← The Monopoly Report
The Monopoly Report
Episode 74: The Attribution Cartel: Why “Privacy-Safe” is NOT what it’s made up to be
00:46:04
Loading summary
Transcript70 lines
- [00:00]
A
This is adtech God and I command you to listen to this house ad. So if you're listening to this show, just know that you've really stumbled upon a giant network of content across advertising, marketing, media, publishing, and of course the people that work in this great advertising industry. So go to market, Check out all of our brands. We have multitude of shows from the Brand Forum, the Advertising Forum, the Monopoly Report, the Ad Tech God, Pod, the Market, and more. We are bringing more podcasts to our network. We are consistently and constantly bringing on new shows. So check it out. Market or search for any of those brands in the app that you use to listen to this podcast. Enjoy the show and see you all soon.
- [00:59]
B
Welcome to the Monopoly Report the Monopoly Report is dedicated to chronicling and analyzing the impact of privacy, antitrust and other regulations on the global advertising economy. I'm Alan Chappelle, I'm outside counsel and fractional CPO to a bunch of tech companies and am the principal analyst at the Chappelle Regulatory Insider, which is a monthly report that shares strategies and insights for digital media worldwide. If your company has a relationship with Forrester, eMarketer, IDC or Gartner, you should check out the Chappelle Regulatory Insider because we are picking up on data points that the other analyst firms are missing. You can find a link to a sample copy of the Chappelle Regulatory Insider in the show Notes this week my guest is Don Marty. Don Marty is the founder of Eludu llc. He works on web ecosystem and open source business issues, including collaborative research on software incentivization and the impact of advances in consent management and tracking protection technology. He has served as the VP of Ecosystem Innovation at Raptive, as a strategist at Mozilla and editor of the Linux Journal. And Don started the California Authorized Agent project at Consumer Reports Digital Lab that led to the development of Consumer Reports Permission Slip service. He is currently an invited expert for the Privacy Working Group of the World Wide Web Consortium. Don also is building a services company focused on what he calls the post surveillance economy, helping publishers and brands get ahead of the tracking wave before it crashes onto them. Today, Don and I are talking about Attribution, specifically a technical standard being developed right now inside the W3C that its backers are claiming to be privacy preserving. Don recently wrote a piece for Ad Exchanger that takes the position that this W3C attribution standard is anything but privacy preserving, that the standard builds on a system that creates direct incentives for more surveillance, more fraud, and ultimately further consolidation. Don and I also get into the research on why premium publisher environments produce better ad performance and why the current attribution infrastructure and is almost perfectly designed to hide that finding. It's a conversation about standards incentives and who actually controls the measurement layer of the Internet. So let's get to it. Hey, Don, thanks for coming back on the pod. How are you?
- [03:32]
C
Oh, not so bad. How are you doing, Alan?
- [03:34]
B
I'm doing fantastic. So I think it's been at least a year since I've had you on this podcast. And so, you know, what have you been up to?
- [03:42]
C
Oh, well, I've written some articles for the Reynolds Journalism Institute, and mostly I'm working on a new services company on the post surveillance economy, trying to help out some publishers and brands that are trying to get out ahead of all this surveillance stuff.
- [04:02]
B
Ah, fantastic. You know, I just saw a research report cross my threshold. I'll throw in the show notes. I don't remember it off the top of my head, but. But they made a pretty convincing case for the value of good privacy practices. And I think, I feel like every once in a while one of those studies sort of pops up. This one seemed to have a little bit more methodological fervor around it or girth to it, which is good. I feel like it's one of those things that privacy people often say and they want it to be true. It's helpful to have some. Some evidence that it is true.
- [04:38]
C
Yeah, definitely. And a lot of the time, the most engaged and most valuable customers are the same people who put in the effort with the privacy tools and settings to make themselves less measurable by conventional ad tech and Martech. So really, by doing the privacy stuff, you're not just avoiding some of the compliance taxes. You're also getting a better idea of who the real customers are and what they're interested in, and not just the more trackable subset.
