Brendan Norman – Classify, Christian Klepp

Brendan Norman – Classify  00:00

I think overall, jobs will change. I think that people will have to spend a lot less time doing a lot of the manual, rote tasks that they’re doing today. You know, kind of in parallel with what we’re seeing in terms of vibe coding and people’s ability to build product really quickly, design new web pages really quickly, like get ship things out quickly. I think a lot of the infrastructure layer tools, or just call them like, like, chatGPT style, cloud based tools, LLMs (Large Language Models), we’ll see a lot deeper integration into existing advertising product. And what that does is it helps democratize the whole ecosystem. So I think it frees up people’s time, you know, to not have to do a lot of the basic administrative, you know, reporting, manual, campaign, optimization type stuff, and it will help service a lot better insights. Ultimately, I think the industry grows, and I think it scales even faster and cautiously, optimistically. I think that we, we will have back to building on the curation piece, and, you know, the advertiser, outcomes piece, publisher monetization piece, user experience piece, I think that all those things will increase.

Christian Klepp  01:07

When done the right way and leveraging the right approach and technology, you can classify B2B content into actionable insights and find the most similar content across the open web. So how can this be done the right way, and what role do B2B Marketers play? Welcome to this episode of the B2B Marketers in the Mission podcast, and I’m your host, Christian Klepp. Today, I’ll be talking to Brendan Norman about this. He’s the Co-Founder and CEO of Classify, a software that organizes the world’s digital content, making a privacy, safe, searchable and monetizable. Tune in to find out more about what this B2B Marketers Mission is, and off we go. I’m gonna say Mr. Brendan Norman, welcome to the show.

Brendan Norman – Classify  01:49

Thanks for having me, Christian.

Christian Klepp  01:51

Great to have you on. I’m really looking for this conversation because, man, like you know, in our previous discussion, besides talking about snow and bad weather, we did have, we did have we did have some interesting discussions around, I’m going to say, AI machine learning, and how that all has some kind of like strong correlation to content. So let’s just dive in. I’m going to start with the first question here. So you’re on a mission to help publishers increase monetization potential and advertisers target the most relevant, curated inventory. So for this conversation, I’m going to focus on the following topic, and we can unpack it from there. So how B2B brands can optimize their own content. And you know, let’s be honest. Brendan, who the heck doesn’t want to do that, right? So your company classify, if I remember correctly. It’s a software that organizes the world’s digital content, making it privacy, safe, searchable and monetizable. So here’s the two-pronged question I’m happy to repeat. So first one is, walk us through how your software does that and B, how does this approach benefit? B2B companies looking to optimize their own content?

Brendan Norman – Classify  03:01

Historically, how a lot of content gets categorized, classified, organized, it’s fairly unsophisticated, and it’s been fairly unsophisticated for a long time, just because, you know, the technology is difficult to do, and we haven’t really had the foundational ability to understand it in a way like a human understands it until fairly recently, and do it at Deep scale. So good analogy for this question is like, if you were having a we were having a conversation just a minute ago about the snow, you know, happening in Canada, and how cold it was and how much snow you got, and, you know, also around the fact that, like you had to shovel your driveway, you have a snow blower you were putting the snow. There’s a lot of different nuance to that conversation. I as a human, and most humans, are able to interpret all of that nuance and kind of positively negatively, understand that there’s a snow blower involved in that snow blower was used to remove the snow historically that conversation, you know, if it was just a blob of text, or if it were a web page, the the basic technology to understand it would have reduced it down to a category like snow or maybe winter, and that’s it, and that’s all the targeting that would have happened to that page. So our conversation, you know, gets transcribed. It gets put on a blog, or it gets put on a news site. The only thing that a machine could understand about it was, you know, snow and then potentially a keyword, tagged snow blower. And that’s all so we took a very different one. One of the reasons why you know that that makes it challenging for advertisers and also for publishers. If you’re the publisher of that content, you’re not able to help advertisers really understand the nuance to like, what are we talking about here? Because maybe an advertiser wants to sell snow blowers for that specific site. Maybe they’re looking to sell ski and since we were talking about removing snow from a driveway, probably not the best application to go sell skis on. What is helpful is to deeply understand all the nuance to like we were talking about a driveway. We were talking about removing snow from that driveway. So we invented, you know, a much better, more sophisticated way to scrape content, classify it according to all of the different, you know, nuances semantic understanding much more like a human would, and then embed all of those different, you know, semantic understandings into, you know, this, this, this file, and then we organize that in a way that makes it searchable and kind of understands all the relationships very quickly. And what that does is it helps advertisers, like if you know, I’m Honda selling snow blowers, which they make, arguably the best snow blower in the market, if they’re looking to reach people that are talking about snow removal from the driveway, they can very quickly see the list of all the different URLs across the internet, and they can build, you know, a deal ID, or they can build a targeting, contextual targeting segment to specifically pinpoint those very specific web pages. And that’s kind of how the technology works, and then also, also why it’s relevant to advertisers.

