had the upper hand from a regulatory perspective. They could just say we don't need to do this, so why would we do this? And number two, their browser, the operating system, they had a million user facing features, products, you name it. And so they could detect and use their identities across various contexts. And so it was much easier for them to not have to do this stuff. And they structurally benefited from it by, you know, not enabling folks to do it. And they benefited from a regulatory perspective. So you kind of fast forward from that simple world where they were in a good place on the display and mobile side to the world today. It's a different world. You know, advertisers are trying to truly advertise cross context and that includes connect tv. And budgets are massively shifting into the TV space. You know, I just ran down some numbers and I can't say where this came from but I think it's like 50 billion just the US market alone in connected TV advertising. And Google is on its back foot in the TV space. You know, they don't have the same toehold they had from a user facing perspective in that, you know, space. They certainly have an operating system, but it's not the dominant operating system. And they don't have a lot of products and features that are deployed across, you know, various devices that are used in the connected TV space. So you know, strategically they thought why are we going to go out ahead of everyone else? Like why are we going to be more conservative than other people in the space on this when we're on our back foot and you know, we're being held to maybe a different standard than other folks in the space. So I think if you put that all together, short story TLDR is they were in a very different spot when they rolled out the initial policy. It benefited them back in the day commercially, also made them look great from a privacy perspective. And then right now they are on their back foot in the TV space and you know, why should they be held to a different standard? And then I think we'll get into it with a discussion around maybe ico, cma, but I actually think it benefits then to roll back this policy and kind of get the regulators to really look at what's going on. And either the regulators have to say, okay, you know, the industry is doing what they need to do and this is okay now, or force the industry to kind of make changes and change the way they're operating. That's the key. Like it's a very long play, but it's a very smart play in my mind. Again, don't know if this is their thought process, but.
Host
No, I think that's right. And I also think, and this is sort of related to what you're saying, that part of what's going on here is that they got bit a little bit by the opaque definition of what fingerprinting is. Because one could very easily have said, no, no, no, CTV is very different. I know, but you use IP address as the identifier. And this gets to my point of like fonts and other things. Like that's not, to my knowledge widely used in the CTV space. It just so happens that, you know, the only identifier in a lot of environments is the IP address. So you now have to use that. But having a vague definition of what fingerprinting is now means that maybe that's implicated. And then that's kind of weird for Google because you can't now say nobody else can do it and yet we're doing it.
Julia Shulman
No, it's a great point. You know, I do see those two sides. So number one is to your point, like some of these tools that they have rolled out could fit in that definition. I also think now it's like a longer term play of hey, regulators, you know, we have these other solutions that we've been deploying to the market. Privacy enhancing technologies, privacy preserving technologies. We've been trying to nudge the industry in a different direction. You know, this is one of the ways to solve for this challenge and also, oh, by the way, like, you're holding us to a different standard than other folks.
Host
And so you touched on this a minute ago. But it turns out that the Information Commissioner's Office, which is the UK's data protection regulator, came out pretty quickly being rather critical of Google's policy shift, which I thought was kind of was interesting. I would hope that the ICO would have a little bit more nuance in their opinion here because I do think that if you don't open yourself to the possibility that that technology can be used in a consented environment, then I'm not sure you're doing yourself a favor in terms of crafting those kinds of policies. But I was curious, you know, there's been a lot of activity between Google, the ICO and the cma. Some of it's in with respect to third party cookie deprecation and the privacy sandbox. But like, you know, I'm more of a conspiracy theorist than you are, I think. But like, was this just part of Google trying to maybe drive a wedge between the two orgs?
