068. Using The C.L.A.R.I.T.Y Framework for LLM Optimization with Tim Lowry

Foreign.

Hello, everyone. This is Emilio Garcia. Welcome to another episode of the Man Gen Studio. With me, as always, is Miranda Jenkins. How are you, Miranda? Good.

Glad to be here.

Awesome. Yeah. I hope that everyone is doing okay. And today we have a very special guest, Tim Laurie, a good friend. And we have had a. I think this is the third conversation that we have, Tim, that we have recorded of sorts, I think so. It's always, it's always a pleasure to have a conversation with you and get some insights in your experience. So really, really excited about today's conversation that, you know, for those of you wondering what this is all about is going to be a mix of SEO and what's going on with AI. But I'll let Tim not only present himself, but frame this conversation today.

Yeah, I'm excited to be here with you guys. Thanks for the opportunity. And I think today we're going to dive into a, Into a topic that a lot of people, if they're, they're not in it, they're going to be getting into this topic and into the weeds of it very soon. So hopefully, you know, our conversation will provide kind of a framework or thinking that they can apply for either their business or businesses that they work with. But. So background name is Tim Lauri, founder of Tiptop Search and Marketing. We're an SEO agency. We work with a mix of e Commerce and B2B clients, mostly national, some international. Been doing this for around 10 years now, and over that time have seen a lot of changes with Google, but we're on the edge of some of the biggest changes that people are going to see, not just with Google, but with all their search behaviors and habits that people are forming. So it's a really exciting time, but it's also another period for some marketers to, you know, perhaps be like, I'm done. Too many changes, I'm out. So hopefully that's not the case. Hopefully today you'll feel a little bit energized after you hear this and can move forward with just more comfort around the direction of where things are going.

Yeah, I, I want to, I want to bring to everyone's attention to basic things first. Yes, I, I have to reiterate that you are dealing here with a person with the expertise and knowledge, you know, that I find in very few people that I work with and know personally. So I want to stress out that I've known Tim for many years and he is the man for, for SEO and search engines and all that. So. But the other thing I want to say, Tim Is that I think I, I talked to you about this in per conversations we have. I felt when ChatGPT launch a few weeks or months after that I was in a very similar situation. When I remember Google became important in my life, I remember exactly how it happened. It was before then. I remember there was this incredible amount of directories where you can, where you could search and the results were basically links as well, but they were based on categories. Right. So I remember it was like, let's say I don't know, out of altavista for example. And so there were so many of them, but none of them had the intent behind on the search that I used to use a tool called Copernicus or Copernicus or something like that. And it was an aggregator. I remember that tool, you know, we'll connect with 10 different engines or directories rather than search engines and they will spill out results. And I've been, I used that for months, maybe a year or so to. And then I remember the first time that I stumbled upon Google and it was, the shift was swift is I just started using it a couple of times and then it got kind of like stuck in my head that well, this thing most often than not give you, give you what you want rather than sifting through all those results. And I remember that that tool disappeared from my radar right after being, you know, a faithful user of it for years. So I, when I started using ChatGPT that, that happened at the beginning. And the only reason I'm always like acceptable about is this going to be a permanent change is because I, I feel the variable that I didn't consider back then is that I was younger and therefore more flexible in my habits. I will say, and I think that now I get older, I feel that it happens that users, you know, say like, oh well, we get excited because we are marketers, but most people haven't even tried it out. And if, if by the time they do, maybe Google catch up and they never change their experience. So it is still in my mind like a, like a gamble what is the direction that it's going to take, but definitely is something very disruptive.

Yeah, it's, it's disruptive and all of the major players are, are in it and they're all indicating that they see this as some component of the future by way of people search, interact and do things. And I think that with Google, the way I look at them is it's like we're all kind of like frogs on a slow boil. So they realize that people Move slowly. So over the last number of years, if you look in, probably in the last two years especially how much the result page has evolved from what it used to look like, the result page has got much more dynamic. There's so many more features. Obviously they've brought the AI overview into there, so they realize that people aren't necessarily going to make this mass shift from something that they formed a habit in. So they're slowly bringing that habit into how you do things. And, and it may not be the people move, but over the space of a few years they form this new habit of how they search, which is shifting from that question answer type format that we've all been used to for so long to more of a conversational format to where you take a conversation for 50% of your journey and then you jump into traditional type search and go to a website afterwards, but you've maybe interacted with AI or some other feature or some other thing to get you along the way. And it started with small things like the people also ask where they're kind of coercing, like here's conversational type things that you can ask to go further. And then of course the AI overview being a part of that. So I think that again, it's not going to be something where every last user that's on Google right now is going to just all of a sudden jump the GPT. But I think that as it gets baked into everything that people use, so whenever it's in the hardware that people use, their Android devices, their Apple devices, in their car, in their home, starts making it something where you just start using it and before you know it, you've evolved as a searcher into somebody that's now using this very different method to find information.