- [05:13]
B
Well, let's coin that right now as the Marty paradox. And because I. I think there's a lot of truth to that and to the extent that one can kind of get behind the scenes and, and suss that kinds of stuff out, I just think is super helpful. You, you've been writing a bunch, and I know that you've shared some of your thinking with me. Attribution and measurement. And I would just start with the premise that, you know, both those things often get overlooked within the larger ad space because everybody just wants to talk about ad targeting. And now agentic. And so you really written recently an interesting piece for Ad Exchanger on a push to create an industry standard for attribution. And so, just so we're on the same page, how are you defining attribution here?
- [06:02]
C
Well, I'm going to apply a general definition to it, which is correlating some kind of an event, such as a person being exposed to advertising, with some kind of outcome, such as that person choosing to buy something. And usually people say impressions when they're talking about the events that are happening and then conversions for the outcome. But it's really a fundamental measurement of does this ad work for the result that I wanted to get out of it.
- [06:35]
B
So how would you characterize the current state of attribution within the ad space?
- [06:41]
C
Oh, it's kind of like almost any other kind of science, where there's the paywalled version in the PDFs and then the very accessible, highly promoted version out on social. And just as you're probably better off making personal health decisions based on the boring PDFs, you're probably going to get better results for ad measurement, going with some of the more scientific ways of doing it. Check out Rick Berner at Central Control, for example. And you're probably going to get worse results when you go with the types of attribution measurement that are being actively promoted by big platforms.
- [07:30]
B
Yeah, and this has sort of been a bit of an endemic problem within the ad space almost since day one. So way back before I was a privacy guy, I very briefly did tech sales. I was at DoubleClick, of all places, and they had this really cool ad measurement product. It was about three times as complicated and three times as expensive as the product offered by a company called Dynamic Logic run by Tom Deerlein and his friends. They kicked our ass every single day of the week to the point where DoubleClick had to sell their ad effectiveness product to Dynamic Logic because nobody really wanted it. So I think they sold it for pennies on the dollar. My point being, though, the problem in the ad space is that complicated is often not our friend. Simple, easy, and if the arrow points in the right direction, all the better. And I'm not sure how you overcome that. We're going to talk about a whole bunch of other challenges. But, like, fundamentally, to me, isn't that really the problem here or is it bigger than that?
- [08:35]
C
And it's sort of a problem with how ad people sometimes fall into thinking about effectiveness. So one ad agency will come up with something that's totally creative and different and gets people's attention. And instead of learning from that, hey, wait a minute, the effective thing to do in Advertising is do something new and different that gets people's attention. You'll see a lot of other agencies copy the exact same thing. And it's kind of the exact opposite of the lesson that you can learn from it if it's available. And so I would agree that there's a simplicity effect and there's also kind of a we don't want to look dumb effect where you'll get some of the big platform company people standing up and saying some complex mathematical assertions. And some of the marketing decision makers don't want to admit that they don't understand the math, so they kind of go along with what the big platform's pushing them toward. And sometimes you need to take a step backward and you'll actually look smarter if you admit to not understanding things sometimes.
- [09:58]
B
Completely agree. And that's actually one of the things I've learned in my career is that you're actually so much better off by playing a little bit dumb and being willing to be called out for maybe being uneducated in a certain area. Because, boy, there's a lot of complexity around here. And I think there's certainly a human nature component in all of this. But your article starts with a description of a W3C working group that is seeking to create a privacy safe standard for attribution. But I think the point being that the standard will actually make real world privacy worse. So walk us through your thinking there. I mean, how does a system marketed as privacy preserving end up creating more surveillance incentives?
- [10:43]
C
Well, the first place to check is the definition of privacy. So in the real world, when people think about privacy problems, they're usually referring to one of two things, either deception or discrimination. So either they're being persuaded that something that's not true is true because of something that somebody out there knows about them, or somebody knows something about them and they're using that to treat the person in an unfair way. So privacy harms are things that philosophers take a long time to explain in some really good and thought provoking material. We link to it in W3C Privacy Principles. But privacy harms as they're felt are way different from the quote, privacy that's being offered by so called privacy preserving or privacy enhancing technologies. What those systems are offering is mathematical properties of the system, such as we can prove using math, that we can't pick you out of a crowd of any smaller than K people. But in reality, most of the facts that I don't want used against me are also things that are true of way more than K other people. The privacy that's being preserved by some of these technologies is not relevant to the privacy as experienced in normal human life or in, in a market.
