B (3:12)

Yeah, I was a beta tester. I think I gave you one small bit of feedback. You fixed whatever it was and this.

C (3:21)

Thing has been awesome.

B (3:22)

I start my day with it. It's a recap of the top news from the companies that I have declared that I care about. And it's good. It's getting better every day. Like, I really enjoy it, man. I'm not just saying I really, really enjoy it.

A (3:36)

I appreciate that. For those who don't know what we're talking about, I created, I sort of, I don't want to say Vibe coded, but I did Vibe code a little bit. I created a new product. It's called maddb AI and it's a newsreader. We pull in thousands of articles every day from every publication we know of, including the blogs and the podcast, but also Ad Age and Adweek and all those about advertising. And then there's so much that I wrote an AI algo that ranks them and associates them with companies and with people. So you can follow people and say, this is someone I'm interested in following. And then it gives you a personalized newsfeed and also sends you emails every day with the latest news on the subjects you're interested in. It's free. I really encourage you to sign up at it's again, MAD DB AI or the Marketing Database. It's not. There are no ads or anything. It's really like a personal experiment. I'm having fun. I'm sort of semi retired. I do the podcast, I code a little bit, looking to figure out what to do with the rest of my Life. So anyway, try it. I will, I will respond to feedback as well if there are things you want or don't like and give us their feedback. All right, with that said, let's jump in. So we've got Scott Spencer from Rewarded Interest.

B (4:48)

Right.

A (4:49)

In a moment. All right, we're here with Scott Spencer, a legend of the advertising ad tech world. He's here to tell us about his new company and tell us a lot about what's going on with privacy and compliance and all that sort of stuff. Scott, thanks for being here.

C (5:04)

Thank you, Ari. I can say longtime listener, first time caller.

A (5:08)

Yeah, we had the DoubleClick reunion episode like a year ago, but you were notably absent, so we'll have to make up for that. But we want to talk about the subject at hand. So you, after a long career, really long career. We'll talk about that in a bit. In, in advertising at Google and DoubleClick have decided to go out on your own with a startup. How does that feel?

C (5:29)

It is not, you know, it was not something I intended to do. After I left Google, I was living in a small town in northern Massachusetts, had a garden. It was very weird for me, but I realized very quickly that there were things that were just still, you know, I kept thinking back to that were left undone that I wanted to go do. And so I started talking to people and trying to figure out how can I start to address the challenges that exist for consumers about privacy. I had built, as you may know, the Coalition for Better Ads. And I saw that that really had an impact on users because the Coalition of better Ads for those who don't know basically did a huge UX survey to find out what was annoying consumers.

A (6:15)

Okay. What is annoying them? Ads.

C (6:18)

Well, certain formats of ads.

A (6:19)

Certain formats of ads, okay.

C (6:22)

These were unskippable ads at the time. Different formats of pop ups, those were still in existence. And by creating a standard that said these are the most annoying and getting industry agreement to say we're not going to use these anymore actually reduced the ad blocking rate. So it was a great thing that, you know, when I was sitting at best use, I'm like, that was cool. We were able to make things better for consumers and make things better for the industry because lower ad blocking rate, more monetization can happen. The industry's happier. Consumers not being annoyed, they're happier. So, you know, basically when I was up there, you know, I kept thinking to myself, there's, there's still interesting problems to solve. And I had really been intrigued by the consumer side of things. How do I start to help consumers? And in a way that's not just beneficial for consumers, but also beneficial for industry.

A (7:14)

Okay, yeah.

C (7:15)

So.

A (7:16)

So you started this company, rewarded interest, and you're getting rid of the cookie banners. Is that the. Is that the elevator pitch? Let's get rid of the cookie banners.

C (7:24)

Well, so it's certainly part of what we're doing, but to understand what we're doing, you have to understand what's going on. So, you know, earlier, I think it was actually early this month, Mozilla did a survey about 8,000 consumers. And in that survey, they found out the number one frustration still today is the use of data without consent, and that only 16% of people felt in control of their privacy choices. So there's a huge challenge still existing where consumers are just feeling like they're being taken advantage of, and cookie banners are broken. Like, that is just not the right interface, the user experience for people to be able to manage their consent. They're annoying. They're not consistent. You don't really. At this point, you're sort of like, go away. I don't want to deal with it. And that defeats the whole purpose of control, because control needs to be informed, and you need to know what you're doing when you're doing that.

A (8:18)

Yeah, there's a real problem. Consumers want control, and the solution is terrible. It doesn't actually solve the problem. And it's ad hoc, and everyone's doing it differently. And as legal compliance issues, it's not a clear bill of health on the legal side either. Right. So. So how are you solving that problem?

C (8:36)

So that's. That's what we set out to solve. We figured there has to be a better way than the cookie banner to be able to create the privacy control. And so we started with the premise of how do we get a consumer able to take back control of their identity? If I think about it as the thing that the. The privacy stuff, it's all this tracking. I want to consolidate that and make that something that they can be able to control. And so by starting with that premise to say, this is my thing, we started to come up with a bunch of stuff. And the idea is you can manage your settings in one place across the web and eventually across other devices as well. You don't need to have consent banners because you already set your settings, and we can communicate that programmatically. All these publishers have great tools at their disposal to get you to say yes. Consumers need a tool to be able to manage those things on their behalf. And so we actively set your settings. We actively block unwanted trackers. You're really in control. And most recently, we're adding the global privacy control, one of the latest control signals out there for any consumers who want to completely opt out. You're in full control with this tool.

