The 30-Page Local SEO Hack Google's AI Is Already Rewarding

Edward Sturm: Caleb Ulku, you are very well known in local SEO and AI SEO and on YouTube, and people have been asking for this podcast for a while. So Caleb, thank you. Thank you, thank you, thank you for joining the show. I'm excited to have you Caleb Ulku: Man, I, I appreciate you reaching out.

I, uh, I watch your stuff all the time. It's funny, I've, uh… My strategist for YouTube, sometimes he'll send me videos like, "Hey, this one seems like a really good one to talk about what they're saying in this video and react to it." And about a week ago, he sent me one of your videos, and I was just like, "Well, I'm, I'm actually gonna be on Edward's, uh, podcast." Edward Sturm: Yeah.

I mean, yeah, Edward, I'm, I'm, I'm hyped to have you. Uh, could you share your background for people who aren't familiar with you really quick? Caleb Ulku: Yeah, sure. So I have, um, uh, I went to undergraduate school in a very small school in New York City called the Cooper Union, which no Edward Sturm: I know it.

I'm from New York. I know it Caleb Ulku: oh, right. Oh, yeah, you said you were, you were, uh, Edward Sturm: Yeah. Caleb Ulku: so yeah, that makes sense.

It's Edward Sturm: Born in Manhattan, raised in Brooklyn, but yeah. Yeah, yeah Caleb Ulku: Oh, born in Manhattan, Edward Sturm: Mm-hmm. Caleb Ulku: Well, uh, Edward Sturm: Close enough Caleb Ulku: the… A lot of people born in Manhattan would make that correction, so I won't hold it against you. But yeah, so, um, I graduated from that school, uh, with chemical engineering, uh, hired onto ExxonMobil, and I moved to Houston, like everyone who hires onto the oil industry.

Uh, I worked for them for 10 years, and my wife and I, had a two-year-old son, and we really just never saw him. We both had jobs that required crazy commutes, uh, long hours. So we actually hired a driving nanny because we couldn't make it to daycare be- or daycare closed. So we hired a driving nanny.

She'd go pick our son up, bring him home, put him to bed. So we basically hung out with him on the weekends, and we did this for way too long. It just felt like normal life, right? And then, blessing in disguise, I broke my hip.

Um, so when I broke my hip, I had to… I was, like, non-weight-bearing during the recovery, and ExxonMobil at the time was building this enormous new campus north of Houston. But, like, the elevators weren't done, the escalators weren't done, so they were like, "You can't come to this campus without being able to walk." Uh, so they said, "You have to work from home." And I worked from home for, um, about a month and a half, and I think the day I went back to work, uh, it took less than a week before I put my two weeks' Edward Sturm: Wow Caleb Ulku: in, uh, because working from home, you know, I was learning to cook.

I was hanging out with my son. I was with my family, and it was just like, what were we doing with the crazy commute and everything? So, put in my two weeks, and we moved to, uh, a town just outside New Orleans, which is where my wife's family is from. Everyone ends up where their wife's family is from.

And yeah, we raised our - My, my son is now 14, so that was a minute Edward Sturm: Wow. And so, and so when did you decide to get into SEO? Caleb Ulku: So when we first moved there, um, plan A was, "Hey, let me go find another job." So I, uh, started to look for other jobs.

It, it's a very service-oriented economy, so I wasn't really interested in that type of a job, and finally just decided, "You know what? Let me go ahead and start my own business." And I didn't know what SEO stood for at the time, so that wasn't my, you know, plan A. But eventually, you know, I was going to chamber of commerce meetings and things like that and trying to explain what my business did, and I would get business owners not understanding and say, "Uh, but can you help people find me online?"

And eventually I just said, "Yeah, sure, I can do that," and started to learn. And, you know, I had this… of my very first clients, um, that I got in SEO, again, had no idea what I was doing. This was a very long time ago. So I did all these SEO things that I'd read about, that I'd heard about, and ended up getting their entire domain de-indexed, which, you know, pretty bad.

Edward Sturm: Yeah, just a little bit. Caleb Ulku: so they Edward Sturm: Kind of bad, yeah. Caleb Ulku: Yeah, yeah, pretty bad. Yeah, that's Edward Sturm: Wait, what did you do, what did you… What, what do you think did it?

Caleb Ulku: Oh, I was doing very shady things with link building. Like, I was doing all the shady things. Yeah. Um, that, uh, and, and I, I guess Google didn't like all the shady things.

Um, so they fired me, which again, right call. That was the right thing for them to do. Um, I did get their domain re-indexed, which was awesome, but they Edward Sturm: How did you get it reindexed? Caleb Ulku: that's fine.

Uh, well, I started, uh, basically disavowing all of the shady things I'd been doing. Uh, and then, uh, the domain did g- end up getting re-indexed after I did that. So that was awesome. at that point, I started buying SEO courses, trying to… 'Cause like, hey, the stuff I'm learning on, you know, reading on blogs and watching on YouTube, you know, obviously I want something a little bit more structured, something that I know works.

So yeah, if you were selling an SEO course in like 2015, I probably bought it. Like, I bought every course I Edward Sturm: Which one was the best? Caleb Ulku: um, a- and started going Edward Sturm: W- which, which, which one, which one was the best in 2015 that you got? Caleb Ulku: Ho- honestly, I think, uh, Matt Diggity's was really good.

Uh, I think it was called The Affiliate Lab. It's now defunct, but that was a solid one. And then to try to fix the shadiness, I bought a course from Matthew Woodward about how to build PBNs, which, you know, we don't really use that Edward Sturm: To try to fix the shadiness, I started learning about how to make PBNs. Caleb Ulku: Well, how to do it the Edward Sturm: Okay Caleb Ulku: I mean, the right way from 10 Edward Sturm: Yeah Caleb Ulku: isn't what we would do today.

Yeah. Edward Sturm: So you got, so you got the courses and then what? Caleb Ulku: So got the courses, uh, and honestly, like I just started applying to jobs on Upwork. Um, you know, I figured out that not good at a sales call where I had to take - talk to a business owner and convince them that SEO would be a good way to get more leads for them.

But if I were on the phone with somebody who already wanted SEO, I was really good at convincing them that I knew how to do SEO, and I could w- make it work for them. So on Upwork, everyone was posting jobs on Upwork already deciding they wanted SEO, they're just looking for the right provider. You know, I, I, um, y- I'm sure you've seen the, like, Wolf of Wall Street scene where Leo DiCaprio is like, Edward Sturm: Yeah Caleb Ulku: Well, I'm the person who's just like, "Oh, look at how well it writes, and look at how nice the clicking is."

I mean, that's how I talk. Uh, so that's how I was able to sell SEO because, you know, they'd already said they want the pen, so I just had to sell them on the features of it. Um, we got a few huge brands, uh, which is great. We got, uh, like Skillshare, Adobe, Stanley Black & Decker from, uh, Upwork.

Speaker: This method of marketing is so effective I had to make sure it wasn't against Google's rules before I kept doing it. It's a form of SEO I call compact keywords. Whereas most SEO focuses on putting up articles to answer questions how, what, when, compact keywords focuses on putting up dozens of pages that sell to searchers who are actually looking to buy. These pages rank on Google and convert so much better than normal that when I discovered this years ago, I couldn't believe this was allowed.

It's less work too. The average compact keywords page is only 415 words. Compact Keywords is a 13-hour deep course on getting sales with SEO. A customer recently said, "Each lesson is dense with information.

You're giving years worth of experience boiled down into 15 to 30-minute lessons with no filler or fluff. I feel like I'm gaining a new superpower." Compact Keywords is about setting up an SEO funnel that brings you sales for years and years and years. It works with AI, it's less work than traditional SEO, and it makes way more money.

You can get it now at compactkeywords.com. Back to the podcast. Edward Sturm: Wow.

Caleb Ulku: the Edward Sturm: You got that from, Caleb Ulku: those Edward Sturm: Upwork? Caleb Ulku: allowed us Edward Sturm: You got that from Upwork? Caleb Ulku: All of them from Edward Sturm: Wow. Wow Caleb Ulku: yeah.

From Upwork. Yeah. Now, budget that we had from all of those allowed us to start doing split testing, right? Like, you know, hey, you - person ABC says to do this, and then this other person says to do this.

Well, let's do each of them 10 times and see which one actually works and which one drives rank. So we did that for a couple of years and started to fine-tune, uh, our process for how we did SEO, how we drove results. All non-local, uh, because everything was from Upwork. So, you know, not a lot of like local plumbers are posting on Upwork.

but yeah, we got - It took a couple of years on Upwork to get to seven figures. Um, since then, I don't do much on Upwork anymore. It's gotten really difficult, uh, even with an established profile. But yeah, you know, uh, I still know a lot of people who have success on the platform.

