The GEO Operating System

Emma Buitendag:

Hey, and welcome everyone. Great to see there are already so many of you in the chat saying hello. Awesome to see all of you streaming in. For those who don't know me, I'm Emma from CMO Alliance. I am the community manager here. So it's wonderful to have you all joining for our live session today. We have got an hour dedicated to the GEO operating system, and there is no one better to dive into this topic than Michael King himself.

So he is joining us as our guest today. He's the founder and CEO of iPullRank, and he's going to dive into this topic a little bit more. We would love it if you could say hello to us in the chat. Tell us where you're joining from. If you want to go a little bit deeper, let us know what you're most excited about for today's session, and we would love to hear from you. You will also see that next to the chat is a Q&A tab.

So please feel free to pop your questions in this Q&A tab throughout the session. We will save some time to get to all your questions towards the end. There's no such thing as a bad question or a silly question. So anything that you think of, pop it in there throughout the session. And then we've got some interactive bits and bobs for you as well.

So really looking forward to spending the next hour with you. And with that, I'll hand over to you, Mike.

Michael King:

Hello, hello, everyone. Thank you so much for joining today. There's so much for us to talk about, and I definitely want to leave that space for your questions and so on. And Emma, thanks for having me. I really appreciate it. So as mentioned, I'm from iPullRank. We are a pioneering AI search agency. We've also emerged from doing SEO for about 12 years and really at the cutting edge of that this whole time.

And so just some real quick stats about us. You know, we've driven over $5 billion in incremental revenue for our clients. We've got over 40-plus people on our team. We've also got over 100 experts in our network for subject matter expertise. This year, we were also named to the Inc. 5000 fastest-growing companies in the U.S. And, you know, we've been around since 2014. That section in the middle is a really, really short list of our clients. We've worked with a lot of the Fortune 1000. And, you know, as far as my own individual stats, USA Today named me a top 10 most influential SEO. Search Engine Land named me Search Marketer of the Year twice in 2020 and 2025. And Forbes also named me a top 20 growth hacking strategist.

But it's not just me. The team that I work with is also world-class as well. And they've all done their own level of thought leadership, whether it's within our clients or within the world as well. And so we were building for AI search before it had a name, you know, before people were calling it answer engine optimization or generative engine optimization, which that title I had a hand in like popularizing. I wrote a blog post for a website called Search Engine Land, wherein I dug into the computer science behind how AI search works. And I made a lot of predictions around things like losses in traffic that people experience, the redistribution of search volume, and all these things that we're actually seeing now. We predicted and built a framework for how to optimize for going back to 2023.

And so a lot of the folks that are popping up now are basically taking the work that we put out there and using that to build their offerings at this point. And so we're also the agency that AI Search actually recommends first. This has been third party validated. This is not just me showing you screenshots of what I see when I ask ChatGPT, third parties have done a bunch of prompts and they found that we consistently outperform everyone else in the space. So, you know, the proof is in the prompting, if you will.

Now, we're going to talk about the GEO operating system today. And basically what we would recommend that you do in your brands and your companies to drive visibility in these platforms and also the third part or the second order effects that you can expect from that visibility as well. But the approach that we take is one that we call relevance engineering. And it's really the evolution of what SEO was going into the past.

And so we think about the different eras of SEO in the late 90s, early 2000s is really about like, how do we rank pages for keywords? And what I mean by that, it was like, okay, how do we shove the keyword on as many times as possible so that we can drive as much ranking as possible? And it was really simple at that point because, you know, if a competitor had a page that mentioned the keyword 30 times, you did it 40 times. You just ranked better. Right.

And then things got more sophisticated as we got into the 2010s. And, you know, Google really hit its stride and developed its technology to be smarter than that. And so then it became a lot more about content links and intent. But now we're in a space where, you know, these results are probabilistic. It's pulling from a variety of different places. And it really requires that you have these cross-functional teams that understand artificial intelligence, content strategy, brand PR, user experience and information retrieval. And you'll note that I didn't say SEO at all when I described that because you can be a good SEO and not really know how anything works.

But we're in a space where all these platforms are building right in front of us and there's open-source equivalence to everything. And if you understand how the technology works, that gives you an advantage on what you can do in these environments. And so that's why we've evolved what we're doing, because the SEO community and space and software and so on hasn't kept up with what all these platforms are doing.

So going into 2027, you're going to need a different model to capture this visibility, capture this traffic and performance and so on. And, you know, a big reason why is because AI search is looking across a variety of channels to compose its responses. Now, if you're in an executive position right now, you may already have somebody doing an RFP, going out to the market to figure out, like, who do we use? What tooling do we use?

