Smarter Ad Targeting in a Post-Cookie World

Tom Nixon (00:01.614)
Welcome back to Bullhorns and Bulls Eyes. Uh they call this Curtis the podcast where precision meets what is it? I’ve already forgotten where promotion meets precision, right? Okay. See, I got it wrong. So, you must have come up with that tagline because I can’t remember it. So, um but the reason I think that’s relevant is because I bet you’d be happy for just a week to take a break from the storytelling and the branding and the authenticity and get down to some real

Curtis Hays (00:11.304)
Precision meets promotion. You got it.

Curtis Hays (00:19.964)
People know better.

Tom Nixon (00:31.07)
nuts and bolts geekery and get into the precision part of promotion. So are you ready for this?

Curtis Hays (00:37.486)
I am ready for it. do really appreciate the precision side, but they do call me bullseye, obviously, for a reason. There is a reason why that’s a nickname. But I’ve had a lot of fun going down the journey of storytelling and branding. it’s been exciting. And I think it connects a lot to probably, hopefully, what we’re going to talk about today because…

Tom Nixon (00:44.59)
This is true.

Curtis Hays (01:03.464)
Even with branding and storytelling, you got to know who your audience is and you got to make sure that your communications is getting to the right audience. So I’m hopeful our guests today might talk a little bit about that.

Tom Nixon (01:12.28)
Right.

Yes, exactly. And the whole notion of, know, the bull horns don’t work unless you’ve got the bulls eyes working along with them, right? I mean, I guess they can work. It just takes a lot longer. So, we’ll decode all of that for our listeners once we bring our guest in who is Brendan Norman, the CEO and co-founder of Classify. If you haven’t heard of Classify yet, we will get into what that is and why it’s relevant to you as a marketer. But you may have heard of Facebook. He helped launch and scale the audience network to three billion dollars of revenue.

Curtis Hays (01:22.354)
Mm-hmm.

Tom Nixon (01:44.704)
it will talk about why that’s relevant as well. So let’s welcome Brendan Norman to Bullhorns and Bullseyes. Hey Brendan.

Brendan (01:51.678)
What’s up Tom? What’s up Curtis?

Tom Nixon (01:53.55)
So, now you know who’s who. You know, he’s the numbers guy. I’m the word guy but I think those things need to work hand in hand. I’m assuming we will learn by the end of today. So, real quick, tell us what Classify is for those who don’t know at a high level and then we’ll get into some of the details.

Brendan (02:09.138)
Yeah, thanks for asking. Well, great to meet you both. Thanks for having me on the show. The best way to describe what we do, Curtis, quick question. Who makes that hat that you’ve got on?

Curtis Hays (02:20.75)
makes it i mean it’s it’s it’s a bailey it’s a bailey it’s about a 350 hat i mean it’s it’s legit i got another one the maid in mexico behind me there i wore that on season two that one’s a little bit more comfortable than this one i’ll take the maid in mexico over over made in the usa for some reason but

Brendan (02:21.682)
Yeah, so the stats in.

Tom Nixon (02:43.916)
Yeah, that seems anti-intuitive or counterintuitive.

Curtis Hays (02:46.632)
Not necessarily, but I think that one’s more handcrafted and even the price on that one goes up closer to about the $500. It all has to do with the amount of animal material that’s in them, the hairs and the leather and all that stuff. But love wearing the cowboy hat. You’ll see me, I don’t just wear it for the podcast. If you see me around town, I’ll be wearing it as well. I just don’t wear it to church.

Brendan (03:09.342)
you

Tom Nixon (03:12.014)
Brendan, I’m guessing you’re asking all of this and letting him go on and on because you want to textualize all of this all of this copy in in speech for some reason to your advantage. Anyways, Classify. Explain. Are you asking the question because it’s relevant to the answer of who Classify is?

Brendan (03:30.014)
Exactly. So, you you’ve got this beautiful Bailey hat, you know, that you wear around town, you wear on the podcast, you thankfully take off indoors in church. And if you’re a Bailey and you’re looking to reach people today, a lot of the industry works based on knowing you, Tom, you Curtis and using cookies to kind of follow you around the web. And when you’re reading about sports or when you’re reading about other, other articles, there hasn’t really been a big investment outside of Facebook and the walled gardens and contextual targeting.

And the reason why there hasn’t been, it’s just really complicated to do at scale. It’s complicated to scrape a website. It’s complicated to read a whole article. And a lot of the technology was pretty simplistic for the last, you know, 15, 20 years since we’ve been doing advertising on the internet. And if you’re reading an article about hats that might get bucketed into like apparel and like headwear, but it doesn’t really know all the nuances to finding this is a Bailey hat, or this is like a hat made in Mexico by some custom designer.