A (9:44)

So I guess this is a bit of a hornet's nest of a question, but why haven't the browsers done this? It seems pretty obvious that the browser should be a single stop for a consumer to tell what they want and then to have it distributed to everything they visit. We have this history where do not track was a failure, where Apple does its own thing, where Google's scared to do any. I guess I'm answering my own question. Shouldn't the browsers do this?

C (10:11)

There is an argument that this is something that should be embedded in browsers. And for the first time, we're seeing other browsers out there. I've actually worked with OpenAI's browser and their extension support's not so great. But look, in the world of agentic browsing, you want a tool that lets you understand what's happening. Because if you think about the analogy for a moment, like if you're the one behind the browser, you want to have control of your identity. But if you're going to hand the browser keys over to an agent that's going to do stuff on your behalf and you don't quite know what they're going to do, maybe you want to have a different identity for that. You need to be able to have that kind of control and you don't want it consenting or not consenting or doing other things that you don't know. That's not a thing you want probabilistic. You want to actually be able to say, this is what I want to do. Here's my settings, and go manage that. So absolutely, this is something that we think needs to be. It's not just in your browser, though. It needs to be almost at the OS level or cross os, because you don't want to have to set those settings when you log into your tv. You don't want to have to set those settings on your phone. So this can be done. Something where you do it once and you now cross device, cross environment, your settings are done.

A (11:21)

Well, there's this interesting debate that's going on in Congress right now where there are competing bills about whether age gating of apps should be the responsibility of the device or the app. And it's heavy lobbying right now because the Facebooks of the world want the devices to know your age. And the apples and Googles of the world desperately don't want to have liability to know people's ages. They want Facebook to figure it out. And it's kind of analogous to what you're doing a little bit very similar.

C (11:50)

As a matter of fact, one of the features we have is we have a minor mode, we built a restricted mode into it. So as a parent, you can install this, you can then lock out your child, your minor from being able to do any consent, because they can't actually consent. They don't legally have that ability to do that as a miner. And so we built it with a similar kind of protection. If you use screen time. So you can put this on and then that user is going to get no tracking whatsoever. All of the privacy pieces are turned on. So absolutely there's this need to give consumers control, give them a tool that can help them manage their digital identity. And part of that digital identity is, am I above the threshold of consent?

A (12:32)

So if you found that the devil's advocate would say that consumers say a lot of stuff, but really they want like, they don't want to spend time on this. They want to just like block everything or don't care at all. So you're giving them all this control in your tool. Basically your tool, if it's not obvious, is like a browser plugin. And it does have a lot. I've been a beta tester from early on and it has a lot of control. What are you finding in terms of consumers, do they care about granular control?

C (13:03)

So it's a great question. And part of our design philosophy is to do this in a way, considering gpr, you can't have defaults to make this as easy as possible for consumers. And so as a result, part of what we have in here, which we'll get into, is we have rewards. We have an incredibly high opt in rate of people just saying the defaulting and say, yeah, I'll happily optimize my settings for rewards. They have. By putting all this control there, most people don't have to do anything. You install the extension, 30 seconds later you're done. And you never have to think about consent again unless you want to. And that's the key because it's not something you want to think about every time you go to a website. But it may be that you hear some news out there about something going on. You're like, you know what? I don't trust this actor anymore. How do I go and block that vendor from ever dealing with it. Oh, I can go do that once and never have to deal with that vendor again. So it's the idea of make it. You need all the control. Not because you're going to use it every day, you don't go and adjust your side view mirrors on your car every day, but you need that control when you want to actually adjust your car because it's not safe.

A (14:10)

So let's talk more broadly about this issue about consent and privacy and all that. What's going on in Europe right now? There are various headlines like GDPR is dead. The reconsidering. No, actually it's the same. It's kind of up in the air. So give us your perspective on the European situation.

C (14:27)

Yeah. So for those who don't know, there is the digital omnibus package that was put forth, which is a significant change proposed to GDPR that would reduce some of its sort of coverage. What is considered personal data, what things are required for consent. It doesn't get rid of them completely. So there's still lots of things that require consent. Require consent banners, unfortunately, unless there's a better alternative.

A (14:54)

Well, that's good for you. They got rid of the consent banner. That would be tough.

C (14:58)

Well, I mean, part of, so part of what we offer is the consent banner, but what we're really offering is that control. And if this GDPR package does go through, there's a lot of Europeans who are going to feel that the rug's been pulled out from under them from a protection point of view and they're going to seek protection. And for the industry, I think it's much better to have a tool that helps solve the privacy problem versus everyone just saying, oh, forget it, I'm just going to get an ad blocker and there's no monetization, content's harmed. It's just not a good outcome for anybody. So what gdpr, what the Omnis package is, it's a set of reforms. It does propose some things, but it's very unclear whether or not that's going to happen. So a lot of people are lobbying. There's a lot of activity going on in Europe, obviously for that. I don't know what the betting, I haven't looked at the betting odds yet, but it's certainly not a true win that this is going to go through.

A (15:46)

Well, so the biggest change in my understanding is that in gdpr, these anonymous or pseudo anonymous identifiers like cookies, are covered as personal data. And I always felt like that was one of the key flaws of the regulation and my understanding tell me if I'm wrong is that the standard change to being able to reasonably identify the user, an ID has to be able to reasonably identify the user in order to be considered personal data and require consent. Am I correct in that explanation?

