AI Search Optimization: Trevor Pyle on Visibility, Trust, and Winning AI Search

Speaker A: Foreign. Hello and welcome to the Beyond B2B marketing podcast. I'm your host, Lee Odin, CEO of Top Rank Marketing. And today our guest is someone who's been working in the SaaS marketing space for well over 10 years in marketing leadership roles at a CRM platform, an events platform and a Gen AI analytics platform. But today he's head of marketing at an AI marketing intelligence platform that we've all heard of called Profound. Of course I'm talking about Trevor Pyle. Welcome to the show, Trevor.

Speaker B: Thanks so much for having me, Lee. Yeah, um, I'm glad we can make this happen after meeting a few months ago in Phoenix.

Speaker A: Yeah, that's right. We, we met at uh, Forrester's B2B summit. Um, and you know, that was my first time and I felt like that was one of the better B, uh two B conferences that I've been to over the last 20 some years that I've been, been attending. I was wondering how, how did the show go for you?

Speaker B: Yeah, I thought it. No, I thought it was really good. I think it's interesting where we've seen this shift in kind of like vulnerability in conferences and kind of just speaking in general, where you used to go to a conference, especially like forced or B2B where it's like, here's what you need to do, here's the strategy, the analysts all have it figured out and this one was a little bit more like, here's what we think, what are you all doing? How does this work? And just like a lot more fluid and conversationally based. And I think what was cool is like as compared with last year where it was kind of like, hey, is this AI thing something we should really focus on? Does it matter to marketers now? It's like, yeah, it does. What are you doing? How are you deploying it within your team? So it was tactical, it was fluid. And yeah, I thought there was some really good sessions so I enjoyed it.

Speaker A: Yeah, I love that collaborative point of view, that perspective that allows sort of like co creation of answers. Uh, as we're all on this journey of trying to navigate the biggest change that's happened since the Internet itself. Right. For those of us who are old enough that, yeah, I mean, you know, when I started back in uh, 97 doing SEO, um, before Google and uh, how things have, you know, where we had to explain what the Internet was a little different, you know, why people even need to have ah, have a website at the same time thing. Uh, and now it's like, well, should we even have A website maybe. That's a question we'll talk about today.

Speaker B: Yeah, for sure.

Speaker A: Um, we have a tradition though on this show that and, and it's a question about your marketing origin story. I'm curious, um, how did you get into marketing in the first place?

Speaker B: Yeah, and I'm sure, I'm sure you get a lot of, a lot of variety there as well. Marketers always come from, from different, from different places and backgrounds, which is maybe what I, what I love about it. Um, yeah, I don't know that there's like a fully aligned path. I mean in college I changed my major about six times. I wasn't necessarily sure what I wanted to do at the end of the four years I was in my last semester and I said, hey, what degree am I closest to? And they said, well, biological anthropology. And I said, great, that sounds good. So I've always kind of had a focus in like understanding societies, how people interact with each other and then kind of like, you know, the actual physiology of that when it comes to understanding biology and just kind of going deep on topics. So with that I went into a very non marketing path of academia actually, uh, enrolling to get my doctorate in kinesiology and was, I remember a few months into the program I was actually dissecting a cadaver and just had this like hit by a truck kind of moment of clarity thinking I don't know that I want to do this. And I think when I looked back at it retrospectively, the cadaver was fine. I wasn't grossed out. Like that was okay. That was, I thought that was pretty interesting. But I just realized that I was on this very, very linear and defined path. We had just had some previous discussions of like in academia this is, this is exactly what you do. Even if you go to a clinical side, this is exactly what you wanted to do. And I've always been a generalist and curious person by nature. So I went into something that I thought would just give me the broadest purview into whatever it was that I wanted to do and that was tech sales. So I started doing some tech sales. Started as an account executive, was building a few front end web, uh, apps on the side, took some coding courses. So after a few stints in sales, moved into a growth role and then eventually moved into a, um, product marketing role. Did a little bit of product management as well. So I think for me product marketing was kind of the culmination of how do people and audiences interact? How do you sell something to somebody, how do you understand the psychology and the technology, the details of how something works and it was just quite interesting to me. So that's really where I cut my teeth. I was at a company called Quantum Metric. Went uh, through all their rounds of funding. Started when we were around 20 million in revenue, left when we were well over 100. Um, and that was kind of pre AI. So you know, scaling a company within three to four years of that was actually fast back then, uh, which funny enough but yeah, send product marketing, sales enablement, leadership roles there and then have been at ah, profound for about the last um, 11 months and it's been very fun, very enjoyable. So started by building out the PMM function here and then quickly started building out the entirety of the marketing function. So as the second marketer hired now the team has uh, grown quite significantly.

Speaker A: Wow. Wow. Yeah, it has been a journey, you know. Interesting. Um, just parallel to my own experience. Uh, I started in Internet, um, in the world selling websites. Selling websites over the phone. Uh, I'm almost embarrassed to say like what, you know. Yeah, that's, that's back when those things happen anyway. But um, had to teach myself how to uh, code with uh, CGI and Pearl and cold fusion so I could make websites that had some functionality and sell them uh, at a higher price point and that got into marketing and of those websites and so forth. So it's just kind of interesting that the sales and coding sort of uh, inputs were very formative. I know for me for navigating you know, how, especially in AI applications, um, um, how to make things work.

