Why "AI Visibility" is critical for your brand with James Cadwallader, Co-Founder & CEO at Profound

Hello and welcome to another episode of Riding Unicorns. Today we're joined by James Cadwallader, CEO of Profound. James, it's great to have you on. Thank you for joining us. So James, we'd love to get into it. You're tackling the hard problem of making sure companies get found within AI search results, plus a lot of other things that we can't wait to dive into. But maybe you could just start by introducing yourself and the business and what you guys have done so far.

Sure. Yeah.

Thanks for having me, James, Hector. It's a pleasure to be on here with you guys. I think, you know, as maybe a precursor, this is the biggest platform shift in the history of the internet, maybe the history of capitalism. You know, if you think about what is the one biggest and most important technology that has been born from the internet, it has been search. You know, it constitutes roughly $200 billion of Google's $300 billion-odd in revenue comes from search revenue, search advertising predominantly. And I think it's been a no-fly zone, by search it's been like a no-fly zone for the last 25, 26 years really where it's just been essentially a monopoly. And for the first time ever, we're seeing that change in this true innovator's dilemma where the world is retrieving information from the internet in a different way. So people now just talk to the internet and it talks back. So you ask ChatGPT a question, and instead of having to go and click through pages of blue links, you just get the answer that you were looking for. And what's happening on the backend there is very interesting, which is ChatGPT in real time is sending a user agent on your behalf to that same index of blue links or that same index of URLs. And the user agent is essentially representing the consumer now and doing that web browsing. That's the backdrop of our business and it creates this huge second order problem for every company on the planet. As now in, you know, we, for the last 26 years, if you are a company with a website, you've been building content for human beings to consume, for human beings to discover. And that's been very focused on storytelling and imagery and, you know, all of the things that matter to humans. But now you You are building content for user agents who are crawling on behalf of a human being. So yeah, and user agents look for very different things. They don't care about storytelling. They're not looking for hyperbole and fluffy sentences. They don't really care about imagery. They want the raw information that they can pull through and use to create a compelling answer with the LLM inside of the application, which is ChatGPT. So really, yeah, I guess like using that as a full circle point, that's the backdrop of our business and a profound— we are building a marketing platform for this new era of the internet. This is a platform that helps you understand whether you show up. It helps you understand how you show up when AI talks about your brand and products, services, competitors. But most importantly, our platform actually helps How do some of the biggest brands in the world create content for user agents, create content that's designed to be absorbed by AI? And yeah, we launched in August of last year. So August 2024. So kind of coming up to, I think, 18 months ago. It's been a wild journey. The team is pushing 100 people now. We work with some of the biggest brands in the world. We're relied on predominantly by the enterprise. We literally work with Fortune 10 brands. And thousands of marketers rely on our platform every week to win this new world of AI search. Yeah, obviously you guys both have sort of investor backgrounds. So I guess in terms of our fundraising history, uh, our first, you know, my co-founder and I met at an incubator called South Park Commons, uh, here in New York City, which was set up by two of the founding engineers of Facebook. We then had Keith Raboie and Coastal Ventures co-led our seed round with a fund called, it was a new fund called Saga Ventures, which was set up by Ben Braverman, who's formerly the CRO of Flexport. Max Altman, yeah, he was previously Zenefits with Parker Conrad, and he famously set up a very successful fund prior with his brother Sam, uh, and Thompson Nguyen, who was formerly at KP. So they co-led our seed round with Keith Rabois. Then Kleiner Perkins and NVIDIA Ventures and Ventures led our Series A. Sequoia Capital led our Series B. And then, yeah, that takes us up to where we are today. So yeah, so there's a long monologue, but that's kind of like the full run-through of our business.

No, James, it's fantastic. And yeah, I mean, you guys have done an incredible job positioning yourselves as probably the number one player in this kind of GEO space. And I know you're now doing a lot more than that, but you literally are, you know, your customers are customers that We and all of our listeners are using every day, you know, they're our favorite tools. And so you've built an incredible brand very, very quickly. Can you tell us in a little bit more detail what GEO is like, what, how you are improving visibility for customers and showing them how visible they are, but also talk more about this content production piece that you're now focusing a lot on.