And so on. But the problem is this is not a mature market. So an RFP is a waste of time. You know, we have been a part of a lot of these RFPs. And what's happened is a lot of these brands are just pausing them after going through all these efforts because what they're getting back is just dramatically different across the board.

There's no consistency in this space right now. You know, what are we even calling it? Right. Some folks are calling it AEO. Some of them are calling it GEO. Some are just saying it's just SEO. And so what I want to talk about today is what you should be doing instead in order to get the right partner or figure out the right solutions for your organization.

But before we get into that, I want to kind of speak to you directly and see where you're standing, because the way that we look at this is we want to meet people where they are. Every brand, every business is in a different space as it relates to their maturity around AI search. And so I want to kick it to a poll. We're going to ask you four different questions so we can get a sense of where this audience is.

So first, let's go with who owns GEO as a channel internally. And let's just get to that poll real quick. Excuse me. And we can get a sense of where people are currently in their organization. It's even asking me. Cool. Looks like we got a good amount of votes in. Most folks are saying that it's marketing owned. After that, we're seeing a lot of people say that it's SEO owned. Really interesting.

Okay, and let's move on to the next one. How complex is your GEO program likely to be? And the poll is in the tab up top to the right where it says poll next to the chat in case you're not seeing it. So this one is how complex is your GEO program likely to be? And most folks are saying either one site with multiple content types or product lines or multiple sites, brands, markets or languages.

And then some folks saying one site, one team, limited content scope. All right, let's move on to the next question. Which best describes your current internal GEO motion? Again, if you're just joining us, the poll is in the upper right, next to the chat. And the question is, which best describes your current internal GEO motion? Okay, cool. So most people are saying we are learning what AI search means for us. Secondarily, we're currently running our GEO program.

All right, and let's move to the last question so we can then jump into this content here. What are your main concerns with GEO? Really interesting mix of people as far as where they're at and you know that aligns with what we're seeing with folks that come to us. Again, everyone is starting from a different space and so we're able to help them based on where they started from.

Okay, so what we're seeing here is the main concern is the competitors are appearing and we are not. After that, we do not know how to measure AI search visibility. Cool. All right. So, yeah, let's jump into this and, you know, help you guys out because we've been doing this work. And that's one of the things that's really exciting for us is helping folks get the results.

So first thing to know is that the clicks are down across the board for these channels. Right. We're seeing that 50% of U.S. consumers are intentionally using AI-powered search at this point. Sixty eight percent of U.S. Google searches are ending without a click. And then 58% lower average click-through rate for the top result on queries with an AI overview. So this is telling us that classic search is melting down to some degree with respect to its ability to drive referral traffic. And on the back of that, what we're seeing is that, you know, the traffic and conversions are diverging. And what I mean by that was historically, if you got more traffic, you got more conversions.

But now what we're seeing as a function of AI overviews, you're getting less traffic. But the users that are clicking through are more qualified. So they're more likely to convert. They've spent the time to do their research directly in the channel. And so anytime they click, they are more likely to want to actually take action. And so we're seeing that the game is getting a bit more fragmented, meaning that it's not just Google as the platform that people are going to at this point. More and more people are going to more and more places, even though Google is still the main stage in two ways.

One, it's still the biggest referral channel on the Web, even though it's reducing its referrals. And people go to Google from these platforms to verify information. The last stat that I saw is about 40% of users that leave ChatGPT go to Google, presumably because they're thinking, Okay, let me verify this information, make sure that it's not hallucinating. But the whole point here is that more and more people are going more and more places to get their information. And AI search itself does the same.

So it's not just about your website. For those of you all that said, you know, you're just focused on your one website and, you know, the content that lives there, that's not going to be enough to get the visibility that you're looking for. Because the way it works, these platforms are performing searches in the background and then pulling in the information like different paragraphs that are then being fed to this pipeline of agents.

And then based on what those agents decide, they send the final series of data points and paragraphs to the language model to generate the response. And so in the background, it looks a lot like this. Imagine the prompt that the user is asking is, what is the best CRM for a small business? It's then going to do something called query fan-out where it breaks it into a series of queries.

And then it's going to say, like, hey, I'm looking for certain things like a pricing table, setup steps, integration lists. And then based on what ranks in classic search for those queries that align with those content types, it's then going to extract components and then re-rank them and send it to the language model to generate the response. And so this differs from classic search. And that classic search would just take the query that the user typed in, and then it would match it to a bunch of different documents.

And then it would give a ranked list of documents back. In this case, what we're having is the query or prompt comes in. It gets broken down into a variety of queries. And then it goes to a bunch of sources to give you that synthesized answer. So going back, when we talk about looking beyond your site, we're talking about a content ecosystem that's being considered here. It's not just your site. It could be anything in your shared, owned or earned media properties that you have to really cultivate in order to be the answer that is provided by the search platform.