Speaker B: I think it just gives you context and helps you better connect with the customer and what they're trying to accomplish. So for me it's always, I just like to have that level of detail with everything.

Speaker A: Talk about. So you've had dramatic growth uh, at ah, profound. Um, talk about your, you know, where you're at now. So you're, you're growing the marketing organization. Tell us a little bit about that.

Speaker B: Yeah, so as, as I mentioned I am a PMM and brand marketer by heart. So we actually started maybe in kind of like a reverse order of what you would do there is started investing very heavily in brand very early on. We knew that this space would have a lot of tailwinds. We knew that there would be a lot of competitive funding, um, there would be legacy players adding on technology similar to our initial wedge product of answer engine optimization. So we invested heavily in brand and product marketing, sales enablement, positioning to make sure that we can go and get the right market segment that we want right away and to create a very strong brand around that because we know that that was the preservation that would allow us to layer on demand gen programs later and have it be as effective as possible. So started out hiring a couple BMM comms folks. Um, was working with a guy named Nick Lafferty, which a lot of you all maybe are familiar with, who previously was at Glean and Loom, was leading growth at Profound. And he was also just very technical. So he's building agents within Profound, within COD for us to kind of help us scale the team early. We stayed very small. We were actually four people when we ran our first zero click conference and none of us were event marketers, which was a little bit of a crazy idea. But it turned out pretty well and kind of became quite a series. Um, so now the team is around uh, 12 people. We've layered on people who are running events and experiences which are very big part of our strategy. Um, we think that we've invested in Brand that we have the liberty to go and show up at events and make them very valuable, useful for ourselves and then have scaled on demanding programs, creative functions as well. But we're still trying to stay as lean and kind of AI pilled as possible.

Speaker A: Yeah, well that's great. Well, you're certainly having an impact and that early investment in brand certainly paid off because in a, you know, in a new space, in a new category, you know, it's uh, first, there's some first mover advantage if you can follow it up. Right. And so, yeah, yeah, that's, that's pretty impressive. Um, you. So at B, uh 2B summit you gave a presentation which I attended and there's um, a comment you made in your presentation that kind of stood out to me and that you were saying that LLMs don't just give answers, they give opinions. And I think that's a really important distinction for people to understand in terms of the buyer behavior and, and uh, you know, know, the impact on, in, you know, intent and outcome. So I'm wondering if you could expand on that distinction between answers and opinions a little bit and maybe you can um, maybe. Does, I'm just wondering is, does that explain why, you know, traffic referred from Google converts at a certain rate and traffic that uh, is referred from AI platforms like ChatGPT converts at a much different rate?

Speaker B: Yeah, I would, I would think that it would. I mean the, the kind of, the entirety of the discovery funnel that a customer would typically go through from awareness to consideration and purchase Kind of just can happen within three turns of a conversation if that. Um, I think one thing that we see with AI answers um, which is quite interesting is you'll ask one very specific question, right? Like what is the best AO platform? How even you know from a retail perspective, how much does like the Nike Fi3 cost? And yeah, you'll get that answer typically, but then you'll get like 12 other tweets of what we would define as something called useful augmentation, which is rationale for this answer. And what that is is pretty much the LLM just revealing the process it went through to make the determination to get to the answer and then just provide you with something valuable because it did send out a bunch of different web queries. Queries and probably scanned and looked at a lot of different information. So what that means is that the job of the marketer is now a lot harder because you have to control for what that useful augmentation contains. Could it contain. It's suggesting competitors to that Nike shoe that I mentioned. Is it suggesting hey, if you. It's a good shoe, but actually maybe you could get something cheaper and better value with this Saccony brand or something along those lines. So I think that there's just a much more of an opinion and a verdict that that's delivered in these as opposed to the old where the verdict had to be made informed by the human. Clicking through links, looking at website landing pages, backing out of that, going and clicking into another link, watching videos. There was just a lot less surface area they could cover. And that's why ranking 1 and 2 was most important all the time. Now it's just making sure that you're cited, that you're present, that the verdict is something that you like. So I do think that buyers are just much more educated when they come to the site. I think. Yeah, I mean, you know, the number is in the high 90s a few months ago is in the 60s of people who are using LLMs to inform their buying decision in B2C and B2B contexts. That's only going to continue to grow. I think the answers are only going to get more valuable and more sourced. Um, so I think we're going to see that being some of that most high intent and important traffic that you want to ensure that you're capturing have

Speaker A: a front row seat to a lot of data. Yeah, um, and, and Profound itself has analyzed billions of AI citations and you see over 150 million real user prompts each month. Um, from what I've discovered And that might be the biggest data set of its kind. I'm, I'm just. When you look at, you know, how buyers are engaging with AI platforms based on all this data, what's the biggest thing that you think that, that, um, B2B marketers are, are getting wrong about that behavior? What are they missing?