Yeah, and, and, and I want to clarify that the other reason I think, you know, is something becomes disruptive, it's not just because of the technology, but how people think about it. Right. Going back to that example, if I think about, well, what got you super excited about Google 20 years ago, that is different today if it's the same tool or even as you're saying, is better in so many dimensions. Right, but because even back then it was just a list, didn't have all the features that we can see right now. And the reason was because my thinking evolved from well, I'm using a tool to find a website to I'm using this tool to get some answers or at least find some information. And I think that's the other thing that AI is bringing to the table that we are shifting our idea of when I go into the web, rather than just compiling information from multiple sources to get an an answer, now we are using it to create things to reason to do work. And those were not the use cases that you will think about of a search engine. And so as much as the, the tools have to evolve, also the way that we see them I think has to evolve. And, and I guess that will be a great intro to what you have, you know, discover and what and are working on. Because for our clients and in, in the case of our listeners for their companies, it's really important to start thinking your own prospects and clients, how are they thinking about search? And that is going to move. Right. So a lot of your current clients right now, meaning the people that is listening to us, most of them will continue to for a long time think about Google and search engines as a way to find answers to some to some information, even if they struggle to couple together the answer. But little by little they will start moving to this idea or concept that we already are experimenting, which is I can use this tool to reason reason along me. Right. Or to, to solve complex problems that require compiling data, researching and all that.

Yeah. And I think the other part of the thinking is again like Google is great, websites are wonderful, but Google is always trying to make the user journey more simple. And part of that making the user journey more simple is not always getting somebody into a website. And people are finding that when they go into a website, they're having to read through all this unnecessary information to try and get to something to find it. It's a complicated journey and it's just got more and more complex as more and more people figure out the algorithms and how to get themselves positioned better. So in some situations you do a search on Google and it's not as easy as it should be. And Google's trying to improve that. And I think that's Even recently like HubSpot and some of the stuff that they had recently with the dropping of their traffic is things were done and they were a great resource and they were done without any form of malicious intent because it's just how you would make something rank. But then all of a sudden Google creates an AI overview that shortcuts somebody having to read an entire article and gives them a link to a PDF that they can just write then and there. Here, download the template for how to, you know, ask for a pay raise. Okay, well I don't need to read this article now. Here's a template with the email already drafted and I just click on it. So it's simplifying that. And I think in certain industries they're going to feel that more and more where like thinking around like SaaS companies, things like that. If you have a product and people have historically found themselves going to sites like Capterra and looking at side by side checkboxes and things like that, that's great. But you have to do multiple searches to keep getting deeper and getting more clarification. Like is this right for my business versus being able to jump into GPT Gemini, one of those other large language models and just giving it a very use specific thing like hey, I am this type of company with this many employees, we operate at this budget. We need a tool that does X, Y and Z, that integrates with blah, blah, blah and you put this very complex conversational type thing in and then it starts giving you the information back like, okay, well here's what we would recommend. This tool integrates with it. So it all of a sudden cuts through 10 different searches that you'd need to do and gives you, here's three different people that could potentially fit the build of this. Here's the pros and cons of each. What questions would you like to ask me? And follow up to that and then you can start having that conversation around a specific one. So you've taken care of the informational aspect of the search. You've now moved more into that navigational where you've identified the brands and then once you've narrowed that down then you can shift your commercial or transactional intent of like I'm just going to go to Salesforce because they're the clear winner. Like yes, their price is higher, but for all the integrations and things that we need, it sounds like they're probably going to be the one that works with all of the rest of the things that we use in our business. So you've just expedited that. And for a company like Salesforce, well then now all of a sudden they're going to find say a look in their Google Analytics that they're getting referral Traffic coming from ChatGPT, from Perplexity, from Claude, from whatever it might be and that person has taken the search engine out of that journey and came into it. So for the clients that I work with and the people that are listening today, this is like, how do we, how do we bridge this? And again, that's a, it sounds like a kind of a big broad example. I'm like, yeah, of course Salesforce is going to be there, but I have a smaller SaaS company that, that I work with and they got their, their first MQL in January and it came from ChatGPT. So like it's something where again, this is, this is not a Fortune 500 company. Like this is a smaller startup of a SaaS company and they, they just got a lead through there and they're now in conversation and that person never once went to Google in their journey. They've already done interview with a person and it's like, okay, we maybe need to take this seriously. Like it's maybe not going to be our full pipeline coming from here, but what are people learning about us over here? What's being said? Are we even in there? And that's where this kind of comes from of like, okay, we got some, we got some work to do because there could be this black box of who knows what out there about your business. So yeah, that's, that's where we ended up.