Christian Klepp  06:21

Thanks so much for sharing that Brendan that definitely helps us give, you know, some perspective into, like, what your software does. And you know, just, I’m asking you this from, from somebody who probably has learned to write one or two lines of code, and that’s as far as my dev skills go. But like, how, how is your software different from like GEO (Generative Engine Optimization), or is there some kind of overlap?

Brendan Norman – Classify  06:46

It’s fairly complementary. I mean, the problem that GEO, you know, is trying to solve, and we’ve got good friends, advisors, you know, like at Blue Fish AI and like, a really cool company, Andre, I worked with him at live rail. He was the co-founder back then, before we got acquired by Facebook, you know. And I think that the problem that they’re trying to solve is going back to that it was just stay on Honda snowblowers. They’re trying to help Honda understand how they’re represented inside of, inside of an LLM or inside of a chat bot. And what they also do is they help these companies restructure their pages for, you know, better representation inside of the other end of like a chatGPT or a cloud answer. So it is kind of SEO (Search Engine Optimization), but for the generative world where we sit on is kind of on a different side of that. It’s very complimentary, though, and we’re deeply understanding content at scale, and that’s helping, you know, the advertiser understand where to position their ad. We’re also just, you know, very quickly, moving into this new space of, traditionally, advertising technology is focused on a human going to a web page, reading that content, reading the article, watching a video, you know, whatever that content looks like, and then helping the right advertisers show up in a contextually relevant way, so that the human will click on that ad, and they’ll go to another web page, they’ll buy the thing, whatever somebody wants to sell. A very recent development, so back up a year or so, you know, chatGPT Claude when they’re out and their agents and their bots are scraping like going out to the web and they’re retrieving information. They’re doing it to train their models to make their models better at answering questions. But now, you know, fast forward to today. They’re actually spending more time just going to content and then using that content to answer a specific question. So like, what’s the best recipe for, you know, creating soft shell craps. It’ll query a couple different web pages. It’ll find that, it’ll retrieve that information and bring it back that that is not being monetized today. And there’s a really interesting thing that we’re, you know, we’re starting to work on, which is monetizing the attention of an agent. And, you know, it’s, there’s a lot to figure out, but it’s kind of like the early days of a web browser, and like early days of search, when humans would go, you know, to a search engine, they would pop in some keywords, or, like, right out of search, and then, you know, Google would look at their entire index of the web, which was an algorithm that was weighted based on the number of different contextual relevancy plus the number of connections between web pages. So a web page that I might have published in geocities.com that nobody else would link to,