And it doesn’t have all the nuances to understanding where you are and stance detection and all the rest of like the really deep insights that a human would read an article and go, yeah, we get all these things. So what we’ve built is a much better way for basically using machine learning to deeply understand what happens on a webpage, all the text, and then some of the other technical signals around that webpage, when there are ads on the webpage to help an advertiser like Bailey hats, who’s getting a huge promo from this.

Tom Nixon (04:55.342)
Yeah

Brendan (04:56.488)
to help them find the right type of content on the web to target an ad again so that you Tom and you Curtis when you’re out reading an article about hats Bailey can put that out against the best the best web page.

Tom Nixon (05:09.036)
it is it more about contextualize not more about but is it solely about contextualizing web content and the digital universe but how does that match up with the is it also about contextualizing the people the Curtis’s of the world to know that Curtis is in market Tom Nexon. Tom Nexon will never buy a cowboy hat because it looks silly on him.

Brendan (05:29.95)
There’s a lot of good technology that exists to kind of follow people around the web. A lot of that technology isn’t quite as privacy safe as we all like to think. So we took the approach of saying, let’s not look at any personally identifiable information. Let’s not look at, you you Tom or you Curtis, I don’t understand who you are. Let’s only look at the content itself. So we’re not looking at cookies. We’re not looking at PII and eventually.

You know, we love to think through the best ways to do that at scale in a really privacy safe way. But a lot of the existing technology isn’t designed to be privacy safe. So we just took the approach of let’s go only focused on the content.

Tom Nixon (06:09.826)
Well, I have a lot of dumb follow-up questions, but Curtis, this is your space more than I as the performance marketer between the two of us. Where’s your head at? Have you used this yet? Or do you understand the applications it might have for your clients and campaigns?

Curtis Hays (06:21.702)
Well, I did go to your website, Brendan, and I did put in some context to the prompt that you have there for one of the brands that I’m working with, which is a high performance boat part, let’s say. And so it gave me some interesting articles about, you know, what people would be potentially reading if they were researching for boat performance.

The context I got from that just real quick was, okay, do I take this and then use those websites to build, say, a custom audience in Google Ads? What does that look like? What’s sort of the next step? So I kind of got that first step in, hey, you’re generating content for me of what my potential audience might be taking a look at.

And there’s two potential ways here. Do I take that information and put it into an ad platform like Google? Or can you connect me with the advertisers? Or I’m sorry, can you connect me with the publishers of those websites so I can be advertising on those websites?

Brendan (07:33.116)
Yes, that’s the right question to ask is like, how do you actually use Classify? So thank you for teeing that up. On the website, anybody who goes to triclassify.com can use a prompt. And we’re all kind of familiar with how chat GPT works today, or you can ask it a question and it just responds back like a human would respond back. We recently launched that new front end, you know, to kind of help illustrate what we do. And in a lot of ways, you know, it’s kind of like talking to somebody and then getting the search results of where to run the ads.

But the next step question is how do you actually action on that? And we’re connected into some of the supply side platforms. So for the marketers who are running big brand campaigns and you’re very familiar with the demand side platform or supply side platform, we’re integrated into SSP side. And then we build what’s called a private marketplace deal ID. So for this specific boat performance part that you’re running, you’d come to us, you build this curated list of

of websites to run, we would build you a custom PMP and then we’d be able to market that out to whatever DSP that you were using to buy those ads. So effectively we’re just creating the list of URLs, pushing it, and then when you run the campaign, it only targets this very specific list of URLs.

Curtis Hays (08:52.646)
And are those CPM bidding models CPC? What does that typically look like? And then the media side, are we talking, you know, mostly display ads or are you guys supporting more rich media and video as well?

Brendan (09:05.8)
Yeah. The cool thing about all that is for us, since we’re on the data provider side at the moment, you know, we chose to just focus on solving this one problem for now and helping advertisers reach that very specific content, whatever bidding strategy you’re looking for. So, you know, if your objective is CTR, CPC or CPM, or you’re driving, you know, down funnel conversions, whatever the actual end objective is, you know, we just help pinpoint your ad towards the right content.

Curtis Hays (09:10.951)
Mm-hmm.

Brendan (09:33.764)
And that’s really where we focus on today.

Tom Nixon (09:37.752)
So, I’m already kind of sold. I mean, don’t even know what I’m talking about but it’s so then but it’s

Curtis Hays (09:42.568)
Do you know why that is though, Tom, before you go into that note is because Brendan started the conversation with why and then went to how. So he, yeah, exactly.