Yeah, for sure. I think something that's been very interesting on this journey so far is, I mean, I think initially I underestimated the human tendency towards heuristic. We want to make sense of the world that we exist in by inferring from the past and from history. So I think everyone wants to be able to put a label on this new art and the closest heuristic is something like SEO, which admittedly is a huge industry. SEO is a $100 billion or $90 to $100 billion a year industry. But I do think both things can be true in that there's similarities there. And I think, I mean, a thing that I would be very cautious to point out is that I think the people that worked on SEO and the people that mastered SEO are very well placed to be the people to be, you know, some of the main players in this new category. But I do also think there's some fundamental differences between this new world of AI search or GEO, which stands for generative engine optimization or answer engine optimization, AEO, or sometimes I would just refer to it as AI visibility. There's a, it's verging on a meme at this point that no one can put a label or a name on the category. I think what's different at the most philosophical level is that you are creating content for bots, not humans. The whole idea of SEO, if you pass it back to, you know, first principles, it's you are creating content that is designed for human beings. It's designed to rank, of course, which is similar to what we're talking about here, but it's designed for human beings to stumble across and consume. The whole idea of this new category of answer engine optimization is that you are creating content that's designed for bots, not humans. And I think that really underpins the core differences in the types of content that you're creating and how you measure it and what the metrics that matter are changing in that, you know, for SEO or search, it was all about human traffic. Here it's, you know, retrieval traffic, user agent traffic matters. More from my perspective. So yeah, we, the way that our platform works is we capture huge amounts of data. We send prompts to the platforms and capture their responses. We have a huge consumer panel of real user prompts that we use to help us understand what people are actually typing into these platforms. We also integrate into our customers' websites to capture that first-party data that helps them understand how user agents are visiting their content and utilizing their content. But that's just the first part. Really what we're doing is using reasoning models to then go and analyze and interpret all of that data and all of those insights and use the data as context to build informed new content.

And what does the, what does that new content look like if you go to you know, a huge company, let's take British Airways as an example, they want to do, they want to optimize their search visibility. How does your production engine manifest? Like, are you putting out, spinning up new web pages on their website with just sort of raw information that these crawlers love? Or are you spinning up content that actually looks a lot like SEO platforms?

Well, let's go through it. So that's it. Let's use that example. Say if it were British Airways. What we would do is we would understand how, let's just say British Airways and they want to focus just on ChatGPT. We go across all of the models. We look at ChatGPT, Gemini, Claude, Deepseek, Meta, trying to think, Groq. So we go across all of the major platforms, but let's just say it's British Airways and they really want to focus on ChatGPT. The first thing that we would do is look at one of our sources of data to understand how people are talking about British Airways or flights, et cetera, when they're asking questions to ChatGPT. We have a database with over a billion real user prompts in that we extrapolate out from to help us understand directionally what people are typing in and how they're searching for things like flights.

Who's the best airline? What's my best airline for London to New York?

Stuff like that. Yes, we'd understand.

You get that data, James, is that publicly available or is it bought? How did you get that?

Yes, it's really hard to find and it's expensive. Yeah, we've gone to consumer panels. So consumer panels, for, you know, anyone listening, are— it's super normal in the marketing world. I mean, consumer panels have been around for about 102 years. Nielsen was the, the first, um, they did it for TV. It was like a panel to help people to, you know, help marketers predominantly understand how consumers were watching television back in the day. And today there's hundreds of different consumer panels that exist, and it's all kind— there's all kinds of ways to capture consumer panel data. It's always opt-in, it's always like they compliant with all privacy and data laws, of course. But, you know, sometimes it is people getting paid a small amount to share their data with one company, with you know, one company across lots of different behaviors. Sometimes it can be in exchange for access to a free product that people will opt in to have their data shared. Sometimes it's in exchange for gifts. There's all kinds of mechanisms to encourage people to exchange a window into how they use the internet as a whole in exchange for something. So we've basically gone to all of the biggest consumer panels. We spent millions of dollars building a patchwork quilt of what is essentially now a mega panel of real user prompts, which is really hard to do because they, as you'd imagine, there's lots of different schema and there's like, it's not one formalized, standardized kind of way in which you collect all of this. So we've had to do a lot of work in a lot of data.

Is the value in the first prompt? Because after that, you presumably kind of know what they're asking at that point.

No, absolutely not. No, I'd say the value is in the whole, in all of it. Yeah, super interesting for, you know, a British Airways to know that when someone talks about British Airways, it's usually, you know, in conversation turn number 15, for example, or these are the other things that they're talking about in that same conversation. So, and I think, you know, the human mind is not great with numbers, but a billion prompts is a lot of prompts. And we add 150 million a month. So we have this kind of real window into a sample of what the world is asking these platforms about. And, you know, we— it's not just a case of buying the data either. You have to productionize all of that data. You have to put it into one big panel and then be able to extrapolate directionally beyond it. And yeah, it's super powerful and it allows us to have a window into an understanding of what people are asking and how they're asking it. And we, yeah, we're super conscious about, you know, making sure that all of the data is absolutely compliant with privacy laws, etc. And we don't expose the raw data in our platform. It's only extrapolations and predictions that we expose in the platform. But yeah, that's a long-winded— Yeah.