C (16:19)

So I'm not a lawyer and in terms of always interpreting laws, it's always fun. But my understanding is yes, in essence the, the criteria for a pseudonymous identifier to be considered personal data is whether or not you as a company can, I can associate personal data with that user. And actually it may even be that you do so. Not that you could do so, but you actually do so. And so that would decrease the burden for those pseudonymous identifiers significantly. And that's a double edged sword because it's good in the sense that great, all these kinds of identifiers can then be used in the industry, but it's not good from a consumer perspective because all of a sudden, wait a second, these, now these trackers, these web bugs, whatever the people want to call them, they're now being used and the control just got reduced. So we, we, we think that it's important to be able to control all aspects of your, your sort of identity. And so even if that changes, that doesn't mean we would say, oh, there's no control, you can still decide to turn them off.

A (17:19)

And the cookie banners, you know, one of the misnomers or common misunderstandings is that the cookie banners were caused by gdpr when actually they were caused by a previous regulation called E Privacy or E Privacy, depending on if you're US or Europe. So the, so my understanding is also the cookie banners are sort of debatable because the E Privacy said you needed consent before placing them on a user's device and this omnibus package may get rid of that requirement. Is that correct?

C (17:48)

So my, again the preface of not a lawyer, but there's a lot of things that require a consent interaction with the user to get their permission to do things. And the consent banners have been a tool for that. Certainly privacy is, was one of the first things. But that required the sort of flyby notice. You could, you could initially was opt out, then it had to be opt in that part of the requirement if the cookie is no longer considered to be personal data is not required. But if you want to do things with that, a lot of things that would be done, you'd still need to get that permission and you have to get that permission at the site of the publisher because there's no other way. To do it. So there's still a need. And that's all in Europe. So keep in mind, in the US California has created a significant need for privacy banners and there's been a bunch of lawsuits that have happened in the US with regards to the collection of data there. So there's another, you know, the U. S. The US Is sorry, Europe is only one regime that needs that. And then the US Has a bunch of things. Then there's Brazil has, Japan has. All these other countries have different things that are requiring this. And the Internet doesn't really like geographic borders.

A (19:05)

Yeah. Okay, so you were standing in your garden thinking about this, decided to do entrepreneurship. How's it going? Like there's no cafeteria, no free food at rewarded interest. I imagine that is true.

C (19:17)

We have no cafeteria, but it's been, look, I love to build things. So for me it's been a lot of fun to build something to help consumers to find a way to try and align incentives across the industry. Because one of the key things we're trying to do here is to create something that is good for users. You don't have to deal with all this frustration. You have much more control. But it's also good for advertisers. Clearly there's addressability. And most interestingly, we found it's really good for publishers and ad networks. Because ad networks, there's a high cost to putting an ad consent banner on 10,000 sites. And if there's a way that you can use some of those low frequency sites, the one recipe, someone goes to use that as an opportunity to get someone to install rewarded interest and then never have to have that user encounter a consent banner. Any of your other 900-9999 sites, that's a huge benefit. So we've had some interest there and a little bit of pull from the industry to do that. And that that also makes me excited, like doing something where we can make make a solution that's better for consumers and better for industry. That's sort of what I wanted to try and do.

A (20:32)

Eric, how. How does the finance community feel about privacy tech?

C (20:36)

On.

B (20:37)

So there's privacy tech and then there's tech that is like, you know, consumer engagement tech. I don't know exactly what we would call it. Right. So you might want to bifurcate it. There's. Yeah, I mean there's been a history of companies that have done things in the privacy space, both large and small. And I think some of the large ones have, you know, just become like these large incent management platforms, you know, and there's, there's a couple of them and then there's a bunch of companies that have taken a stab at what Scott is building with rewarded interest, which is, I mean the name speaks for itself, right? Rewarding consumers for data. And you're having a privacy angle. We've yet to see a breakout there, but it is such an enormous market that there's always the attraction of smart operators like Scott to get after it. Because if, if a company figures it out, you have an opportunity to, you know, do something that's enormous. So I think, you know, because of the fact that there haven't been breakouts, in my experience, it's been challenging for these companies that are consumer facing to like, you know, really raise a lot of money and scale very quickly. But that doesn't mean that there won't be a success in the category.

A (21:54)

Well, you heard it here first, Scott. So if you're looking to raise money, you know Eric's email. So let's. So Scott, you and I have known each other since I entered the ad tech world. I didn't report to you, but I remember like my first PRD at DoubleClick. I had to get approved by you and you were like, I don't know, it's rich media, who the hell knows what this shit is? More or less was your reaction. So let's set the record straight first. How many years were you at the combined DoubleClick Google? Because you're close to having the record.

C (22:25)

I don't think I had the record. So it was 22 years from DoubleClick to when I left Google.

A (22:31)

That's amazing. Okay, true or false. Did you invent rtb?

C (22:36)

Well, it depends if what you define as rtb. What I definitely did invent was what we call network proxy bidding.

A (22:45)

No one's ever called it that.

C (22:47)

Well, so the. That's the patent. That's what's on the patent. The.

A (22:53)

Okay, you didn't invent it, but you have the patent. Okay, I understand how it works along.