So ultimately, you need to organize your efforts around your customer journeys. You've got to be thinking about, Okay, what content do we have that's not on the site? What's going on in social media? What are we doing in our digital PR? You've got to also be thinking about, you know, what are the brand considerations here? We've got to be thinking about different technologies, things like the universal commerce protocol, agent to agent protocols, MCPs, things like that. We've got to basically align across the organization. This isn't going to be a single team thing. You're going to have to build Tiger teams across these different groups in order to make this work.

And so your GEO program is going to need four connected parts. One, who's the owner? Who's deciding what we're doing? Two, what is the strategic direction that we're going to focus on? Three, how do we execute? And that could be a combination of people and technology. And then four, how do we measure so we get evidence that this actually worked? And the way that we break this down at iPullRank is into two separate but overlapping capabilities. One is called relevance engineering, and that's really defining what do the machines and the bots need from this content and this content ecosystem.

And then resonance design is where we're defining what the audience needs, like how do they need this content to be positioned? You know, historically, when we've talked about search, we've talked about thinking of the bots as like a secondary audience and maybe even a tertiary audience. You know, they often would say, don't create for the bots. At this point, you have to create for the bots because the bots are actually consuming the content more than people at this point.

I believe it was a few months ago, according to data from Cloudflare, that bots have overtaken the amount of traffic on the web versus people. And then the resonance side is really about, Okay, how do we make things that are going to make people actually take action? So within that framework, you still need to think about the people side of it from your organization. Like, what are we going to have them do and who are the people that are doing the things? So, of course, you need the executive sponsor. That's the person that's like the trailblazer saying, Okay, here's what we're doing. Here are all the resources that we're going to commit to this and here's how we're going to make it happen. Then you need an accountable program lead.

You know, ideally, that should be someone that has come up out of the SEO space and the reason why is because a lot of the skills from SEO are transferable to this and frankly, SEO has always needed to touch a variety of different functions in order to be effective but historically that just hasn't been allowed to happen within organizations it can also be someone else right like there are other folks that have been cross-functional that can support this work as well like strategists, project managers, program managers, things of that nature.

And then you need a cross-functional working group because you need people from brand, you need people from product, you need people from, you know, content, so on and so forth to all come together to work through this. And they need to be working within their groups to facilitate things happening as a function of this GEO program. And then you need activation owners. Who are the people, boots on the ground that are actually doing this work?

Now, ultimately, there are two models that we're talking about here, or excuse me, one model, but two ways to scale. So if you're a mid-market brand, it probably looks like this. You probably have a marketing lead, a small working group that's cross-functional, and then one buyer journey that you're starting with. And then on the enterprise side, you probably have, you know, again, that executive sponsor and that program lead as the owner.

And then you want centralized standards and cross-functional brand or location teams. And you want to start perhaps with a few brands or a few regions. But, you know, below that or supporting that would be your agency partners and also your technology. So your agency partners are going to really help you ramp up on this quickly because, you know, if you're working with a group like us, we've been doing this for three years, whereas organizations are just coming around to saying like, hey, this is a priority.

Let's figure out how to do this. So in-house will likely be a cold start, whereas working with an agency partner, you can get a lot of the learnings that we've already capitalized on. And then technology, one, is going to support the measurement, but a lot of the technology that's supporting AI search at this point is also AI native and has things like you know, repeatable workflows and things of that nature.

So the operating cycle looks like this. You know, we want to first identify the goals, the challenges, the readiness, the resources, and timeline. And that really comes in that assessment phase. Like, where do we stand? Then we want to prioritize what the activation is going to look like. And then we actually activate on it. So the prioritization, of course, is going to be the most important thing because you could do everything. Right. But, you know, we want to make sure that we're doing the things that are most in alignment with our business objectives.

And then we've got to make change management a big part of this operating rhythm, because the reality here is that this is not just a thing where you can say, Okay, SEO team, do this with your own devices and then come back once you've got the visibility. So you're going to have to change a lot of processes across the organization to get this done.

And again, this has been true of SEO historically. It's just that we got to the space where people just assume like, okay, SEO is just meta tags in the corner and go get some links or whatever. But again, because we got to do things like digital PR, we got to do things where the product needs to be changed. We got to do things where brand is considered. We've got to have consistent messages that are being placed across the web in social media and in your digital PR and so on. This impacts everyone in the organization and they're going to need to do things differently.