It's just harder than it was, you know, five, 10 years ago. Edward Sturm: Something that's working in SEO right now that a lot of SEOs still aren't doing Caleb Ulku: Yeah. So it's really And a lot of this - So the way that we did local or non-local SEO, right, is we would have the money page, the page we were trying to rank, and then we just built all the supporting content for that money page. Well, in early 2023, uh, we pivoted completely away from non-local to local SEO.

Uh, we did that, I'm sure, Edward, you, you know the timing. ChatGPT came out in November of '22. So, you know, I don't know. I've, I've talked to a couple of other SEOs, and they had a similar realization because, you know, by the time 2022 had come around, uh, we'd watched Google keep people on the SERPs for years at this point.

One of my favorite examples, uh, is the search term mortgage calculator, right? Where you type mortgage calculator, and then there'd be a bunch of results. You'd click on them, and they had mortgage calculators. Well, Google decided that they would put a mortgage calculator on the SERP.

for users, horrible for the businesses that had spent all the money to rank for mortgage calculator because suddenly their traffic went away. So in the first couple of conversations with ChatGPT 3.5, it's just like, man, Google is going to be able to take this information, this technology, and keep people on the SERP forever. So we pivoted very hard away from non-local informational type searches and said, "We're gonna do local SEO."

Right? However advanced AI is, it's still a couple of years from being able to replace people's water heaters. So when we started doing local SEO, we did it in the same way that we did non-local. So, uh, the money page for local SEO, we started treating to be your Google Business profile.

So then we thought, okay, everything on the website is supporting content to establish more trust, more authority for the Google Business profile. we started testing this concept where whatever your categories and services on your GBP, we would build the website to exactly mirror the GBP categories and services. So now we call that the Core 30 because it - for most local businesses, it ends up being around 30 pages. And we have found that works so well because when your website matches the GBP exactly, uh, it builds so much trust, not just with Google, but with the AI systems as well.

So I think that's the simplest way for - Uh, and the local business owner can do a lot of that themselves, uh, you know, by just AI to write the content. Uh, and the local business owner knows obviously enough about that business to be able to get the AI to write something unique and interesting about each of those categories and services and structure the website to match the GBP. I mean, for 80% of markets, that alone will significantly increase rank and the number of calls you're getting, even if you don't do anything else.

So I think that's like the low-hanging fruit. And when somebody comes to me and says, "Hey, will you do SEO for us?" I mean, that's like one, build out the Core 30 And, you know, month one results almost always just look fabulous because it's such a simple thing and, you know, the machines don't know how to interpret messy signals the way that humans do. So we're just trying to create the same signal that makes sense, your website to your GBP.

That was a really long answer to that Edward Sturm: No, that's awesome. Could you give, uh, could you give like an example of what that would look like? Caleb Ulku: Right. So, uh, I, I do, yeah.

So your homepage is going to be based on the GBP primary category, right? So, Edward Sturm: Yeah, yeah. Plumbing is good Caleb Ulku: running ads for local businesses, we ran ads for plumbers, so we've had a lot of plumbing clients. So my example always goes to plumbers.

So say, say plumber. So for a plumber, your primary category is clearly going to be plumber, then your GBP landing page, which is usually your homepage, you're going to base that around, like, plumber city name, right? Like plumber Houston. And then on your homepage, you're going to talk about the secondary categories and then a couple of what we call, like, core services, services that you really, really care about.

So for a plumber, your secondary categories might be drainage service, bathroom remodeling, and then the core services that you really care about, most every plumber I've ever talked to, the services that they want more calls for are water heater replacement and main drain line replacement. So we'll do an H2 for each of the secondary categories and then an H2 for the core services, and then a paragraph, you know, maybe 50, 70 words underneath that H2, and then we link to a page where that page is about bathroom remodeling or where that page is about water heater replacement.

So that's gonna be the H1, remodeling Houston, and under there, we're going to talk about all the services that live under that secondary category on your GBP, right? Your - When you set up your GBP services, you set up your categories, and then you set up services under each category. So the services under bathroom remodeler, we talk about on that page, and then each one of those links to a page about that specific service. So maybe, you know, you have faucet replacement.

So that's a service under bathroom remodeler. You'll create a service page for faucet replacement Houston, and then the bathroom remodeler page links down to that service page. So that creates this exact same hierarchy structure that your GBP has Edward Sturm: do with businesses that have multiple locations or multiple cities? Caleb Ulku: Yeah.

So for that, uh, there are - that is one of the two reasons why your GBP landing page might not be your homepage. So, uh, essentially every location should have its own GBP landing page. Uh, but if you're taking on a new client or you have an existing business and you've set it up so that all of your GBPs point to your homepage, we don't move them to an internal page until we run the rank map, right? And I'm sure you know this, Edward, but if you - if your landing page is your homepage and then you move it to an internal page, your rank is going to plummet.

So we never do that if they're ranking, right? So we always look at the rank map, and if the rank is horrible, uh, or they're relatively new, then every single GBP gets its own GBP landing page. It's basically just a copy of the homepage. Um, and we build out the Core 30 for every single location.

So if you have five locations, that becomes the Core 150. But to keep in mind that I love multi-location businesses because as you're building all of this out, the domain is getting so much topical relevance and authority that, uh, you'll hit a point when you're adding, you know, location number four, lo- location number five, depends on the competitiveness, et cetera, et cetera. But eventually you'll hit a point where you'll create a GBP landing page, you'll point the GBP to it, and then you'll be ranking reasonably well just because the domain has so much authority and relevance for plumbing, you don't need the Core 30 anymore for this new location.

And then we just go straight into producing geographical relevance content Edward Sturm: Okay. What's the biggest, when it comes to Google Business Profile, what, what, what's the biggest mistakes that you see businesses making? Caleb Ulku: Yeah. So, uh, categories and services, right?

Um, the most common businesses we look at, they'll have one category, maybe two. Google allows up to 10. There's this myth floating around that if you add multiple secondary categories, you're diluting the ranking power for your primary category. We've tested this over and over and over again.

We've never seen - We've only seen positive results from adding more secondary categories. Uh, and then the other mistake is your services, right? Uh, it, it - and the mistakes we see there are everything from not enough services to keyword stuffing, right? So we know that Google works based on entities.

It doesn't use keywords anymore. So the days where if you had a personal injury lawyer, you needed to do car accident lawyer, car accident attorney, uh, car wreck attorney, car wreck lawyer, right? So now we have four services that describe the exact same thing. Google knows those are the same thing, and keyword stuffing your services, I don't know that it ever worked, but it certainly doesn't work today.

So that's where we spend most of our effort for a new client, is we're going to come in there, take a look at their categories and services, and make sure those are correct and optimized before we start building out the Core 30. And it's not hard, right? Like, but it's just something that most businesses don't do. Um, then the other aspect is just making sure every single box is filled out.

The most common one that we don't see filled out is, uh, holiday hours, because Google doesn't let you put in holiday hours forever. You can only put in the holiday hours for like the next two or three holidays. So you need to log into your GBP every few months and update your holiday hours for whatever holidays are coming up. So I mean, those are the, the biggest mistakes that we see, but by far it's get your categories and services correct.

Edward Sturm: How are you structuring the URLs when you're, when you're, yeah, when you have these, like, individual pages, when you have multiple locations? What does a website look like in terms of URL structure and, and, and internal linking? Yeah, internal linking and hub pages and all of this stuff Caleb Ulku: Yeah. So the URL structure, like I care very little about it.

So whatever my developer finds easiest to work with to keep track of all the pages and everything, that's fine, right? 'Cause Google doesn't build its image of your website based on your URL structure. It builds what your website looks like based on your internal linking structure. So I care a lot about internal linking and very little at all Edward Sturm: Do you think, do you think that, that Google looks at URL slugs or parent or parent folders Caleb Ulku: I don't think Google pays attention to parent folders.

Uh, we will almost always make sure the slug is the service entity that we're writing for, and that's just because it's pretty common that a lot of your links coming in will be the naked URL. So having the keyword in the slug means you can put your keyword in the anchor text without being shady or Edward Sturm: So wait, so, so, okay, so you would do, so you would do, like, forward slash, um… Okay, give me, give me a service. Give me a service Caleb Ulku: Sure. So it would be like, you know, domain.

com/waterheaterreplacementhouston Edward Sturm: Okay, so you would do, you would do domain.com/houston/waterreplacementhouston Caleb Ulku: No. Well, if, if, if, if, if you have subfolders, then we would just do water heater replacement in Edward Sturm: Yeah, that's what I, that's what I meant. Yeah.