C (22:57)

With George Hayley who co invented it with me. But so what we did was if you going back to like early on Figuring out the DoubleClick ad exchange, we built it on a merger of what was then DFA and dfp. And at the time we used this thing called the internal redirect, the inread. And the reason that I was focused on the inread because it had something super cool that nothing else could do in reds could say no. So if you were to just ask A third party for an ad. The tag went out, an ad had to be delivered. But an in red was cool because an in red could say no, meaning.

A (23:35)

The publisher could say no, the advertiser could say no. Okay.

C (23:38)

The advertiser could say no. So the publisher would say go call dfa and if DFA had an ad, great. And if it didn't, it would say don't have an ad and it would fall back and then you would have DFP would go to the next ad and it's queued to go select an ad. And the other cool thing was it passed all the parameters. So DFA didn't have to go refigure out the geolookup. It didn't have to refigure out any of the targeting elements. It passed the parameters.

A (24:04)

Yeah. So just for context, for those people who don't know what we're talking about, we're talking about the buy side of DoubleClick and the sell side of DoubleClick interoperating on an ad server level. Before exchanges really were a thing, it was called an in red or internal redirect. So just give a little more context. Go ahead, Scott.

C (24:20)

Yeah, and so when we use that in red, that's how we built the original exchange. Right. So we've used that in red as the mechanism. We added price to it as a parameter. And so now we could do with the auction side of things. Now fast forward a bit and DoubleClick gets bought by Google. We have to rebuild the entire DoubleClick ad exchange on the Google stack. But until then, there's AdWords wanting to bid on the double click ad exchange. And the only method we had was an internal redirect. So we needed to come up with the external way to do that. And so Jorge and I sat down and figured out a way to proxy the bid call from DFP in order to go to AdWords with all the parameters associated with it. And eventually that concept of being able to pass that information over to a third party, third party, to the two different infrastructures with the bid pricing and all that. That was what we then said, hey, others want to buy on this? Great. We created, converted that into that external proxy bidding thing and then we renamed it Real time bidder.

A (25:31)

Who is the first buyer? Do you remember? Who is the first impression bought by RTB outside of AdWords.

C (25:37)

We opened it up, if I recall, it was like there were like 14 buyers that were immediately buying on it. It was not one. It was when we opened it up because we had, we had a Bunch of buyers who were working because we actually give. There was a buy side of addicts at the time. I don't know if you remember this.

A (25:52)

But yeah, you had to like upload a tag and stuff, right?

C (25:55)

Yeah, it was, it was a full buy side little thing because there were no bidders, there was no anything else. And so we actually build a full buy side. So all the people were using the buy side were like, well, I want more robust solution than this buy side thing. And so we had a bunch of buyers ready to go and when we turned that on and they made it available, they were all able to plug in.

B (26:13)

This is fascinating. I tested this out, this question out prior to recording. I asked Gemini and I asked ChatGPT who invented RTB. They wildly different answers, by the way. But the one thing that both said was there was this moment in the early 2000s where multiple companies and multiple people were all working on this thing. Gemini actually gave a very thoughtful answer and named names. It was like Brian O' Kelly at Right Media, Jason Knapp at SDC, Dr. Boris at iPon. Right. Like all kind of like, we're working on this thing at the same time. You certainly deserve a mention in this. Based on the DoubleClick stuff.

C (26:58)

Yeah, for better or worse, the AI systems, they base on what information's out there on the web. And we never went through and sort of talked about the creation of RGB or the signal at the time. I mean, the funny part was it was super organic because for us we had to connect Google and DoubleClick and the data centers, they're different locations and all that stuff. And so we had to come up with this just low latency way to pass the information across. And I remember. So they're like, you gotta use Protobuf. And I'm like, I have no idea what a Protobuf is.

A (27:35)

But yeah, for the record, you know, I wrote a book about this though, you know, Scott unfortunately wasn't able to speak to me during the course of the book because of his ongoing involvement in litigation. But what I found was Rajiv at PubMatic thinks he was one of the first people. He says he literally wrote out the spec on a Word document and sent it like by email to various people. And there were some early transactions with like Turn and invite. He didn't say he was first, but he was, he thinks he was quite early. And. And o' Kelly believes that he sort of hacked something together related to ebay as an advertiser buying on, on sample websites on like Mike on Ads, which was a blog and some things like that. So. But I think it's true that it was extremely diverse group of people working on it. It was all in like the 2009 ish timeframe. 8, 9. And that is the history. I'll ask one more question, Scott. Okay. I know there's sensitive topics, you've been on the stand and things like that, so feel free not to answer. But we're here, we're 15 years or more like 17 years after the invention of RTB. Would you any regrets? Would you do anything differently? If you could talk to yourself in 2008 and say actually this RTB thing should be a little different, what would you do differently?

C (28:58)

It's a fun question. The process of creating the exchange and the integrations all was fairly organic. It was not like there was a master plan behind things. I remember anxiously tracking every day the number of transactions that went through the DoubleClick ad exchange on a spreadsheet. There was no automatic reporting even at the time. And it was just exciting to, you know, to figure out a way to take what had been working so well with CPC based ads in the second price auction and move that over to display. You know that I was just really excited about moving that thing forward and being able to create the transactions. At the time most publishers hated it, right? So you have to keep in mind that when I built the DoubleClick ad exchange, publishers were like this is terrible. You're going to hurt my direct sales. That's all they cared about in terms of regrets. There's always things you go back, you're like, if that could have been done better. But I think my biggest regret is not taking more time to step back, watch and be excited about what was happening. Was so in it to build it that didn't get a chance to really understand just how amazing the transformation was and celebrate with the people there the amazing work that was done to change an industry. So technically things will happen the way they, I think they sort of unfold more organically than people think. My biggest regret is just that there's some amazing, amazing people forgotten to time who were responsible for this transformation. That it's amazing.