Julia Shulman
I sort of, that's a great conspiracy theory if it's true. Like, yeah, you and I, let's, we should talk offline about writing a book over time on, on some of this stuff years down the road. But maybe, but honestly, again, I think, I think where they are in the TV space, it is a pivotal moment and they needed to make a move and they're saying, hey, we look, we think we have a solution that either might help the industry from transparency and a consent perspective, you know, this is a solution we could all use and, or maybe it's a solution that enables the industry to operate without, you know, needing to enable consent. And I see you nodding because there's been some conversations about, you know, whether user choice is necessary when, you know, some of these pets are being used. So I think this plays into their longer term game. I don't think that they were necessarily trying to create a, like drive a wedge between the two regulators, but I think from a timing perspective they needed to get out there and I don't think they're surprised that the ICO came out. And I think it plays into them that this is sort of driving a wedge between these two regulators and they're kind of saying, hey guys, like you've put us in this very challenging position where, you know, everyone's coming at us from an antitrust perspective. But you know, from a privacy perspective, we've been trying to the industry in the right direction. And we've been spending and investing, I'm assuming, hundreds of millions of dollars on some of this technology. And you're telling us we can't deploy it for competition reasons. So what do you want us to do here?
Host
Yeah, well, that's really the crux of the challenge here is sort of like you can't on the one hand tell Google that you have way too much control over the marketplace, and then with the other hand say, well, wait a minute, you're not exercising enough control over the marketplace.
Julia Shulman
Totally. I mean, come on. Yes, it absolutely is. And so Google's in a really challenging spot. I know we all have friends that work over there, and, you know, they're, they're really trying hard to come up with some solutions that work for the market, you know, help their company at the same time. And then they're just stuck between a rock and a hard place. And I think this is a really good and interesting move by them because it is the CTV space. Like, this is really where fingerprinting is broadly used. Fingerprinting. You know, we can have the debate over what fingerprinting is, and they're not. You know, I don't think anyone would try to claim that they're in a dominant position in the TV space. So it's, it's not a bad spot for them to have to have this debate.
Host
Yeah, so, so jumping over into the, the TV space, I mean, you took the general counsel and chief privacy officer role at Telly recently, and it struck me, and this is completely from the outside looking in, but they seem kind of unique. In the CTV space, you, you seem to have adopted the model used by the wireless carriers where you're providing the TV for free and then you're subsidizing it on the back end with either like a research panel or advertising. I mean, am I thinking about it the right way? And what is telly, I guess is what I'm coming up with.
Julia Shulman
Yeah, let me, let me. And you'll have to come over, you know, see that, see the device in my place at some point. But taking a step back, high level, your device is different from a traditional tv. And it's just a couple things. First, it's dual screen, which no one has that out there today. So there you've got your kind of traditional theater screen that you think of we all watch TV on, which is just one major screen and then a nice sound bar. And below that is actually a second screen, which is a smart device, so that comes preloaded with games With a virtual assistant, multiple widgets, you name it. And so this overall device is really not just a tv, it's a device, a smart device for your home that also, you know, happens to have a theater screen associated with it. And as you said, we provide that for free to our users, a transparent value exchange, and we support that through advertising. So, you know, from a privacy perspective, user comes, signs up for the device, you could go do this yourself. They fill out a survey. So they respond to us with their interests and, you know, information that advertisers are interested in. And while they're doing this multiple times during the signup process, before they even get a device, we are reminding them about our data practices, we're asking them to consent to our data practices. We're reminding them that if they don't want to participate once they get a device, they can always return the device to us. And then we remind them again during setup. Of course, we hope everyone loves this device, but anytime they can give it back. So it's a really cool, really interesting model, and we've just had great market interest. We've had our users really love it, and we're able to move fast and just constantly ship. Really cool, interesting new features. Again, not just the tv.
Host
Yeah, no, very cool stuff. So turning to the broader CTV space, how do you see that evolving? Will the IP address continue to be the primary point of addressability for most in the CTV space, or will you see things changing?