So we've got to start from a shared series of definitions and standards. You know, we've got to establish the vocabulary. We've got to say, Okay, here is how we get the messages to be consistent across all these things. Here's the tooling that we're going to use. Here are the KPIs that we're going to measure against. And then we've got to train people so they can actually be conversant in this and also be able to do this work.

And then you've got to review on a regular cadence. A lot of our clients do it on a monthly. But if you're doing this internally, I would actually say perhaps a biweekly basis is a bit better. And then you want to have clear escalation paths for where things go off the rails so that we can make sure that we're being consistent in this effort and we're driving the results that we expect.

Now, as far as what the program needs to fund, one, training and review, that's going to be a big part of this because we are introducing a lot of new tooling, a lot of new methodologies and things of that nature. And then we also need to fund measurement because measurement is perhaps one of the biggest things that's misunderstood here and does need to have an additional series of tools introduced in order to do it effectively. Then program leadership.

And again, that can be a function of, you know, a series of new hires or it could be an agency because, again, there's been a lot of learnings in this space. But, you know, brands have been very slow to move. And then the last component is going to be implementation capacity, which can, again, be a function of technology and/or people or agencies. And, you know, the technology really helps out here because, again, we can scale things through agentic applications through the various tools that are out there.

And so there is also going to be the aspect of human involvement. So the human in the loop component of this, too, where you are, again, standing up these tiger teams, which should be led by people that have a bit of experience in here. And, you know, you need to actually be able to free them up so they can actually do this work.

So if you're looking to get started here, there are three projects that I would recommend that you kick off before the end of the year. One, you want to set a baseline. So you want to measure where you currently stand. That's going to require you getting a tool in place. There's a variety of tools out there. The one that we recommend and we partner with closely is Profound, but there's many tools out there that do similar work.

And then the second project is expanding where the brand shows up. And what that really means is figuring out where the gaps are in the content that you have and also where that content lives. Because, again, it's not just on your website. It can be in partner sites. It can be on affiliate sites. It can be social media content. It can be the conversations that are happening on Reddit and LinkedIn and content on YouTube. All of these places are where you're going to want to expand across that content ecosystem. And the last thing I would recommend that you can get started with pretty quickly before the end of the year is piloting agentic workflows. Again, because the tool that you would likely buy for your measurement also has the workflow management within it. You can stand up a series of those right away that can do some of these AI aspects for you.

So like it can be content creation. And I'm not saying all your content creation should be thrown to AI. But there may likely be some assets that you can spin up quite quickly using internal data or internal content that isn't otherwise published. And then also for some of the work that physically needs to be done, like the implementation work, you can leverage agentic workflows for that as well.

Now, as far as measurement, you want to measure what your leadership can actually act on. So visibility, you know, are we showing up? Do we have to share of voice? And also, is that is where we're showing up accurate? You know, citation accuracy is a big deal here because if these platforms are saying things about you that are not true or not accurate, you're being presented in a poor light. And that's, you know, just a brand safety issue. Then you also want to measure capability.

So can your teams actually implement these things effectively? And so that way you can understand, like, Okay, we got all the strategy, but did they actually move forward on it? Performance, is it driving demand? Is it driving pipeline revenue? and we need to talk about that very specifically, but it is something that should be measured here. And then also measuring execution. So there's the capability aspect, like can your team actually do it? Do they have the tooling and so on?

And then are they actually executing on it as well? Now on the measurement, a key thing to know is that these platforms are primarily brand channels. There's a performance aspect to it, but for the most part, 80 to 90% of what's happening in those platforms is users learning information and not actually acting on it. So you need to measure accordingly. And the way that we do is we stratify it into three different layers.

So performance measurement is what it's always been. Right. Like what does the referral traffic look like? What does it do? How does it convert once it lands on your site? And again, that's going to be the smallest component of what you're going to see as far as performance. And then there's the channel metrics. So how are we showing up? What's our share of voice? What's our citation accuracy? What's our citation rate? Things like that.

And then there's the input metrics. That layer is all about what we can affect and what did we ship? So we can look at things like how many content assets have we created? What's the bot activity look like to those? What are the rankings for the sub queries that these platforms are running in the background? And there's a variety of other metrics that we have internally, some of which are proprietary that are predictors of what's going to happen in the channel metrics and then also the performance metrics.

So look at it as a cascade. The things that your team is doing to impact the input metrics then hits the channel metrics, which then hits the performance metrics. Now, as far as tooling, you know, there are three layers of thought there. You know, the tool needs to measure the impact of things, and also it helps to move the team's program forward.

So the tool quality is definitely something that you need to dig into when you speak to these vendors that you're considering, because, you know, the methodologies differ across each of those platforms. And so if these tools that are measuring your AI visibility are just running the prompts once a day, that is not telling you much of anything because these are probabilistic environments.