A (30:46)

Yeah, it's a two sided coin of working for Google. You make a lot of money, but no one knows who you are and you Never update your LinkedIn. All right, this is an awesome conversation. So if you want to get Rewarded Interest, what's the URL?

C (31:00)

Rewarded Interest.

A (31:02)

All right, go to Rewarded Interest. Download it for free. I've been using this as a beta. Big fan. And we'll be back in a minute with the news of the week. Scott, thanks so much. This podcast is brought to you by audiohook, the leading independent audio dsp. Audio Hook has direct publisher integrations into all major podcast and streaming radio platforms, providing 40% more inventory than what could be accessed in omnichannel DSPs. What's more, audiobook has full transcripts on more than 90% of all podcast inventory, enabling advanced contextual targeting and brand suitability. Audio Hook is so confident that in addition to CPM buys, they offer the industry's only pay for performance option, where brands can scale audio and podcasting with peace of mind, knowing they are only paying for outcomes. Visit audiohook.com to learn more. That's audiohook.com.

B (31:55)

Okay, we're back folks, with the refresh a bunch of stuff. Today we have. Oh my God. A scoop with Google at Ari found meta and Chinese ad fraud, some AI and announcements. YouTube versus Netflix. It's all over the place. But before we do it, Ari, what you got?

A (32:14)

Quick correction from last week. I told the story that the coach of the Michigan college team was caught by doordash when he doordashed a plan B. Turns out that was a disinformation. Misinformation on the Internet. I fell for it. My bad. Sorry.

B (32:28)

You got.

A (32:29)

God, I got. Got. My bad.

B (32:31)

Wow. Man, what is this world coming to? Okay, let's move on.

C (32:36)

All right.

B (32:36)

Before we even go to the news, I think we need to talk about something.

A (32:41)

We do.

C (32:42)

Let.

B (32:43)

Let me pull up something. For the people that are watching on YouTube. So for those that are on YouTube, you see this. For those that are not on YouTube, we're looking at a screenshot. The first frame of this TV ad that vive. Are you. You recently were on the. On the board of the company. I'm sure you're an investor. Did. And posted to X. It is a frame by frame copy of the. Of the American Eagle, you know, viral ad from earlier this. This. This year. And Arthur has blonde hair. He's wearing prosthetic female body parts in like head to toe denim. And again, it's a total copy of the American Eagle ad.

A (33:34)

Arthur, he's got Sydney Sweeney's boobs on somehow. How is this possible?

B (33:40)

You said it. I did not. He has facial hair, but he does kind of look like her.

A (33:48)

He does.

B (33:48)

4.5 million views to date. It has all of the X meme accounts jumping in, doing more. I mean, hats off to Arthur for one of the most creative Things we have seen, I think, ever with ad tech, advertising. Oh, my God, this thing is moving.

A (34:06)

It's incredible though.

B (34:08)

Audio's off. We're not going to get copyrighted.

A (34:10)

I thought it was AI. I thought it was AI. But it's prosthetics, it's makeup. He did the full production. I think he said in a thread it cost fifty grand to produce this ad. And this is what happens when French entrepreneurs move to the us. They lose their brains and I just love it.

B (34:28)

Well done, Arthur. More of this, please. This is fantastic stuff. Scott, did you have a chance to, to see it?

C (34:36)

Yeah, I just, I just saw it there. I like the tagline. Streaming TV has great leads. That, to me is definitely the winning part.

B (34:44)

It's a total copy of the, of the American Eagle. Sydney Sweeney had. Well done, Arthur. Again, more of this. All right, you got scoop. What's going on with Google?

A (34:53)

Yeah, this is a, this is a kind of interesting scoop. Not because of the implications, but because of like. Well, because the implications, not because of the reality, which is Google just quietly changed a bunch of things that they've been asked to change in their court cases before a ruling came out or in reference in response to the EU rulings. So one of the most controversial product changes that Google made in the course of its ad tech saga was this thing called Unified Pricing Rules, or upr, where they disabled the ability for publisher customers to charge different amounts for different exchanges. So typically a publisher might have a higher floor price for Google Ad X than they do for like Index Exchange because they might have like a deal with Index. Or Google's adx is known for having higher take rates, so they might want to, you know, charge a bit more. And Google removed that ability in like 2016, 2017 is a big aspect of the court case because, you know, from a legal perspective, removing a feature is the antitrust equivalent of increasing a price. It's the same kind of concept where monopoly power allows you to do something, something like that. Google argues this is a nothing burger, but whatever. So this week Google just removed it. They changed the name from Unified Pricing Rules to just Pricing rules and they removed the feature. So with no fanfare, with a very minor blog post, they also took their rev share on open bidding, which is their server to server connection to other exchanges, from the usual 5 to 10% to 0% in the EU, or at least for EU based publishers. And I did some reporting on it. And what's happened in the background is that Google responded to the EU's findings of their monopoly about in September with various proposals. These are two of the proposals which they just executed. Good for them. And then the third, obviously the third thing is that the EU still wants them to spin out adex and they don't want to. So that's still contentious. And obviously that is the main crux of the DOJ's case that we will get a ruling on in January or February. I think the next shoe to drop is when ADEX starts bidding into pre bid instead of directly into gam. And that's something they've agreed to in the US but the question is, are they working on it now? Will they just do it whether or not the court requires? So there's good news for publishers, good for the whole process of unwinding this mess. And it's kind of funny the way it just came out.