Speaker B: Yeah, I think there's a few things. I think sometimes you can fall into the trap of like this amount of data and this amount of real user prompts is very hard to make sense of. Um, and this is not necessarily to toot our own horn, but like we've invested in a very significantly staffed data science team. Like our data science lead came from Cambridge Analytica and was at Uber. And when you get this volume, you have a bunch of different, very long queries and questions that are kind of like rambling and very different than a traditional keyword search of best running shoes. Right. So to kind of like normalize the intent under, underneath that, to make sure that it's, you know, of course, like you're actually getting to the nugget and the kernel of the question, um, is, is very hard. I think that you can find some form folks out there will have proxies, but oftentimes that's pulling from SEO heuristics which are just fundamentally different from a data structure perspective, or just asking ChatGPT to estimate the volume for a particular keyword, which is just kind of this like weird recursive thinking. So I do, I do think that it's important to have this, this real user data set. Um, I think the thing that, that people can get wrong sometimes is you want to ensure that you're like meeting the customer at the right stage of the journey. Like the intent is going to change based on the question that they're asking. Right. So if it's something more commercial in nature, you probably need like a long tail specific landing page to answer just that and to ensure that it's cited. So it does lead to a lot more very specific and focused, um, content generation. And that's kind of something that's only possible with a jitsic agents, uh, agents for the most part. So I don't think that anybody's necessarily getting anything wrong. But if you're not basing your entire kind of like measurement structure in practice on what your customers are actually asking, whether that is from a data set like profound has, or it's even pulling from gong, or it's pulling from customer surveys, you can be in a like, dangerous position of building measurement around the Wrong. Actual infrastructure.

Speaker A: Yeah. You know, I mean the answer, just the switch towards answering buyer questions as the focal point is a thing that um, amazingly a lot of companies are just starting to wake up to. You know, I mean, maybe they've had this egocentric view of the world where they're like, okay, this is what the market needs to know about our product or service and that's what we're going to create content around and so on and so forth. But ultimately it's understanding what is it that buyers need to know? What do they want to know? How do they express that? Right. Um, this is at the center of a framework we use called Best Answer Marketing, um, where we have inputs like what you just talked about, whether it's gong in sales conversations or whether it's uh, customer interviews, CRM data, even some SEO data. Maybe people also ask and query fan simulators, stuff like that, just to get an understanding of the kind of language that they're actually using. And then you can reconcile that obviously with, you know, whatever your, your go to market. And I mean your, your content marketing, um, generation and content marketing planning is going to support. Um, so there's this sort of, um, I don't know, it's, it's you know, kind of like brand demand. You know, is it, uh, do we go after AI search or do we go after Google search? I mean, um, you know, Forrester says 94%. I think you, you alluded to this business buyers are using AI in their buying process and at the same time Google claim their search usage is continuing to rise. And so as B2B marketers confront, you know, these signals about, in their process of prioritization, do is this something they have to choose between, um, you know, optimizing for traditional search, which says it's growing more than ever, or optimizing for AI discovery?

Speaker B: It's such an interesting position that Google's in, right, because they have this like very much cash cow of a, of a business that they're also trying to make a good experience for their customers too. Like their customers are looking for AI overviews, they're looking for simplicity. They don't necessarily want to click and fumble through three or four or five, six links. I mean, maybe sometimes they do if they're in their second intent and they're getting specific because they've kind of narrowed their, their competitor set. But so I think that they're always going to have some kind of like, interesting outward positioning and where they're saying search traffic is growing. But I think there's a couple of like maybe caveats to that. They are seeing large amounts of traffic that will continue to grow. But a majority of those customers are now seeing AI overviews at the top. And I think probably what they're seeing is the depth of people then clicking into sponsored links. Clicking into the top ranking pages at the bottom is probably reducing even people clicking into the citations forming in the AI overview. Now that's still going to matter in different parts of the funnel. I think the unfortunate thing about this distribution shift in AI search is that marketers just have to kind of do everything all at once. Now in a much broader scale, if you think about like the human capacity to search through and build a verdict a bunch of around a bunch of different information, it's much smaller. One query and one web agent now can go and scan thousands and thousands of sources on your site, off your site. 3rd PD Party G2 reviews, podcasts like you now have to be everywhere and make sure that your brand and sentiment is maintained there positively. So you kind of just have to do both, um, across all channels. And I think a profound like we have one, two core beliefs. One is that answer engines will become the most important distribution shift of um, probably our lifetime. And then two is that marketers will use agents to do their work better, faster and at higher quality. And that's kind of the only way that you can keep up with the pace and the expectations here. So really you have to do two because Google AI overviews are increasingly used and increasingly important. And then people depending on the segment will still rely on quad perplexity, OpenAI, et cetera.

Speaker A: Yeah. You know, I wonder about uh, looking back me when social media was becoming more popular and search and social media had this sort of people would find in one channel and then validate in the other. And I wonder if that behavior exists between going to Google and then going to like ChatGPT, you know, or do I discover something on ChatGPT and I'm like, yeah, this is good. I think I'm going to go gut check that over on Google. Yeah, you know.