Christian Klepp  09:50

wow, GeoCities like…

Brendan Norman – Classify  09:54

Throwing way back remember the days of like writing like HTML and you know, creating that, you know, looping in some type of image because nobody else had linked to that, like personalized page that you built, it would never get shown up. And, you know, the top 20 or 30 or probably even couple 1000, or maybe even 100,000 search results. So their algorithm was about contextual relevancy, plus the number of links that other pages that had to your page. And then they started to include advertising in that. So early days of ads in search were literally anything, you know, it’s any advertiser that wanted to advertise to you, and they were just kind of choosing the highest price, trying to figure out, you know, how do we make money? And then it evolved into much more contextually relevant ads and sponsored post or sponsored advertisements. So now you know, if you’re searching for, like, what’s the best, you know, LLM or chat bot, you’re probably going to see a sponsored ad from, you know, Claude and Perplexity and chatGPT. Now you’re also going to see the search results underneath those. What’s changing about that kind of rapidly is how we’re influencing because humans are spending less time going there and doing that, and also within Google, Gemini is also surfacing some AI summary quickly and kind of superseding that, creating a chatGPT experience inside of Google, which is a brilliant way to do it also. But a lot of human interaction with the web now is humans going to chatGPT going to cloud asking questions and kind of treating it like we used to treat search back in the day. So influencing that, influencing that agent, going out to the web and sitting in between. That is another really interesting way that you can help an advertiser tell that story, not necessarily to a human but to the agent who’s retrieving the information and then bringing it back to the human,

Christian Klepp  11:56

Right, right, right? And if we’re talking about content, it’s, you know, doing it in such a way that the content shows up in the AI search.

Brendan Norman – Classify  12:04

Exactly.

Christian Klepp  12:05

Because everybody, everybody’s got those now, right, like Google or Bing, or whatever, they’ve got the, they’ve got the AI summary at the at the very top of the page, right when you, when you, when you key in something.

Brendan Norman – Classify  12:17

Yeah.

Christian Klepp  12:18

Okay, fantastic. I’m gonna move us on to the next question about because we’re on the topic of optimizing content. So what are some of the key pitfalls that like B2B Marketers and their content teams? What should they be mindful of, and what should they be doing instead?

Brendan Norman – Classify  12:34

That would be actually a better question for some of the GEO companies and something like more SEO focused companies about how to specifically optimize like your content. It’s a great question. I haven’t spent as much time, you know, deeply thinking through that. And the problem that we’re trying to solve is more of, you know, at scale, what is the semantic understanding of like, how somebody has built their page and or construct the video, as opposed to advising them on what they should do? You know, to think about it in a way that’s either more engaging. I would pivot that question more to the Geo and SEO focused folks, yeah, but super high level. I mean realizing that now web has two primary users of traffic. There’s humans who are bouncing or reading a, you know, web page or watching a video. But there’s also agents. And now the scale is like, changing very, very quickly. So you know, in the next year, two years, everybody will have lots of agents, kind of doing things on the back end for them. And, you know, we believe that, you know, in the next what, 6,12,18,24 months, Agent traffic will surpass human traffic on the web. So realizing that there’s these kind of two layers that one, humans see a web page and nice pretty pictures, and, you know, they see the layout great, but also having a web page that’s optimized in HTML, markdown, JSON, in ways that agents consume that, and then also knowing the different types of agents. So the cool thing that we’re building right now, in addition to this content graph of all the content, which is effectively like a understanding all the context between the content. It’s a mouthful, an agent graph that helps to inform this is an agent coming to my site. So in a lot of ways, it’s very similar to the folks who over the last decade or so, have built these identity graphs or audience graphs, and they know that like you, Christian versus me, Brendan, they’ve got some profiling on us. They understand our search history, our retargeting, our purchase intent, a lot of things that they’re appending to like you as a specific profile or an IP address. The rapid evolution of all this is mapping out the land. Landscape of different agents, where they come from, and then the personalization of these agents, and basically applying a lot of the similar logic that we’ve used for identity graphs and for audience graphs towards agents to help understand, how do you modify the content on the back end that humans never see, so that when they’re retrieving information, interacting with the content they’re doing it, you’re presenting in a really thoughtful way that drives like the answers and the results that you want to

Christian Klepp  15:33

right, right? No, absolutely, absolutely. And in our previous conversation, you talked a little bit about contextual versus audience targeting. So and I mean, I’ve asked you this back then, but do you think one is better than the other, or do you think that they can work together?

Brendan Norman – Classify  15:50

They should absolutely work together.

Christian Klepp  15:52

And why?