Okay Caleb Ulku: Yeah, but normally we don't do Edward Sturm: Okay Caleb Ulku: unless the developer or client Edward Sturm: Yeah, 'cause, 'cause so, so if you have multiple locations, then you're not, you're not separating the URLs by location. So d- does that mean that you don't… Caleb Ulku: usually Edward Sturm: do you have, do you still have a hub page for the location? Do you have /houston, /anchorage, /brooklyn Caleb Ulku: We might have like this /Anchorage serving as the GBP landing page for the Anchorage location.

But in terms of like all those subfolders and everything, I mean, like I'm not gonna go create that. If it already exists, I'm not going to undo it either. I don't really care very much either way Edward Sturm: And that's because the most important… You don't care because… So, so what are the key, the key elements to putting the keyword in for you? Caleb Ulku: So the keyword, ideally it should be in the URL slug, uh, but that's the least important of the really important places to put it.

The m- really important places to put it are the title tag, the H1, and then the first paragraph. We also wanna mention it very early on in the first paragraph. Uh, a very common mistake I see among so many local business websites is their first paragraph of text is a history lesson about their business, right? And I've, I've, I make the joke, right?

No one is typing plumber Houston into Google because they're interested in reading history lessons about the local plumbers, right? So that first paragraph, such a missed opportunity if you're not talking to the person who just typed plumber Houston into Google, and w- you know, why you're the plumber that they should call, right? Like, nobody cares about what year you were founded and, you know, why your grandfather went into plumbing and why he immigrated from the Old World.

Like, nobody cares about that, but that's where people start talking about on local business websites. But yeah, so title tag, most important, H1, and then in the first paragraph, we'll start talking about it as we talk to the searcher who landed on the page. Also important, right? Because as… You know, there was an interview with Sundar Pichai, the CEO of Alphabet, that came out, I guess, probably about a month or so ago now, and one of the things he talked about is that he sees Google evolving to become a home for agentic So I think what that means, you go to google.

com and y- instead of a search box, there's more like a prompt box, and you type in what you're looking there, and then Google sends a bunch of agents out to find the information and bring it back to you. So we're - That is reaching a point where we're no longer writing websites for people, we're w- writing websites for the AI agents to summarize those websites for actual people. So again, very important when you're writing website content to make sure you're writing it for someone who is searching for what you're looking for, whether that be the agents or actual Edward Sturm: Has that changed the way that you do SEO?

Caleb Ulku: Not significantly, um, because the, you know, the right way to do SEO has always been to structure the website in such a way that Google can understand it, and we always wrote for the algorithm, and then, okay, let's also make sure we're writing for conversion for the people. you know, if you have the best website ever that's directed to people and you're ignoring the algorithm, it doesn't matter because no one will ever see it, right? Uh, assuming you're not running ads or something like that.

Um, so that like, make sure it's structured properly, make sure it's written for the algorithm. That stays. Um, and, you know, whether a person is reading it or an AI is reading it, it doesn't really matter as long as you have the information on there. So the biggest difference, right, is like a messy website or a messy structure, people have the patience to figure that out, whereas an AI doesn't.

And on the flip side of it, the AI can read your entire website and pick out the most important parts of it and give it back to a user, whereas most users will spend, you know, 20 seconds on your site before they decide whether to call you or not. So that's where we get into like attribute match- matching, uh, which is something that we weren't super focused on before AIs, but now we know, right? So if somebody searches for like lead main drain line replacement, the AI, whether it's ChatGPT or Claude or Google's own Gemini, the AI is going to want to see that you're talking about lead main drain line replacements.

Well, if you don't talk about that, if you just talk about main drain line replacements without using the word lead, you might not get recommended. Um, so that's, uh, that attribute matching is something that's relatively new, uh, when we're looking at content, building out a website for, for local Edward Sturm: Yeah, I always like, I like to look at other pages that are, that are ranking for keywords that I'm targeting, and I see the language that they use. Like, basically, like, for me, I'm a, such a good SEO, and I think you're, y- you, I'm sure you're the same, where you can go into any niche and within a, uh, like a week or two understand the niche well enough that you can actually write pretty expert content in it by just, by taking notes of the, the types of language that appears throughout pages that are ranking for your target keywords.

And then you make a list of these. You, you have your target keyword. You know where the target keyword goes. You have a hierarchy of where the keywords are going, what's most important, what's least important, and then you can build your pages like that.

What do you think about, um, about AI images, both on, uh, the website and in a Google Business Profile? Caleb Ulku: Yeah, so the - what you just talked about, we call that, uh, we call that the open book test, right? So if you wanna rank for whatever service entity you wanna rank for, the best way to get started is to type that into Google or type that into ChatGPT, and it will tell you, "Hey, here are my favorite pages about that topic in the world." Uh, and then look at what they did and just do a little bit better, right?

Like add a little bit more information that they don't have. Uh, and there are a variety of ways that we, uh, focus on to do that. But yeah, exactly. The, the point is to write local SEO content as the local SEO business - or not, as the local business owner expert.

Edward Sturm: Oh, actually, another great question is, is are you making, uh, do, do you try to make all the pages unique to the locations? Are you giving extra information based on location? Or like how do you think about, how do you think about, uh, about doing that versus something like doorway pages? Caleb Ulku: Yeah.

So every page we write for local SEO is hyper-focused on that location, right? So even, you know, I talked about lead main drain line versus main drain line. So if we had a plumber client in the suburbs of Boston, talking about lead main drain line replacements would be very important because that's a common issue in the suburbs of Boston. But for a client in Southern California, we would never talk about that because they don't have lead main drain lines in California.

The hoses are a lot newer and the, you know, they use plastic main drain lines. Um, so those, you know, local differences are so important for local SEO. Um, and we, we actually saw an issue, I'm sure you remember, is, is like in early May where Google started mass deindexing AI slop, which, you know, fabulous news. So when we build local content, uh, we, we build it for - There's four different types of local content that we write.

Uh, so we write category content, we write service page content, we write topical relevance content, and then we write geographical relevance. of those four types, when Google started mass deindexing, three of them, no problem. But the geo content, we started to see some deindexing. when I say geo content, what we'll do is we'll run a local rank map for the target service entity, and everywhere that you're not ranked in the top three, uh, especially if you're ranked like four, five, or six, we're going to write hyper-local content about landmarks in that specific area, right?

So, again, using Plumber Houston as an example, we'll write content, you know, Plumber in River Oaks, Houston, Plumber near Montrose, Houston, uh, Plumber near, uh, 610 West and I-10, Houston. So we're finding these hyper-local landmarks like lakes, zoos, city halls, colleges, whatever, uh, and we'll write content about a plumber in that region. Well, for a v- a very competitive space like Houston, we might end up with a couple dozen of those, and it's challenging to make plumber Montrose be a significantly different article than plumber River Oaks 'cause they're so close to each other?

But if you know Houston, uh, Montrose, uh, is a very different neighborhood than River Oaks, right? River Oaks is very fancy, estate s- sized homes. So you could write about the differences for being a plumber in River Oaks 'cause they are different than Montrose. But how do you do this at scale, right?

what we started doing is just an insane amount of research. So, uh, we would pull, like, US Census Bureau data, we'd pull Reddit data, we'd pull all the data and information that we could find. We started hooking into the CRM so that we could pull conversations and leverage all of this data to write content about, "Hey, why is being a plumber in River Oaks different than being a plumber in Montrose?" Uh, and we did that, went back through the de-indexed content, and added all of this additional research, and we got a lot of it to be Edward Sturm: Yo.

Caleb Ulku: now it actually is Edward Sturm: Great Caleb Ulku: additive, right? We're not, we're not writing the average of the content that already exists with redundant information. We're - Now we're adding new information, right? River Oaks, uh, plumber is, you know, you have to worry about walking 200 feet back to your truck because the front lawn is so big.

Um, and yeah, Montrose is, you know, you might have a fiveplex or something that's two stories tall in Montrose. Uh, so yeah Edward Sturm: Yeah, that's awesome. That is awesome. And, and so how about AI images?

'Cause you see this, you see this from lo- all, all sorts of businesses trying to put in AI images, um, but you also see it in local. So what do you think about, yeah, both on a Google business profile and on the website Caleb Ulku: So obviously, we always prefer real images from the client, but clients are stunningly bad at providing real images. It's just like, take a bunch of photos Edward Sturm: Yeah. Caleb Ulku: and text them to me.

That's all you need to do. Uh, but so bad at it. So AI images are better than no images, but I would much rather have real images from the, the client. Edward Sturm: And, and do you… Oh, go, go on Caleb Ulku: Go ahead.