And then you're taking, so you're taking this information, then what are you, how are you producing the content?

How are you?

Yeah. So we went on a bit of a tangent there. So basically there's a few stages here. So we're going back to British Airways on ChatGPT. The first thing we're looking at is how do people talk about British Airways and flights in ChatGPT. Then what we're going to go and do is we're going to go and send prompts to ChatGPT and look at how it responds. How does it respond when I ask, you know, what's the best airline to fly from JFK to London Heathrow? Does British Airways get suggested? How does it get suggested? What else do these models say? Like, does it just say— it's obviously not going to just provide a link. Like, it's going to say something. Like, you know, ChatGPT loves to opine. We know that, right? It won't just say British Airways, it will say British Airways because of their great service and their, the attention to detail and the fact that they don't have long lines, for example, it says things alongside. So we're capturing everything that ChatGPT says and we're doing that thousands of times a day to understand in higher volume because these models are probabilistic. So if you ask the same question 10 times in a row, you'll get 10 different answers from ChatGPT. So we capture all of those responses in high volume and we understand whether British Airways is showing up, how they're showing up, and then more importantly, why. Because when ChatGPT or all of these answer engines provide an answer, they have to go to the internet to build that response. Or like 8 or 9 times out of 10, they do. If it's a really simple question, a static evergreen question, like how do I tie my shoelaces? The model will just use the pre-trained LLM. But for anything with some sort of like commercial intent, like, you know, what's the best airline or, you know, what's the best running sneaker for a marathon is one that gets thrown around a lot. Like the models will always rely on the internet because it doesn't just inherently know the most up-to-date information. So it's using the web, you're sending a user agent to the web to get that information. So we're able to understand where did ChatGPT go to build that response. Did it go to britishairways.com? Did it go to Skyscanner? Did it go to my— the Points Guy, a bunch of random websites? Did it use Reddit, Wikipedia? Where did it go? And that varies from category to category, business to business, industry to industry. You will use different combinations of that. So think of it like this. We're capturing all of that data, everything I've just mentioned. And then this is the important second part. We're giving it all to an orchestration of reasoning models and saying, you know, with the prompt, these reasoning models, hey, here's a ton of context of how ChatGPT answers these questions and where it goes to build these answers. Go and look at the commonalities. Let's go and look at the patterns. What is it across these 10 websites that are being used to answer the questions like, what do you find? What are the, you know, what's the schema of the website?

What's the structure and the company's visibility?

Well, using AI, you're then able to build insights and suggestions and recommendations of what you can do. And that again varies based on the category, the product, you know, what, what, what it varies from, from industry to industry, business to business, but quite often that will be a combination of let's create content for britishairways.com based on the content that we see is performing best. It could be let's look at earned channels like YouTube or LinkedIn, or I mean, I guess that's kind of like operator channels like YouTube or LinkedIn, you know, social channels that the models rely on a lot, or it could be earned, you know, going to third-party blogs or affiliate sites, etc., or even places like Reddit, Wikipedia. So yeah, you're using our platform to build those content strategies and create that content and, uh, then distribute that content or publish it.

Super interesting.

And James, with the greatest of respect, it feels like quite an obvious business idea, but then, but then a very complex one to execute, um, because you've got big players, big datasets, challenges. There's a lot going on there. Were you ever put off by the fact that maybe, maybe this was just, you know, this was going to happen and maybe it was too obvious? Or was that the driving force to be like, this is inevitable and we're going to be the ones to do it?

No, it wasn't obvious when we started. It was not obvious. It was, it was so—

no one else was talking about it.

There was like rumblings, but it was very, very earth. I remember feeling like I was going crazy because it was, I mean, I shouldn't say it wasn't, it was obvious to me and my co-founder Dylan at the time. I was really, I was, I was, I remember vividly feeling very confused. It's almost like some version of like gaslighting or something where you're like, Well, if the biggest monopoly in the history of capitalism has built its entire business on clicks, and this new technology means that we don't have to click, then surely that matters. Surely that's gonna completely disrupt marketing. And you'd mention it to people and they'd be like, yeah, yeah, interesting. Yeah, I could see how that could happen. But that was it. Like there wasn't, and you know, we, we really, kind of, it felt like a leap of faith in the beginning of 2024. I mean, we started building Profound when ChatGPT didn't even use web search. So this is when ChatGPT, you'd ask it who's the president of America and it would say, you know, as of my last training data and get it wrong.

Yeah.