Tom Nixon (09:50.615)
Right. That’s why I’m. Yeah, that’s why I’m tracking. Now I want to know what. So I’m ready to put this thing in my cart and check out what is that? I got to ask what’s the what’s the the revenue model? So as a customer, what am I paying for?

Brendan (10:07.356)
Yeah, you pay a very small data fee on top of, what you’re normally buying. So say you want to launch with this hat campaign, the boat campaign, you would normally set aside a marketing budget to run ads. And a lot of that goes to the publishers themselves. When the ad shows up on their page and somebody clicks on the ad, when somebody sees the ad.

Some of that is picked up by the demand side platform. Some of that is picked up by the supply side platform. Then we take a very small fee throughout this whole process. So on a CPM basis.

Tom Nixon (10:40.174)
Cool. So, I mentioned Facebook at the opening and explain how relevant is what you learned building that platform on Facebook. How relevant is that to where you ended up to launching, co-launching Classify.

Brendan (10:54.664)
Very relevant. I joined a company called LiveRail in around 2014. And then we got acquired by Facebook to help build out their advertising technology strategy for off of owned and operated properties. And we had a couple of years before GDPR, the big European regulation around privacy protection and cookies. And every time now you see this accept cookies, decline cookies, whenever you see that consent, a big part of that is because of GDPR and the Europeans focus on privacy.

also it was prior to the Cambridge Analytica thing that happened, at Facebook when kind of lost a lot of trust, think with, with a lot of users, but, digressing the audience network specifically. And when I joined, we were just building out what’s called the audience network. And that was taking all the great complexity and sophistication of Facebook’s ad targeting and allowing that to extend off of.

Facebook and Messenger and WhatsApp and Instagram for the rest of the mobile web, sorry, mobile applications and mobile web. And we scaled that by integrating an SDK out in the ecosystem. We collected a lot of signal and then we helped, if you’re buying an ad on the Facebook platform, we helped extend that off of Facebook. And to answer your question specifically, what did I learn? One of the biggest things is it is important who you are as a person.

But it’s also really important what you’re looking at. Because when you’re reading an article about something, or you’re watching sports, or whatever you’re learning about on the web, your mind is already thinking about kind of that product or that service. And typically, that’s the best time to serve you an ad. So it has a lot of extra downstream benefits, because the conversion rates go up. It’s better value for the advertisers. The publishers end up making more money. And it’s just a better user experience if you’re seeing an ad that’s really relevant to the thing that you’re looking at.

Tom Nixon (12:49.644)
Yeah, absolutely. That’s something you’ve been talking about Curtis too is that once the informed buyer is so much further down the sales funnel and closer to an action than somebody who is doing a very general search, you know,

Curtis Hays (13:01.636)
Right. And Google’s tried to get smarter about this, particularly what Google’s been talking about since launching Performance Max and now what they’re doing today and trying to get the learning about audiences from your first party data, as well as really that offline conversion loop, which in e-comm sort of happens in real time. Like, you know, somebody buys a product in e-comm, you can send that conversion data encrypted back to Google. It says, hey,

Okay, I understand Google understands then that users may be prior behavior if they came through Google as the channel before they converted and it’s going to know some information about them. In the lead gen world, which is where we do a lot of work, that’s not as accessible because somebody comes they fill out a form that doesn’t necessarily mean they were a good lead. And it doesn’t mean that they actually completed a checkout that they bought your services.

60 days later, 90 days later, six months later, right? So this offline data of giving Google that data and now the algorithm has intent-based information based off of your audience that you can give these signals. And I see Google and what’s happening with AI moving away from…

bidding on keywords, we say, hey, don’t focus on keywords, be focusing on your audiences. And the reason for that is the search terms and all those things are getting more complicated from the user side. And for me to come up with a list of search terms that any single user is going to use in the platform is ridiculous because they’re like you, smart Brendan, putting up a prompt, right? You didn’t say, hey, give me…

you know, just put in some keywords or list your product here. You actually have a full on prompt that gives context, right? And that’s what Google is trying to get to is this context ultimately, so it can match your ad to a buyer at the right time. Right? And that’s what that optimization is all about. And it sounds like that’s where you guys are headed as well, Brendan.

Brendan (15:04.06)
Very much. You know, we’re also releasing an MCP server. Are you guys familiar with this new multi-prime tech?

Curtis Hays (15:08.776)
Yes, yes, we’re, yep, yep, we’re working on model context protocol. So think APIs, but think APIs in the world of LLMs. And it gives the ability…

Tom Nixon (15:11.062)
Well, hold on. It’s an acronym. So I probably am not. What is it?

Tom Nixon (15:21.996)
You just use three acronyms. So hold on. All right, go ahead. I’m following but I fear our listeners stop. Anyway, go ahead.