Speaker B: Absolutely. Because that's the human part that we still have. And you do want to make your own verdict and opinion at the end of the day. But maybe you're asking about a category of technology ABM platform, something that you need to go and purchase. Help me get to the consideration set. Now I'm gonna go and go through the organic listings and find these specific landing pages that I'm curious about and actually use my fleshy eyeballs to Go and view that and make the verdict there. Right. So it's definitely a combination and there's a big dark funnel that's hard for marketers to account for there.

Speaker A: Yeah, for sure. And I think multi channel visibility is part of know, architecting that visibility. Answering those questions in the different channels where your audience happens to be looking. I think it adds up to a sort of consensus. Right. And when you have continuity of how you're cited across these different channels, then that makes it, that's kind of a, uh, isn't that a signal that an LLM is going to pick up on as it makes a decision about what to represent as an answer that consistently across all the sources that it's crawling or, or maybe it's augmenting um, with a real time search and it finds these are the entities that consistently appear in a particular context. Um, and then those entities are what gets delivered as answers. Seems to be the case.

Speaker B: Correct.

Speaker A: So um, this question is kind of a big one. A lot of folks are talking about tools, platforms like Profound and others that represent you know, tracking. Like, you know, there's this idea that like in from a, uh, search engine days, right, you have a ranking report or something like that where keyword data is pulled in from actual user queries and that sort of thing where it's obviously it's different with an AI tracking platform. So you know, these are in some cases the prompts that people are tracking are educated guesses, right? Or they're based on keyword research or you know, something that they decide that they're going to track so they put that in, um, versus what people are actually searching on, um, is often the case. And I'm just wondering how, how do you think or if that is the case, um, how can B2B marketers, you know, use AI tracking tools most effectively to connect the dots between what they're putting in market and what impact they're having. In other words, you know, how will they know that the things that they're doing are actually having an impact, um, as it relates to real buyer prompts and searches?

Speaker B: I think in reality like, you know, so much in our world is just nuanced and it's about being reactive here. And I think this one is nuanced as well. I think it's, I think it's a cocktail. I think one you do have to have the foundation of what we talked about of like real user data, real user prompts. Try to understand from a volume perspective what is your audience in the segments that you care about actually asking and what's the intent behind those? Is it commercial in nature, is it generative in nature? Exploratory, educational? That's very important for you to understand where those prompts kind of sit within the funnel. So that's important. And there's kind of very few tools out there that have access to that type of data set. Of course, us being one of them. I um, think another one is pulling in keyword data as well. That is still a very good proxy. And why it may not be like a long formed question. It's kind of more of like a topic level thing that you can maybe like wrap some prompts underneath. Right. So like some keyword data is good as a proxy and then just getting customer input. Gong calls, surveys, interviews, what your team thinks, always talking to customers and understand what are the questions that you're asking. When it comes to our category and the problems that we solve, we partner with somebody called Listen Labs. They do really great work and you could do surveys to your customers here. So I think the like kind of more old school traditional marketing of just getting in front of your customers and asking questions now that it's like less abstracted into a, ah, three word keyword and more in like just a conversational question. They may just be using whisper flow to be asking questions. Right. And trying to discover new products or services. Um, so I think it's just kind of a mix and blend of all those things.

Speaker A: Yeah. And I think at a minimum it's the relative measure. Right. It's okay. We've defined what it is that we're going to track and then we're going to be able to see relatively from a uh, certain cadence, monthly or whatever it is how we're moving the needle on those defined uh, prompts or queries. And you can always add, you can always subtract, um, in the absence of actual buyer, you know, unless anthropic or OpenAI starts actually providing, you know, some sort of data, uh, directly from what it is that people are typing in. Um, you have to enter, you have to uh, forecast it. Uh, you have to predict it, so to speak.

Speaker B: Right, yeah, yeah. And I think one other thing you have to do is you have to close the loop on the work that you're doing. Right. So if you're tracking prompts and you see that a certain prompt or maybe a set of prompts within a certain topic is lower invisibility. So you put a content plan to put in some very well structured answer forward like information, rich non commodity content Pages together. You then want to see if those are increasing agentic traffic. So ensuring that you have a way to see if bots are indexing, crawling, training and citing the content that you're putting out there to answer specific prompts. And then if those people are then converting down funnel, so having some semblance of like an agentic analytics or agent analytics is what we call it, to then see if people are actually citing coming to your site and what content is getting cited by both humans and um, bots and agents.

Speaker A: Bingo. Exactly. You know, AI search is, is obviously top priority. I mean it, it, I, I mentioned before we started recording, you know, every, every company that I'm talking to right now is mentioning AI as one of the, um, outcomes, even if the program isn't search focused at all. You know, you've argued that whether people, you know, companies are asking, you know, are we showing up? Right? And, and, and you've argued that that's just the first question they should be asking. Um, so, so I'm wondering what, what else should they be thinking about when it comes to their, their visibility? What is it that they should be tracking?