- [12:36]
B
But also if the privacy solution requires one to have a PhD in data science in order to understand and perhaps be comforted by it, recognizing that most people, myself included, do not have PhDs in data science, how valuable is that? Just generally speaking, that's a lot of
- [13:00]
C
the research that came up at FTC Privacy Con. There were some really good user research papers stating that people are actually less comfortable having some inferences about them processed than they are about some of these individualized facts or, or individualized tracking. And I can give you those links for the show notes. It's some good research.
- [13:28]
B
Yeah, there, there's some, I'm probably familiar with a couple of them, but there's a bunch that I'm sure my audience and I haven't gotten a chance to check out, so we'd love to have that. Why is so much of the effort in the ad space around privacy seemingly focused on the wrong things?
- [13:45]
C
I think a big part of it is the impulse toward centralization on the part of some of the large platforms. I know everyone is sick and tired of hearing yet more tales of Google Privacy Sandbox, but the idea of taking centralized control of a large industry is very appealing to some of the decision makers at those companies.
- [14:17]
B
I get why they want to do it. Here's what I have trouble understanding. It feels like those efforts are often supported by people that one would otherwise refer to as privacy advocates. I mean, like a large number of the folks within the W3C I think legit are trying to find good privacy solutions. How would you characterize the thought process of someone who really, really does legit care about privacy being co opted into one of those types of programs like the Sandbox?
- [14:53]
C
Yeah, it's. It's sort of a boundary testing thing. So when Privacy Sandbox was first announced, Google came out and said, hey, we're going to do this mathematically restricted processing on people's web usage. And that will be something that means we don't have to do fingerprinting. And of course later on in the privacy sandbox process, Google ended up saying, well, fingerprinting is back on the menu. So the appealing offer of well, if we do this privacy preserving math, then we won't do some other thing, third party cookies, fingerprinting, whatever, that's an appealing offer to start with. And then I think more privacy people are seeing now that that's not the final offer that's testing your boundaries to push the browser toward a more intrusive state.
- [15:56]
B
So I always thought that part of Google's strategy with the sandbox was to recreate something that at least from the outside trying to look in seemed very similar to what Apple does. I mean, you had cohort models and noise injection and maybe the goal there was to decide, well, if the world craps on what we're doing, maybe they look across the street and take a closer look at what Apple's doing. And maybe I'm completely off.
- [16:25]
C
And Apple of course, has their own advertising business as well. So if you're selling iOS apps, for example, you have a choice of putting that money into social search or into ads on Apple's own app store. So there's not, there's not so much of a bright line between Apple being for privacy and then say Google and Meta being for tracking, as a lot of the online arguments would have it. There are all these companies have their own advertising operation and their own motivations to have some advantage or moat in access to data.
- [17:15]
B
So we talked a minute ago about research and there's a pretty consistent body of research from comscore and Moat analytics, newsworks, I think even the trade Desk. And the research is showing that ads perform better in trusted premium contexts. Now you refer to this as the halo effect. And so if attribution were genuinely accurate, it would surface that same type of finding. But it seems like you're confident that that type of finding is going to elude us because of the nature, I assume the nature of the data processing within what you've referred to as the attribution cartel.
- [17:55]
C
Yeah. In a sustainable advertising medium, you get a kind of feedback loop going on between an ad supported context, such as a TV show, a magazine or a website and a brand. So a brand will spend money on advertising, they will build up their reputation in some ad supported context. That context gets the ad money, they can put it into more expensive content. That brand builds up the reputation enhancing properties of that context, which gives more brand value to that brand, which means they have more money to spend on advertising and so on. There's kind of a, a positive feedback loop between ad supported context and advertising using brands in a sustainable medium. Unfortunately, what we have with ads on the Internet now is a commodification loop where you'll have a system such as the best known example would be Google Performance Max, that is searching out the least expensive possible ad supported context for the most expensive possible app. And so rather than placing those Ads in the, in the higher value contexts where you get that kind of positive feedback loop going, those ads end up chasing cheap eyeballs and the brand loses their opportunity to build brand equity and the publisher, the owner of that ad supported context loses their opportunity to get more money to invest in content.