I love what you're describing because the, the, the takeaway that is important for, for everyone listening to this one is that yes, the medium, the technologies continue and will continue evolve and will continue to evolve, but the steps that we use or we follow to research about things that interest us haven't changed. Right. That, that process that you're describing is, is going to continue to happen. The only difference is where it happens. And that's the part that we have to adapt to.

Right.

Is that people still are going to ask around, do research, compare, you know, try to find the information that really caters to their use case.

Yeah.

But instead of you saying of multiple searches and website visits and maybe going to Reddit and going deep into that, and some of do, some of us still do, but maybe at the first stages you just summarize what is already written, which is pretty much what you are asking the AI to do for you.

Yeah, you're just, you're having a conversation. And like a lot of people, I've went through the various emotions of AI from whenever it first came out where I was like, I hate this, it's stupid. And then you do all of the kind of quirky things where you're making it do silly stuff. And then you start realizing the business use cases where it's really tactical and starts helping. But now as we evolve, you know, a couple of years down the road of it being more mainstream, you start realizing like, okay, people are using this legitimately in their day to day of like things that they're asking you have conversations with people and you learn like, oh, wow, they changed their search habits entirely. They're still going to Google at some stage, but they're asking questions somewhere else or getting information somewhere else. And as we took to the models and started testing things for our clients, that's when we started realizing like, it's not an apples to apples. Like what's being said about you as a company in Google is not necessarily what's being said about you as a company over here. Because they have totally different ways that they scrape the web, that they crawl, that they get information and how they bring that information together. So then that set us off in kind of a learning journey of like, okay, well we know how to get people to rank on Google and we're still Google to the core because that's still the largest, that's the whale that still gets the largest amount of business for people. However, if in two years from now, if this represented 25% of the pie for a customer, well then that's enough to be noticeable where if you're not doing something, we don't want to be in that catch up position of trying to figure this out. So with that said, I'm happy to kind of dive into the framework. You can ask questions or whatever way you'd like to, to go about this. But yeah, it's, it's been a fun journey.

Let's do that, let's do that. Let's dive in.

All right, cool. All right, so essentially the framework we have created is called Clarity. So it's an acronym and we'll just walk through each part. Please ask questions because I don't want it to be a monologue of me just like going through here. But yeah, so essentially the C is for crawl. So this is a very manual part of the thing, but it's just going in and testing to see does something come up whenever you type in your business name? I guess does your business even, even surface? And if it doesn't surface, then it's trying to check and make sure there's nothing blocking that. So it's the same thing you would do with Google. Can I actually be crawled by Google, by Bing, by the various bots that are out there, or, or do I have something in my robots Txt that's preventing me from being able to be crawled by any of these AIs? And if so, has that been done intentionally or was that just something that was done out of a knee jerk reaction a year and a half ago of like, I don't want these Things to get my information. So I'm going to close them out. And now it's like I forgot about it. Yeah. And now it's like maybe actually I do want them to get my information. So there's different tools that are out there, there's that you can test for that, but there's about 20 something different bots that crawl and everything from chat, GPT, Gemini Llama, which is Facebook's Claude Perplexity. They all have their own agents that will crawl and find that information. So the first step is really simple. It's crawl. Can, can they actually discover and find the information or are you blocking them out? Then we move.