Speaker B: Yeah, I mean we work with a lot of companies and since we have such an interesting cross section of folks from the biggest Fortune 10 companies in the world to the very popular notable scale ups like Ramp, who have effectively a data science team that sits on their marketing department, they can always be kind of testing and optimizing and pushing each other forward. So I think like last year it was just kind of everybody was just approaching this, this very big problem of hey, are we showing up? Are we visible? CEOs were asking ChatGPT about their brand and then texting screenshots to CMOs left and right. Right. Like always. Fun even, even I still get it. So if that makes anybody feel better, um, it's a constant battle, right? Like models are changing. You're not just ranking for one algorithm, you're trying to sit across a bunch of different models depending on, on what, uh, tools that your customers use. So visibility is important, but the more important thing is that that useful augmentation or that opinion that's injected into the rest of the answer and the verdict, that's really important. And that comes from two places. It comes from one being accurate and then two, from that accuracy. What type of sentiment and judgment does the answer form? So I think if you're starting somewhere, it's getting prompt tracking in place and then ensuring that things are just accurate like the answers that are coming back that you're seeing are accurately reflecting the product and service that you have. Like a good example is Southwest. If I were to ask, we ran this prompt a few weeks ago, what's the best airline for group travel? It tells you Southwest is very good for group travel. But then it says, also, I'm just looking at the response. If you want, you'll want to get there early. They do want assigned seating. Actually changed that months ago. But there's still all of this content floating around. So it's very important for the Southwest team to understand. Okay, where is that being cited from, for example? Right. And this is not just a Southwest problem. This is somebody, something that everybody has, especially larger enterprises with, with a bit of content sprawl. So it's very understanding, important to understand. Where are the facts actually coming from? How do we ensure accuracy? And then from there you can start to layer sentiment on top of that to ensure that the verdict is positive about yourself and your brand.

Speaker A: Right. That opinion. Right. That opinion is, is going to be. It's a powerful thing to the context of opinion being delivered as information. Um, yeah, so, yeah, that makes sense. So people in the industry don't seem to be able to agree on what this is called. Right. And I know these acronyms aren't necessarily the same thing because AEO versus geo, AI Search Optimization. I find myself saying AI Search more than anything. But I'm just kind of curious. Um, you said that the biggest misconception, while, you know, this is sort of a context for my question, the biggest misconception is, um, that this is just SEO with a new coat of paint. And, and, and also that the winners treat it as an extension of SEO, not, not as a replacement. So what is genuinely new about AI search and AI search marketing or optimization? And, and you know, what's, what's carrying over from SEO? And also, what do you call it?

Speaker B: I think everybody knows the thing that we're referring to, so I'm generally okay with any acronym. We call it Answer Engine Optimization just because we believe that the consumer is using these things to get an answer. For the most part, that is what marketers care about. Are we? The answer is our brand the answer, as opposed to geo, which is more generative in nature. So yes, we call it Answer Engine Optimization. And yeah, I couldn't agree more. I think that prototypically too, the SEO is the best steward for this shift in kind of the way that marketing teams are organized and orchestrated. Right. So they're technical in nature, but they understand content, they deeply understand the buyer, they're like looking into research around what customers or what questions customers are asking. So like from a skillset perspective they are the best and now they are the most adjacent to this as well. Because the things that still matter are technical crawlability, high quality content that's easy for a human to read but easy for a machine to parse out. And it's actually like original and it's not commodity and it has specific facts and details and isn't fluffy. Like there's brand mentions, all of these things and then there's like page structure of the URL and making sure that it's crawlable. All those things are still very important both for like the traditional Google crawler as well as the new answer engine crawlers. I think what's a little bit different and why many SEO teams are now tacking on AEO or AI search to their team holistically and getting more resources, which is really cool to see, um, is that a lot of this now matters off site too. Like your own domain is very important, but that's not always the answer you're trying to drive people to. Sometimes it's a third party mention or a media article or it's G2 reviews or it's your partner providing influencers to you. Yeah, influencers are huge. LinkedIn social is massive now, right with Reddit, uh, for example. So like yeah, the job got harder. But that also means that you know, this off site work that you're doing, working a lot more closely with PR and the comms and the content team, uh, is really quite important. So I think that that matters.

Speaker A: Yeah, absolutely, yeah. You know, SEOs, I uh, mean I came into marketing through search, um, and have had my toe uh, in the water of search for the 25 years we've had our agency. But know SEOs as a group are you know, some of the most interesting people. And from having both left and right brain capability, you know, the lateral thinking, um, that that is necessary in order to chase this black box of, you know, there's no cause and effect, uh, available to you, uh, to understand really how search works. So maybe that's narrowing a little bit, uh, with some of the lawsuits Google's been involved with and the disclosure of how their machine actually works. But um, there is a creative side to search, I think that plays an important role. And um, you know your, your own research at profound, um, has pointed to you know, original non commodity content, proprietary data, first party insight, you know, like, like real lived experiences and perspectives as what outperforms in AI answers So, you know, what is that kind of content? You know, why does, why does that kind of content win in AI search so much? Um, and do you think this makes a case for the idea of doing original research, like, or uh, creating novel content? And original research is just one type of that?