And so what you want to get a sense of is the distribution of visibility per day. So ideally, they should be running like each prompt 5, 10, 20 times a day so they can show you like, Okay, here's the rate that you showed up. And then these tools should also be helping you actually execute on this work as well, helping you make decisions and improvements that are shipped so that you can get value from that program, driving the demand, driving the visibility, driving pipeline and revenue.

And so when we talk about this project of expanding the surface area of your content ecosystem, again, that's thinking about like, how do I get more content in the places that these AI models are looking at? So, of course, getting more indexed in Google and the variety of other search engines that are at use here. It's not just Google. It can be Bing. It can be some of these services like Exa and Tavily. You want to make sure and also Claude is using Brave Search.

So you want to make sure that you have more surface area in all these platforms. And that could also mean like doing content like podcasts, getting in the news via your digital PR, making your video content. Video is still the easiest way to be an answer in AI search because there are a lot of situations where these queries are run and they're looking for videos, but there's no video available.

So it defaults to whatever other content type is there. So if you're the one that has the video on the subject matter, you then, you know, get the higher likelihood of being fed to these language models. So ultimately, one of the projects needs to be like, Okay, let's figure out where our content gaps are and spin up a learning center. And then also think about, you know, where do we distribute this content to other places?

How do we get more active in Reddit and things of that nature? And so as you're considering, you know, which company to partner with, there are four questions that you should be asking. One, how do you measure? Because a lot of people haven't figured out that layer of measurement that we're talking about, where we have a deep understanding and how the different layers interact with each other. And a lot of them are just being like, oh, yeah, we'll give you share of voice and, you know, referral traffic. And that's that.

And then you also want to ask them, what will you change? So there still needs to be a level of analysis in this sort of thing. Like, I can't just be like, Okay, we'll change these random things. But there is a methodology that you should expect from whoever this partner is. And with us, a lot of it is like, okay, again, how do we create that surface area in alignment with those sub queries that we know these platforms are looking at?

And then also, how do we set up your measurement very effectively so you have visibility across this entire cascade? But a lot of it is still going to be very bespoke strategy on how we get there. And then another question is going to be who will get it live? So how do we actually implement this work? And again, we have a lot of methods for how we do it, whether it's going to be in an automated way.

But also coming from that SEO background, we have a lot of experience of like, Okay, how do we actually implement these things? So we have a clear path for what teams need to be involved and so on and so forth. So if whoever you're going to consider doesn't have a clear answer to that, then, you know, you're going to get a bunch of paper pushed back and forth and nothing is actually going to get done.

And then what will we own after 90 days? So that's really important. These sorts of programs need to be building infrastructure for you, not just like, Okay, we did an audit of some kind and, you know, we ran our agents and then we're leaving and you have nothing. These programs should be building infrastructure that your organization can continue to use. And to that end, here's how we build and run agents to support the work that we do for our clients.

So everything starts from strategic planning. We want to first review the goals and the workflow as it stands. And then if there are augmentations that we need to do to those workflows to make them work in an agentic environment, we would, of course, do that. And then we want to map these agents that we're thinking of building to different opportunities because an agent isn't always just like doing one task. It can be a series of agents working together.

So let's say we're talking about content creation. You may have an agent that does the research around the keywords and prompts and then an agent that writes the brief and then the agent that drafts the copy. And then, you know, an agent that judges the copy to make sure all the things in the brief were accounted for and so on and so forth.

So you need to think about, like, how do we want to make all this stuff work together as well? Each agent is an interchangeable part. From there, we build out an agentic operations plan. Then we develop those agents for you based on your own knowledge layer or context layer, as some of the platforms call it. And then we build the whole agent library. We evaluate and then we QA it.

And then we deploy the agents and train your team on how to use them. And then there's all agent ops from there to make sure things are continuing to work as expected. And so here's just a quick little catalog of the types of agents that we build to support this sort of work for our clients. So there might be an owned audience builder that figures out which organic pages and AI-referred visits turn into subscribers and accounts. Or we can build personas and synthetic buyer journeys. Or we can build like a knowledge graph for you. Or we can identify sales questions that buyers are asking and, you know, in the calls that they're tracking in Gong and then build prompts automatically from that. Or we can, you know, look at competitors and so on. Any group that you're working with on this sort of work should have something similar that they can give to you and/or build out for you to support the deployment of your work at scale with AI.

And so when we're doing these agentic marketing programs, let's say we're building 48 agents for you or your group. We're going to build this whole library and then build workflows around it. And so a lot of it is, again, that strategic planning where we're doing an assessment and then we build operations plan. At the same time, we're building out that knowledge layer for you.