Curtis Hays (15:30.579)
Well, hey, okay, so let’s say in the AI world and Brendan, maybe you’ll have a better explanation for this than me, but okay. So APIs though, were essentially, if you wanted two applications to talk to each other, you had to build a bridge. So let’s say one was talking Spanish and the other one was talking English. Now we’re talking coding language, but I’m going to use that simplicity. You had to have a bridge in between a translator. And that was, that was basically an API.

Tom Nixon (15:35.786)
AI I get I got AI.

Curtis Hays (15:57.851)
And all of that had to be structured. You had to basically create these one to ones. Well, MCPs basically are more sophisticated way where you have like AI in the middle and it’s easier to translate the languages, you know, on both sides. So basically every platform is going to have an MCP server that allows all the AIs to essentially talk to each other across the internet. And we can all tap into them. These MCPs.

and do custom agents, create automations and all these things, sharing data across, not just one-to-ones, but sophisticated one-to-many’s across the internet. yeah, how did I do, Brendan? Anything to add?

Brendan (16:42.206)
No, you’re hired. You’re hired.

Curtis Hays (16:45.788)
Hehehe.

Brendan (16:51.558)
Not at all. No, mean, Curtis, thank you for that deep exclamation because I think you nailed exactly the value proposition. In a lot of ways, it just makes it more accessible. And when we think about, you know, contextualizing content, when we think about, you know, building this new prompt based front end, our goal is to just make this accessible. And right now we’re integrated kind of for the more sophisticated marketers and advertising agencies where we work with some folks that are larger, they’re able to, you know, afford it. They spend enough.

to be able to have a seat at the trade desk or a DB360 or some of the other larger like DSPs. But the goal is to make it accessible and democratize it everywhere. Google and Facebook both work with somewhere between 10 to 20 million unique advertisers. And it’s everybody from the largest hold companies and brands, down to folks that are running 50 bucks a month. And we just don’t think it’s fair that we should only be focused on this really big upmarket.

largest agencies and brands, think everybody should have the ability to run an ad because there’s value. So the short answer is MCP allows us to provide a lot of that agentic experience everywhere and it makes it accessible. So that’s exactly one of the reasons why we’re investing heavily in it.

Tom Nixon (18:08.258)
Yeah, love it. So, anyone can go to triclassify.com to check that out. I wanted to ask you about, I was teasing about the acronyms because now that I think about it, we we traffic in them here with the peso model, the race model, right? Ida. Yep. so, but you introduced a new one, I think, new to me anyway and a LinkedIn post last week and that was Rosé. Explain what Rosé is and why that’s important for advertisers to consider.

Curtis Hays (18:23.45)
Ada. Yup.

Brendan (18:36.572)
Yeah. So the night before I was supposed to do this panel, ironically about sustainability AI in the advertising space on a yacht, which somebody called out the irony of doing it on a yacht.

Tom Nixon (18:46.232)
Wait, now I’m listening. All right. In listeners know why, but go ahead.

Curtis Hays (18:47.752)
Thanks.

Yeah.

Brendan (18:53.33)
I know I hadn’t thought about it at the time because it was a stationary yacht. It was just sitting in the dock, but still not very sustainable to have a yacht to begin with. Rose is kind of the drink of choice of Cannes and all the advertisers in the south of France. was my first time going and seeing the spectacle of being there on the cassette. And the night before I was brainstorming with some friends and we were drinking Rose and I was thinking about the words in Rose and I was thinking about

Really the fact that sustainability is deeply important to all of us. I think we all care about the planet, but we’re also building businesses and typically it’s more of a nice to have. And I think a lot of the sale about doing anything that’s carbon accounting or emissions or anything that is sustainable focused tends to be more of like a, should buy this thing because you should know about it and think about it as opposed to really tying it into specific business objectives. And the parallel is to like fleet management systems.

And there’s some really interesting fleet management systems out there that have figured out ways to track, you know, efficiency of miles per gallon routes. And they’re doing it and they’re making, you know, they’re actually, they’re saving money because they’re so focused on trying to minimize, you know, route times, how much gas they’re buying. But the benefit is really the fact that like we’re using, yes, gas and there’s fewer emissions going out. And I was thinking, how do we do that for the ad industry? And I.

saw this interesting stat recently. I’m not sure if it’s still true, but the ad industry is around 3 % of global emissions when you account for all the data centers and all of the electricity that’s used to just run this giant real time bidding structure for digital ads, including Google, including Facebook. Airlines are 2.5%. So every time you’re seeing ads, you’re running ads, you know, it’s actually less good for the planet. It’s worse for the planet.