Yeah, I wanted to say there that, yeah, I agree with you that I tell, I, I tell clients all the time that for your brand there's a conversation going on. You know, there's, there's nothing that you can do about it for not happening. The only thing that is truly in your control is are you part of the conversation or you are out of it. Right. Because this has happened over and time and time again when you say, well I don't want people to say bad things about the company on Facebook or in this other social media network. Well, if you're not part of it, you have even remote control or at least influence what is being said about you or be part of it and do something about it. So I think yes, the approach of I don't want to be crawl or areas like ideas like those is not productive because and at the end of the day the conversation is going to happen. It's just that it doesn't find it in your website or your digital assets is being found in what other people is saying about your.

Exactly. And that's, that's the scary part because if, if you've struggled with reputation management and I've, I've worked with some clients before where like trying to do that cleanup on Google is hard but all of that stuff is now like the fingerprints of it there still remain. And for some companies that's now bleeding into AI because that isn't cleaned up. Or even something more simple in theory. But in practice, sometimes difficult naming conventions where you call your company one thing on one social media platform, you then use a slightly different variant to the company name and another one and then you have a short reference to your company and then like having, having a convention across all of them where it's like all of our things use the company name and X, Y, this way and this way and this way. So then whenever the AI goes and crawls. It understands this is all the same entity and brings all of that together versus fragmenting it into a lot of different things. And that's something where people are finding on a lot of the personal stuff now on AI overviews where you'll have an AI overview and it's a dumpster fire where it's like, well this is me as a person, but this is an image of somebody else that's not me. And here's something else because there's this name and it's used in lots of different ways and you have all of your different profiles and handles and your full name. Exactly. And it just becomes this just. Yeah, just a hot mess of like stuff. So I was reading a case study the other day of a guy where he went out and basically anything where he could change, if he had been on podcasts, if he had been on YouTube interviews, if he had done things like went and got his name fixed in all of the different descriptions across all of the things and about six days later it fixed the AI overview to where they were referencing the correct person for all of the different things because he had uniformed how his brand was being mentioned for him personally. So I think for companies too like being mindful of like how your brand is represented because if these different agents crawl, like if you don't have a strong entity tie between all these things, they're not going to understand that. But moving from the crawl aspect, then we go to learn again, we're sticking with very manual things in the beginning here. And that is where you're going to start discovering what these large language models know about you. I would say, you know, don't go feeling that you have to do every last one, like do Chatgpt, do Gemini, they're kind of the, the 600 pound gorillas in the room, they're the big ones. So get an understanding of like what they know about you and just test brand related queries. So you know, what is Company X? What does Company X do? What are the solutions that they offer? Are they a good company to work with? Like just start putting these questions out there and keep track of that. Put it all in like a Google sheet or where you can reference back and even just grab some of the answers and then just have a conversation about your, your company and then start identifying like what are the gaps in information, like what things is it saying or should be saying and it hasn't said. Is there any misrepresentations in there? Is there something where it's talking about A solution we were known for 10 years ago, but we don't even sell that anymore. So why is it still talking about this? Like, how do we, how do we change this misinformation? And basically you're going to try and understand what it knows about you as a company and do it from a little bit as a bias of like, you're in the company, but then shift that to where now I'm somebody that's wanting to work with our company.

Yes.

What are the questions that they're going to ask? Are they going to ask are we a good product? What's a better alternative to this? What should I know about them? How is the leadership at this company? Like, think about the things that somebody, if they're going to spend. And again, if you're in, in B2B, like you could be selling a six, seven figure widget at the end of it. So people are going to ask some pretty detailed questions, especially if they are going to be getting into bed with you where it's one of those types of relationships that you become heavily integrated into how their business operates. So, so they're going to start asking questions. So take that same positioning and you want to learn as much about what these AI tools say and just document, like I said, the gaps and the misrepresentations, those are the things that you want to make note of because that's where you're kind of framing a little bit of that roadmap of like, okay, we have work to do based on what we're doing there. We then go from the learning to analyzing. So yes, we're going to look at those gaps, any misrepresentation. But then we're also going to look for like, what is the sentiment around our brand? Are we seen positively as a company or is it kind of a negative output that we're getting? And that's some of those reputation things. Like, are people like, is it, yeah, it's a good company. However, you know, Glassdoor says, like, don't even consider working with this company because of this. Like, where are those things that are out there? Or they've got a horrible Better Business Bureau score or something like that. Because again, all of those places, they're being crawled, that information is being retrieved and it's being brought back about your company to create this profile in the AI of like who you are. And then the only thing that you want to do is kind of think about your website as you look at the misrepresentations, the gaps. Your website should be a sales Rep for you. So what does your website not currently have as you look at that information? So you're saying like, okay, well, they're saying this about us. And then if you go to your website and then have that realization on your website, actually, we don't actually state this is what we offer or this is how much it costs, or this is who we work with. Well, then of course it's not going to have that information or it's just going to scrape it from somewhere else. Like you said, like, if you're not talking about it, somebody else is talking about it. So you might as well make your website be your sales rep. So, like, you're looking for those things that are not on your site that you've discovered as gaps. And if your site is not addressing it, that's your, your first protocol. It's like, let's go to our website and make sure that our messaging is clear and specific to what we do, who we do it for, why we do it, what it works with. We may be creating more landing pages just based on what we discover. But again, this is your opportunity to start addressing some of these things.