C (37:29)

Yeah.

B (37:29)

Why would they just go ahead and do this? Do you have any kind of 4D chess thoughts around it?

A (37:36)

I think this one particular thing, Unified pricing rules, was so obviously wrong that they knew they were going to have to remove it. So why not? I don't think they lost. They gained more leverage of goodwill than they lost in holding, holding this card.

B (37:49)

Got it, Got it. Scott, want to take a pass on this, I assume.

C (37:55)

Unfortunately, I think until, until there's any rulings, that's something I should probably not not comment on.

B (37:59)

No problem. All right, we'll move on. We got something juicy here. So Meta. So Reuters did a piece on the. The scam ads coming on Meta largely, you know, via China and other other sort of kind of non US operators a couple weeks ago. They article today that basically estimates that 20% of Meta's Chinese revenue is coming from scams and gambling and other banned content. So I'll walk through it. So they got a hold of some internal dots that estimated $18 billion. Basically more than 10% of Meta's revenue comes from Chinese advertisers targeting foreign consumers. Reminder, Meta, you know, is banned in China. So this is just like Chinese advertisers targeting Everybody in the U.S. and around the globe. 19% of that amount. So $3 billion was linked to scams, gambling, other bad stuff, including illegal content. Now this is where it gets a little weird. So Meta, apparently internal docs allegedly crackdown. The spend went down. The spend went down by half and then abandoned the crackdown maybe because the.

C (39:13)

Spend went down by half.

B (39:13)

I don't know. All allegedly. This is a weird thing because if we'll link to it, it's a very long article. It's A very interesting article. The Chinese market is very opaque. So meta works with 11 resellers, big, big companies who then work with intermediaries like agencies and networks and God knows whom, who. You know, basically it's the wild west and you know, they start doing all, all this bad stuff and it gets very, very difficult to track because you can't have a target on the back of one of these 11 resellers. So seems like it's part of the complexity of the Chinese market. You're more apt to see these types of things, but it's a mess and it's a lot of revenue.

A (39:53)

Yeah. This is continued damning reporting from Reuters. If 10% of the advertising from the second largest advertising company in the world should be illegal, is illegal, should be banned. And I wouldn't say they don't do anything about it, but they take a reactive stance. They wait for complaints, they wait for things to come to their attention instead of proactive. We had Rob Leathern on the show maybe six months ago or more and he basically was saying that it's overwhelming the number of new ads they get every day. They don't check every single one. It's reactive. So I'd be interested in hearing, Scott, your position since you did work in trust and safety at Google. How hard is this problem? How active? Maybe not specifically Google, but what's your perspective on this problem?

C (40:42)

Yeah, look, so this is not good data. This is, this sounds like a real problem that they have. And when I was managing trust and safety, the goal was trust. So if you have this much bad ads running on a system, that's going to mean anybody who's monetizing through you is not going to want that those ads to run. They're not going to trust you as much in the long term. So it's in Meta's interest to clean this up and to make sure that they've got a good environment for the ads. What's interesting is there are a lot of ads that come into any system on a daily basis, but the technology for assessing ads is also kept apace. And so one of the questions that I would have for that team is how much are they using AI and ML to pre screen those ads and looking at a combination of heuristics and other things to understand that. But we knew that the number of hops that you had between ads directly proportional to the probability of it being fraud. And so there's a lot of heuristics that can be brought into that even in China, to go and protect so I'm sure there's some tough conversations happening in that trust and safety team to clean that up. At least I hope that there is.

A (42:01)

What about the point of view that the more barriers you put between an advertiser and a spend, the bigger impact you have on revenue, even for the good guys?

C (42:11)

Well, there's always a trade off, right? If, if you get rid of a bad actor, you lose revenue.

A (42:17)

Yeah, yeah. But let's say you're willing, let's say you're willing to do that, but I'm talking about the good guy. So, so I've heard from this perspective especially it's actually especially on the publisher side, like we want to onboard as fast as possible and the more we try to get rid of the bad guys, the more the good guys suffer and it slows revenue throughout the whole company.

C (42:35)

Company. That's, it's true. Right. You're going to get some good actors that are going to be restricted when you're trying to protect against the bad. That's always a trade off. It seems like in this case it's probably too far to one side, but you're never going to be perfect. And really where you set that dial is going to be what matters. And frankly it's sort of how much do you put value into the long term value of your audience, of your consumers? I've got a website, it's gonna be around for five minutes. Great. I'll take all the crappy ads I can get and if those users get hurt, so what, I'm gone. But if I have a brand that I really want to invest in and want those consumers to come back, then I can't have the article that says, oh, I got malware because I went to this site or you know, I clicked on an ad and it ended up being for some guy who robbed me. So it's, it's always a balancing act. Clearly, you know, sometimes things go off the rails to one side or the other, but it also requires a lot of investment. It's like it's, it's not a, an inexpensive problem. At one point, I believe the Trust and Safety team was the single largest user of ML at Google to try and deal with this.