And then we build that agent library. And then we build you an agent evaluation suite just to make sure that things are working as expected. And then there's a whole training and SOP handoff to your team. So that would be the third project that we recommend for you. And again, you don't necessarily have to be doing these things with us, but these are the opportunities that we're recommending. that you capitalize on as a pilot through the end of the year.

So you're set up for, you know, 2027 success for AI search. And from my perspective, all three of those are better than running an RFP because the RFP responses are going to be so far all over the place. You're not even going to be clear if whether or not people can actually do this work for you. Whereas in this case, you can just go after one of these three pilots and they can be bespoke work that are not super expensive, that you can get some infrastructure on the back of and then you can ramp up an engagement further from there.

So what I would recommend for you is that first 90 days be establishing the baseline, doing that prioritized work, get the changes live and then figure out how you want to expand from there. And so I have a quick case study that I want to share with you for work that we had done, you know, around this time last year going into the new year for a self-storage brand.

So they came to us and they said, like, hey, our board said we got to get this visibility score up for AI search and it has to happen by February. And so we were able to actually increase their AI overview visibility by about 300%. And there were also a lot of other improvements in other channels. So in AI search visibility broadly, it went up 34%. Referred sessions went up 32% and then referred revenue from AI went up 21%.

And so with them, we did our whole AI search pilot for them where we do an identification of prompts and the queries that are behind them. We do an omni media content audit where we're looking across their content ecosystem to figure out where the gaps are. We do the same for the content plan. We build out a measurement plan for them. We do a whole technical assessment and then we also give them a strategic roadmap.

And so you'll note that we started, you know, towards the end of the year, building out the strategy for them. And then there was a bit of a delay for the implementation just because that's how big companies function. And then as soon as they got that implementation going, they saw a dramatic increase in performance. So what this speaks to is that need for clarity on who's going to be the implementation partner when you work through this sort of engagement. If your teams aren't ready to implement, of course, you're not going to see significant impact.

But if they are ready to implement, this space reacts a bit faster than what you saw in classic search in that a lot of the retrieval is happening in real time in platforms like ChatGPT and Perplexity and so on. So when you make changes, you tend to see reactions to those changes very quickly. And so if you are considering like what your roadmap should be to get prepared for 2027 and you want to get started now and you're looking at your internal teams, I would say spend between now through the end of the year, setting goals, identifying owners, identifying baselines and then building out your budget and then staff up for a focus pilot going into Q1. And then, you know, Q2 through Q4 is when you expand based on where the results and the team capacity supports it. And the reason why I'm saying that it's going to be difficult for you to get anything kicked off between now and the end of the year internally is because internally it's not realistic that your team has the skill set to do everything.

They just haven't had the reps around this just yet. And it's going to take time for them to build out a viable strategy and so on. But if you go the agency route, you can actually see results between now and the end of the year because they have the reps. They know what to do or ideally they know what to do. And then that way you can get a result by the end of the year and figure out how you want to expand in a structured way going into Q2 or Q1.

And then Q2 through Q4 is still expanding further based on those learnings for the preceding five months. But ultimately, you need to make five decisions here. What are the outcomes that you're trying to achieve? And again, you're going to figure that out based on the baseline of where you stand. Who's going to be the owner for the GEO program? And then how are we going to align this with the customer journey so that we make sure that we are doing things that have business impact, not just AI search impact? You're going to want to identify and assign resources.

And again, you're going to want to measure to make sure you're getting the value out of this that you expect. Cool. So quickly to just wrap this up, you know, before we get into the Q&A here, we have actually written something that we call the AI Search Manual. And it's about 25 chapters of pure fire on everything you want to know about how AI search works.

You know, there's also a quick start. So one chapter that you can read as an executive and, you know, know what you need to know to then inform your team. And then you can share this with your team so they can get all the details, no vagaries on what to do here. And you can grab that AI Search Manual from this URL right here.

And then one more thing to know about us at iPullRank. We have a conference called SEO Week, and it is the premier AI search conference. Tickets for sale right now. So you can go ahead and grab those early-bird tickets. I think some are still available. And it's definitely the show that you want to send your team to if you want them to really know what's going on in this space. And with that, I just want to say thank you and turn it over for questions.

So let's see what we got here. Okay. Q&A. All right. Yeah, I'll just go through these questions. On the case study, while it's understood goals, KPIs may be unique to that brand. Sure. Yeah, I mean, every brand that we work with is going to have their own unique KPIs. But one of the KPIs that we stand up is going to be around that visibility.