No, I was gonna say, if we are using AI images, we're going to make sure to add metadata to those AI images so that they look, feel less like AI images, if that makes sense Edward Sturm: you talk about that? Caleb Ulku: Yeah, so I mean, we'll put the geo coord… So if you take a, a, a photo with your phone, Edward Sturm: Exif d- Exif data. Caleb Ulku: to be added to that. Edward Sturm: Exif data Caleb Ulku: Yeah, Edward Sturm: I made a, I made an episode about that, but r- do you know Jay Conley?

Caleb Ulku: I don't know Jay Edward Sturm: Oh, yeah, he is, he is, uh, uh, he's, uh, very deep into local as well. He did a crazy study on EXIF data. Uh, but yeah, the study, uh, anyway, his study said that EXIF data doesn't actually make a difference. Um, but, uh, it's - I think it's a v- I think it's very interesting that you, that to try to… I never considered trying to put it on AI images to, to make the AI images look more convincing Caleb Ulku: Yeah.

Well, I mean, you want it to feel like a real image, Edward Sturm: Yeah. Do you, do you ever get, um, do you ever try to get clients to make videos? To record videos? Caleb Ulku: Yeah, videos Edward Sturm: Videos are amazing.

Caleb Ulku: Uh, especially Edward Sturm: bet your YouTube channel is ranking for a lot. I said I bet your YouTube channel is ranking for a lot Caleb Ulku: It - One of my, uh, largest sources of traffic on YouTube is Google Search. Yeah. And what's really… Okay.

So, and you know this, Edward, back, back, back in March, Ask Maps was rolled out where, you know, you go to maps.google and you talk straight to Gemini, obviously no one is doing that right now. It's like, like five people are. But, uh, I take it to be a signal for where Google wants to take the Map Pack algorithm in the next, who knows, six to 18 months.

'Cause right now, the Map Pack algorithm hasn't meaningfully changed in over 10 years. Uh, but Google adding Gemini to the Map Pack algorithm is obviously a massive shift, and Gemini is the only frontier model that was natively trained on video content or - because obviously Google owns YouTube. Um, so Gemini loves video content, so if you can feed Gemini video content, really powerful for a local business trying to get Gemini to recommend them, 'cause so few local businesses do video content.

So 100%, that's something that we tr- we try… When we're publishing local content, uh, we try to do for between a third to a half of the articles that we publish. Edward Sturm: like that a lot. Yeah Caleb Ulku: if, if the, if the local business owner can't or won't or doesn't wanna do it, no problem. Generally, we'll just use AI to generate those videos and then put those on YouTube.

I mean, no one watches them. They're not for people to watch. Just like, you know, my first couple dozen YouTube videos, no one watched them except my video editor. Um, but it, it's for Gemini.

It's not for people to actually Edward Sturm: then will you put, since they're not for people to see them, will you put those on the website or no? Caleb Ulku: Yeah, absolutely. We'll, uh, upload the video to YouTube, uh, and then we'll embed that video link onto the actual Edward Sturm: wow. Okay.

Caleb Ulku: time Edward Sturm: And what are you using, um, to, for the AI? I- is - Are you doing AI avatars? Caleb Ulku: No, uh, we're just using, uh, it's a service Edward Sturm: Mm-hmm. Caleb Ulku: and, you know, we can feed Pictory like a two or three-minute script and it does a pretty good job.

Like, I mean, they're not, you know, they're not gonna win any awards, but they, they're not embarrassing if somebody Edward Sturm: Yeah, that's cool. W- something else, um, I don't know if you've tried… This is probably still hard, but I don't know if you've tried asking, like, if, uh, if a business owner doesn't wanna take video, maybe you could get them to just record something. You give them a transcript, and you're like, "Look, just read this as a voice message. We'll do the rest."

Caleb Ulku: Yeah, and then use like an AI Edward Sturm: Yeah, then you could do, you could do, you could do that or you could just have B-roll. Yeah Caleb Ulku: Yeah, it's a good idea. We haven't tried something like Edward Sturm: Yeah. Caleb Ulku: it Edward Sturm: Uh, are there, are there other, are there local SEO tactics that still work despite everybody saying that they're dead that, that you hear a lot?

Caleb Ulku: I mean, the way we do local SEO is incredibly simple, right? We're primarily just focused on writing quality content that is informationally additive. Uh, we source quality links. Like, you know, we tell all of our clients to join the local chambers of commerce.

How many they join is based on how competitive their space is. Uh, to rank a personal injury lawyer in New Orleans, we once joined, like, 17 chambers of commerce. We joined every chamber of commerce within 80 miles, um, which was awesome, uh, 'cause it worked. Uh, we'll look for sponsorship opportunities, you know, like l- youth sports leagues.

For hundred bucks, you get a link from their website, usually very powerful local websites, like local charities. Again, the personal injury lawyer in New Orleans, we sponsored the, uh, the Bayou Bash, which Edward Sturm: Wow Caleb Ulku: festival. They were looking for sponsors. It was a couple hundred bucks.

Awesome. Um, so yeah, like, that's the type of stuff that we're trying to do in Edward Sturm: How do you find these, these Caleb Ulku: Every Edward Sturm: opportunities? How do you find them? How do you find… Yeah Caleb Ulku: them?

Uh, we have a prompt that we give to Gemini. Um, Gemini is so much better at finding them than Claude or ChatGPT. I don't know why. Uh, but if you give the prompt to Claude or ChatGPT, it, it's kind of middling, but Gemini does a phenomenal job of looking for these local organizations that are looking for sponsor, sponsors.

And, you know, so we ran the prompt. One of our clients, they do, uh, house cleaning in Austin, and we ran the prompt, UT Austin was doing a TEDx Talk, and they were looking for sponsors, and the minimum sponsor was, like, 250 bucks. And for $250, we got a link from utaustin.edu domain to our house cleaner.

Uh, like, I, I would've paid thousands for that. Like, their rank went up, like, three or four numbers overnight just Edward Sturm: That's what's up Caleb Ulku: And for t- yeah, for $250, plus the business owner got to actually go to the TEDx Talk because they were one of Edward Sturm: That's applicable to any business. That's cr- that's crazy Caleb Ulku: 100%, yeah, 100%. I couldn't b- like, I was just, I was, you know, Gem and I found it, and I was just like, "Oh, a TEDx sponsor.

They're gonna want thousands." But it was like, "$250? Edward Sturm: So, okay, so walk, you know what? L- like, walk, let's, let's, let's switch it up.

Let's say you have a SaaS business owner and they wanna build backlinks. Walk, tell them what to do to find backlinks Caleb Ulku: So SaaS is different, right? SaaS is non-local, and we haven't done non-local Edward Sturm: Yeah, but the, but you're finding these opportunities that are still applicable Caleb Ulku: Yeah, you're right. Like SaaS getting the backlink from UT Austin, I, I, I'd happily do that for a Edward Sturm: For any business, for, for e-commerce too Caleb Ulku: Absolutely.

Yeah. So I mean, we'd probably run - So, so the, the prompt is designed very specifically to look for local sponsorship stuff. I'd probably just take the same one and delete a bunch of the local references and just go find me sponsorship opportunities. Um, I haven't tried something like a chamber of commerce membership for non-local.

Uh, it's still a powerful link, but for local it makes sense why it's so powerful. Um, I don't know if I would - Like, it might be interesting, certainly worth testing if, you know, go join 50 chambers of commerce for your SaaS and, and see if your rank gets any better. I mean, that, like 50 of them, you know, most chambers are between $100 and $300 to join depending on the size of the city. So I mean, you're talking it's not that much money, five grand.

And yeah, that's a lot, but for a SaaS who's like trying to make a go of ranking nationally, that's a relatively split testable, uh, budget to be thinking about. I, I would, would try it Edward Sturm: Where do most agencies waste time in local SEO campaigns? Caleb Ulku: Uh, writing writing blog posts that don't provide topical relevance, yeah? Um, Edward Sturm: talk about, yeah, I, I w- I want, also wanted to ask you your thoughts on just, like, d- blog SEO and informational keywords and top, top-of-funnel searches.

But yeah, so okay. Th- they, they… Most SEO agencies, local SEO agencies are wasting their time on doing blog SEO that doesn't provide topical relevance. So, like, what d- what does that look like? Caleb Ulku: Yeah.

So a great example, again, plumbing, is if you write a blog post, "Top five ways to winterize your water heater," right? That is not providing topical relevance for a plumber who wants to provide plumbing services. So had, uh, a client came to us… Well, a prospective client came to us, and they wanted local SEO, uh, a LASIK eye surgeon in downtown Chicago. And they were getting 80,000 hits a month from Google Search, right?

They were paying for SEO, and the agency they were paying was, you know, objectively doing great. 80,000 is a lot of traffic to come from Google Search. But when you went and looked, uh, in the search console to see where they're getting their traffic, it was stuff like, you know, "Top five exercises to keep your eyes healthy," or, "Best foods to eat to keep your eyes healthy." It was stuff like that.