- [19:47]
B
Like I get the theory. Has this been sort of demonstrated through research? Is that some of the research you're going to share in the show notes?
- [19:54]
C
Yes, that is all that great research coming out of those multiple sources.
- [20:02]
B
Got it. And you know, I probably should have said this a little bit more clearly at the beginning. You know, there's a set of standards that are being created around attribution within the W3C. This whole thing started, it's kind of as an offshoot of what I'll call the privacy sandbox discussions and debates. And now the program is just called Attribution. I think it was called something else before that, Private attribution ipa.
- [20:28]
C
IPA was a different one. This one used to be ppa and now it's just Attribution.
- [20:34]
B
Now it's just attribution. They wanted to try to keep it simple. So you've got a number of entities who are trying to create this standard within the W3C. They're saying, yeah, but Don, you don't care about privacy. Why not? Like this is, this is making the world better for privacy. And what I'm hearing is that actually the research demonstrates exactly the opposite of that. Consumers, many of them, because I don't think they're a monolith. But many of them don't necessarily feel that the target that this group is collectively aiming for is going to scratch the itch of the privacy itch that there, that people really care about. Is that sort of the gist?
- [21:17]
C
That's part of it. And unfortunately it's much worse than that because the attribution system as it's currently proposed creates incentives to do more and worse privacy harms. I can walk you through that. It's based on.
- [21:38]
B
Yes, please.
- [21:40]
C
It's based on the fraud properties. So they say. Fraudulent registration about impressions is a particular concern with the Attribution API. Because impressions are stored only on the device, it is not possible to apply server side intelligence to identify fraudulent impressions and exclude them from attribution. So fraud is an open, unsolved problem. Now, if you look at fraud, at first glance it looks like, well, that's just the problem for the legit sites. Some legit site is showing somebody an ad and thanks to their reputation or halo effect or whatever, they help make a sale, but the attribution for that sale gets stolen by some fraud site. So fraud looks bad for the legit publisher. The problem though, and why the attribution proposal is bad for the user too, is how would a fraud site go about stealing attribution? Well, they would find somebody who's just about to buy, sneak an ad in front of them and they've got credit for that conversion.
- [23:00]
B
Right, that's sort of the oldest trick in the performance marketing book.
- [23:03]
C
Yep. And the attribution proposal hides that old trick. Some really sneaky ad tech player who's figured out how to listen to the mics that are on smart TVs and figure out who's talking about buying stuff. Well, that player, they can't go out and sell smart TV mic conversations directly. They'd be on 404 Media in no time. But what they can do is launder that data through the attribution system by using it to place those conversion stealing ads.
- [23:42]
B
Wow. Okay. The companies involved here include what meta, Google, Apple, Mozilla. These are some pretty smart companies collectively. They must know this. So is what you're asserting that they know this and that they've opted to address different problems than this and just sort of leave this as a known unknown? Or is this, as they say, a feature and not a bug?
- [24:10]
C
Well, there's so much fraud and noise that adding a little bit more fraud into the system is not a big news story. It's a dog bites man story. If you look at things like the FBI putting out a warning about Google Search or Meta getting kicked out of IAB Sweden for their fraud problem, those are a lot bigger and more obvious stories than attribution fraud.
- [24:45]
B
Right. It still isn't something that's super covered other than maybe. You know, I hear from Dr. Foe every once in a while, but it doesn't seem like it's one of those things that cracks through the cacophony of noise that is the ad space. Is that sort of the problem?
- [25:00]
C
Sure, sure. It's like if some mud gets spilled in a nice clean restaurant with white tablecloths, that's going to get noticed right away. But if the street's already muddy that then nobody's going to notice it.
- [25:16]
B
Yeah, we have a little bit of problems of that nature operating in Washington D.C. these days and perhaps we'll just leave it at that. So you've used the phrase attribution cartel to describe the metas, Googles, Apples and Mozillas sort of operating together inside the W3C and it struck me as a, as a pretty strong term. So I would invite you to make the case. So what does that group have in common that earns that label and what's your sense of what you know, the Microsoft's and the Amazons, other, you know, big tech companies, are they taking a position on this standard?