I want to, I want to add that between the L and the A, I love it because very specific to B2B brands. I, I talked about this quite often. It's not even my idea. But is we. We tend to forget that buyers in B2B situations, obviously they want value, right. That difference between the problem they have and the solution that you deliver for the price that you offer. But as you go downstream at the point of purchase, then a variable that is missed in marketers is the risk. That value gets adjusted for risk. Because as you were saying, if you're buying hundreds of thousands of dollars of product, even if the value is there, someone's career is on the line if you can deliver. And unfortunately, that's something that is hard to assess. It's the most difficult part, especially when you're buying services or things that will be, you know, delivering the future. And that's where your buyers will search around. Right. Or ask these tougher questions to assess how risky you are as a provider or a brand. And that's a big gap that I think is important for, for marketers to. As where you, as you were saying, right. As you go through that learn and analyze stage that you ask those tough questions that are acceptable of your company ability to actually deliver.

Yeah. And that's like the, the response to something like that is in, okay, well, maybe we need testimonials, case studies, White papers, like, things on our website that back up like, oh, wow, they've worked with this reputable company and this is what they've done and this is how they done it and the outcome of that. And that's also backed up by, you know, a video interview with somebody that works for that company talking about their experience of working with the company. So you, you have to really, like, take this information and back your marketing plan into it. Of like, okay, what are the things that we haven't touched before? Because we don't as a company, like, you've seen this before, I'm sure. It's like, well, we don't really talk about that. Like, whenever we're in the conversation with somebody, we'll cover these things. And now it's like, well, what if you don't get the opportunity to be in the conversation?

Yes.