And their rank map for, like, LASIK eye surgery Chicago looked pretty ugly they were focused on these informational non-local searches. So when we took over, uh, their homepage had used the word Chicago once in the footer Edward Sturm: That's crazy. Caleb Ulku: right? Edward Sturm: Yeah Caleb Ulku: Uh, so we obviously rewrote a lot of their homepage to focus on Chicago.

We started focusing hard on building geographical relevance, so we'd write all this content about these different neighborhoods. You know, I talked about the Montrose, uh, River Oaks example. Very similar except, you know, Wrigleyville and near the Willis Tower, et cetera, et cetera. Um, and their monthly traffic has gone way, way down.

They're now at, like, 30 to 40,000 a month, but it doesn't matter because their rank map is green across all of Chicago. So they're actually getting calls because they're showing up for local search. So that's the challenge, right? If you're a local business, you don't care how much traffic you're getting.

You care how many, how many local people are finding Edward Sturm: I could talk more about that, that case study on, on what, what was done Caleb Ulku: I mean, that, that I like - l- so w- yeah, we redid the homepage, like I said, to focus on local, and then we just started building geographical relevance. So when we're looking at a rank map, there are, um, a couple of things that we're looking at. The first one is, and, and it's percent top three is the metric that we care most about, right?

So, uh, 'cause if you're in position four, that's the first loser. You're completely invisible. No one ever clicks more businesses. So what percent of this rank map are we in the top three?

So the first question is, you know, is this rank map the right size? So we look at the top competitors. Uh, usually it's not the client 'cause, you know, this is the first rank map we've ever run. Uh, so we'll look at the top competitors to see what their percent top three is, and our goal is to see something like 60, 70 to 80, 90%, right?

If it's much below 60%, then we know that rank map is too bi- too big, and we need to rerun a l- a smaller rank map, right? Uh, our rule of thumb is we can do at least as well as an existing competitor, but we're never going to tell a prospective client that we can do better than has ever been demonstrated in their area, right? So we're not gonna run this giant rank map where the top player is 30% top three and show the competitor and say, "We can dominate this space." Uh, and depending on the competitiveness, the - how large the city is, the rank map might be huge, and it might be really small.

Uh, like we'll run rank maps everything from, you know, five, 600 square miles all the way down to like 30 square miles, depending on how competitive the space is. Always looking for the top players to be in that 60, 70 to 90% top three. obviously, if the top players are 100%, then the rank map is too small. We need to rerun it larger because they're ranking outside the, the rank map.

So, uh, we need to zoom out and, and run it again. so that's the first step to make sure that the rank map is the Edward Sturm: What, what software are you using? Caleb Ulku: uh, usually we use - Not usually. We use this software called LeadSnap, uh, to run the rank maps.

Uh, there's a bunch of other rank map tools, but we started using LeadSnap because they have the, uh, really great API connection with citations. A lot of other tools do now also, but at the time, they were the only one that did, so that's why we went with them. Edward Sturm: Okay, so yeah. So next part Caleb Ulku: So next part.

So, uh, then we're looking to try to cross, we call it the topical relevance threshold. the topical relevance is convincing Google that your business entity actually does that service entity. We're not trying to establish geographical relevance just yet. We're just trying to say, "Yes, Google, we are a plumber in Houston."

That's it. we'll, um - we, we, we say that when our client's top 3% is roughly half of the market leaders, we've established that trust with Google, right? So, you know, player's in the 80% range. So once our client is 40% top three, cool, topical relevance threshold met.

And we build topical relevance. The first step for topical relevance is, of course, the Core 30. And if we're in the competitive enough space where the Core 30 is not enough, we've built out the full Core 30, and we still haven't hit that topical relevance, then we start building supporting content. But we build supporting content that supports their service entity, right?

So we'll - it, it would be something more like what, what to think about when you're hiring a plumber in Houston, not top five ways to prepare your water heater for summer, right? So it's always focused on that service entity and providing relevant content for the service entity. And keep in mind, right, we might have four or five different rank maps depending on the target services. So for a plumber, we're always gonna have three: plumber, uh, water heater replacement, and main drain line replacement.

So we're doing this process three times for each of those three main services we care about. not gonna run rank maps for, like, unclogging a toilet because most plumbers don't want that call, so, so we won't track it. And most of the things like that, if you rank for plumber, you're gonna rank for a lot of the stuff underneath it anyway. once we hit that topical relevance threshold, then we start building geographical relevance, and that's where, you know, let me find the four, fives, and sixes, uh, where they're close to the top three, and then I'm going to build hyper, hyper local content about a local landmark around that dot.

Uh, because - And it - Obviously, it's important to use Google Maps for that 'cause you wanna make sure Google recognizes that landmark. And, uh, w- now we do it automatically with, like, a Google Places API call. Uh, for the same theory, we wanna make sure Google recognizes the landmark, and if we're calling it with the Google Places API, then we know Google recognizes it because Google gave us that landmark name. and then we'll just write about, yeah, replacing main drain lines around that lake in Medford, Massachusetts Edward Sturm: doing internal linking with these posts?

Caleb Ulku: So, uh, it depends on the competitiveness. For m- larger, like more competitive spaces where we might have dozens of supporting content, we're going to create like a supporting content hub. and then we'll link to that hub either from the GBP landing page, maybe the footer, and then that hub will link out to all of the supporting content pages, and the supporting content links back to the service page we're trying to support, whether it be the water heater replacement service page or the plumber primary category page.

the reason for that is, you know, on the GBP landing page, if, if we have a couple dozen supporting content pages, we don't want, you know, 25 links from the GBP landing page. So that's why we put that hub in Edward Sturm: I want to talk about AI. How has AI changed your SEO strategy over the last 12 months? Caleb Ulku: So, so far it hasn't done a lot to actually change our strategy.

Uh, it has significantly reduced costs. Uh, like our cost to produce an article, and when I say produce an article, like it's obviously to write all the content, it's to place it on the website, do the internal linking, source an external link. It's do all of that stuff. Uh, our cost to do that is like roughly 80% lower than it was in mid-2022.

Um, and AI has allowed us to write higher quality content, right? 2022, we couldn't legitimately spend a couple of hours doing research for every single article we wrote, but now we can do the equivalent of hours of research with the right data calls and just have AI assemble everything and pull all the research and assemble an outline, and then start writing. And we get a level of quality that would've taken hours, uh, in 2022. Um, so, so far, uh, that's been the biggest impact is that we can write higher quality content at scale than what we could at 2022 at a lower cost.

we are preparing for this world where AI agents are primarily the ones reading everything. So I'm in the process of working through adding a fifth content type for local SEO. Uh, I talked - You know, right now we have four, and we've had four for 10 years. Um, we're looking at, uh, building out this fifth type, which we're calling trust.

And the reason I feel like it's different now for like a plumber, the average user would spend 30 seconds on your page, so you didn't really need a bunch of content about why that user should trust you because they weren't reading it, they weren't looking at it. But now we can make the assumption that an AI agent is gonna read everything on your page. So having that trust content is suddenly quite valuable. So, uh, that's what we're working on.

So it's gonna be like pricing information, uh, specific stories about things that have gone wrong in the past. We find that stories about things that have gone wrong are really, really good at building trust. A lot of businesses try to hide that. Um, a lot of information Edward Sturm: Wait, talk, Caleb Ulku: know, Edward Sturm: talk a- talk about the, talk about, talk about that a little bit.

Oh, yeah, that too. Yeah, these things. People wouldn't expect it. People wouldn't expect that.

So, so why this might not be for you, things that have gone wrong. So you, so you're saying include that? Caleb Ulku: Yeah, 100%, because, right, the… So we know, of course, that when somebody opens up Google and searches for something, it's a personal search. Google knows your search history, and it's gonna show you things that it thinks you will like.

Well, ChatGPT is that, like, times a bajillion, right? It knows who you are. You've had all these conversations with it, and you're not gonna open up ChatGPT in Houston," right? You're gonna open up ChatGPT and say, "Hey, my garbage disposal stopped working.

Uh, can- do you know somebody who can try to f- fix it tonight?" And then ChatGPT is gonna go read all these different websites to find someone who can fix it tonight. So if you have content on your website about, you know, you respond, how you can fix garbage disposals, and maybe you have a review, even better if you have a review on your GBP, uh, that somebody said, "Hey, I called them about a broken garbage disposal, and they were there in three hours." Uh, like that, those- that attribute matching is incredibly important.

So telling ChatGPT your service is not for, uh, is incredibly helpful at getting ChatGPT to recognize, hey, per- my user, I know that my user is none of these things, so this seems like a g- would, would be a good fit for my user Edward Sturm: And so you're finding that the SEO that you're doing is getting recommended - is getting your clients recommended a lot within LLMs? Caleb Ulku: Yeah. So, um, so we're trialing a bunch of that trust content. Before we roll it out as our normal process, of course, we trial it.