Speaker B: Yeah, yeah, no, it's a great, it's a great question. And um, I mean the kind of overall arching answer is yes. And I'll kind of get into specifics of that. Like even dovetailing off of the last question, um, like it used to be like SEO, you could get a page to rank and it's like, nice, there it is. And it's, it took some time and it took a lot of domain authority and backlinks and all of this to make it rank. And then it just kind of stays there and that, that page will just print for you in that traditional modality and for that channel that still exists and is very strong, that will still work. But LLMs are trying to be authoritative and credible and believable at the end of the day. And they are smart. They know that there's a few things that they need to do that they need to like, bring in kind of like very detailed rationale for the answer and why they're making the determination of the answer. They also want to make sure that everything they're providing is very fresh. Like over 50% of content that is cited is, you know, less than 13 weeks old. So they want to see, okay, this is the best of list for this quarter, not last quarter. It's not outdated. So it's very important for them to uh, appear or pull in fresh content because it just gives them more authority. And yeah, when it comes to like using your first party research and being as objective as possible about everything that your product does, the services that you offer, it's very important because then that allows an LLM to itself become believable by using those points to justify its claim and justify its rationale. So it's just much less of a kind of binary surfacing engine and it's very much looking for things that will make it believable effectively and provide a good answer. So that's why we've really focused on first party research data, trying um, to open that up as much as possible and then just being as explicit with uh, what our product does and just being very like detailed with any type of content that we're providing. Because I think that humans want that at the end of the day. Um, but I also think it serves LLMs really well, well, trust is such an important thing.

Speaker A: Uh, it's harder to earn than ever and it's more important than ever. And obviously, you know, I think two of the characteristics of winning content in AI search and even in any kind of discovery is um, the notion of proof and consensus. That, you know, there's continuity or consistency, as I was talking about before, of how you're represented and then there's also proof to back it up. And that's not just for the LLM, but it's for the people too. It's like, you know, when you think about it in terms of search, you know, what good is ranking? What good is having visibility if people don't believe what they find? Right. So, um, I think that sort of seems to be baked into the way, um, an AI platform is going to surface answers. Um, and so that's incumbent on the marketer to help give the platform what it needs to represent you the way you want to be discovered.

Speaker B: Uh, yeah, and that, just to like expand on that, I mean like, that's a play that we ran very early on. Like we invested heavily in internal research because we saw that the space was exploding. People were asking lots of long tail queries around what should I be doing about Reddit? What type of content should I create? Does it need to be fresh? Does it need to be uh, this, this or that, Any, any kind of like delineation of questions that you would see around the tactics of aeo. So we based a lot of our first party research using the prompt volume data set that we have to just put out our perspective and what we're seeing in the data on that. And we saw that that was some of the most heavily cited pages on our website for the prompts that we were tracking. So yeah, I think it's quite important.

Speaker A: So when you, I think you mentioned that, you know, content found in AI search is recent content, like I think your own research showed it's like ah, 13 weeks old, um, or less than 13 weeks old. So what does that mean from a content production, you know, standpoint? Like is that, does that mean making more content or is it being, organizing things around refreshing content or both?

Speaker B: I do think that it is making more specific content. Like we are in a world of marketing right now where um, like attention is kind of the moat that you have and brand is, is the moat that you have as a marketer. And um, I'll credit that to my, my friend George at Ramp. Um, um, I think they do a very good job of this where you're just kind of capturing the zeitgeist continually. So I do think there's an aspect of creating a lot of content and that that doesn't always need to be like something that's LLM crawlable, that's more like video and media houses and how B2B marketing teams are kind of becoming media shops. But maybe that's a different, different podcast. But I think that it's, it's really hard to do. Like you have to use agents to be able to do this effectively. Um, so what we will do is we will build optimization agents that are consistently on scheduled runs, utilizing Profound and going out and crawling competitors, crawling the latest news and seeing if there's anything that exists in our corpus, our knowledge base, our site map, which we've mapped into Profound. It can and should be updated. Um, it's connected into our Slack instance and connected into what we're doing on the product side so that we can see, hey, we updated this part of fact check or of agent analytics, go and update these three pages or we did a case study, go and add it to the social proof of this particular page. So like having kind of a content optimization partner or agent that exists within your team is very important. And then using something like Profound, you can actually have it write and suggest the inline changes, publish that to staging and your cms and then you can kind of just review and publish that. So I think consistent updating and optimization of content is more important than net new generation. But that is still quite hard to do.

Speaker A: That's great advice. I think, I think a lot of folks are realizing this. I know a lot of large enterprise companies are starting to open the, you know, open themselves to shift, uh, in how they approach large scale content optimization that way. And that kind of effort requires a certain skill set or person and technology and specific to a role. You at uh, Forrester B2B summit, you talked about a role or the rise of a particular role called marketing, uh, engineer. That's someone who understands marketing obviously, but can also build workflows and agents, um, and perform automation tasks. Is this something that you see on the rise for most B2B companies or just specialist companies or certain use cases? And also how do you see marketing organizations changing, you know, because of this type of role emerging?