And really this whole category only opens up because of the web search that the models are using, like trying to influence the pre-training of LLMs is. Far less tractable. It's much harder to do because really you're just trying to influence like it's based on what side of the bed like Sam Altman and Greg Brockman get out of to an extent.

And then so it was obvious and you almost sort of pained by that inevitability that this was going to be a massive thing, which I think is a brilliant feeling to have probably all that, you know, it's two ways, but a brilliant feeling to have. Weird feeling. Yeah. But then you've just talked us through how it actually works, and you've broken down each step of how do you actually get to a meaningful outcome to have meaningful results. Where does that kind of first principle thinking come from for you? Where did you kind of learn those skills? Yeah. Is that your superpower as a founder, to go, big problem, I'll work it out?

I don't know. I mean, I've never met Bryan Peterson, but you know, I, I've always been a fan. I think he's an impressive founder and I, there's something that he, he referred to himself as a business junkie, which I found like very relatable as a, as a concept. I enjoy like left to my own devices. If you, if you know, let's say if Profound vanished off the face of the earth and it was my one job was to relax and enjoy myself. I would find myself listening to podcasts like this, reading books about the great— it's David Senra at Founders Podcast is an investor in Profound. And I've had this conversation with him before, which is, it was like meeting, you know, an idol when I first met David. And he is a fantastic guy, but I just listened to his voice so much because I love, yeah, I love business. I don't know what it is, like a sport. And I enjoy it and I've been running my own companies since I was very young. I've never had a real job. And yeah, this feels like a culmination of everything. I feel very lucky to be here. Yeah. So yeah, when, when the universe as well, like, you know, offers up an opportunity like this where it's like, hey, this is a, you were right. This is a huge platform shift. You're well placed, well timed. You go all in, you give it everything, you reciprocate you reciprocate the opportunity by offering yourself almost in a sacrificial way.

What does that mean for you? What does going all in mean for you and what you've done with Profound?

I think it means that you can't work harder. I think it means that you give everything, right? You leave no stone unturned. You work as hard as you can. And, you know, thankfully I'm 36, so I mean, I've been working very hard my, my whole life, frankly, especially as a founder of businesses. You know, I've got a very thoughtful fiancée who puts up with it. And it means, yeah, what does it mean? It means that, you know, you're working 12 to 15 hours a day every day. It means that you're in the office on Saturdays and you're working on Sundays. It means that you don't do much else except work. You know, I don't have a good response when people ask me what are my hobbies these days. There aren't any hobbies left. But yeah, it's, it's, I, people sometimes ask if it's fun. I don't know if fun's the right way to describe what this feels like, but it's very, uh, if you, you know, the Maslow's hierarchy of needs, sorry, very self-actualizing, I would say is the, is the way to explain it, which is quite energizing.

You talked about going against, uh, well, you said how SEO has actually quite a good, is quite well placed to win in this, in the AI space as well. Like, what, what do you mean by that? And how, how can you use your sort of startup mentality and the infrastructure around that to, to win?

I, I've got thoughts on this, and it's interesting. There's been, you know, not, not unanimously, to be fair, as well, but there's been some pushback from SEO folks who I think it comes maybe from a place of fear, or I don't know where it comes from entirely, but this kind of pushback from SEO people being like, oh, it's I don't know, a kind of reluctance to see this as—

Can I add something there, James? I heard an SEO expert describe GEO as SEO. They was— they were like in denial. They were like, well, GEO is just SEO. It's just making sure you rank top so that when the agent goes and searches, you're, you're already ranked top, which is SEO.

Yeah. And it's like, you know, it's not—

I, I was— was he wrong? Was he wrong?

There's— it's kind of also true that there's a ton of similar muscle groups, and it's like entirely true. And it's like, I'm not— I wouldn't go toe to toe and say that there's no similarities there. And it's like, it is the same muscle groups. But I'd say the point that I'm making is that I, I also believe that is truly a different category, and it's only going to get more different as you are building content for machines, you are building marketing for machines for the first time ever. Not for machines to rank, but for machines to consume. You're trying to change the opinion of machines, not humans. You're trying to influence machines, not humans. It's primitive and from first principles, it is very different. But this is my point and this is my kicker and I'd definitely implore you. In fact, I'd, I'd ask very politely that you don't edit this out. I think I'm like so pro SEO. I think SEO teams are like they have all the right muscle groups to be the marketer. They've, you know, SEO's quite, for quite some time have been the sort of forgotten child of the marketing stack, despite it being a very important part. Most businesses still rely quite heavily on search and SEO, but the SEO teams have been this kind of neglected part of the stack to an extent. It's not so sexy. It's got like this kind of like, you know, there's this kind of something that's attached to it that's not so sexy. But I think this is the opportunity for those SEO teams to become the hero of the marketing stack. I think that over a 50-year time horizon, the only thing that will matter in marketing is how does AI talk about your brand and products and services? I think that will be the most important factor of marketing. And I think SEO teams have this interesting combination of skill sets where they're very analytical. It's not super— it's like this combination of like there's some creative thinking, it's very analytical, it's very data-driven. They're used to this kind of loop of monitoring, measuring, creating, and learning over time, and it's very similar muscle groups. And I think— and yeah, of course it does also, it requires indexing and ranking on search. So there's, there's I think they're, they're so well placed, these teams to dominate here that they should be celebrating this as their, a big opportunity.