Tom Nixon (20:40.162)
Wow.

Brendan (20:48.254)
than it is if you’re flying around. Anyway, it’s a long-winded backstory to say, as I was sitting there thinking about Rosé and I was thinking about how to tie these two concepts together. Rosé return on sustainable efficiency is a metric. And I also love dad jokes. So I love tying things into just kind of silly humor. Whether or not you end up ever seeing that in a dashboard somewhere, it’s okay if it’s not. But to reintroduce that as a concept to get people talking about and hopefully.

comes a stepping stone for folks to have that discussion. And there’s some really cool companies out there like Scope 3 that are doing a lot of brand suitability, brand safety. And then they’re also really focused on carbon accounting. We’ve partnered with Scope 3 on a couple of cool things. And their big focus is to help reduce carbon in the advertising industry. Step one was, you know, understand how to track it. Step two is tying that into specific business objectives and metrics.

Tom Nixon (21:47.149)
love it. Well, if you’re going to bring dad jokes and almost talk about Yacht Rock, then you’re going to fit in just fine here, Brendan. But Curtis, was reminded of your marketing efficiency ratio measuring two different things, but it’s kind of conceptually the same thing like what’s happening versus what are we getting back, right?

Curtis Hays (22:04.36)
Right, right. No, I bravo to you, Brendan. I mean, this is a real concern right now. I’ve I know here in Michigan and I family in Wisconsin and there are concerns about the number of data centers that are being built and everybody’s electric bills going up because of these new data centers and those types of things. And to think about it as digital marketers and advertisers and our contribution to that. And

There’s a lot of waste in the industry. So really, you know, how do we find efficiency within our industry that reduces that carbon footprint isn’t something I hadn’t thought about previously. And bravo to you for bringing that up.

Tom Nixon (22:47.714)
But I love the fact that Brendan said you may explore this in the pursuit of a revenue return, which is what we’re doing, right? When we’re trying to reduce route times and fuel and all that stuff, but the outcome will also be, you know, some efficiencies, sustainability, etc. So I love it. Brendan, do you have any case studies? mean, I don’t need a formal case study, but examples of this inaction where a company actually saw real results.

Brendan (23:16.478)
Yeah. Um, some of the early ones were just so outsized. think it, it’s almost disingenuous to talk about how good they were. Um, there are, there are outliners and over time, you know, as we’ve added in more standard segments, so still a lot of folks can just grab a segment, a contextual segment off the shelf, mapped into whatever IB category that they want. And then we also have some seasonal segments that sit out there. So we try to, you know, get ahead of things that are sports related or seasonal related.

And then the real power is in the custom ones. When somebody comes to us and says, find us this really curated list of URLs. But on average across the board, mean, we’ve been taking a look at all the campaigns that are running and it’s about a 52 % lift in CTR and the click through rate according to just industry average benchmarks. So when we think about adding these, you whether it’s just standard off the shelf things where we see a decent lift.

or whether it’s like really precise custom lists for your hat, for example. Sometimes those are like many, many multiples of what somebody was previously doing when you add in this level of contextual focus.

Curtis Hays (24:26.834)
Mm-hmm.

Tom Nixon (24:27.284)
That’s pretty impressive. Let me ask you this. just asked that or just tripped on something. Do you philosophically as a person that’s one side and does the your platform from a technical standpoint prioritize or put greater value in niche audience content communities where there’s the depth so it might be niche. It might be you know not huge massive audience, but the depth.

of the people in that community like somebody who goes to a website about cowboy hats every single day. It ends up buying one every third visit, you know, because they just love them so much. That’s the silly example. But you know what I’m asking, right? So what’s your take on that? And does your software does your solution fit? You know, factor that in at all.

Brendan (25:09.596)
Okay, so I think I just, I’m going to make an offer for Curtis for our CMO and maybe, maybe you Tom for a new, we already have a great CPL. but to run on our product team, that was a great suggestion and a great question. You know, we kind of launched with just looking at the actual content itself and trying to really deeply understand that looking at, you know, a lot of different signals. Like per web page where it’s a hunt, many hundreds of like different attributes that go into each web page. And we’re thinking about classifying it.

In addition to that, I think there’s a lot more signal that we will start to get smarter at and the systems that we’ve designed start to get smarter over time. And that signal that you just mentioned, you know, around the strength of like a community, I think is a really valuable signal. So the answer is yes, thinking about it. No, it’s not an active part of our algorithm right now, but it should be.

Tom Nixon (26:01.025)
good. Well, we could talk the position and salary and all that stuff once Curtis is off because alright, Curtis. you gotta have questions.