- [25:56]
C
Well first, how do you define the cartel? There has to be a rule for who's in, who's out for the attribution cartel. That bright line is whether or not to seek consent. And so all of the attribution cartel companies are saying that our system is something that doesn't require consent from the user. Our system is something that isn't required to honor opt outs or objections. So really that sticking a flag in the ground and saying our stuff is a no consent zone that really defines the membership of the cartel. And as far as Microsoft goes, I think Microsoft has been much more cautious in communicating about stuff like this and they tend to be more forward thinking and avoiding some of the issues. So if you look at some of the stuff that Julie Brill has written about Microsoft doing GDPR everywhere, about Microsoft doing CCPA USA wide, Microsoft is more willing to consider the downstream impacts and not just jump straight into this we don't need consent Leroy Jenkins thing.
- [27:28]
B
Yeah, it helps that that the ad space has not really ever been their, you know, their, their core revenue stream.
- [27:36]
C
Yeah, yeah. And they have, they have situations like hey, what if a contract developer in Canada is working on a European code base and the data is about people in the usa so they know they have a bunch of partner companies and by the time they start drawing on their whiteboard about how are we going to do GDPR when different companies working on the same data and the same code are in different countries, then that forward thinking attitude results in saying too many edge cases, we're going GDPR everywhere.
- [28:17]
B
Well, and you've also pointed to the W3C as being a source of some of the problem here. And the W3C is supposed to be a pretty neutral standards body. And you've written, and I'm not sure I disagree with this, but you've written that it's become a bit of a destination for big tech forum shopping. And so I'm curious, how did that happen? What drove the W3's lack of a coherent antitrust policy? What does that cost the broader web ecosystem?
- [28:49]
C
Yeah, if you look back at 2019, 2020, the point in time when Google was looking for where to release their privacy sandbox stuff at that point, W3C was still part of MIT and W3C had an extremely basic competition policy. So the alternative would have been something like IAB Tech Lab, which launched Project Rear pretty much around the same time. But IAB Tech Lab had a really pro level antitrust policy that would have made a lot of the kinds of discussions Google wanted to have on Privacy Sandbox. It would have introduced a lot more friction.
- [29:39]
B
Well, that was one of the challenges, I think, from day one of Privacy Sandbox, because it felt like Google and really the folks working at the W3C were only too willing to pretend that this was an industry stakeholder process, when the reality was is that Google had created a forum for getting a limited amount of input into a tool they were building. I'm not sure what they were doing was necessarily bad, but it felt a little bit disingenuous to hold it out as the first thing when it so very clearly was the second thing.
- [30:11]
C
Yeah, yeah. And it could have gone a lot worse than it did. Fortunately, the Competition and Markets Authority in the UK started paying attention to that process. So I think we've got James Rosewell to thank for raising the issue.
- [30:33]
B
Yeah, I don't think they get enough credit for that.
- [30:35]
C
Yeah, the CMA reports really checked a lot of Google's homework that they had been making assertions about in the W3C process.
- [30:47]
B
So you have pointed to the Tech Lab, the admap, as a workable alternative to this attribution proposal. Now, so for listeners who haven't dug into the technical specs, what does Admap do differently and, and why is it that its architecture makes fraud harder rather than easier?
- [31:06]
C
Well, I think it comes back to the definition of privacy. If you're looking at the very extremely mathematical theoretical definitions that people like to throw around at W3C. Well, ADMAP doesn't have those. In ADMAP, you do have a record of the person who's being matched, and that record is present on both sides of the matching process, both on the publisher side, as in who would have seen the ads? And on the conversion side. What AdMap gives you, though, is it's compatible with Right to Know, which is a right that we have under CCPA and state privacy laws. Now, of course, Cal Privacy has pointed out that getting through Right to Know may not be as straightforward as it could be. But if you have some kind of a research program organized by an NGO or an academic research group, then you can use those right to nose to spot some of the problematic data practices that would be hidden by a more complicated and heavier math system.
- [32:37]
B
Do you have a sense that folks are actually doing that? Because I see a lot of requests and it seems like there's a component where people just want to extract their pound of flesh. There's a component where people don't understand how stuff works and are legit frustrated and trying to figure it out. And heck, most of the ad space certainly doesn't help itself or consumers with that. But I don't think I've seen a ton of requests that, at least as far as I can tell, are trying to dig deeper to really understand how the space works.