And what are you going to do? I mean, there's a, there's a whole, like the whole startup, well, the whole business world is kind of debating who's going to win AI, you know, when it comes to agents, is it going to be the startups or is it going to be the incumbents with the data and the distribution? And you know, that, that is true from Salesforce to ServiceNow, you know, all of these incumbents who are starting to get their act together and build agents, like, and it's relevant in, in your space as well. Like you're, you're speaking highly of the SEO industry, but like, what's, what's your advantage as a startup? Why will you win?

I mean, I think we've already built this advantage in that You know, if you were a UFO hovering over planet Earth right now and you had to select one company with the most data around how answer engines or AI interfaces with brands on the internet, Profound is unequivocally that company right now. Yeah, we work with all of the biggest brands in the world. We've, you know, we've had that first mover advantage and we've not taken anything for granted, meaning that we've built out an engineering team that has shipped hundreds, maybe thousands of features at this point that I think are just going to be— it's very hard to catch up with us in this space.

And James, you said the most important thing will be how the AI— how does AI talk about your brand? But I think there's that old, that old stat of like CMOs have the shortest lifespan of any C-suite. Position, like 3 years or something. And obviously you talked about SEO being the non-glamorous team on the marketing team. And that was probably because other things took over, like influencer marketing and things like that kind of became the shiny toy for a bit. But they've— SEO has remained important and now it's coming almost full circle back around to, to GEO or AI visibility. So I agree with everything you've said, and it does feel like there's just an inevitable wave of importance that your brand needs to rank with, with AI visibility. But what happens if there is another, you know, shiny toy? Will the AI visibility just still be— it's still going to be important, right? It's going to be like the SEO. It just might not be as sexy as it is at the moment.

Yeah, I would just point towards the kind of the, the size of the platform shift. The biggest monopoly in the history of capitalism has been built on people clicking links.

Yeah.

And people are no longer clicking links. I believe they will stop clicking links. I believe you'll just speak to the internet and it will speak back to you. And even more than that, again, we're all creatures of heuristic, myself included. And I think we're not factoring in the, that you will shop within ChatGPT. ChatGPT will use software on your behalf. ChatGPT, ChatGPT's user agents will be the end user, the actual consumer of software. So if you are, let's use, I mean, you mentioned Salesforce, for example, you know, I believe that at some point in the future, Maybe not this year or next year, but OpenAI will essentially navigate your Salesforce instance on your behalf. And then whether ChatGPT decides, you know, when you're standing up your first CRM, whether ChatGPT, which will become your co-pilot, I mean, I guess that's why Microsoft called it that, in this decision-making, whether it decides to suggest Salesforce or HubSpot or Atio as that CRM will depend on its ability to navigate and use that software. So it's beyond just discovery. It's like the actual ability to navigate and use software or the ability to purchase goods due to, you know, the infrastructure. So I think we're underestimating just how different this will be to information retrieval based search.

Yeah. And, and then when Google, they bought, they, for them to build out their ad platform, they built a company called DoubleClick. I think it was about $450 million, which now seems absolutely minuscule considering it's the sort of backbone of the biggest ad platform probably in the world. I don't know. But so, so they brought in third-party ad tech to kind of then become Google Ads. Will OpenAI have a sort of its own ad platform and, or will they acquire, or, you know, how's that all going to work? Are we going to start to see sponsored responses and things like that? You know, one thing I've thought of is why certain words aren't already hyperlinked and there's already a cost per click through, you know, where if British Airways is mentioned, It should be hyperlinked so I can go to ba.com and then it's, there's a cost per click of 30p or whatever and 8 and a half hours.

And you see the end destination is just, hey, do you want me to book the flight for you?

Yeah. So I guess then it'll be defunct in a year when they allow it to all be agentic and shoppable.