Curtis Hays (26:09.618)
Well, makes me think earlier today, and I want to maybe relate to the marketer who’s kind of in the position that Katie and I were in this morning who, you know, really struggle with the platforms and, you know, kudos to the platforms for taking privacy seriously. And, but what it’s done to the advertisers is it’s limited what we can do inside the platforms. And so we’re sitting here today.

you know, trying to figure out how to serve our customers in a platform like Facebook that, you know, doesn’t let us get down to the niche area that you were just talking about, Tom, and in some of those clients who are really niche to say, yeah, we don’t have a whole lot of money to spend, so we can’t go wide. How do we go narrow, the platforms don’t let us go very narrow, especially in some special categories, right? And so,

It just becomes a real challenge inside both Facebook and Google in order to do that. I mean, it seems like, Brendan, you guys are kind of an opportunity, at least from a branding perspective, to build brand awareness and some level of consideration, more top of funnel, to be able to do some of that, reach those audiences, get them to click, and then we insert them into some remarketing campaigns or those types of things with YouTube or Meta.

And I’ll just add with that, similar to what you’re doing, I’d say most of my most successful Google ad campaigns, as far as YouTube goes, are placements campaigns, where we curate the channels or the videos that we know our target audience is watching, and that’s where we put our ads. So it seems like a smarter way of targeting versus…

Tom Nixon (27:58.095)
That was kind of part-

Curtis Hays (28:02.182)
Hey, I’m interested in boat owners versus like, no, who are the people who showing off the boats or getting real technical in YouTube? And then we’re going to make sure our placements are there.

Tom Nixon (28:12.876)
that sort of part and parcel of my question I think because of the not maybe I should have I’ll claim it wasn’t or I’ll just confess that it wasn’t couched in what you just said which is the privacy if they’re gonna limit what we know about the people maybe you need to take the the the cues or the signals from the content because

My example would be for me, so I don’t know if people can see my bass guitars in the background. I’m a huge bass guitar guy, right? There’s a website that all of us nerds go to called talkbass.com. Now you’ve probably never heard of it if you’ve never played bass guitar, right? So now we’re talking very niche. But if you went to that website and you saw the content and you were selling a bass guitar, there is no question that you would rather be there than anywhere else. And it’s very niche, but it’s very deep. So the content can tell you that even if you don’t know.

about me as a person and whether I went there or not, right Brendan?

Brendan (29:08.208)
Exactly. And what’s even to go even one step deeper into what you just said, you know, across that whole website, if you think about some existing contextual technology today, it might assign the entire top level domain and or a sub domain against a specific category about bass guitars. But when we, you know, scraped and looked at all the pages inside of that entire domain, you know, we’re able to help the specific bass guitar manufacturer pinpoint

These are the specific pages that we should target. And maybe they want to target a competitor’s page. Maybe they want to find areas inside of that entire community that are talking about a very specific issue. And maybe there’s two pages. Maybe there’s 2,000 pages. But we can help provide that very specific, your point, very specific pinpointed URL level. These are exactly what pages you should be running.

Tom Nixon (29:57.687)
you. Well, we’ve reached the hot

Curtis Hays (29:59.145)
Yeah, and there’s the waste. There’s the waste Tom. But that’s that’s the thing, right? So here, you’ve got this huge website, and you don’t just pick that domain and your ads serve on every page on that domain, you actually get to serve your ads to the content that the people are actually reading about that relate entirely. I mean, I get it. I get it, Brendan. And that would be just like me, looking at a channel that says, yeah, this guy covers boating, but he’s got two videos where he really dives deep into this product.

that we’re interested in selling to people, we’re going to advertise on those two videos, right? And we’re going to be a bumper and, you know, a 30 second, 15 second before that video starts, right? I mean, that’s where there’s power. So.

Tom Nixon (30:43.032)
Yep. I was going to say, all right, so we’re good. We have now reached the hot take portion of the podcast. So I’m going to start us up. I’m going to ask you to, for especially you, Brennan, you’re the expert. Curtis just wears a cowboy hat, hot takes on the ad industry. So, but I am going to offer my hot take, which is not like a hot minute ago, we were saying we no longer buy the medium. We buy the person, right?

It’s so now this feels like a very sophisticated tech forward way of kind of doing some old school targeting, which is go back to buying the media, but be really smart about what the media is and what the content on that media is. So maybe it’s not going back to advertising in your local TV news because that’s such general content, but it is going to talk based on so my hot take. But is this sort of kind of old school meets new tech?