- [33:13]
C
Yeah, and if you're doing a research project on right to nose, let's say you're at an NGO or you're an academic and you've managed to get funding to do some of this research, then you're going to get a bunch of people signed up and going to give them a standard process to go through and something they can cut and paste into the form to try to get their info from a bunch of companies. And that's honestly a lot of work. And NGOs and academics are on project based funding or grants or some student is doing the research for their dissertation and they want to graduate on time. So you don't really get the kind of sustained effort on right to no's unless one of those projects lines up as far as funding and, and available researcher time.
- [34:15]
B
Yeah, I think that that makes sense. And yeah, I guess that explains why why I haven't seen too many of those out there. So I want to shift gears. You've, you've written that state legislators need to pay as much attention to privacy enhancing tracking schemes as to traditional tracking. And so what in your opinion, does a good regulatory oversight of something like attribution actually look like? You know, what should say, for example, the Cal Privacy team be asking for that? They currently aren't.
- [34:47]
C
You first of all need to start from the real world privacy harms and work backward. So there's a lot of research and enforcement that's around kind of gotcha research like oh there's a mistake in your JavaScript and this cookie got set to this thing when it should have been set to this other thing that's kind of suboptimal because you're not finding the data practices that led to those harms. So rather than saying we're going to enforce based on did this specific identifier get transferred from one company to another, then the best way to write the laws is in very general terms Focusing on human harms. And that's also the best way to research and enforce those laws.
- [35:52]
B
I would agree with that. But I would say that oftentimes that's just not the case. They tend to focus on very ethereal things. Or, you know, right now the Cal privacy folks seem to be very, very fixated on consumer based rights. And we'll see. They may very well be proven right here that the majority of Californians or a or significant plurality of Californians really think this stuff is important. But that seems to be what they're betting that they're sort of betting on.
- [36:22]
C
Yeah, yeah. And right to know is incredibly important because if you have a state regulator that will pay attention when companies aren't doing right to know properly, then that facilitates a bunch of those research projects that might otherwise have been impractical to do. So you've really got, as a person trying to avoid privacy harms. You've got the state government working for you, you've got independent researchers and software developers working for you, and you've potentially got lawyers working for you. When those three groups kind of go off in their own directions. If a software developer decides I'm going to go off and write, not track and completely ignore the law, or if a lawyer says I'm going to sue somebody over a JavaScript mistake and not pay attention to whether there was any concrete harm, or a legislator says I'm going to copy California and go off and do another CCPA clone law even though we know we can do better there, that's problematic. Those, those three protectors of privacy need to do a better job of working together.
- [37:46]
B
Completely agree. So this attribution standard, let's just say hypothetically it gets deployed into the major browsers before anybody has the time or energy to coalesce around an alternative, what does the ad ecosystem look like in three years? You know, who's going to win, who's going to lose?
- [38:05]
C
Well, first of all, the attribution tracking numbers are going to come out very closely matching whatever ad placements are being output from Google Performance Max or Meta's algorithms on Facebook and Instagram. There's absolutely no way that these large companies are going to build an attribution tracking system that says, here's a report on which ads are effective. That report is going to match the recommendations that those companies made to advertisers in the first place. So there's very little surprise or suspense in what the attribution reports are going to say as a result of that. From an advertising allocation point of view. In the short term it ends up as more of a duopoly kind of decision. Instead of say in the late 70s when you had a bewildering array of eight bit microcomputers, it's going to be more like, oh, are we getting a Mac or a PC? What percentage of our ad money are we going to put onto Meta? What percent are we going to put onto Google? It really would end up turning the entire advertising industry into a much less interesting space to work.
- [39:40]
B
So what can folks listening to this do about that? Is it a question of providing feedback into Cal Privacy? Because I know they're very open to feedback. Is it just continuing to work within Tech Lab to see if you can get that standard kind of pushed over the finish line as it were, or is there something else they can be doing?