Brendan Norman – Classify  15:54

The reason, the reason is, you know, knowing who you are is a very important piece to the puzzle. Like, and if you even take a step back, like, what’s the whole point of advertising? Like, the whole point of advertising is storytelling, so that a brand or a service or a company can help market their brand service to the right person they’re trying to sell them something. The cool thing about the internet is we all now have this, you know, basic shared awareness that, like, there are certain things that are paid for on the internet, certain types of content that are gated. I might buy a subscription to The Economist, you know, I pay Claude a certain amount of money, a lot to be able to use it, you know, a lot and chatGPT, and then a lot of the web is free. Facebook is free, Tiktok is free, Instagram is free, LinkedIn is free. But the economics, it’s very expensive to run these businesses, so they have to, you know, support it through advertising. Ideally, you know, there’s a couple of ways to think about it, and there’s one camp of people on the internet who think that advertising is a necessary evil or a last resort, you know, we just cram it in there and make some money. There’s another camper of folks who actually think that it can be additive to the experience. And one of the reasons why, you know, it’s kind of a meme, and you always hear people talking about, you know, I didn’t need this thing, but I saw an ad for it on Instagram, and just had to buy it because it was really cool. The reason why that exists is that their advertising is phenomenal, and the targeting and optimization is phenomenal. And why it’s phenomenal on the back end is it knows a lot about you know me, who I am, what I’m interested in, based on my history, what I’ve been engaging with, where I’m spending time, you know, what I’m looking at, but it also knows specifically when I’m looking at that thing, you know, it might have a framework of saying, Brendan, really, you know, likes these types of skis, you know, he’s interested in, You know, a couple other, couple other interesting products, but the best time to serve each one of those products might be different, and it’s different depending on what I’m looking at, what I’m thinking about in that exact moment. And to kind of align these, these different graphs, graphs of intent, contextual understanding, and then audience, you know, the best time to serve me an ad for a new pair of skis is when I’m reading an article about skiing or something about the mountains. You know, it’s not necessarily when I’m reading about the Warriors, because I’m not really thinking about skiing when I’m reading about basketball. So to your point, the most effective ads are when you’re combining those two sets. It’s great for the advertiser, because I’m much more likely to click on it and go check out the skis. It’s also giving me a better experience, because it feels more native to the overall content that I’m reading. And that’s why it’s so important. It shouldn’t be an afterthought or a necessary evil or a last resort. It should be something that is intentionally thought about the entire design, because it can, it can actually be a cool experience.

Christian Klepp  19:06

Absolutely, absolutely. I mean, you know, you’re talking to somebody that started his career in the in the advertising industry, so, yeah, I’ve heard that one before, and what you’ve been describing in the past couple of minutes sounds to me a little bit like time of day marketing too, right? Because you’re you know, are you the had a guest on, like, a year ago who talked about this? Right? Is, is Brendan, the same guy at eight in the morning and one one in the afternoon and seven in the evening? Right? There’s different different times of the day, different mindset, different motivation, different reason for being on your device or looking at, looking at specific type of content, right? But it is interesting, right? And it’s interesting and sometimes a little bit scary, how, um, how quickly the algorithm picks, picks this stuff up, right? Like, for example, last year, I was researching a lot on Japan, because we went there, right? Family trip and whatnot. And. And that’s what I kept seeing on Instagram, right? Like, because I was looking up specific temples and whatnot and and today I got another push. Like, would you like to invest in a temple that’s an on island in the Sea of Japan, right?

Brendan Norman – Classify  20:12

Like, sorry, did you invest?

Christian Klepp  20:17

No, I did not. But it was just, it was just funny that I got that ad right, like, it’s, like, Okay, interesting, but like, it’s so like it not, was not on my radar at all, right,

Brendan Norman – Classify  20:29

Yeah,

Christian Klepp  20:29

Okay, great. From your experience, and you talked a little bit about it now in the past couple of minutes, but like, from your experience, how can leveraging AI agents improve efficiency and save marketing leaders time?