Brendan (31:31.847)
you’re you’re spot on and that is how you know it used to work right if you wanted to buy a newspaper ad or a magazine ad or a TV ad you would look at that specific where it’s actually running the placement inside of that but we didn’t have the volume of what there weren’t billions and billions and billions of magazines and TV episodes to run so you know it’s a very complex thing to do at scale and that’s that’s been the biggest challenge so yes I mean you’re exactly right we’re taking a very old idea

and we’re applying it in modern scale across the web and being able to give that same level of granularity or even actually much deeper granularity based on that original concept.

Tom Nixon (32:10.146)
Yeah, back then the ad kit or the media kit would just say, you know, we have this many listeners and they have this they’re in this demographic and they have this sort of income, which is all good. But that’s what we were limited to back then. So what about you, Brett? Any hot takes in terms of what people are doing wrong? There’s an adoption of a lot of new gimmick. I call them gimmicks because who knows if they’re going to have staying power, but there’s so much new popping up. A.I., you know, all sorts of things. Have you seen sort of the sheep going in a way in a direction that maybe you would

caution them against.

Brendan (32:41.886)
don’t know that I’d caution anybody. I look, in marketing, you have to test. Like you’ve got to run and figure out what products work well for you. And there is never just a one size fits all channel or tool or anything that just kind of like, you know, I think Zach a year or two ago famously talked to the ad agency world and said, just give us money and we’ll give you outcomes. And I think that scared a lot of people. And I totally get where he was going with that. But it’s also not necessarily understanding how

think about wanting to have some level of control over when they’re running a marketing campaign. But the hot take right now, I think, is I don’t know if you guys follow the Trade Desk and Index Exchange, kind of recent back and forth, but the Trade Desk is this very large DSP, publicly traded demand side platform. And they recently branded all supply side platforms as resellers, which really created some complication in the supply path optimization flow. And it’s

creating a lot of challenges. So what you’re seeing on the supply side platform side is a lot of pushback on the trade desk right now. So there’s this kind of big, you know, demand side versus supply side. It’s getting a little spicy in terms of how folks are thinking about the end of the day, our goal is to drive advertiser value and publish a monetization through the lens of a good user experience. And it sounds like there’s a lot of additional kind of noise going on right now. So it’s been an interesting thing to follow.

Tom Nixon (34:08.064)
Yeah, interesting. All right. Spicy. Speaking of spicy, Curtis, you have any hot takes or

Curtis Hays (34:12.872)
I do. I actually want to talk about something real quick that’s maybe a little spicy and that’s maybe the reduction in traffic that publishers are seeing due to zero click searches and AI. So are there concerns there from the publishers? They’re just not getting the traffic that they used to get. Now know the traffic they likely are getting, particularly if that traffic does come from an LLM, is highly qualified traffic. So they are getting that traffic and these are great pages.

I think your opportunity to convert there is maybe better than it’s ever been. But the struggle still exists that we’re, publishers are getting a lot less traffic to their websites than they used to.

Brendan (34:53.916)
Yeah, it’s a really good question. And I’ll be totally honest, I do not have all the answers. But I’ll say, you know, some of the parallels are things like when we were only running magazines, and then all of sudden, the internet was born. Publishers had to quickly adopt and figure out, how do I look at this in addition to what I’m already doing? And, you know, I think it would be silly to assume that all open web

Curtis Hays (34:59.368)
Nobody does. It’s okay.

Brendan (35:19.198)
content will just go to zero or traffic will go to zero. There will still be web pages. People will still visit that. People still read the newspaper. People still watch news at night. But it is getting more fragmented and the user experience of going to a chat GPT or going to Gemini or cloud and interacting with that as opposed to going to Google search and looking at the links, clicking on a web page that is rapidly changed and is going to continue rapidly changing in that direction. So

You know, without having all the answers, I’ll say that having already gone out and looked at most of the open web and built some really sophisticated ways to kind of help make sure an ad in whatever format, whether it’s directly through us, through a supply side platform or demand side platform, or eventually through some other type of generative AI, know, prompt based, LLM platform. However that works, you know, our goal is to make sure that we’re delivering good

value to an advertiser by just being deeply aware of how folks are thinking about content and where they’re actually engaging.

Curtis Hays (36:26.148)
Right? No, and the reality is like the example you gave up, Tom, like, when you can create a community, like I’m sure you’re, you have with this, this base website, and base players, and they’re getting information and technical, they’re not, they’re not going to chat GBT, they’re not, chat GBT doesn’t have the community, they’re still going to this website to consume that information, partly because the community is there, right? And they can, they can get that feedback, they can talk to real

Tom Nixon (36:26.21)
Yeah.

Curtis Hays (36:53.756)
real experts potentially or get that expert advice from content that’s on that website. What are your thoughts there?