Uh, what's the right to put in place to write the content in such a way that it gets picked up? Uh, but honestly, on top of that, there are a couple of other things that we started doing regularly. Uh, one of them, for all of our clients, we do a monthly Google News press release. Um, just a press release to make sure it gets picked up by Google News.

Um, Edward Sturm: What are you using, Caleb Ulku: valuable to Edward Sturm: are you using for press releases? Caleb Ulku: We're using a service, it's called PR Underground. It's the least expensive one we found that regularly lands on Google News, and that's all we care about. Like, if it lands on Google News, I want the cheapest service.

That's what I'm trying to get. so yeah, we knew that helped with Google rank, uh, but you know, it's not super inexpensive. It's, you know, 80, $90 for each one. So for most local businesses, we didn't need to do that.

Uh, but we find that if they're regularly mentioned on Google News sources, ChatGPT is much more likely to pick it up. Um, so we'll do that every month for every client now. We add it like, again, helpful on Google, but uh, we started doing it for ChatGPT. And then for clients in larger areas, uh, like larger metropolitan areas, we use another service.

Um, boy, I think it's PRASAP, like press release ASAP, they have the ability to get you into a lot of local publications. Like, you can publish on the, you know, whatever the local Kansas City newpi- newspaper is. You can pay like a couple thousand dollars and get a feature article in the local Kansas City newspaper, and again, it's helpful for Google rank. It's even more helpful to get recommended by the AIs.

Um, so that's why we started adding stuff like that. But I mean, it's all local SEO. Like, there's nothing that we do for AI that doesn't also help Google rank, Edward Sturm: Uh-huh. Are there any, are there any, uh, local marketing tactics outside of SEO that you just really love?

Caleb Ulku: No. Edward Sturm: Okay. No, 'cause I like, sometimes I see, I see just local marketing in general that I think is so cool. Like, you have, like branding for example, and you have like consistent branding, and you have these big bright red vans, for example, and y- the way that, the, the way that your technicians dress and, uh, a- and the way that, that people behave or talk on sales call.

Like, I, all of that stuff, you know Caleb Ulku: Yeah, you're right. A lot of the branding is very cool, and, you know, meeting people's needs when they call you on the phone, like being nice to them and, uh… 'Cause Google is Edward Sturm: Oh. Caleb Ulku: your Edward Sturm: Mm-hmm. Caleb Ulku: right?

Half of the US smartphone market is Android. It's not like Google created Android out of the goodness of their heart. Like, that data feeds the Google search engine to make local search better. So we've seen this over and over again, where if, you know, say one of our plumbers goes on vacation and doesn't answer the phone for two weeks, their rank map looks noticeably worse after two weeks of not answering the phone.

Uh, 'cause Google is - You know, we call it goal completion, it's something that Google has been doing for over 10 years, especially in the non-local space. But even in the local space, if, you know, uh, the, uh… For non-local, the easy thing to understand is ping-ponging, right? So you, you type something into Google, you click on the third result, you hit Back, you click on the second result, then you close your browser, right? That's a really bad sign for the third result because you didn't meet the searcher's needs.

The same thing is true for a local business. Somebody calls you, and, you know, you were really mean or rude to them or whatever, and then they hang up, and then they call the second competitor, like one of the other GBP results. That's a bad sign for you because Google knows, hey, this user had to call someone else after they called you. So 100%, like when, when we bring on a local client, one of their contractual agreements that we put is that they have to have a minimum, uh, pickup rate, where they can't have too many missed calls, 'cause we can't rank someone who's not gonna answer the phone.

Um, and not just because, you know, the ROI is really hard if they don't answer the phone. That's obviously true also. But it's more that, like you just can't consistently sit in the top three if you don't answer the phone and answer the phone, like, in a real customer service-centric sort of way. Yeah, so I, I would consider, you know, nice to someone on the phone, making sure they don't call one of your competitors after they get off the phone with you, that is 100% local SEO.

Um, So and, and I mean, there are cool things that a lot of people have done in the local space. Uh, you know, the, the Google, uh, uh, LSA ads, like those are very cool. Uh, we don't do 'em, and I don't know how to do 'em. Um, whenever a client asks, I, I usually refer 'em to someone Edward Sturm: Who do you refer them to?

Caleb Ulku: but yeah, I mean, there are a lot of, uh, you know, there are a lot of really cool Facebook ads that I've seen, uh, local businesses run that have really helped to get, uh, business, but again, Edward Sturm: Who, who do you, who do you refer, who do you refer them to for LSAs? Caleb Ulku: Uh, we have a - I have a couple of, uh, other agency owners who focus on, on those aspects, who don't do SEO as much. They'll focus more on paid traffic Edward Sturm: Are you- Yeah Caleb Ulku: very different.

Like, one of the challenges, so a lot of local business owners, if they haven't done a lot of marketing before, they're used to referral traffic. Uh, and when somebody calls you as a referral, it's a very different conversation than when someone calls you as, hey, they found your GBP online. And then it's another step change in how challenging the conversation is if they found you through an ad. So one of the challenges of the running ads is to have success with running ads, what we have found, and it's not true for everyone, of course, but for a lot of local businesses, they need support and help in terms of, you know, how to handle a sales call, how to handle follow-up, which is important for organic, but less so, and not important at all for referral-based.

But it's really hard to have a positive ROI with ads unless you know how to have a sales call. And most of the time, you know, those Meta ads are just running to a receptionist or to the business owner, and then they complain that, "Oh, your leads are horrible." Well, maybe, but also you're horrible selling your service on the phone, and it's not their fault. They've never done it before.

They've never had the training and experience to talk to someone who found them from an ad. Um, so usually, like, a much higher price point, and you need to be able to provide the coaching and listen to the sales calls, and we just don't have the, um, infrastructure set up to provide that. But yeah, a, a lot of - Like, one of the, the guys that I know that I refer people to, he, uh, literally has a staff that will… Like, he'll run paid ads, and his staff will Edward Sturm: Wow Caleb Ulku: and, and sell the, the person calling, and then once they're sold, passes them on to the business owner.

yeah, very cool business model. He's obviously much more expensive than hiring somebody to run your ads, but he actually gives you sold leads instead of, you know, a random call from, from an Edward Sturm: What would you say to somebody who wanted to have an AI answer the phone Caleb Ulku: So what we would say, like, so right now, having an AI answer the phone is better than not answering the phone. Um, and some of the AIs that can answer the phone are, are quite good. but you gotta watch the data, right?

If, if people are calling and your AI is people to hang up, uh, then you need to get a human to answer the phone or get a better AI. Because again, yes, obviously that's costing you the customer who called, but that's also impacting your rank position. If somebody called you from an organic listing and then hung up on you and called someone else, that's pretty bad. So, uh, you know, if you don't wanna pay for like a human to answer the phone, I get it.

Humans can be expensive. Have an AI do it, but watch the data, listen to the recordings, get the - make sure the AI is getting better and better until you don't see people, you know, hanging up on your AI Edward Sturm: Do you - Are, is there any software that you would recommend? Caleb Ulku: Um, n- I - we've used a handful of them. Some are, are, are good, uh, some are less good.

Usually, it comes down to the prompting, uh, and the information that you're giving it. So, I'd hesitate to say, "Hey, we use this one and it's great," because, like… So now that a lot of my leads come from my YouTube channel, uh, we service people in a wide variety of service niches. And, you know, if we had, you know, whatever, pick one, HVAC, right? Um, we don't have the data to be able to set up an AI agent to answer HVAC calls.

Um, I have the data across thousands of calls that I can set up an excellent agent for plumbers, because we've had so many, like well over 100 plumbing clients. Um, so I mean, that's sort of, it, it's really in it up right, and y- you just need data to be able to do that. So yeah, uh, an HVAC person, I would say, "Yeah, we can't help you set up an HVAC AI call center, but we, we can help a plumber set up an AI agent because we have the data, we have the prompting to be able to do that."

And so am, am, am, am, am, am I Edward Sturm: Yeah, yeah, yeah. But, uh, yeah, and how do you, how about, uh, how about like an agency on the other side of the world or something? Caleb Ulku: Oh, um, we haven't Edward Sturm: A call agency, s- somebody to answer the phone Caleb Ulku: Yeah, I, we haven't tried that. Um, I mean, I can't imagine it wouldn't work as long as, one, the people answering the phone were trained well to be able to handle those types of calls.