Speaker B: It's, it's something quite interesting. The, the way that it actually originated, um, was when we were, a few months ago we were selling into our customers and they're like this is great. We're excited about Profound. Okay, so we know that the PR and comms team needs to use this and we know that content and SEO and ao. So even outside of budget, who's the, who's the owner of a platform like Profound. And as we've released just past our, we have, you know we kind of started in this analytics world and now we're building agents for the marketing team holistically, yes to capitalize on aeo, but also there's agents that can help you with paid search, with landing page generation, with PR and comms outreach. Like we believe that we're going to be the workbench for the modern marketer, just kind of full stop. And that means that you need somebody who understands at a pretty deep level all the workflows that, that that marketing team or that department or individual is going through and can build agents and automations around those. Now you could say let's just have all of those people get like very technical and build them on their own. But I actually think it's a force multiplier to just have somebody in house who's very, very good at this and has context and could build agents for them and have that person be the stakeholder and have them refine it and then by them using the agent they're actually getting more fluent with AI. So yeah, we actually with our customers kind of co coined uh this role and it's just really taken off. We've seen dozens of companies start to hire for it. I think just a couple days ago, um, Figma's hiring one, Stripe's hiring one Cloudflare Okta. So I think it's a very important role for B2B uh especially scale ups. I think we'll start to see B2C grow it uh, more as well. I've got one on my team. So the guy I mentioned, Nick Lafferty is our founding marketing engineer and he actually built that content optimization agent which within profound that I was talking about. So he's kind of like squirreling around looking for inefficiencies and where people are being slowed for arbitrary reasons and just building automations around that. And then we just have the cool effect of him also like looping right back to our product team as well. So our marketing team instead of sitting next to sales we sit right in with the product org um so we stay very very close to them not structurally but like physically in our office.

Speaker A: I'm definitely um, agree with you. That's absolutely a role we're going to start seeing um, on the agency side for sure and certainly on the client side as well, looking at the impact of AI and marketing, like, you know, the marketing organizations are changing, you know, it's becoming. And even CMOs are expected to have more tech savvy definitely than what they've had in the past. And so if, if AI keeps, you know, compressing the buyer journey. Right. And delivering pretty much everybody, everything a buyer needs within one chat.

Speaker B: Right.

Speaker A: Um, especially if you think about how Google search interface is supposed to be working here soon. Um, what, what does that mean for traditional demand gen? Um, what does that mean for brand marketing in the future for B2B companies?

Speaker B: I, I know I come from the brand side and I have met very many, very many friends in uh, the demand gen leadership side. I do think it's going to be a little bit harder for demand gen for a little while. I think the direct response kind of like dollar in, dollar out stuff that we're used to seeing is just going to be a little bit harder. But and I think that there is going to be a rise of brand marketing because I think again as I had mentioned, like attention is kind of the moat that you can create as a marketing team and then also like these things that uh, I've always been kind of. Yeah, I guess we should have somebody responding in Reddit or somebody like reaching out to this, you know, particular publication or making sure that we're looking at our off site community engagement. Like that's very, very important now. Um, so I do think that it'll be a little bit of a power balance in terms of like brand and demand there. But I still do believe that demand will have to think about how to capture the customer at the end of that compressed funnel because they're not going to at least now click buy for you know, an enterprise level contract within ChatGPT. But that's when the consideration set starts. So you have to be hyper targeted on like surrounding that account as soon as you see referrals coming from ChatGPT or whatever because you know they're a lot more informed and then maybe even um, like continuing the modality, like it probably doesn't work to cool. You came from a very, very informative chat. Why don't you go and process this very long white paper. Probably having a chat experience on your website itself that's like very valuable buyer forward. We're going to have to get a lot more like forward with our products and not hide them behind as many lead forms. Like I think people are expecting information at their fingertips so we'll have to just, just kind of change that, so there will be, there will be a little bit of a shift. I think step one, if you're coming into the role, regardless of which team or you know, flavor you're sitting with is like, just understand what AI is saying about your brand. Like that's the most important influencer in the world right now. So you have to understand that to

Speaker A: start speaking about, you know, impact on demand gen and changes people are making. Um, we've had companies ask us, you know, do you think we should stop gating for the purpose of providing more information that can be crawled by LLMs? It's like, well you have to make that call based on the purpose of your, of your content. But I know in our own research reports we publish um, research reports around influencer marketing and thought leadership and we started ungating those and that's been great. But also the next version, um, we're just finishing capturing Data on our 2026 Influencer Marketing Report for B2B and the next iteration is going to have an AI interface meaning that you can interrogate the data. We want to have an experience where people on their own terms can go ahead and ask questions and get answers and you know what I mean? Because uh, that just doesn't exist so much. IBM is a great example. IBM um, has a group that's focused on thought leadership and they created an interface like this. It's pretty, pretty useful and inspirational to me. We can do that too. I can use Google AI Studio or something to create uh, or Claude or whatever to create the interface. Um, so what's some advice for someone who's listening or watching that is like, okay, there's slot information, um, and they're, they're, you know, it's coming at em from every direction. Expectations, fewer resources. All right. If you, knowing what you know now, you know, if you joined a mid market B2B company uh, tomorrow and, and you had a mandate to improve AI search visibility, where would you start?