But like I said, we stratify it into the input, the channel, and performance measures. And so we're never just looking at the one thing. But in the case of that brand that we were speaking to, they specifically were like, we need to get this number up because that's what the board said. And that was specifically around AI visibility. But more importantly, you know, we did also move the needle around revenue for this brand as well.

So we got their AI-referred revenue up 21%. So, again, we're going to look at a series of KPIs, no matter what the situation is. And, okay, how important in your view is book publishing to drive AI visibility? So publishing a book creates a lot of signals across the web that will be picked up. I wouldn't say it's like a super important thing to do, but if it's something you are doing, you can capitalize on it in AI search. And what I mean by that is when you publish a book, there's like an Amazon page, there's a Barnes & Noble page, there are a variety of different sites.

You know, it could be like Goodreads or there's as many sites that your book becomes a part of the feed. And then also there's just the natural users talking about the book as well. And then the digital PR around the book. So it creates an event where there's a lot of content created across that content ecosystem that creates more surface area for you and your brand if it's associated with the book.

So I wouldn't say you have to publish a book to get visibility, but if you do publish a book, it will support you, your brand and those entities related to it. How do you get all the agents to work together? I currently have agents, but I am only allowed to use one at a time. So I need to know more information about where your agents live, but there are tools that allow you to orchestrate agents.

So I'm just going to speak to Profound because it's top of mind for me. But the way it works, you build a workflow and you say, Okay, I have agents one, two, and three that do tasks A, B, and C, right? And so you would say, okay, agent one, do this thing. Now, agent two, do this thing. And if you wanted to do things like concurrently, you can also set that up and then say, okay, once these two tasks are done, feed it to this other agent.

So there are a variety of tools out there that allow you to do this, right? Like one of the more popular open-source tools would be n8n. But the whole idea is that once you've got your agent set up, you can then set up a workflow where they do things either concurrently or one after the other. And the input for, excuse me, the output for one agent is the input for another agent.

And then you basically just say like, this is a series of steps that I want you to do. And then give me an output that lives in Google Docs or in a database or whatever it is. Okay. Oh yeah. And also, you know, some of your agents can be plugged into something like Slack or Claude or whatever. Where are we at? I love this organized list of agents.

In an ideal world, I do it in order, but as a startup with limited time and budget, what would be the most efficient order to start from? E.g. do I do video content first and website content and podcasts, etc. So that should be driven by research. And what I mean by that is, you know, you want to identify what prompts your audience is looking for and then figure out what queries those prompts are triggering in the background.

And then you want to see what ranks for those queries and how you rank for those queries. And so you can see there's tools that will show you like, hey, ChatGPT ran all these queries and then it selected these pieces of content to feed to the model. So what you want to do is compare what you have versus what's selected. And then that'll tell you, like, Okay, here's what I need to create. It's like, Okay, they're looking for videos. I don't have a video.

So I'll make a video for this query. Or, excuse me, there's a lot of text content that's being considered here. and I don't have a page that speaks to this or there's a bunch of conversations in Reddit that I'm not a part of. So I need to have my team, my social team contribute to those conversations. So there's no blanket thing for me to say here like, oh, just make more content on your site. No, you need to figure out where the content is being pulled from and then where the gaps are in your content ecosystem and then prioritize based on the conversations that you want to be in that are most valuable to your organization.

Please could you explain the methodology behind the statement? Bots are overtaking humans reading slash searching. Reading what? Where? The web at large. So what Cloudflare reported on was how much traffic across the web is human-based versus bot-based. And bots have overtaken human traffic. And that means everywhere, like going to your website, going to social media, every place that people are looking at content, bots are going more than actual humans.

Standard question, will the deck be provided? Yes, absolutely. The slide mentions CMS and Slack integrations. Can this also integrate with Asana? For example, creating or updating tasks, assigning owners, tracking recommendations to review approval and reporting on implementation status. If yes, is that included in scope or an additional bill? So if you're talking about us, like, yeah, absolutely. We do these sorts of integrations with whatever you've got. And it would be a part of the scope of what I talked about.

But anyone can do this, right? Like, again, if you're using existing tools like Profound, like n8n, like anything that's out there, they have connectors to most things. Asana tends to be one of them. ClickUp, like all the popular project management tools are typically available. Slack integrations are typically available so that you can trigger these workflows from directly there. So, yeah, you should be able to do that.

Are all bots created equal if the web team blocks reporting from bot traffic? Are we losing valuable GEO data? Absolutely. Not all bots are created equal. Some bots are training bots. Some bots are malicious actors. Some bots are for rendering content. Some bots are for real-time retrieval based on what the user is asking for. And what is blocked needs to be a business decision at this point, because if you're blocking yourself from these platforms, you're effectively killing your visibility for these platforms.