Julia Shulman
This is a great question. So if this space continues down the same trajectory it's on today, which is all these various streaming services, so much fragmentation across the ecosystem. So kind of multiple different operating systems, different content providers, streamers, you name it. Really not wanting to adopt a standardized identifier. And you compare that to a display or mobile ecosystem where there are just a handful of browsers and two really dominant mobile operating systems, maybe around the world there's another one or two TV space that doesn't exist. So let's just start there. And you're nodding. So multiple operating systems, they all have their different AD IDs at an operating system level. And then from an app perspective, so all the different streamers, you name it, they have, you know, their own qualms about sharing the operating system identifier, their own identifiers. And so everyone's trying to figure out a way to do everything from fraud protection, use cases we talked about earlier up through targeting. I think the industry could either be driven by regulators or frankly, more importantly, by commercial interests to align around a solution. And so that solution could be cooperation on a shared ID solution. You know, we see UID 2, we see, you know, some other ID solutions that are out there, or it could be a standard os, look at Trade desk, what they're doing with Ventura today, or it could just stay, frankly, a mess. And, you know, we could just continue to see IP address or, you know, a number of different ID solutions that kind of come together and maybe, you know, regulators say IP address isn't going to work and you've got to think about something else. And we could talk about that in a second. I think the other interesting thing though is if the industry changes the way content consumption works in tv because all of us consumers were really frustrated by having to go to 14 different streaming services and forgetting which ones were signed up for, et cetera. I do think that might help solve the issue because you're going to end up having whole bunch of these different streamers, a whole bunch of these different content providers cutting deals where things are more consolidated and consumers can go to, you know, one or two different services. We're kind of back to the cable model in the first place. That might, you know, drive a solution and make it a little bit less fragmented.
Host
Yeah. One overarching comment. As I say, we can never underestimate ad tech's ability to remain a mess for longer than anybody thought possible.
Julia Shulman
And so, I mean, and you like how many times I was actually having this conversation last week with someone who's newer to the space and you know, they, they had a theory that the ad tech market, I mean, we all say the ad tech market is a lot like the financial market and the financial market, you know, 6,070s even into the 80s, was very similar to the ad tech market with like a million different broker brokers and dealer dealers and just all these intermediaries. And there was a lot, you know, there was a lot of fragmentation and it just became efficient over time. And they said, well, you know, that's, that's what's going to happen to the ad tech market. And I kind of looked at this person, I said, look, we're not trading billions of dollars. Ad impressions are, you know, like maybe CTV at most, you know, 30 bucks CPM. There's just too many incentives in this industry to keep it opaque and for people to make money by, you know, staying opaque. And I just don't see those efficiencies playing out over time. I don't know, you know, what your theory is, but I just having been in the space now for a While commercial incentives are not driving that direction, either from an ad tech perspective or from a CMO or marketer perspective who just need to hit their metrics. And as long as they can say they hit their metrics, they don't necessarily care. You know how that's happening behind the scenes.
Host
Yeah, there's a whole cadre of like ID less solutions out there and some of them are pretty interesting and pretty cool and like there's a part of me that kind of wishes them well, but ad tech is not going to adopt those other than via kicking and screaming and being dragged into them. And so I think folks are waiting for third party cookies to be deprecated once and for all. Although, boy, who the heck knows when that's happening. And so it's sort of interesting like the, you know, you've got some kind of cool things on the table that probably you're getting short shrift because you know, they're only what, 60%, 70%, whatever the number is as effective as the status quo. And you know, until they get either get closer or the status quo comes down, it's going to be really difficult to kind of move, move that kind of stuff forward just because you're right, the incentives aren't aligned. Have you looked much into the Video Privacy Protection act in connection with the CTV space?