Speaker B: Yeah, no, it's a, it's a great question. I hope before you join that company you asked a lot of questions to the CEO to understand that they know that marketing is shifting and changing right now and that there's going to be a, have to be a lot of trust, um, and that to do good marketing you're going to have to take swings with budget and that the funnel is going to get a little less like here's this program, what it generates, this program, what it generates. Um, so I hope you kind of start there because even in investments in AI search like There's a lot of returns from it, but there's still a lot that everybody's figuring out as well. Where I would start is just like, let's assume baseline, get visibility, go talk to your customers, ask questions, look at prompt volumes, understand what are your customers asking at each stage in the buying funnel, and start to deploy prompts and interrogate answer engines on those exact prompts within the models that you care about, the regions that you care about, and start getting that data back. The second thing I do, ideally on that same day or within that same hour, is deploy something like agent analytics on your website so that you can then understand, okay, what bots are actually coming to my website, what are they citing, which content is getting picked up and which content is not getting picked up so that you can start to kind of prioritize your content strategy from an agent perspective. And then you start to look at citation sources and maybe pick off those top three of like, these are the surface areas we need to really go and dominate and invest in. If you're vertical specific, maybe there's like, you know, a Financial Times or a Forbes or a publication or something like that that's very like, specific to your particular vertical. Or maybe there's a lot of discussion about what you do or your space on Reddit or a, uh, different community. So you need to see what your major citation drivers are, good and bad sentiment, so that you can go tackle that and make sure that you're getting ahead of it. So I'd probably just do those three things right away.

Speaker A: Great advice, great advice. Pretty much our last question here as we wind down. Well, second to last question, it's a, uh, community question, um, that I didn't, that you don't know about, but it's pretty straightforward. Um, this is from Sam Berbrick, uh, who's an account manager at Top Ranked Marketing, and her question is for you, Trevor. What do you think will give companies the most competitive advantage the next five or even just two years given how much the search landscape is changing?

Speaker B: Yeah, it's. I, uh, don't want this to sound like everybody else, but it is whoever builds the most effective and proficient agents that their team can use. But I think what's important here is, is that these agents need to be ultra context loaded on your brand, how your team works, on your product, on your customers. So whoever can build the most contextual, relevant and effective agents that their team can use is going to be the most effective. I don't necessarily think that marketing teams will get smaller. I Think that they will just start to punch way above their weight and be very important when it comes to kind of just everything, modern distribution. So like build agents well, build them with high context, make sure that your team can use them and is proud of the, the output and the quality and that kind of like matches the craft that you want to represent as your brand. Um, and you'll be, you'll be in a good place.

Speaker A: Yeah, that's a great answer. Um, and I think, you know, research in its various forms is going to be a great input for that context. Right. Whether it's that, that CRM gong data, whether it's question research, whether it's original research from the marketplace.

Speaker B: Right, Very much so.

Speaker A: Ah, yeah. So, um, our last question is about a non marketing thing, meaning that imagine marketing just wasn't in the cards for you at all and you could be doing anything else. What would it be?

Speaker B: It's a good question. Um, I spent, I live in New York now and I love it. My tiny daughter sleeps in a closet. But that's okay. I think that I've learned that that's fairly standard. So I'm good with it. And it's still an expensive closet. Right. It's a big nice walk in closet. So I do love living in New York, but I had spent a lot of my time previously in Colorado. Very, um, much enjoy being outside. I would probably be either a ski, like a backcountry ski guide or a mountain bike guide in the summer. Um, something between that and owning a wood fire barbecue restaurant. Because I am from Austin, so it depends on the day.

Speaker A: Oh, you are? Okay, yeah, yeah, yeah.

Speaker B: So one of those two.

Speaker A: Wow. Ah, that's awesome. I got a chance to live in Austin for about a year down. I lived in the Circle C Ranch area.

Speaker B: Yeah. Yeah.

Speaker A: Um, went to the Salt Lick all the time.

Speaker B: Drifted big open pits there. Maybe not scale, but more of like a Terry Blacks level scale.

Speaker A: Uh, sure, sure.

Speaker B: Yeah. Some good offsets and post. Yeah.

Speaker A: Well, I hope we, uh, run into each other in Austin sometime or somewhere else. There's great barbecue available. Um, that'd be fantastic. Well, super. Uh, where's the best place for people to connect with you?

Speaker B: Definitely LinkedIn. Yeah, I mean that's a big distribution channel for us. It's kind of just where a lot of marketers are existing right now. Um, and we do try to publish a lot of research there. A lot of different perspectives, host good conversations. So I'm there, I'm in New York as I mentioned. So if you're ever around Union Square, shoot me a note on LinkedIn. Happy to grab a coffee. Something along those lines. Um, but yeah, that's uh, that's the place.

Speaker A: Fantastic. Yeah, I'm waiting for someone to say, to not say LinkedIn when I ask that question. I know, I know it's kind of a, kind of a gimme. But um, I appreciate that. Well, thanks so much Trevor. This has been fantastic. I'm sure everyone listening, watching has got a lot of value out of the advice that you've shared and I really appreciate you being on the show.

Speaker B: Yeah, yeah, anytime. Thanks Lee. Thanks so much for having me.

Speaker A: I want to thank you for tuning into the Beyond B2B marketing podcast. Uh, make sure you subscribe so you can stay tuned to our next guest. And remember, there's no better time than now to become a best answer brand.