So, but I guess my question is, do you really want to go to ba.com and mess around there? Don't you just want to say, don't you want it to just say, hey, there's a flight at 7 PM.

I just want to know that I'm maximizing my Avios points, to be honest. They seem to say, well, it'd be worth this much, and then you actually get to it, it's about a tenth of that. But so, so my, I guess my question is, what, what will OpenAI and some of the big platforms' role be? What will, how will they support or build to innovate around the sort of branding side of what LLM responses provide?

They'll ship ads and the ads will be the most effective form of advertising in the history of humanity. Yeah, absolutely. Yeah. There's, there's a reason why Fiji would, you know, has been brought in there. I mean, it's kind of known, but maybe slightly unknown that Fiji really like Fiji was at Facebook when really they didn't have their ads model figured out. And Fiji was the product manager on Facebook mobile newsfeed ads, which was their, the inflection of monetization, which essentially catalyzed Facebook to become the business that it is today. Yeah.

I remember when it became feed-based and it changed everything and we had interstitial native ads and it was. What we all now know is like the Instagram feed is like literally invented basically by this sort of change of advertising model.

Yeah, well they had the product, the, the consumer adoption was there already, but it, it hadn't really been cracked. And I think it's the, it was this shift to mobile that Facebook executed so well where the, every advertiser on the planet woke up and said, hey, we need to target people on mobile. And that opportunity was far more nascent and Facebook were very well placed to become a bastion of opportunity for advertisers who wanted to target consumers on mobile. And Fiji was at the epicenter of that. Fiji drove that forward. And that's, you know, I think, you know, that, and then you look at the work that she did at Instacart as well on, you know, on the commercialization. And it makes so much sense why she's gone to OpenAI. She's one of the best in the world at ads and Yeah, I think people are underestimating just how powerful advertising will be if it's, once it's executed correctly, we will have, I think, hyper, I mean, personalization is a bland, overused term at this point. I mean, like true, truly personalized advertising that is served up to you in a conversational interface. At exactly the right moment where you are most interested and most likely to want to respond. The amount of intent that ChatGPT, the amount of intent that a platform like ChatGPT captures from those conversations is unreal in comparison to, you know, it's incomparably incomparable amounts of intent in a ChatGPT conversation versus like a forward Google search.

100%, because ChatGPT is like your planning zone. So you go in there to kind of riff and plan and like flesh out maybe just a seed of an idea, whereas Google, you could just be searching Rolls-Royce and looking at pictures of them, uh, that you're not intending to buy one. So there's a deeper level of intent there, and the whole Google business ad model is based on intent. So super, super interesting. Can we pull the pin on the grenade and ask the question, will Google die? Is this the start of the end for Google?

No, I think again, you're showing your nature of heuristic-seeking human beingness there. Like, we're so used to it being monopoly, the monopoly, that I think we're like, well, who's going to win this monopoly? Well, what if it's not a monopoly? It doesn't have to be a monopoly. I think Google—

and I guess they have Gemini, so that's one of the—

I mean, Google, yeah, Google is so well placed right now. They have the whole stack. They're the only player that has the compute with the TPUs and they go all the way up the stack to the consumer application.

Every ad relationship on the planet.

Every ad relationship on the planet. They're threading the needle very well. They've caught up to an extent from a technology standpoint. It's maybe, you know, I'm sure OpenAI are going to release another model. I heard rumors of a new model called internally called Garlic or something today that people were talking about. But yeah, nevertheless, they've, you know, they've reached some level of Google has reached some level of parity on the technology and they own the whole stack and they're very good at ads and I definitely, definitely would. It would be crazy to write Google out of this equation. But yeah, I hope it's, I just, I hope it's not a monopoly. I think monopoly, there's a reason why monopoly, you know, this is me talking as a red-blooded capitalist as well. I do think there's a reason why monopolies are illegal and I think it's because they're net bad for consumers. I think it's net bad for innovation and yeah, I think it would be great to see this new world of AI search or AI, yeah, AI information just to kind of, yeah, be a little bit more fragmented just to see, you know, we have Google, we have Gemini, we have, sorry, Google Gemini, we have ChatGPT, we have Claude, we have Grok. I definitely wouldn't even write Elon and Grok out of the equation here. You have the open source models like DeepSeq, you know, I think people are writing off Meta, but there's a chance that Meta could play a bigger role here. So hopefully we see more fragmentation. Our business doesn't rely on it. It's not existential to Profound, but I think just generally for the space and for consumers, I think fragmentation would be good.

And James, you, you, you're clearly like super ambitious and really smart in this space. And you guys have done a lot of good things in a short, short period. Like, where do you want to take Profound? Like, what, what's the end goal here in the vision?