Julia Shulman
Yeah, you know, I think any of us in the CTV space need to understand, understand it, kind of understand the requirements and understand how the industry is reacting to those requirements. So, you know, for folks who are listening, it's the Video Privacy Protection Act. And I, I think we all say, AKA right, the Blockbuster act or the Video Rental History Act. And this is all driven by, correct me if I'm wrong, you know, someone's nomination hearing and his porn history, Robert Bork. Right. And his, his video rental history, you know, was made public and, and there was, there were some unseemly videos in that history. And so it requires consent to share video viewing information. But again, this was drafted way back in the day when we were all walking into blockbusters in our little local video rental stores to rent VHS tapes. And historically, if you kind of talk to folks across the industry, there's been a lot of differences in how people have reacted to this. And I'll go back to our earlier conversation. It's not necessarily reactions because of the privacy requirements. I think people, early days saw the commercial benefit of this requirement from a streamer perspective, content provider perspective, you name it, where they said, oh, we can hide behind this and you know, we don't need to share this information with our various vendors or for selling programmatically. This is something that we think is pretty valuable information about our users. And so even if we could get consent, we don't necessarily want to share it. So I think there's a spectrum. You see some people say, you know what, I'm going to get consent for this. I want to sell mostly programmatic or you know, I just don't want to, I don't want to have to deal with this. Let's just be open and honest to make sure we've got consent. 1 Then there are folks that will maybe share it with limitations. So they'll have like a couple of different vendors that they'll share it with, but they won't sort of more broadly share it. And I think that you kind of see a number of other people these days making it available through clean rooms and through more controlled environments for certain use cases. But it enables them to control the exact use cases that people run against this data. So they can't build enhanced profiles. And then there are some who just won't mess with it at all. And you know, they don't even want to log it, they don't use it. It doesn't really play into their strategy and you know, they're not using it or sharing it in the ecosystem.
Host
Yeah, so I was sort of struck. I was out at the Tom Deerlein gala last night and so it was a whole bunch of old school, mostly business person. I literally might have been the only privacy person in the entire room of a couple hundred people. I had two different folks on the business side telling me that the VPA is going to be world ending. And I was really struck by that. Not that I necessarily agree with it, but boy were they, they were worked up. They. And I think there's a couple of.
Julia Shulman
Publishers you've gotten inter. Yeah, I mean this, this is, that is probably my gut is this is, that's more website focus. Right. They're probably getting hit with like the Metapixel and, and that just for folks listening. So the VPI like we contend with it rightfully for sharing like true viewing activity information. So what are people watching in their apps and fast services and you name it. But then there's another world in which there are class action plaintiffs, attorneys, etc. Out there bringing claims against websites, apps for embedding, you know, third party technology that technically shares small information about like a video running on their website, for example. And that's probably, you know, what they're Most focused on. And yeah, to your point, like, those are. We don't. I don't think we have this on our agenda to talk about some of the class action stuff that's going on. Wiretapping, you name it. But those are really challenging because it's early days. You know, there are statutory damages associated with them, and it's a lot of money to defend yourself against this stuff.
Host
It's funny when, whenever I start thinking about class actions, you have to say, like, how creative a lot of the plaintiffs are there because, like, they have convinced not just one judge. This is now widely accepted that a website playing a video is a videotape manufacturer. That's pretty wild. I mean, that somebody. Now, that took a lot of years to get there, but that's. They're persistent and they're very, very, very creative. So we had, we had Lisa Hutnik on the POD talking mostly about sipa, but we ended up having a VPPA discussion. And I actually think I may end up covering it in the newsletter because it seems like it's becoming an area of concern. I wanted to get your thoughts on how you see choice and consent evolving in the CTV space. And our address is even stable enough to be that mechanism for recording privacy choices.
Julia Shulman
Yeah, I mean, this is a challenging one. I think we've all been talking about this. I think you and I were on the NA board at the same time when we put out the Beyond Cookies guidance. Was it. Was it 17?
Host
So, you know, back in the Groveman era?
Julia Shulman
Yeah. Oh, gosh. Was it?
Host
Yeah.
Julia Shulman
Okay. So are they stable enough? Look, I think there are some providers out there today who have built out solutions and have thought through this, and it's not perfect, but they have some stuff that they can do. I think there are, frankly, others that have no idea how to tackle this challenge today that don't know how to handle unstable IP address, don't really want to nuke entire office buildings and blocks, and they don't know how to manage rotating IP addresses. So I do think there are solutions that can be built, but they are hard. They require industry cooperation. Arguably they require like, ISP cooperation and other cooperation that is very, very tough to get. So, like anyone out there, and I don't have a perfect solution, but I do think if people really banded together and thought through this one as an industry, we could probably build, you know, something that meets kind of basic needs. But for now, we're. We're all on our own. And it is, it is a tough problem.