Brendan Norman – Classify  20:47

Ooh, there’s a couple different ways to think about that. So you know, part of it is this new agentic framework for how existing tools, you know, advertising and marketing tools, will communicate with each other today. You know, it’s fairly complex. You know, if I wanted to go build a contextual targeting segment to help one of our brands that we work with find the right contextual or inventory to target contextually, I would have to work with them. We build a targeting segment. We would upload that into our one of our SSPs, we would build a deal ID, you know, they would connect it back. And there’s a lot of different pieces that happen along the way. And each one of those pieces you have to go to, you know, a UI, I’ve got to go to a dashboard, I’ve got to push that thing in. Some of it happens through an API, but a lot of it happens like going to a whole bunch of different web pages to make sure this stuff all works. So stuff all works. What’s cool about agents? And I’ll unpack this, and then I’ll go to the more of the consumer focus side too. But what’s really cool about agents using, you know, things like the ACP framework from the Agentic Advertising Org., the ARTF (Agentic Real Time Framework) from IAB Tech Lab is they’re kind of built on some of the existing frameworks that allow humans to use natural language to communicate between these different systems. So there’s still the back end pipes of API pushing data or pulling data from one system to another. But on top of that is more of an agentic framework that allows, you know, a human just to use some prompting, like in chatGPT, to make a request, you know, that talks to a back end system. So that’s one part of the agentic framework for like, you know, how to think about this through the lens of advertising and marketing. And then the other side is, you know, more of the consumer focused. There are so many interesting and very quickly growing tools you know, that you can start to plug in, into Cloud, into Cloud code, and to building things that just rapidly accelerate development of different products and your ability to analyze data quickly. I think in the next, you know, 6 to 12 months, we’re going to have a totally different landscape for how people are buying like trading media also, you know, one more final thought about all of this is that a lot of the sophisticated tooling and pipes that we have are only accessible towards the largest advertisers today. And I think that you’ll pretty quickly see a democratization of the ability for anybody to just buy programmatic ads, whether you’ve got a $20 a month budget or a $20 million a month budget. Now, the ability to similar types of tools to access the right content across the web will start to be available towards a lot more folks outside of the existing, you know, kind of ad tech ecosystem.

Christian Klepp  23:55

And I might be stating the obvious when I say this here, but that’s a good thing, isn’t it, because, I mean, I, again, I came out of this industry, and I know that, like, you know, if you wanted to advertise in the New York Times, for example, right? Like, how expensive that would be, or, or anything that was print, right? And then they migrated all that to digital, and then it still wasn’t, it still wasn’t affordable. It was, it was cheaper than print, but still not like, exactly like, you know, yeah, I wonder, wonder if they’ll be worth the investment or not. And then now you have this, this push towards the democratization of all of this through AI and machine learning and, and I do think that you know, for all the the scare mongering that you know people are doing now with, with, oh, you know, all this stuff around AI, I do think that that part certainly will be advantageous to to B2B companies and to marketing in general.

Brendan Norman – Classify  24:49

Great. I mean, yeah, optimistically, I think I’m excited about the entire landscape changing because it does a couple things. It allows for much more contextually relevant ads. I know right now there’s only, let’s call it to the magnitude of like, 1000s, 10s of 1000s, maybe hundreds of 1000s, of campaigns and or brands that are able to use these pipes to reach the largest publishers. And all of a sudden you expand that out. You know, I think between meta and Google, they each have somewhere between 15 to 20 million unique advertisers on their platforms, and what that means is, you get really hyper specific ads. And it also means that, like, I might get a local ad for my hometown here for some restaurant that’s launching a promotion that I might only get here, and I might only get to your point, maybe not in the morning, but I’ll get in the evening. There’s a lot of different data sets around my identity, you know, the psychographic profile, contextual understanding of what I’m reading at that exact moment. And what it does a lot of things. It helps smaller brands get more traction, get more visibility. It also just helps improve the publisher experience, and like publishers, make more money. And then the user who’s consuming that content, reading the web page, watching a video, also has just a better experience. And then the other layer of that will continue to just go on, this narrative of agentic, tension, but the agents who are reading that content, watching that video for an end user. On the other side, are also able to interact with advertising content that’s very contextually relevant to the content that they’re consuming again, and it’s good for the storytelling of the advertiser and good for monetization of that publisher too.