I think that the internet is changing. Really for the first time in 25 years since its inception. This is the first real, I mean, you could say cloud or mobile or social, but I think information retrieval is changing for the first time since the birth of the internet. And it's going to create an opportunity to build a huge marketing platform. A platform that the modern marketer, the AI-first marketer, will rely on to create all of their marketing for this new age of the internet. And yeah, that's, that's what we're trying to do. We're trying to build a marketing platform for an age of super intelligence. Basically, you'll create marketing, distribute marketing, measure marketing in one place.

Is visual marketing dead? Like, what about— yeah, in, in person marketing department's gonna be trimmed down, it's all going to be focused on agents, you know, it's obviously not going to happen overnight, but I presume there is still a space, right, for visual marketing and relying on all of that.

Yeah, you'll do all of it in Profound is speaking to, you know, what we see as the opportunity. I think visual in AI search right now, I think with intellectual honesty, it's not the, it's not a visual medium right now. But I think that's more just due to the nascency or the infancy of these products. I think as they mature, we're going to see more visual components come into platforms like ChatGPT and Gemini, and it will be generative. They won't— it will generate new UI or video content or imagery in real time as you speak to the model. And I think where— and it will be the same paradigm as what we've seen with just standard, you know, text retrieval now, where it will have to get that information from somewhere. So how you provide these models with that visual ammunition, I think will become again, P0 for marketers and you will need technology and a platform to do that. So yeah, it speaks to the breadth of our opportunity. This is going to be the last business I ever build. I'm not doing another one. This is like the amalgamation of everything. And yeah, I would like to, to build a platform that every marketer, I think marketers will need to become more technology driven. More, more technical, more— they will use AI all day, every day to create marketing. That seems so obvious to me. That seems the most obvious insight ever. And Profound will be that platform where they do that.

And do bigger companies win or smaller companies win? I mean, in the influencer space, we had companies like Gymshark and stuff like that come out, which were almost like influencer first, and they took on— let's just say they took on Nike as a, as the, the new marketing channel, created new brands. Will this create new brands or is it better for the British Airways and the Apples, the incumbents that are kind of established big companies? Obviously those are quite hard categories to go after as a startup, airlines and iPhones. But yeah, big companies or small companies, who wins?

I think it probably supports the long tail more. You know, I think we saw that with the internet itself has been better for SMEs or you know, small independent businesses because anyone can, you know, enter the global stage more easily. And I think agent, if you think of the concept of blue links, right? Or the, or Google SERP rankings, the value has always been in the being in the top 3 or the top 5 blue links, right? But why? Why does it matter that you are in the top 3 or top 5 blue links? Not a rhetorical one. What do you think, Hector?

So, so I would say that the link, the index is based on, you know, how many backlinks, how much visibility you have on the internet. So underneath the ranking is actually something similar to what matters to you, which is, you know, how much of a footprint is there of a brand online and thus what LLMs are trained on.

And so as a company, let me rephrase my question as a company, as a brand, as a marketer, why does it matter that you're in the top 3 of the top.

Consumers are searching for something, they're clicking on the ones first. They can't be bothered to scroll.

Yeah, you just said it. They can't be bothered to scroll. Agents don't care about scrolling. They, they, they—

is there some form of social currency happening there of sort of like, if Google are saying these are the top ones, then I sort of inherently trust them a little bit. And so, you know, they've scored there for a reason. So I don't need to scroll down because I'm just getting further away from the ones that Google say are the best anyway.

Well, maybe, but you could also say it's just the brand, you know, it speaks more to the fat head of the internet. You know, the incumbents, the big players, the ones with tons of budget who could create all that. But like, this is what I'm saying. I'm answering your question that I think we've been constrained by the laziness of human browsing, meaning that I just can't frankly be bothered to go past the third, fourth, fifth link. On the internet. So that is where the value has been, that real estate. User agents are far less limited. You know, maybe you could say like some level, you know, there's some sort of like inference cost of an agent going through 50 links, but you know, if you apply some sort of form of Moore's law over time, that will go down and it will become easy for a user agent to reference, you know, 500 links on the internet to build its, and it will be more likely, I mean, put it this way. And I'm maybe over-laboring this point, but James, let's say, let's say you gave me and Hector the same brief and you said, hey, you, you two have a competition and it's who can find me the best pair of sunglasses based on this brief. I get to use 5 pages of Google. Hector can only use the first 3 links. Who do you think is going to give you a better answer?

Well, it might be the same answer, but I, But if you said to me, by the way, I've gone through 5 pages, I'd go, well, that your answer must be a better answer then.