So it's really about like, what does your business care most about? Is it getting visibility in these platforms? Is it, you know, you being a part of the process when the user is trying to do something? Or is it that you care more about people going to your website so you're going to block everything? Those need to be business decisions at this point, not just search decisions. I can't tell if I got all the questions or not.

Let's see. Okay. We already used Semrush Enterprise for GEO research prompt, ranking analysis, and page optimization briefs. How does this program complement that work rather than duplicate it? Which parts of the workflow would your agents own versus remain in Semrush with our internal team? So there is no tool that does everything that we do. There's a lot of overlap in the things that Semrush Enterprise does. But, you know, I haven't seen the tool yet that looks at the prompts, identifies the sub queries, gives you the rankings for the sub queries, looks at your content, analyzes it and tells you where your gaps are. Those are all things that we're doing because, you know, the tooling out there available doesn't do it yet. And then.

So as far as like how things would work, if you're using Semrush Enterprise, we would leverage Semrush Enterprise for what it does, and then we would build around that for agents that will do the things that we would typically do. And then you guys would have like an instance of all that that you use long term. It wouldn't be like we're not a software company. We're not trying to get you to subscribe to that. We would support the maintenance of those agents, but you would be the ones that use them.

We're talking about AI visibility, but what about new AI models that are starting to interact directly with websites such as Astra 6, Muse AI, etc.? This means visibility will increasingly depend on a website's operability. In your opinion, will this change the AI visibility landscape? Absolutely. And this is one of the reasons that I've said that this is not just SEO, because these platforms are going to start integrating with a lot of other things.

And then it becomes more of an engineering problem than an SEO problem. And so my prediction is that we're moving into a new era of media that I'm calling machine media. And it's going to be far less users going to websites, which we're seeing in the trend that I've already talked about. And then there's going to be these platforms and also apps that don't currently exist that are going to go out to the web and proactively get the content for you and then remix it into the format that you want as a user.

So the interoperability of these websites and platforms is going to be really important. And so these protocols that are being introduced, like MCP, A2A, UCP, ACP, all these sorts of things are the beginnings of that. And so getting ahead of that is really adopting these protocols as soon as possible. And again, it's going to be far less about the website itself. It's going to be the website is just one of the available presentation layers of the data. And we're going to serve the website content in a variety of different ways. And these platforms are expecting enrichment of that content for each platform. And ultimately, we're already seeing the beginnings of this because, as an example, Google has something called NotebookLM. It used to be called NotebookLM. And you could feed it videos or text content or audio content and then tell it to make me a video about that or a podcast about all the content I put in there.

So you can expect that we're going to get to a point that that's going to be the primary way that users consume content. What else have we got? Okay, being that SEO is... Wait, where is that? Okay, being that SEO is always a true foundation of visibility can you share if paid media has been factored into any of your strategies or beta tests example we are running ChatGPT ads for clients so um we aren't doing that and one of the reasons why is that i'm not a believer yet that those ads actually work um you know i think ChatGPT has had the race towards providing that but by their own admissions and some of their court documentation They said like no matter what they do, like no matter how big they make citations and so on, they can't get users to click on them.

So I'm not... I haven't seen evidence anywhere that running media on ChatGPT actually does anything. But as that matures, we will definitely dig into it. And we may be an agency that offers that as a service once we know that it actually works. Okay, regarding the agentic marketing program, you will create an operations plan, integration, knowledge layer, and agent library, but what will we still need to implement internally? For the 48 agents, can you outline the ongoing work by role and estimate the time commitment needed to action and maintain the recommendation? I feel like this is a discovery call now.

So any of the agents that we build are going to be unique to you, like based on your situation. It's not just going to be like, Okay, we've got a bunch on the shelf here. Use this subset. We often build net new agents that we haven't built before. And it's going to be based on use cases that matter to you, your brand and your business. And the time commitment for actioning and maintaining the recommendations.

So the agents are meant to deploy themselves. Like if there's any sort of human review that we want to set up, we certainly will do that. But it's not a thing where you need to maintain what they've done. They can maintain what they've done themselves. I think we've answered all of them. Yeah, we have. Okay, cool. Well, first of all, thanks, everybody, for joining us.

You know, my contact information is right here if you want to reach out for any additional questions. And also, if you are interested in potentially working with us, of course, you can use that QR code to book a discovery call. Slides will come to everyone. And again, we've got plenty of information on how all this stuff works. So if you go to iPullRank.com, you'll see some of the most cutting-edge information on how to make AI search work. And like I said, I'm Mike King. Very appreciative of your time. And yeah, looking forward to hearing from you.