Tom Nixon (36:59.682)
That’s exactly right. I think it’s not that in that case, they want a simple answer to a simple solution. They want to understand the conversation around a style of play an instrument, a pedal board or whatever, because they’ll contextualize what is important to them based on all of that. They don’t want the AI to contextualize it for them. So which this is my kind of final takeaways. I think going back to your point that there will still be content websites and people will click through to them is that two things. One is

You said how fast this has evolved from going to Google and looking for 10 blue links. Just my own behaviors have evolved like that in a heartbeat, and I’m an old dude based on my hairline alone. So things have moved quickly. However, even I can say like even when I am looking for a quick answer to a quick question, I start to then look at all right, here’s the AI response, but there’s a link that goes into more depth, which really answers the question that I came with. And I find myself clicking on that link. So it’s a different path.

But I think it validates that the best content is still going to win. was companies out like you with like yours out there trying to contextualize all that content. That’s the reason why you still have to win at the content game. Right.

Brendan (38:11.454)
It’s the reason why, you when you think about, I don’t know, when Netflix blew up and then we started to see a whole bunch of fragmented kind of digital distribution around TV shows and all these different channels. The content, at least according to me, beginning wasn’t great. It was tough to find more high quality content. And all of sudden now we’ve had some really incredible production companies and story writers and actors. And I actually feel like there’s a lot more really good content.

Tom Nixon (38:30.115)
Mm-hmm.

Brendan (38:39.634)
But it’s driving people towards higher quality things than it is away from a lot of the noise. with any type of new platform change, Instagram made everybody a photographer. And it’s great. It kind of allowed the democratization of photography and the sharing. And I think it elevated everybody else’s collective photography skills because we’re all looking at pictures now and it kind of made us all a lot more aware. So I think we’ll see the same kind of a thing with this new step change in AI. But obviously nobody has all the answers.

Tom Nixon (39:09.75)
Right. Well, Curtis, you might not have all the answers. Any final thoughts on this topic takeaways?

Curtis Hays (39:16.646)
Yeah, well, I would say for my customers, I mean, it’s just thinking about your content then, you know, whether you have a business website or you are a publisher that you even more, what we’ve been saying is you have to know who your customer is and you have to make sure you’re writing good content for your target audience. And it can’t just my hope is that the LLMs and the search engines, you know, figure this out to weed out the stuff that isn’t genuine, that isn’t good. And that

the publishers who do create good content and do create communities, they get to win in that, you know, they put in that hard work, right? And they’re working with companies like you, Brendan, who, you know, are serving up ads and those types of things to monetize the work that they’re putting in to get the payoff for what they’re doing there. And they’re not just gaming the system. So my takeaway is that I feel like

Tom Nixon (40:12.536)
Mm.

Curtis Hays (40:15.208)
There are a lot of people who’ve gamed the algorithms and gamed the system. And now we are in this position where we do have this waste. It’s like, how big is Google’s index at this point? How many trillions of web pages exist across the internet that Google has to index? 90%, maybe 98 % of which is just garbage. And so how do we weed through this is my…

I don’t know if that’s a hot take, but it’s really made me think about that today. Yeah.

Tom Nixon (40:47.47)
we’re at final thoughts now, so it doesn’t have to be spicy. But I agree with me, the best content win. think you’re not going to be able to trick even the Google algorithm anymore. So, all right, Brendan, anything you want to leave our listeners with? Of course, we will link to triclassify.com. Any final thoughts for you?

Brendan (41:07.098)
I just really appreciate you guys making the time for this great conversation. And it’s really cool what you’ve built. It’s really cool. Like the listeners, I think a lot of the content you’ve brought on just gives such unique, interesting perspectives on new tools and new ways to think about marketing. Because at the end of the day, like there is no one playbook or one size fits all. It’s all about kind of figuring out what makes the most sense for you and your brand and your budget.

Tom Nixon (41:25.165)
Yeah.

Tom Nixon (41:29.006)
That’s a great final thought. So, this was a very enjoyable episode. I learned a lot. I understood about a third of it. But no, I in all seriousness, it was really opens my eyes. I will include a glossary of terms for the dumb out there like me. I’m going to have to work on it. Curtis, you’re going to have to explain all this to me after. There we go. Why did I think of that? It’s because I can’t even spell AI. All right.

Curtis Hays (41:46.898)
They could just take the terms and put them in their chat GPT.

Tom Nixon (41:53.953)
Everyone. Thank you once again, Brendan. Great meeting you. We’ll be following you around the internet and expect great things out of Classify and everyone else. We’ll see you next time on Bullhorns and Bulls Eyes, the podcast where Curtis. Boom.

Curtis Hays (42:05.66)
Precision meets promotion.