Yeah. And I can use the first 3 links that Hector looked at as well.

Yeah. It's a better rationale to like that you've covered more ground, haven't you?

I've covered more ground. I've been able to look into the long tail. So I think all of this is to say, to go back to your original question of does this, is this good for small businesses? I think absolutely, because I think user agents are more likely to be able to not just go to the fat head of the internet and explore the long tail, which is where a lot of the beauty of the internet is contained.

And it goes into that deeper intent piece as well. If we're conversationally digging down into what we actually care about, uh, we're going, we're going a few layers beyond just the one Google bar. So even just the interface changes the depth of intent as well because it's multi typed messages rather than a single prompt, uh, which Google is essentially a single prompt. Super interesting. This has been so great. Every conversation is obviously wildly different. James, obviously we've had you on as a founder, but it's been a lesson on platform shifting to AI, to the world of marketing, how that's all going to change. You've obviously touched on at the beginning the success you've had on fundraising, and it's, it's No surprise really when it's this massive, massive opportunity. And the way you broke down how you've actually come out with a, a meaningful product that's driving meaningful results for some of the biggest brands in the world is, is amazing to hear. So thank you so much for sharing all of that. We've got our final two questions. So the first is our future unicorn prediction. So if there was another company that you thought could go all the way, uh, who would they be?

So we've been using a company called Monk.com, M-O-N-K, like a Shaolin monk, Monk.com for all of our accounts receivable. So essentially getting there, they've built AI that gets you paid, that gets your invoices paid. And yeah, it's been super transformative for our business. I'm friends with the founder, the CEO, George Carden. They're just an incredible team. George was like formerly at D. E. Shaw and a professional poker player. And yeah, we, we knew each other from South Park Commons and it's just, they're, they're an incredible team and they're building, yeah, they're using AI to actually, you know, to, to solve a real problem, which is for any business, you know, we work with thousands of brands and getting paid is, is like an annoyance of every business on the planet because you send invoices and it's, you have to follow up and they've built incredible software that helps. You do that. So yeah, monk.com would be my predict next unicorn.

I feel like invoice chasing is one of the best uses of AI because it removes— if an AI agent is chasing you, it's like, there's no sort of social, like, do you mind paying my invoice? So it's a great—

and then I'm oversimplifying as well. They do small accounts receivable. They cover the whole gamut.

Yeah. I wonder how much that domain costs. That's a great domain, monk.com. Dinner party guest game. So if you could have dinner with any 3 people, who would they be?

This one killed me. It's a spoiler to the audience listening. I was given these 2 questions just upfront to think about, and I really struggled with this second one. I think my conclusion would be just because this may be at surface, it might just seem like a really obvious, like uninteresting answer, but I think with the context that it would be just a really interesting conversation to watch unfold. I would love to sit at a dinner table with, I would have Peter Thiel, Sam Altman, Elon Musk, and maybe Demis Hassabis on the, around the dinner table. I think because of the notorious Elon-Sam beef, I think it'd be super interesting. And obviously Peter Thiel has a bunch of Genesis there as well as does Demis. So I think that would just be like a crazy conversation just to sit back and listen to. So that would be my answer.

Yeah. Awesome. Obviously big names. I think they've all been mentioned at some point on the show. I'm actually, I'm just looking it up. Demis has not been mentioned. Can you believe that? That is crazy. In 230+ episodes, no one has ever had Demis on, which is blasphemous. So you've solved that for us, James. Thank you.

There's an awesome doc called, I think it's Thinking Games. Demis is just like one of the most slept on pure, you know, pure geniuses of our time. And I think, yeah, he's not the best at PRing himself and I think he's one of the most fascinating characters of today. Um, yeah, I would recommend highly the, the documentary Thinking— I think The Thinking Game.

Awesome, great pick, because I am so shocked that he has never been mentioned before. Um, well, thank you so much, James. As I said, it's been a really fascinating conversation. I think some people will be hyper-aware of the problem, some people may have just heard about it or heard someone mention that there's this opportunity to kind of rank within AI and everything, but to have it explained and, you know, put the magnifying glass on what this is and why it's such a big shift and why it's important to every brand in the world is, is so fascinating. So I've absolutely loved it. And, uh, thank you so much for coming on and sharing your Rider Unicorn story.

Awesome. Thanks for having me. Appreciate you guys.

Thanks, James.

That's it for this week. Thanks very much for listening. To stay up to date with the latest episodes, please follow or subscribe on your favorite podcast platform. We also have a newsletter called Reading Unicorns, which is another great way to get every episode direct to your inbox. Please tell your friends about it and engage with us on social media, and we'll see you on the next episode.