Christian Klepp  26:38

Absolutely, absolutely. Okay. So how can high fidelity curation? This is the next question, right? How can high fidelity curation make B2B companies more sustainable? And if you can just provide an example,

Brendan Norman – Classify  26:54

Curations like, it’s such an interesting term, but you know, effectively, it’s just, it’s helping to use the word and the definition, the definition in the word, curate the right inventory to run an ad campaign on, and curate the right inventory and audiences. So it’s a really important part of the business. I think it involves a couple things. It involves front end targeting, of knowing who’s the back to that question, who’s the audience, and then what’s the right content, and then it also involves a lot of ongoing optimization. And I’ll say that there are some some interesting companies that that are really good at curation, who are building out the right automatic tools to think about more real time optimization, and it’s something that the really big social media companies do very well, like they’re constantly looking at lots and lots of signals when they’re running a campaign, and they’re looking at inventory and stitching together based on the signals that they’re acquiring around. Why certain campaigns do well, to your point, you know, when we’re testing that, selling that pair of skis to Christian, we’re testing a lot of things. We’re testing what he’s reading, you know, we’re testing maybe time of day. We’re testing, you know, where he is. There’s a lot of different elements on the back end that they will ingest and understand and then refeed into that targeting and optimization algorithm. And I think that that is one of the cool things that AI to use, like the air quotes, AI will help enable the processing of a lot of this data to just be a lot faster, be a lot more cost effective, and a lot of these systems that you know previously have been not accessible to the ad tech ecosystem, just because we we operate at such a crazy scale of 10s, hundreds of billions of requests and impressions and transactions that happen every single day. It’s very cost expensive if you’re processing all of that data and all these different signals, with the advancement of how the model cost is getting a lot less expensive, very quickly, not just from an LLM perspective, but then the foundational layers and the infrastructure layers, like we’re doing contextual intelligence as an infrastructure layer. There are inference layers that all kind of sit underneath the LLM and help inform an LLM understanding of that content. As those costs start to decrease, you’ll start to see a lot better performance from curation, just because, you know, it’s not as cost prohibitive, and we’ll be able to find that balance in terms of economics.

Christian Klepp  29:45

Yeah, yeah, you hit the nail on the head there. Because, you know, I was just writing this down. You said faster, more cost effective and in my head, and you said it, it’s like, and at scale, like, you can scale this stuff faster, like, when I when I think back, like, years ago, when we, when we launched an ad campaign, and, you know, just the amount of effort, like, for the print and then the cost into, you know, the media placements and all of that and and just alone for like, one city, just just the amount of investment that was involved in all of that, right? Just think, thinking about that. It’s like, gosh, and then now you can scale all of that, like, even faster, because it’s because it’s digital, right? So it’s just such an incredible evolution. Like, I’m getting just as excited as you are man, I’m like, for this next question. Brendan, I’m not sure if you’re the type that likes to do this, but I need you to look into the crystal ball for a second here, right? Because we’re looking at, like, stuff that is, you know, the events that are yet to come, if I’m gonna that, make it sound a little bit suspenseful, but, um, the future of digital advertising, like, how do you think that could become less fragmented and more optimized with everything that we’ve talked about in this conversation.

Brendan Norman – Classify  31:04

Yeah, I caution against, like, having any, any specific predictions, and more of, like, a framework for, I mean, for me, at least, yeah, more of a framework for how I think overall, jobs will change. I think that people will have to spend a lot less time doing a lot of the manual, rote tasks that they’re doing today. And, you know, kind of in parallel with what we’re seeing in terms of vibe coding and people’s ability to build product really quickly, design new web pages really quickly. Like, get ship things out quickly. I think a lot of the the infrastructure layer tools, or just call them like, you know, the like, chatGPT style, cloud-based tools, LLMs, we’ll see a lot deeper integration into existing advertising product. And what that does is it helps democratize the whole ecosystem. So I think it frees up people’s time to not have to do a lot of the basic administrative, reporting, manual, campaign, optimization type stuff, and it will help service a lot better insights. Ultimately, I think the industry grows, and I think it scales even faster. And, you know, cautiously, optimistically, I think that we, we will have back to building on the curation piece, and, you know, the advertiser, outcomes piece, publisher, monetization piece, user experience piece, I think that all those things will increase, and I I’m hopeful that with the integration of just better technology, embedding AI into a lot of these systems, it’s going to help steer us towards having better experiences across any type of Publisher content. I think that the advertisers will see better outcomes. I think that the people that are in this industry will get to think more creatively about how they’re, you know, building better creative storytelling, better reaching the right people with those stories. And my hope is that it just continues to expedite and grow the overall industry.