Like, what if AI is the person having the conversation for you and you're not equipping them with the answer? Like, you would never send a sales rep out into the field and not equip them for what they have to be able to answer. So it's the same thing where it's like, it's a sales rep that's out in the field. And if you're not equipping them, then they're going to just do the best job they can do with whatever they can gather together and kind of throw out there from, from the ether of the web. So, yeah, we move from. So the crawl, the learn, analyze and we move to respond is essentially then the next part of it. So the R. And that is where we look like we're kind of talking here, like closing those gaps. So with that, it's what can you do on your own website initially? So what needs to be improved there? Do I need to rewrite copy on my homepage, on my solution pages? Do I need to be more clear? Is there missing areas where I just don't have any information on that? Do I need to add a really robust FAQ to the website? And that would be a great way to combat lots of these different questions. How do I mix the types of media that I have on there? So it's not just all writing. Like, let's get a video testimonial where we sit down with one of our bigger clients and we talk about how this has changed the dynamic of their business because they now have us as a partner. And. And again, that variation, like the AI crawls that it's not just them relying on text. They will crawl video, they will crawl audio like they can, they can interpret all of that. So it gives lures and depth to what you present out there. And there's again a different context in a conversation, in an interview with somebody in a testimonial like that than there is just writing a testimonial. Because now there's the sentiment of the person's tone and their voice and how they're conducting themselves, which creepily. Again, the AI can understand and interpret all of these types of emotions in a different way that brings it into there. So having that mix of different media of how you do things is going to help for it. And then if there's things where you know, even your about page, like simple things like that, like when was the last time we updated it? Oh yeah, we actually have a completely new C suite that's in our organization. Maybe we should update our about page. Things that you tend to just kind of throw on the back burner until the next time you redo your website. It's like, don't let those things expire. Is there awards that you've received? Are there things that you've done? Have you got any, you know, publicity media? Have an amazing like media and reviews page that includes some publicity of places where know you've won industry awards or whatever. It is like all of that, like showcase it because again, if they find it on your site and then they can basically find that on another site and kind of cooperate the story between the both, then that's really strong for the brand whenever that information is out there. Um, so yeah, you're really like the respond is the big part. Like you, you get back with your questions and that could be you looking at your next three, four, five months of work because you gotta rewrite a bunch of stuff and go back to the drawing board about like, clearly we're not clear enough about our offering because what I'm reading does not match what we're trying to say. So let's go ahead and clean that up. So yeah, all right. Then from there we then move to integrate. And as we think about integrate, it's like, what are all of the things that are touch points that integrate into your site, but also the integrations that these AI models go to. So if we take ChatGPT largely funded by Microsoft, and Microsoft basically lets LinkedIn become a playground for ChatGPT to be in. Google obviously has YouTube, but then they also have a deal with Reddit where they have exclusive access to Crawl and Reddit. So it's thinking about like where all of these other integrations that the AIs train and learn their information from, and what are you doing in those? So does your LinkedIn business profile and the people that work for the company, are they doing things that are reflective of the business and the brand and what you do, or does that feel like just something that's kind of off to the side and an afterthought? And it's not really clearly messaging in there, or somebody that works for the. The company and they're a leader, but they talk more about their hobby ofGOLF on LinkedIn than they talk about what they offer. So it's like, no, you gotta bring these things, because then the AI starts to learn the entities and tie these things together. Um, are there review sites and forums where things are being said about you that you need to address? Is your Wikipedia page up to date? Or do you need to update information on Wikipedia? Because nobody else has went in as an editor and brought it up to date. So the last entry was, you know, 2008, and you've had, what, 18 years of evolving of your business since, like, the last entry in Wikipedia. So all of those things, you now start going out and you're trying to identify like, okay, we've got our website. Good. How do we now copy this information across these other places so that it's a cohesive message about who the brand is, what we offer, what we do and who we do it for? And you're really trying to strengthen all of those places. So, again, like, this is not an easy lift, but it's also something where don't just think of it as an exercise, like, well, this is just for AI. And what if AI never brings us a lead? This is just making you better overall. Like, nobody's ever going to say you've done an awful job being more clear on your website about what you offer. Like, that's clarity is always something that people look for. And if there's anything that's out there that isn't clear or ambiguous or negative about you, then ultimately it's just a good thing in general to make that better. But the flip side is you're doing it with the understanding, like, all of these things are going to eventually influence the AI models and language models and whatever else are out there to be able to better represent us. When somebody goes into conversational search mode versus traditional Google answer mode with that.

I love that because, yes, I keep using this quote more and more about that. We should focus on the things that won't change, not only on the Things that change, and I think for our space, things that don't change is that, as you were saying, it doesn't. It never hurts to get better at positioning, refining the positioning of your company, it will never hurt you. It doesn't matter the technology or the medium that you're using to do that. And the other one is it will never hurt you to improve your reputation again, it doesn't matter if it's traditional media and through PR efforts or if you're doing, you know, more traditional digital advertising efforts or in this case using generative AI. Is those two things right? Getting better at refining and really being clear on your position, positioning and managing your reputation across, across different spaces, it's something that will not get out of style.

No. And I think if you're like, if you're in marketing and you have to try and get budget or buy in from the various stakeholders, like that's a great way to be able to frame this. Where it's like we want to do this and here's one of the benefits. But ultimately this is going to just make us be better online for anybody that interacts with us as a company because people will be able to find more current, more relevant information about what we offer versus having to potentially lose business or do a lot of educating with somebody that calls and then we disqualify them because they realize they're no longer the client fit for what we do. So yeah, it could help with that for sure. So then moving from all of that, we then go to train. So there's a, there's kind of two parts to the training. There's training for yourself and then there's training for the AI. So the AI and large language models, they crawl differently than Google. Google is always crawling, always discovering, always finding new things. Especially if you're a large website or a media company, like they're, they can find news instantly. Smaller sites, not as frequent, but still they'll discover in a matter of days or weeks. Whereas with the large language models, there's a large period that goes by between the language model updating and it's training on a set of data and then that updates. So you've got that training window for, to really be working on and then document whenever. So let's say version 4, 5 comes out of GPT and the next couple of weeks here. Well then after you've done some of this work, go back in document and find what has changed. What is it saying differently? What does that look like? Okay, these are still the gaps we're now going to go back and work on these things and know that whenever version 5.0 comes out, hopefully it will have discovered more things. But you might have months at a time between these training components. And then the training for yourself is starting to think about whenever you're creating things, you need to create them to be more entity dense so that it can draw associations. So like, if you look at a lot of the answers and things that surface, many times in the AI, it ties together concepts and entities really closely as it's answering something. So whenever you're creating things on your site, you want it to understand and tie the different entities between who you are, what you offer and who you do it for, all tied together as you're answering things. So there's kind of that training of going through and learning how do I rewrite? In a way which again, it's going to be ultimately better and clearer, but it's doing it in a way that large language models easier understand what we're doing. And that's something that we've been doing for about a year or so now. Like even in our ARC blog, content writing. Like we've been writing content where before, if it was, you know, a broad, like what is X? And you're trying to win that what is query? There would have been a way that we would have answered that historically to get like a featured snippet. But then that changed whenever AI overviews came out because it's a different bot that crawls for that and it looks for different stuff. So we have to restructure some of that content in a way that those bots favor versus what the traditional Google bot would have favored. And it's typically tying in the concepts and entities into those answers and things that you're writing to. So again, trying to stitch as much together to make it really dense so it can come in and just get something and then be like, oh, that connects to this, and this connects to this offering. And this makes sense with this that they do. And the more you can do that, the more helpful it is for, for whenever it starts to drain off your data.