- [40:00]
C
Well, participating in IAB Tech Lab is always good. If you are a member of W3C, then paying attention when there are competition problematic events going on is always a good thing to do. And I would say get involved in the political process at the state level. There's way too much sort of coordinated lobbying or semi astroturfing by Google and Meta around state privacy laws when they come up. So if there's a privacy bill hearing, then small businesses and business organizations in that state will get communications from Google and Meta saying come out and support us against the privacy law. And that's not really advocating in your own interest for a legit business. Advocating for more legit advertising and appropriate privacy protections is a better approach to take at the state policy level.
- [41:17]
B
Well, and as you and I occasionally go back and forth on LinkedIn, I think you've made you said something that I thought was particularly insightful, that a lot of the regulatory effort today is focusing on the privacy harms created by interoperability rather than the privacy harms created by the mass collection of data by a small handful of companies.
- [41:41]
C
Absolutely.
- [41:42]
B
We need to figure out a way to change that rule set.
- [41:45]
C
And the real privacy harms that can happen behind the firewall at those big platform companies are much worse than the kind of stuff that gets addressed by a lot of the nitpicking of ad tech and cookies and JavaScript in the browser. Look back at the ProPublica stories that started in 2016 about housing discrimination on Facebook and how long Facebook was able to hold out with the discriminatory housing ads, even in the presence of a lot of regulatory attention.
- [42:32]
B
Yeah, good point. So Don, this has been a great discussion. Where can people find you these days?
- [42:37]
C
You can find me on my website and I'll put a link for the
- [42:42]
B
show notes okay, Fantastic. Don, Marty, this has been a fantastic discussion as always. We'll have to have you back on again soon.
- [42:48]
C
All right, thanks Alan.
- [42:51]
B
That was a great discussion. A couple of moments from this conversation that I think are worth emphasizing. The first is what I'm going to call the Marty Paradox. And yes, I coined that on the pod. You heard it here first, folks. Don's observation that the most engaged, most valuable customers are often the same people who've taken the most deliberate steps to make themselves less measurable by conventional ad tech. Which means that if you're an advertiser that is optimizing purely for trackability, you may be underweighting your best customers. I also want to mention the fraud laundering mechanism Don described. The attribution proposal that's being pushed within the W3C as it's currently structured anyway advocates for the storage of impression data only on the device. Now, local storage certainly can enable the owner of the device to have more control, but there are trade offs. Don notes that one of those trade offs is that it makes server side fraud detection impossible and the practical consequence is that a bad actor who has figured out how to identify purchase intent signals, for example Via smart TV audio through behavioral inferences, well, that bad actor can effectively launder that data through the attribution system by using it to place conversion stealing ads. In retrospect, I wish I had asked Don to expand upon his data cartel comment. From a public policy standpoint, I don't believe it's helpful to require a consent for attribution or measurement. And even the EU seems to be considering whether or how to grant exceptions for certain uses of data under the Digital omnibus. But if the premise is for the W3C to create a standard which enables big tech and or browsers to engage in attribution or measurement outside of the regulatory framework of privacy choices, then the W3C needs to be able to justify their rationale for doing so. And the current justifications around air quotes improving privacy are vague and should be viewed much more suspiciously given their clear anti competitive impact of enabling one set of actors to engage in attribution and measurement without friction while effectively denying critical data to the rest of the marketplace. And finally, the admap alternative being put out by the IAB Tech Lab. Don's case for the ADMAP over the W3C attribution proposal in addition to everything else I've talked about, it comes down to to the compatibility with data subject access requests. Admap keeps a record that can be audited. The W3C proposal by design cannot be audited, and that strikes me as a pretty meaningful distinction. We have a bunch of other fantastic guests coming up on the Monopoly Report podcast over the next few weeks. Next week I'll have Rogina Gosnevi, who is a PhD student from MIT who will share her research that claims that that the very rules designed to protect consumers from privacy harms can simultaneously accelerate industry consolidation. I'll also have journalist turn entrepreneur John Batal sometime in early May. Please subscribe to the show@monopoly-report.com or on Spotify, Apple, YouTube or wherever you listen to your podcasts. And thanks for listening. Some Follow the Noise Bloomberg Follows the Money Whether it's the funds fueling AI or crypto's trillion dollar swings, there's a money side to every story. Get the money side of the story. Subscribe now at Bloomberg.