B (43:49)

Makes sense. Just to underscore it, this hasn't been confirmed by Meta, but there's Some reports that 15 billion ads per day run across all of the Meta properties. So it's operating at a scale that is just like ML can only solve this, right? Like there's, you know, there's just such massive scale that it's like impossible to, you know, have a. Have a team on it. All right, awesome. Two things this week. And you know, this is interesting, Ari, because one of the things that, you know, you've said a few times over the course of the past year or two is like, where is the innovation coming from? You know, the scaled platforms, the scaled companies who should be advancing the cause in AI. And you know, these two announcements this week are from large companies. So the first is from Pubmatic. So Pubmatic claims they ran the first agenta campaign with Butler Till and shout out to Butler Till, Butler Till. They were in the ad CP announcement. They've been like all over this stuff. Indie agency up here in Rochester, New York. So here's the explanation. So using adcp, Butler Till submitted a campaign brief in plain language to Quad. Pubmatic's agents then took over, basically translating the brief into a media plan and setting up and then executing. So it starts in Butler Till's AI tool, they didn't name it. And then ultimately Pubmatic takes over from there. So there's no DSP involved in this one from what I understand.

A (45:18)

Yeah. And I wrote this in my coverage of adcp, my sort of critique of it, not using that word in a negative sense in that it appears that this is going to open opportunities primarily on the sell side. The buy side is going to say, here's what I want. You don't need a lot of tooling for that. The sell side needs an intermediary because individual publishers don't have scale, can't do performance. So this is actually a really big opportunity for sell side intermediaries, be them SSP or swivel. I know you invested in. I had a holiday party conversation with some swivel folks who were saying they're like basically running campaigns or very close to running campaigns. And if I was on the sell side, I'd be jumping all over this. I don't know the mechanism, if it's going to be guaranteed buys, audience buys, performance buys. And that's also what I wrote in my critique. But I'd expect a lot more of this.

B (46:13)

Yeah, agreed. This is super interesting, Scott.

C (46:16)

You have a POV on this one? Yeah. So to me, the world of agentic inventory creatives, this all hearkens back to like early days of display. And what was needed then, I think is needed now, are intermediaries to come in and figure out how to make that friction lower. Ad networks were one of the earliest things that came up to try and bring that inventory together, help the Buyers buy it. I think we're in the need now. We're going to need players to come in and figure this stuff out and make it so it's really easy and effective for advertisers to get the ROI they want and for publishers to get monetization. And they're gonna have to do that messy middle stuff to make that work. Great opportunity. Who knows who wins now may not be the winners long term. There was a huge shakeout in, in that world, but I think we'll follow the same pattern. There will be the performance networks of Agentic World for a while until it starts to become standardized and mainstream. I don't think we can jump to the. There's some standards required to sort of start things, but I don't think it's going to jump to where it lands for quite some time.

A (47:21)

Also, Eric, we didn't cover in the notes, but the Agentic people created a new association, the Agentic. I forget what's it called?

B (47:28)

Agenticadvertising advertising.org check me on that one if I'm wrong. And interestingly, Randall Rothenberg is either involved or leading it. So you want somebody that knows how to get an industry standard in place. He is your person.

A (47:43)

Definitely. There'll be some very lengthy keynotes at their first conference.

C (47:48)

I think that was a huge boost of credibility for them, a huge benefit for them to bring Randall in. He's a great person for doing this. He has gone through that arc many times. And so congratulations, Randall. Congratulations. That are, that's, that's exciting.

B (48:02)

Absolutely.

C (48:03)

And then on the other side, and.

B (48:04)

This is interesting, I think it's interesting because we just talked about how this is a big sell side opportunity. This comes from a buy side company in Filion who owns a number of assets including Media Math. They launched something called the Agent Connector. So this lets AI agents basically plan, buy optimize without A dashboards and then B to optimize in any system. So it is like, like it's kind of like a buy side solution without the dsp. Is that your take on it?

A (48:35)

Yeah, I mean I think that the buy side has a lot of workflow that is AI ineligible and being cross channel makes a lot of sense. Let's tie this to the other news item that we haven't gotten to yet, which is fluency raising $40 million. So it's really similar. Right. Fluency offers cross channel optimization using AI. That's at least what they claim to offer. And the difference between planning and optimization is kind of a dotted line in this stuff, I think Buy side really wants this going beyond a once a year media mix model and instead saying this week, how should we move budget? What should we do? What's working, what's not is great stuff. AI should do it. It's not that hard. It's not that hard. There's a lot of companies doing it and you really can't move much besides budget. You know, you can't optimize very much when you have channels like Meta, Google, Search and Trade Desk. What can you mix between those three? What learnings go across them? Not a lot. It's budget and it's maybe flighting and a couple other things. So I think this is really valuable, but it's sort of hemmed in by what it could ultimately deliver.

B (49:47)

Yeah, that makes sense. I mean, on the other side, I'd love to see again these scale companies start to push forward. And maybe the launch of ADCP and MCP was what we needed to see this adoption.

A (50:00)

I think we're going to be overwhelmed by the number of press releases about AI enabling legacy ad tech companies.

B (50:06)

You think ces, we might see one or two.

A (50:09)

We might see a couple at ces.

B (50:11)

Maybe.

A (50:11)

I'll keep count. Let's do a counter. How many times I hear I'll have like one of those things you keep in your hand to count stuff. Every time someone says AI, I'll go.

C (50:21)

I hear AI is going to be big. I'm not sure actually, but you know.

A (50:24)

I hear it, I hear it.