Yeah, I, I, I'm going to use this very blonde analogy, but I imagine it as, you know, for me, search engines are like a, a librarian that somehow is capable of reading all the books that he has on the library. Right. But when you come in, you say, you know, I want to know the answer to this. And the ability of that librarian will be that I think this book in this page is the closest to the answer that you're looking for. Right. And you just learned that after some feedback from the user saying well, that's not the answer. Or can you bring me this other book? It learns to say, well, typically when someone asks this, this is the page of the book that I typically will offer. And for me, the way that you create content for. For AI is more like a librarian that actually is again superhumanly to read all the content and you ask it a question and based on all the things that they have read from that library will say well, my answer to this is this. And so it changes completely how you set up content because or, or. Or knowledge. Because as you said, it only. It's not only what you have to say about the topic, but what that model knows about the rest of the pool of knowledge that it read at the time.

Right.

You said.

Right.

It's like the, in next year, the next version, we'll read the library again, all of it. And some books will be new, there will be more books, some of them will say some things slightly different. But whatever it learns about your book is not only what. What he knows about you, but what he. What he knows about the rest. As you said, I think there's the. Making a. Stronger associations with the rest of the knowledge pool is more important because of that, right?

Yeah, it's 100% like the. You don't want to just leave it open to the large language model to just make assumptions and try and like stitch things together because that's when er and mis. Misinterpretations, all these things can get in. So you're trying to eliminate that by giving it very clear structure to information so that it understands exactly how all of these things relate and tie together. And if it's going to talk about you, it can do that. And I saw an example like somebody the other day was saying how, you know, whenever they now write again, that was their BIOS and things like that. For whenever they're doing the large language models, it's just a very clear like my name is X, I do Y and I do it for W. So it's like that is just. It's consistently X is then to Y is to W. So that if somebody, our large language model was to talk about that person, it's like what does such and such do? They know exactly what they do, what do they work for? They know. It's like just X, Y. It's always the same consistent time of like those three entities of things together. So there's never a question of like does this go with this? It's like, yes, those go together. They're always associated and linked as a thing. And that's kind of how you would go about that with your, your business, where it's like this is this business and this is the solution that we do and this is who we do it for. So that that's consistently just a clean narrative. No matter where it is. It's always like this company does this for these people. That's it. It's not complicated, it's just very clear and you repeat that and you just make it make sense.

Positioning 101.