And, and second, uh, and I don't mean to come across as, you know… Second, Edward Sturm: Accent. Caleb Ulku: they should speak Edward Sturm: Yeah. Caleb Ulku: thick Edward Sturm: yeah, yeah. Yeah Caleb Ulku: A lot - Uh, we've seen it, like even with, uh, plumbers in the United States, if they have a very thick foreign accent, uh, they struggle to close calls compared to, uh, an equivalent plumber who doesn't have that.

And, and, you know, we can lament the state of the world, but I mean, that's what Edward Sturm: Yeah. Caleb Ulku: data Edward Sturm: you, how do you approach reviews? Caleb Ulku: Yeah. great question.

So we used to, uh, like, so the, another thing that's, uh, in the, in the contract, um, we rarely set up any sort of review management type software because the conversion rates, the action rates on those are so low. What we're always going to try to do is to get the business owner, uh, f- or the service provider from whatever phone, whatever phone number they're used to talking to the customer on, uh, we'll get them to text the customer from that phone number and ask for a review.

Um, the take rate on that is, like, 40 to 45%. That's 10 times higher than a generic, you know, review management service. Um, but we did change after, uh, when Ask Maps started to come out. we started to, you know, dive into it, and I've seen several other data dives on where Ask Maps is getting its attributes from, and one of those sources is, of course, GBP reviews.

But Ask Maps seems to care a lot about non-GBP reviews. So the two changes that we made is first, instead of having our client, "Hey, hey, could, could you… You know, if you enjoyed our service, could you leave us a review?" And then the GBP review link. That's all we used to do.

Now we try to rotate review platforms, uh, so a GBP review, a Bing for Business review. ChatGPT uses Bing, so the Bing for Business review is good for ChatGPT. Uh, we'll do a Yelp review. We'll do an Angie's review, and then we'll come back to the Google review.

And then on top of that, the ask is also different. So now the ask is, "Hey, could you tell us about what happened and what the outcome was?" Instead of just, "Could you leave a review?" Because what we want the user to put in are the attributes, right?

We want the user to say, "They came and fixed my garbage disposal at 7:00 PM." Right? That's what we want. Uh, so, but you can't ask for that.

Um, Google says you can't ask for that, uh, so we don't. We just say, "Hey, could you talk about what happened and what the outcome was?" Yeah. So reviews are more important now than they used to be because of the AI agents.

The AIs will read all of your reviews and look for attributes Edward Sturm: What are mistakes that you see people making with review collection? Caleb Ulku: So, um, even beyond just fake reviews, which I hate. never used fake reviews. That's a really good way to… Yeah, people are getting like… Anyway, enough about that.

Edward Sturm: No, no, talk about it. Talk about the, talk about the results with the fake reviews. Talk about what you've seen Caleb Ulku: Yeah. Well, I mean, uh, Google will remove a lot of them, then there have been several stories about people whose main focus are fake reviews, uh, getting caught and I can't remember, by some government agency, getting sued by some government agency, uh, 'cause fake reviews are, you know, they're… there are a lot of things that SEOs do that are maybe against Google's terms of service, and at least as of now, Google's of service isn't law.

So you can break Google's terms of service, and if you get caught and Google gets mad at you, then, you know, whatever. Um, you might get your entire client website de-indexed. Um, but, you know, fake reviews are actually against the law. So that's a very different bar, uh, beyond which often they just don't work very well because wants to see reviews done by pe- like, somebody who lives 500 miles away and posted a review of a plumber just doesn't make any sense.

So, I mean, it's hard to… So anyway, uh, so fake reviews is the big one. Another one that we see all the time is review gating. I'm sure you s- see this too, um, where, you know, it's hard to get caught, but we've seen people get caught. Review gating is against Google's terms of service.

Uh, if you get caught review gating, they'll typically scrub all of your reviews, which is pretty brutal. Uh, and again, it's hard to get caught. A lot of people still do this, but, you know, we, we try not to do something risky like that, so we don't review gate. And it's almost funny, right?

So we've seen that the best conversion rates, uh, generally happen when your rating is between, like, 4.7 and 4.9. Uh, a perfect 5.

0 rating generally has fewer calls. So I have given advice to clients before to, like, be really mean to the next few people who walk in. Like, "You have a perfect 5.0 rating.

That's too good. We need to get some one-star reviews 'cause you'll get better conversions, so just be really mean to the next few people who come in to get, get, get some one-star reviews." So if you get bad reviews, that's part of doing business. Just get more good reviews.

Yeah. Um, and then I mentioned the review management software, which, yeah, it's easy, but the action rates on those are generally, like, 5%. Uh, so you just need… Yeah, just whatever phone number the customer is used to dealing with, have that phone number text them and, and ask for a review. You'll get 10 times as many Edward Sturm: Did you see how, uh, Google is asking, "Does this business offer rewards in exchange for reviews?"

Caleb Ulku: I did see that, yeah. Hundred - And obviously that's also against their terms of service, so, uh, a lot of… You know, I've been to a restaurant and they'll be like, "Hey, if you leave us a review, we'll give you a free bowl of queso." Edward Sturm: Yeah Caleb Ulku: And you gotta be careful with that, 'cause if, if too many people say yes, then, uh, you know, Google might come and remove a lot of your reviews Edward Sturm: you're also, you, you also talk about, uh, about how to build up an agency, right?

Caleb Ulku: Yeah, Edward Sturm: Yeah. It, uh, so if somebody lost all of their clients tomorrow, how would they rebuild, Caleb? Caleb Ulku: Yeah. Uh, good question.

So a lot of that depends on how fast they're trying to rebuild, right? I know this sounds kind of silly, right? But best way to get clients, in my experience, is to have started a YouTube channel, like two years ago. Uh, and you know, obviously starting today is better than not starting, but it takes time to build up.

I'm sure, yeah, you've, uh, you've been there, Edward. It takes time. Uh, but when it starts working, it starts working like, you know, nothing else that I'd, I'd ever tried. beyond that, uh, you know, going to every single chamber of commerce meeting, um, not trying to Edward Sturm: Yeah, that's huge Caleb Ulku: the people who attend chamber meetings t- are typically local business owners, and it's typically the same group of people who attend over and over again.

So just go, listen to them talk about their business. Everyone loves talking about their business. Um, maybe run rank maps, email them rank maps. Uh, uh, email them, "Hey, I got a couple of ideas.

Change your title tag, change your H1." Um, so many local business owners - So we - Uh, I have this series on my channel where I do like SEO in real life. So I'll literally pull up Google, I'll do a Google search, and then I'll just talk about usually the number one ranked business, the number four ranked business, the first loser, and then somebody who's ranked in the 20 plus range. And especially like the people who are ranked in the 20 plus range, most common title tag I see is the word home, because that's the default title tag for so many website builders, uh, for the homepage.

And literally, if you run a rank map and then change the title tag to be, you know, Plumber Houston instead of home, you're go - and then rerun the ti- uh, rank map, you'll see improvement with that 30-second shift. Um, so just make friends with the people who attend, talk to them about how you do SEO, give them free advice. You know, make these changes, do these things. And again, it takes time, but you will be able to grow a local SEO business doing that.

Um, and I suppose a third way, when we started in local SEO, we just started running meta ads. Um, but then, you know, you need an ad budget. Uh, we probably had spent five, 10K in ads before we had developed an offer that worked. Um, I had a sales staff, like people who are answering the phone, taking these calls.

Uh, obviously, if you do the calls yourself, you can avoid that, that cost. Uh, but it takes a huge amount of time, right? Like my, my sales guy, he was doing, you know, f- four or five calls a day, um, every day, six days a week for months. Um, so just it's, it, it gets challenging, it gets difficult, but, you know, we landed - Uh, once we came up with an offer that works, uh, we landed like, uh, 97 plumbers in the next six months.

Um- Meta ads can work, but you have to budget for it. You have to do your split testing. You have to figure out the offer. Uh, overall just, you know, being present, you know, building actual relationships, uh, where business owners go, uh, that's probably the approach I would take if I were starting over because assuming, you know, starting over, you might not have 10K to drop on split testing Meta ads to find something that works.

And, you know, it's even training on yourself to get better on the phone, right? I talked about how hard it is to close people who call from an ad, and that goes for the agency owner too. You have to learn how to get better on the phone. You have to listen back to your phone conversations.

Um, you know, now you can give transcripts to AI and a- ask AI for advice. You know, "What did I do wrong? What did I do well? What do I need to improve on?"

Um, all of these types of aspects. Yeah. Edward Sturm: Let me ask you, what, what do you think separates a six-figure SEO agency from a seven-figure agency from an eight-figure agency? Caleb Ulku: Yeah.

So interesting. So figures, you can usually get to five or six clients, right? I mean, that's $8,300 a month. Um, so a six-figure SEO agency is very easy, can be a one-person show.