B (50:26)

Well, thankfully there's a new tool, tool called MAD that's helping me track all of this stuff. Yeah, absolutely. So, all right, let's, let's talk about YouTube versus Netflix. This thing is like coming to a head out of, out of nowhere for two companies that say they don't compete.

C (50:43)

So.

B (50:45)

YouTube I think leads the the market in terms of podcast downloads, podcast streams. Netflix just signed a couple of big podcast deals. So iHeart, they signed a video podcast deal with Netflix. Barstool video podcasts are going to Netflix. This is definitely a shot across the bow to YouTube who obviously has YouTube music, but. But it's a place where you will watch the video format of podcasts.

A (51:12)

Yeah, YouTube had a dedicated podcast or Google podcast that used to exist. They shut it down about a year ago, started pushing everyone to YouTube. YouTube supports taking a RSS feed from your favorite podcast studio and putting it in. It is widely discussed in podcast circles that it's great for discovery. Whereas Spotify is really bad. And Apple Music, they're both really bad. For discovery. So people are investing more and more time. There's. It's interesting These two deals, iHeart and Barstool, both very conspicuously said they're exclusive with regard to video podcasts. You can still get your audio podcast anywhere you want to listen to it, but if you want to watch the video, you have to go to Netflix. I think these are super interesting deals. I'd love to know what Netflix's plan is, if they're going to create different products out of this or if it's just going to show up as a card on your Netflix when you're at home and you just start watching whatever, Barstool or whatever, you know. So I think this is, this is pretty interesting.

B (52:13)

It is. I watch a lot of podcasts on YouTube. It's very natural to YouTube. It's become a habit for me. I'm sure it's a little bit of a stretch to think and maybe it's because I don't watch these podcasts.

C (52:25)

A little bit of a stretch to.

B (52:26)

Think I'm gonna use my Netflix time to, you know, kick back and watch a pod when I could be watching like something much different. But again, maybe this is just a habit that needs to. Needs to get birthed.

A (52:38)

I mean, maybe if I was Netflix, I'd be thinking maybe I need a dedicated app like on the phone to have a podcast listening app that's, that's got a mixer of audio and video. Maybe, I don't know, I, I don't think it's enough just to put it on the home screen.

C (52:51)

I mean, I think it's interesting because YouTube's YouTube is really focused on the discovery problem. They've made it easy to discover new content, related content, new formats.

B (53:00)

That's why I watched so many pods on, on YouTube.

C (53:03)

It's going to be a challenge. It's going to be a challenge to, for Netflix to do. It's not an impossible challenge. Maybe they'll do with a different app, maybe they'll start to format. But look, we'd be remiss if we weren't first to say shout out to Neil for making the Times CEO of the year.

A (53:17)

Oh yeah, Unbelievable.

C (53:18)

Lovely picture there. That was fantastic.

B (53:20)

Oh yeah, yeah, absolutely.

C (53:22)

Neil for many years. So it was exciting to see.

A (53:24)

So you reported to Neil for, for quite most your time at Google, right?

C (53:29)

Yeah, I mean I reported to him for Times, not reported to him other times, back and forth many times. But I've known him for, for a long time. He's a fantastic person and just a Great strategist, a great CEO. So it's great to see him getting the recognition there.

A (53:45)

Absolutely. It's quite a run. And I reported to him for three or four years. He's a great guy.

B (53:50)

Speaking of Neil and YouTube, so you got podcasts going to Netflix. The Oscars are coming to YouTube.

A (53:59)

Bit of a unbelievable.

B (54:02)

Starting in 2029, the Oscars moved from ABC to YouTube. Talk about a bellwether of the state of legacy media. It is incredible what we're seeing. It really is.

A (54:15)

Yeah, it's incredible. Coup. I mean, we've seen a football game here and there. You know, we've seen, you know, a little bit of content, some, you know, mma, whatever. But to have this like, tentpole event, one of the most important broadcast events of the year, go to YouTube. I mean, I'm looking forward to the day where my whole family is like, how the hell do we watch this on the TV? Because that's what's gonna happen in early 2029 to me. But it's a big deal.

C (54:43)

Well, one of the things that if it's gonna go to YouTube, there's no longer a broadcaster that is doing it. So the rules are off in terms of what gets said there all of a sudden.

A (54:52)

Let's.

B (54:52)

Let's see what happens now that we're gonna.

A (54:57)

I just want to make one more point, which is like we're heading into this Netflix Discover, Warner Brothers deal and a lot of the conversation is like, oh, well, Netflix has a monopoly. They're such a big streamer and they're smaller than YouTube, significantly smaller than YouTube by any account. They're the second largest streamer, not the number one streamer, which you could still argue shouldn't be allowed to buy that company. But the conversations, especially on Twitter, like Tim Wu, who's a former Biden administration official, saying that they're not competitive is just a joke. I mean, they're head to head the two most important companies in streaming competing.

B (55:36)

Absolutely.

C (55:37)

Absolutely.

B (55:38)

All right. I think that's a good place to end it.

A (55:41)

Yeah, absolutely. This is a great conversation. Scott, thanks so much for being here and giving us your insight reminder. If people want to download it, it's at rewarded interest. And we will be back next week with our last episode of the year. Our year end wrapped. So please tune in to that. Thank you for subscribing to marketecture. New interviews are added every week at marketecture TV and your favorite podcasting app. Thank you for listening to the marketecture 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 marketecture 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.