Host
You know, maybe we get Together with we bring it Andy Dale and a couple of others and maybe we start to work on something like that. I think it would be worthwhile.
Julia Shulman
Should, would that, would David Weinberg kind of come off of his year long sabbatical and get back at the table and roll up his sleeve.
Host
He'd dial in from Tibet or wherever in the world he is these days.
Julia Shulman
But yeah, I mean, I don't, I'm, I'm laughing but oh man, PTSD from TCF days and all that other stuff. And I know you've got war rooms from dnt, et cetera. So for, you know, the business folks listening who think we're just throwing out acronyms and sharing inside jokes, it is no joke to work on an industry solution for, you know, some of these privacy related requirements because there are so many commercial and strategic interests at the table, it isn't very, very, very hard to find consensus. And it's not just ad tech that has to come to the table and come up with these solutions. It's really a solution that has to be deployed by publishers and advertisers and some of their other vendors as well.
Host
Yep, no, great point. And, and I think that probably gets a little under, underappreciated frankly by the business community just how, how contentious a lot of these types of discussions end up being.
Julia Shulman
I mean, yes, I think I aged about a decade in the two to three years of working around the clock on tcf.
Host
So this has been a great discussion and I really appreciate you coming on. But before I let you go, I would love to hear what is your semi secret hobby or passion?
Julia Shulman
Oh, goodness. So believe it or not, I was a competitive archer. I started as a kid back at the Parrot Trap Lake summer camp that I went to in Tennessee. No way. Yep. Yes, I was, I really, you know, it's probably in the same bucket as surfing or you know, anything that any of us do to just sort of have a singular focus, get your mind off of things, not, you know, certainly not have a device or something around you. Obviously back in the day when I started this, I was a kid and devices didn't exist. But now if they do, it's a really, really good way to just get out there, clear your head and have to focus and, and really kind of perfect your art at something and I laugh. Maybe one day I'll be like Geena Davis and you know, try out for the Olympics. I don't know if you know that she is a professional archer and actually just missed making the US Olympic team. I think in 2000.
Host
No, I didn't know that.
Julia Shulman
Yeah. Yeah. So it's kind of a fun hobby also, because you don't have to be, like, 20 years old and in the best shape of your life. You just have to be able to focus and, you know, still be able to see to a certain degree.
Host
Well, if. If the zombie apocalypse ever happens in New York City, I'm. I'm hoping that you and I still remain friends. We can, like, you know, I'm not sure exactly what I'd bring to the table, but, boy, I'd sure want you on my side. We can use the archery skills.
Julia Shulman
Let's do it. Let's do it.
Host
Julia, thank you so much for coming on. This was a great discussion.
Julia Shulman
Yeah, thanks for having me. Thanks, as always, for asking the right questions.
Alan Chappelle
Well, thanks.
Host
We try to have fun here.
Alan Chappelle
That was a great conversation. I always love talking to Julia. She's just so smart and so on it and just a lot of fun to talk to. We've got a bunch of other fantastic guests coming up on the Monopoly Report podcast over the next few weeks. We've got Anna Milicevic of Sparrow Advisors, where she'll be talking to us about reimagining the digital media space without HTTP cookies. Please subscribe to the show@monopolyreportpod.com or on Spotify, Apple, YouTube, or wherever you listen to your podcast.
Podcast Summary: The Monopoly Report – Episode 14: Julie Shullman on Google's Device Fingerprinting Policy Shift
Introduction
In Episode 14 of The Monopoly Report, host Alan Chappelle engages in an insightful conversation with Julia Shullman, a seasoned ad tech expert with extensive experience in legal and policy roles at prominent companies like AppNexus, Triple Lift, and Tele. Released on January 22, 2025, this episode delves deep into Google's recent policy shift regarding device fingerprinting and its broader implications for the Connected TV (CTV) landscape.