Brendan Norman – Classify  33:17

That will be my hope as well. All right, get up on your soapbox here for a little bit. What is a status quo in your area of expertise? So anything that we’ve talked about now in this conversation, what’s the status quo that you passionately disagree with and why? Oh, you must have a ton.

Brendan Norman – Classify  33:44

I definitely do. I mean, you know,

Christian Klepp  33:48

just name one, just one,

Brendan Norman – Classify  33:50

Like in any industry, you know, there’s always, there’s always the early adopters, you know, there’s always the kind of like the middle stack, you know, there’s always, like, the laggards. There’s definitely, you know, a smaller, but growing quickly, minority of folks who are really leaning into, you know, I’ll just call it AI, and then the agentic web, and there’s a lot of discussion right now in ad tech around like, what that means? I’m still hearing that. There’s a lot of skeptics who are kind of making fun of it, or, you know, trash talking about different protocols. Fine, like those are the folks that are absolutely going to get left behind. And I think a lot of those folks on the soapbox in the next 6 to 12 months will look back at, you know what they said, and we’ll all kind of say that didn’t age well, and you were not building this stuff. You weren’t fingers on keyboard or hands on keyboard. Vibe marketing, vibe targeting, building stuff like shipping new product and testing and iterating. What I what I don’t think, is that the really big platforms are just able to be super nimble and adapt to a lot of these new frameworks quickly, totally like the pipes will continue to stay there. I think that there will be startups that are more nimble, that can build and ship things, you know, proof of concepts, prototypes, get things out, learn from them, fail, iterate, and then start to scale meaningful businesses without having to rely on a lot of the existing infrastructure that exists today. Do I think the trade desk is, you know, going anywhere? No, do I think that they will, like, continue to be a valuable piece in this ecosystem, absolutely. And I think that they will ship things. I think that they’ll enable the industry like to build on top of of the pipes that they’ve already built. And at the same time, I think a lot of that rapid advancement will come from startups who are kind of proving that, like they don’t necessarily need the existing pipes and channels to be able to at the end of the day, you know, this whole ecosystem is about helping an advertiser surface their ad against the right content for a human or for an agent. And there have been a lot of folks kind of sitting in the middle for that space for a long time. One of my favorite stats, soapboxy stats, is that if an advertiser puts $1 in to the open web with a programmatic web, 35 cents comes out to a publisher, so 65 cents is being taken by some combination of middlemen, you know, who are collecting a margin for, you know, different services, also some version of fraud. There’s a lot of things that happen in between that and what I’m again, cautiously optimistic about, you know, like the big picture, AI, of facilitating, is the ability to reduce that margin so that, you know, advertiser puts $1 in. A lot more of that dollar comes out towards the publisher, I think big social media, you know, it’s around 70 cents comes out. So they take, you know, somewhere between 25 to 30 cents, which is kind of the value exchange of providing the services, all the targeting, all the technology that goes into supporting that, you know, as a more fair exchange. So I think what a lot of the folks on more of the startup on more of like the front end of the frontier tech in the space we’re excited about is getting to reduce a lot of that inefficiency and a lot of that margin in the middle, and helping more of that dollar show up towards the publisher where it should.

Christian Klepp  37:34

Boom and there you have it. Man Brendan, this has been awesome conversation, so thanks again for your time, please. Quick intro to yourself and how folks out there can get in touch with you.

Brendan Norman – Classify  37:45

Yeah. Brendan Norman, CEO co-founder at Classify, please. You know, hit me up on LinkedIn or shoot me an email. Check out our website, which is, you know, www.tryclassify.com. I’m happy to connect. You know, if you have questions about advertising from a publisher side, from an advertiser side. Love to chat about it.

Christian Klepp  38:06

Sounds good. Sounds good once again. Brendan, thanks for your time. Take care, stay safe and talk to you soon.

Brendan Norman – Classify  38:13

Cool. Thanks, Christian.

Christian Klepp  38:14

All right. Bye for now.