You know, maybe they have a developer or something like that, or a, a VA. Uh, but you know, 90% one-person show. And how they got clients, you know, referrals, the chamber of commerce stuff I talked about. You know, just very local, serving their local community, uh, doing the SEO things, right?

um, getting to six figures is really about learning how to be consistent with prospecting and learning how to get better on the, on the phone and, and closing sales. So now growing from six to seven, obviously massive difference. Now you need $83,000 a month. if your client o- is on average 2K, you need, what, 42 clients in order to hit that figure.

So going from six to seven, uh, is learning how to grow a team, right? Who do you hire? How do you hire? How do you train people?

Uh, like you figured out how to prospect. That's how you got to six figures. Now you need to bring a team on board and prospect even more, but have your team do a lot of the prospecting, have your team manage clients, have your team, uh, do a lot of the fulfillment. And going to eight figures, I don't know, Edward, I, I'm not there yet.

Edward Sturm: Be there soon Caleb Ulku: Maybe. Um, but yeah, so i- i- s- growing above seven figures, a lot of it has just been, you know, how to make sure that you're not the bottleneck, right? And 'cause even at, even at eight figures with 40-odd clients, you can still be involved with every single client. You can't talk to all of them, uh, obviously, but you can still be involved.

You can, you can be at the strategy meetings. You can review what, you know, the plan is for every single one of them over the next month. But as you start to grow 50, 60, 70 clients, you just, you suddenly have dozens of clients that you've never even heard of the client before, right? You, you have no idea what's happening, and your team needs to be just that much better to be able to manage them without even you knowing that they're a client of yours Edward Sturm: Last question, Caleb, and thank you again.

You were awesome. I hope we do many more of these episodes. This was so much fun. Uh, last question.

Five years from now, what will the best SEO agencies be doing that the average agency isn't even thinking about today? Caleb Ulku: So, okay, great question. I have to start by saying, you know, I mentioned the Sundar Pichai interview from a couple of months ago. One of the things he said in that interview, which was mind-blowing to me, is that, you know, obviously Google used to plan five, 10 years out into the future, they don't do that anymore.

Now Google's long-term planning is 12 months in the future because AI is changing so much that when they plan more than 12 months out, they have to change the plan anyway because AI did X, Y, and Z, and it got this much better, so that changes their, you know, their two, three, four-year projections. So my first answer is, man, Google's worth $3 trillion and they can't plan more than a year out. You want me to plan five? Um, but beyond that, like even thinking 12 months, 18 months, 24 months, um, it really feels like going toward a world where websites aren't visited by people very much anymore.

Um, that right now we write websites for Google's algorithm, we write websites for the LLMs, we write websites for actual people. And it feels like that's going to collapse to where we're just going to be writing websites for AI agents. And, uh, the AI agent is going to get whatever relevant information they need and then give that information back to the user. Uh, so everything that we'll be doing is going to be optimized around agents, uh, being able to easily see, understand, and regurgitate the information that we're generating, which, you know, uh, quite different from writing websites for people.

And I've heard people talk about, you know, websites being like infinitely customizable. Uh, we're seeing some of that now where you can build AI-generated websites, uh, that serves as your Facebook landing page, and then based on… You know, you can set - You set up Facebook ad campaigns that are running only to a very specific demographic, and then you set up your landing page to be focused on that demographic. Um, so that like ratcheting up so that websites are more and more custom.

But, you know, that seems like a stepping stone to people just not visiting websites anymore. Um, so I think that's going to be largely what we look at in the future 'cause… I had a conversation with someone else and they said, "Man, do you really think SEO is just gonna be content and mentions in five years?" And if you step back, right? So if you're ever really bored, uh, having trouble falling asleep, uh, Larry Page's original PhD dissertation from 1999 is available online.

It's public and you can read it, and he sets out the foundation of Google search algorithm. And all know this today, it seems obvious, but the incredible, uh, flash of genius that he had was to use links as a sign of your, how quality your website is, right? Uh, more links from quality websites means your website must be more quality. Quality websites don't link to spam.

so that, obviously the foundation of Google search algorithm. Humorously, they tried to sell that algorithm to Excite for like- $150,000 and Excite turned them down. Um, so, um, from that foundation, Google has spent the last 27 years to… 'Cause y- they, they work hard to get their own users, right? ChatGPT, Claude, Google, they work hard to get their users.

So now have to build a machine that can recommend a trustworthy business to its user based on nothing but what that machine can find online. So yes, SEO is going to… Optimizing for those machines, whatever machine it is, whether, you know… And, and I don't subscribe to GEO, AEO, whatever soup you want. I, I, I… It's all search optimization, whether it's optimizing for the LLMs or for Google or for Bing, it's all optimizing for search. And for all of those tools, they're all trying to figure out, "How can I decide whether this is a trustworthy business based on nothing except what I can find online about that business?"

'Cause that's how they work. That's all that they can work on. So even in five years, the question is, how do we optimize so that your online presence is trustworthy enough that whatever that machine, whatever you wanna call that model, that one is recommending you because of how it sees your online presence? I don't know.

I Edward Sturm: That's a great answer. I, I, I wanna ask, aside from the backlink stuff that you talked about, are - do you go hard with digital PR? And yeah, you talked about press releases, you talked about backlinks, but are you doing any- anything else? Caleb Ulku: Yeah, so you know, it, it, so the, the PR Underground for the monthly Google news reports and then the PR ASAP, uh, they're a little bit - Well, not a little bit, they're quite a bit more expensive.

Uh, but for the larger metros, yeah, $2,000 and you can be in the Dallas Star or whatever the Dallas newspaper is called. Uh, really, really good for the LLMs for building that level of trust because, right, uh, Joe the Schmo doesn't get an article about him in the Dallas newspaper. Um, and you know, I, I, I mentioned briefly earlier how much I love multi-location businesses because it's much easier to optimize for Google's algorithm if you have multiple locations. It's even more powerful for the AI systems because a lot of, like, local businesses, a, a lot of their ranking is based on proximity, right?

That's a major rank factor, just how far away you are from the searcher's location. But Gemini, the AI systems, they don't rely on proximity nearly to the extent. So if you have a 50-location business or a thousand-location business like Roto-Rooter, how much easier is it to build trust for Roto-Rooter versus Joe the Plumber, right? So that's one of the challenges as you go to LLMs making all the recommendations is they like businesses so much more because they can find so much more information about them.

So Joe the Plumber, uh, s- is going to struggle to compete with Roto-Rooter, uh, unless, you know, the sooner they start getting their name out there, optimizing their online presence, the better it is for Joe the Plumber, uh, when - yeah, right now, local businesses, 90, 95% of their traffic is still coming from the map pack, the map pack algorithm, same one that hasn't changed in 10 years. Um, that's going Edward Sturm: Maybe the answer, maybe the answer is like what are the, what are SEOs not thinking enough about right now is vi- is viral linkable assets.

And everybody thinks that a linkable asset is a calculator. It's not just a calculator. It's a stunt. It's, it, it's, it's something that's newsworthy.

It's, it could be, it could be a research report. It, it's, but it's something where people are going to write about, they're gonna, they're gonna write about the linkable asset, sure, but they're going to write about your brand. And when you have something that's viral, that means lots of people are writing about it and making videos about it and talking about it Caleb Ulku: Yeah, 100%. Like, what could you do in your local community to get your l- like maybe it's not the Dallas newspaper, but like for example, um, my business a couple years ago, we had an article that was about my business in the Huffington Post.

And we live in a relatively small city, so I reached out to the business editor of the local newspaper and said, "Hey, I thought this was pretty cool. I thought you might be interested in this." And then they wrote a feature about my business because it was featured in the Huffington Post. So if you're a local plumber or a local business owner, you know, what can you do that might get you noticed even by your relatively small local paper?

Like, all of these mentions are incredibly valuable to building the online trust basis Edward Sturm: you are awesome. Thank you so much for coming on the show again. This was, uh, so much fun. You shared so much great stuff and, uh, yeah.

Thank you Caleb Ulku: Yeah, it was a blast, Edward. Really honored that you asked. Thank you so much. Um, I'll, I'll, uh, I'll have to have you on my channel and we can have, uh, more Edward Sturm: That would, Caleb Ulku: back when Edward Sturm: that would, yeah, that would be fun.

And, and where should everybody find you? Caleb Ulku: So luckily I have a weird name, Caleb Ulku, uh, U-L-K-U, so you can find my YouTube channel, uh, my website, calebulku.com. It's, it's, it's hard to get me confused with someone else.

Edward Sturm: Once again, thank you. This is episode 1,085 of the podcast. 1,085 days in a row doing this show. Let's go.

If you watch this on YouTube, thank you so much for watching. If you listened on Spotify or Apple Podcasts, thank you so much for listening, and I will talk to you again tomorrow. Bye now.