Understanding Device Fingerprinting
Defining Device Fingerprinting
The episode begins with a comprehensive exploration of device fingerprinting. Julia Shullman clarifies:
"It's really just a unique identifier that's created by collecting a number of details about a user's device... it's used to create this distinctive fingerprint that can be used to identify the same device across different contexts." (02:04)
She elaborates that fingerprinting encompasses various device attributes, including operating system, browser version, screen resolution, installed plugins, time zone, and other hardware and software characteristics. This method allows for tracking users even when traditional identifiers like HTTP cookies are absent.
Thresholds and Accuracy
Julia addresses the ambiguity surrounding what constitutes fingerprinting versus regular browser tracking:
"Most of the industry really don't want to use fingerprints that probably go below like an 80% threshold of accuracy. Because why are you, you know, what are you getting out of that?" (04:18)
She emphasizes that while some vendors claim near-perfect accuracy, the industry generally favors fingerprints that offer at least 80% reliability to ensure meaningful data without excessive false positives.
Google's Policy Shift on Device Fingerprinting
Initial Policy Rationale
Alan prompts Julia to discuss Google's initial stance on device fingerprinting and their recent policy reversal. Julia provides a strategic analysis:
"The decisions are made for two reasons. First and foremost, privacy in a while. But I think the more important reason that they made this decision initially was for commercial reasons... they made the right call 20 years ago or... they benefited structurally by not enabling folks to do it." (06:37)
Originally, Google's policy against device fingerprinting aligned with privacy concerns and commercially advantaged them by maintaining control over data without needing to employ fingerprinting techniques.
Reasons for Reversal
Julia posits that changing market dynamics, especially in the CTV space, have prompted Google to reconsider its stance:
"Strategically they thought why are we going to go out ahead of everyone else?... they're on their back foot in the TV space... it benefits then to roll back this policy and kind of get the regulators to really look at what's going on." (06:53)
Facing increased competition and a fragmented advertising ecosystem in connected TV, Google may be seeking regulatory favor or striving to level the playing field by allowing fingerprinting to enhance their competitive positioning.
Regulatory Interactions and Challenges
The conversation touches on Google's interactions with regulators like the UK's Information Commissioner's Office (ICO):
"I think if you don't open yourself to the possibility that that technology can be used in a consented environment, then I'm not sure you're doing yourself a favor in terms of crafting those kinds of policies." (12:02)
Julia suggests that Google's policy shift may be an attempt to influence regulatory perspectives, advocating for privacy-enhancing technologies while challenging the current definitions and standards related to fingerprinting.
Regulatory Landscape and Implications
Video Privacy Protection Act (VPPA) and CTV
The discussion shifts to the Video Privacy Protection Act (VPPA) and its relevance to the CTV sector:
"It requires consent to share video viewing information... historically... people early days saw the commercial benefit... some [companies] say, oh, we can hide behind this and we don't need to share this information with our various vendors." (25:22)
Julia explains that the VPPA, originally designed for video rental services like Blockbuster, now impacts how CTV platforms handle user viewing data. Companies are grappling with consent requirements, balancing privacy obligations with commercial interests in data-driven advertising.
Regulatory Criticism of Google
Alan highlights the ICO's critical stance on Google's policy shift:
"I would hope that the ICO would have a little bit more nuance in their opinion here... was this just part of Google trying to maybe drive a wedge between the two orgs?" (12:57)
Julia acknowledges the complexity, suggesting that Google's actions might inadvertently create friction between regulatory bodies by pushing boundaries in privacy and competition standards.
Choice and Consent in the CTV Space
Challenges in Implementing Consent Mechanisms
Alan inquires about the evolution of choice and consent within CTV:
"How do you see that evolving in the CTV space.... is it stable enough to be that mechanism for recording privacy choices?" (26:53)
Julia responds by acknowledging the technical and collaborative hurdles:
"They require industry cooperation... it's really a solution that has to be deployed by publishers and advertisers and some of their other vendors as well." (29:30)
She underscores the necessity for unified industry efforts to develop effective consent mechanisms, given the fragmented nature of the CTV ecosystem.