Exactly. Yeah, and positioning 101. It's simple, but it's just tying those entities together so it always sees it. And then from there the final part is in yield. So like all marketers, we always want to actually know do things actually perform and what do they do? There's slowly some tools coming out into the marketplace. They're becoming available. Some are only available for like enterprise size accounts, but there's a great one called Noah Tua that's out there and you can put your questions in that are important to your business and then start tracking against that to see how it is. So it goes out and it fetches these based on whatever the current update of the language model is and it'll let you know like is your brand mentioned? So it uses again, it's not like traditional ranking, like well, where do we stack up on this keyword? And so on and so forth. It's more if somebody was to ask, you know, who should I work with? Or I need a CRM software, like is our brand name in that list? Is such and such a company a good company to work with? Yes, you're mentioned in there. And here's the thing. So it pulls in some of that. So it's good to be able to have again some form of questioning, prompting, querying that you're tracking against. Especially some of those important things where like these are the non negotiables. Like if somebody asks this, we must be present in that answer and then that will inform you on again the things that you need to create. So if you have something that's there and you want to show up for it, then you need to reverse that back into your plan of like, okay, well if that's the question, what do I need to create on my website and what do I need to create off my website and what should other people be saying about me? So that whenever the large language model goes out, I'm more likely to show up for that. And some of it is going to take time. Like, that's not something, especially if it's a, a really big ask of a thing to show up for. It's just like traditional Google, like I want to be position one for this big shiny keyword. Well, if you're a new company and you're competing, competing against companies that have been doing what you're doing for, you know, a decade plus, and they have this long history and established track record of doing that, you're not going to dethrone them overnight. So it's kind of that North Star of like everything that we do works towards getting us visibility for that. So yeah, again, that's the framework that we're, we're using. It's been really helpful to start to get a clearer picture for our clients and being able to help them to understand like, what do we look like out there? And I think the, just having knowledge is part of the part at the start because it takes away some of that fear and ambiguity around it where it's like, okay, we're not, we're not a ground zero. There's some information out there about it, but there's some of these things could be better. And that's where you, you start working on. It's like, what could be better. And then as you get a little bit more comfortable, you start getting more advanced in what you do. Um, but the nice thing is like, you have control over one of the biggest areas that it trains and that's your website where you can really lean into that and use it to really hone your messaging and bring, bring clarity to what you do.

I love it.

I love it.

So, yes, if I got to summarize from my perspective how I understood it, CL&A allows you to diagnose, right, Starting with going to the LLMs and, and, and learning about it. And then R, I and T allowed you to do something about what you have learned and, and take action. And obviously at the end of the day, as you see, as you say, there's a little bit of feedback loop and finally you want to measure success with the why with Jill. I, I would like to ask you, Tim, for, for those of our listeners that maybe are in this particular situation right now, they're trying to respond using your words. What will be a good expectation? What, what can they expect in terms of time investment, in terms of how long does it take to start seeing impact on the language models, what you've seen so far?

So it's the the level of investment and work is going to somewhat be variable based upon how much, sorry, how much you have to do on your website. Like if you're, if you're messaging and things are already very clear, well then maybe there's not as much to do. But then there might be just these massive gaps where like all you've done is your website and there's nothing about your company outside of your website. You've never done any off page efforts, you've no PR publicity, there's no brand work that's been done, there's no mentions of you. You basically a ghost outside of your website. Then you have a totally different type of effort that you need to do. But from a timing standpoint, if you start creating and doing these things, I would say like this is one where right now like set your expectations at 12 months before you really see it. And that's just based on the fact that like we don't know how many times in a given year you're going to see a large language model update. So just kind of set your clock on like started today we're going to do that. It might update once during the year, it could update a couple of times during the year. So we can have these micro goals. But it's not the same as Google where we know they're going to crawl every day. We know that they do typically like four large core updates throughout the year. But then we also know that they have the ranking systems that run in the background that are always making just micro improvements and adjustments and suppressions or improvements upon websites during those time. So it's just, it's a different model. So I think it's something where there's a little bit of patience that's required and knowing that like we're building these things, we're not going to see it in the same way that we see it in Google. But ultimately we're making a better, a better website and a better off page presence for ourselves in what we're doing. And that will also reflect in the information that Google has in other places to make sure that like our messaging is really like dialed in and locked across the board for all of that.

Yeah, love it. Well, this has been wonderful. I think so much insights that we are going to get so many clips, social media clips out of this. So all they have to say is thank you Tim for joining. If, if you know, marketers are there, want to learn more about your framework, about you and tip top, what would be the best place to.

Yeah, I would say find me on LinkedIn would probably be the best one. Yeah, like all the other ones, like X and everything else. I think everybody's just kind of burned all of those accounts, so, yeah, just find me on LinkedIn.

Fantastic. Well, thank you, everyone, and thank you, Tim, for joining.

Yeah, thank you so much for the opportunity, guys.

Yeah, we really enjoy it. And as always, you know, let's. Let's do another one some point in the future.

It sounds good.

Thanks, Emilio. Thank you, man.

Thanks, Miranda. Thanks.