Potential Solutions and Industry Cooperation
Julia emphasizes the need for collaborative solutions:
"I do think if people really banded together and thought through this one as an industry, we could probably build, you know, something that meets kind of basic needs." (29:10)
However, she remains skeptical about the likelihood of achieving consensus due to diverse commercial and strategic interests.
Future of CTV and Ad Tech
Fragmentation vs. Standardization
The conversation explores the fragmented state of the CTV market compared to more consolidated ecosystems like mobile and desktop:
"Multiple operating systems... different content providers, streamers... trying to figure out a way to do everything from fraud protection up through targeting." (18:07)
Julia suggests that without standardization, the CTV advertising landscape may remain chaotic, with various identifiers and methods hindering seamless data usage.
Ad Tech's Resistance to Change
Alan compares ad tech's complexity to financial markets, while Julia disputes the likelihood of similar efficiency gains:
"Ad tech market is a lot like the financial market... but... too many incentives in this industry to keep it opaque." (20:48)
She argues that unlike financial markets, where transparency can lead to efficiency, ad tech's vested interests in opacity and data control impede similar progress.
Prospects for Privacy-Enhancing Technologies
Despite the challenges, Julia remains open to the potential of privacy-preserving technologies, albeit recognizing their current limitations:
"There are some providers out there today who have built out solutions... it's not perfect, but they have some stuff that they can do." (27:45)
She stresses the importance of industry-wide collaboration to refine and adopt these technologies effectively.
Personal Insights and Conclusion
Towards the end, Julia shares a personal anecdote about her passion for competitive archery, highlighting a facet of her life beyond the ad tech realm. This segment humanizes the technical discussions, providing listeners with a glimpse into her multifaceted personality.
In wrapping up, Alan and Julia reflect on the complexities and ongoing challenges within the CTV and ad tech sectors. Julia reiterates the importance of industry cooperation to navigate privacy regulations and competitive pressures, while Alan appreciates the depth of the conversation.
"It's really a good and interesting move by them because it is the CTV space... not a bad spot for them to have to have this debate." (14:35)
Key Takeaways
Device Fingerprinting Defined: A method of uniquely identifying devices using multiple device attributes beyond traditional cookies.
Google's Strategic Shift: Google's reversal on device fingerprinting policies is likely driven by evolving commercial interests and challenges in the CTV market.
Regulatory Pressures: Interactions with regulators like the ICO and compliance with acts like the VPPA significantly impact ad tech strategies.
Consent Mechanisms in CTV: Developing effective consent solutions in a fragmented CTV ecosystem remains a major challenge requiring industry-wide collaboration.
Ad Tech's Inherent Complexity: The ad tech industry's resistance to transparency and standardization poses ongoing hurdles to efficiency and privacy compliance.
Future Outlook: The CTV advertising landscape is poised for potential transformation through technological advancements and regulatory interventions, but substantial obstacles persist.
Notable Quotes
Device Fingerprinting Definition:
Julia Shullman: "It's really used to create this distinctive fingerprint that can be used to identify the same device across different contexts." (02:04)
Accuracy Thresholds:
Julia Shullman: "Most of the industry really don't want to use fingerprints that probably go below like an 80% threshold of accuracy." (04:18)
Google's Initial Policy Rationale:
Julia Shullman: "They made the right call 20 years ago... they benefited structurally by not enabling folks to do it." (06:37)
Regulatory Influence:
Julia Shullman: "They have these other solutions... you're holding us to a different standard than other folks." (06:53)
Ad Tech Fragmentation:
Julia Shullman: "Ad tech market is a lot like the financial market... but there are too many incentives... to keep it opaque." (20:48)
Conclusion
Episode 14 of The Monopoly Report offers a nuanced examination of Google's policy shift on device fingerprinting and its ramifications within the CTV and broader ad tech ecosystems. Through Julia Shullman's expert insights, listeners gain a deeper understanding of the intricate balance between privacy, regulation, and commercial interests shaping the future of digital advertising.
For those interested in the evolving dynamics of big tech and antitrust issues in the advertising economy, subscribing to The Monopoly Report via Marketecture.tv is highly recommended.