Jesse Cunningham’s 50-Microsite Statewide SEO Experiment

0:00This video is going to be your masterclass in how to build a Microsite Live. Right now, we're going to do 14 steps, 9 props. We're going to get into it. 7 pieces, bottom to top. We're going to literally build this together. So if there's mistakes along the way, I'm sorry, but we're going to do this on the fly based upon how I like to build websites nowadays. Now, Microsites are a different beast. We'll talk about it, but there's 7 pieces, bottom to top. Number one, a phone that rings, competitor recon before the structure. Private GitHub. We're going to do LLM Wiki plus a Claude Code over SFX.

0:30We're going to do our own custom theme on WordPress and we're going to use Kingston or Hostinger for WordPress hosted. When you break them down, all these steps, it's not terribly complicated, right? This is, this is certainly not rocket science. So what this is based on, this is the Microsite structure. I call this the star method. I've done a video on the star method recently, but basically this, a Microsite is to help a main website. So most business owners have their primary website and they can rank really well in their primary location, right?

1:00Here, you rank here, right? But as they get away from their location, their rankings are going to fade because that's how Google's algorithm is. So these tertiary spots, these secondary spots are Microsites. So the star method, which we're building a Microsite for the star method in this video is to be able to dominate a large area and get many, many leads. And this is what it looks like in real time, right? That would be as if you do it in Miami. This would be as if you do it in Indianapolis. Now, before we get into the building, what I want to show you this, I just posted it. Where the heck did it go? All right. So in the community here, I posted a win. I like having wins in the community. And this is what it looks like when you have Microsites that are working, right? Take pause right here to the left is my inbox. So we have leads coming in every single day from these websites and we resell them to business owners. That's the whole point here. So we are more or less the middle man. We are the provider of qualified leads. This is like home advisor and Angie's list have been doing this forever.

2:00The cool part is the reason I posted this like to the left is nothing new. That's a good lead flow. But down here you can see the call duration for one of these that just came in was eight minutes and 50 seconds. So you better believe that was a nice qualified lead that came in from one of these websites that I'm going to show you how to build right now. Now we're going to use this as the primer. So we're going to tell Claude the whole mission first. But what we're doing here is two workers. So there's one boss and that's going to be me. I make the decision and Claude is going to be the worker, right? It's going to read each instruction. And I have

2:30many different prompts that we'll use along the way that you can use to to build this out yourself. This is Kingston and what we're going to do is install WordPress to begin. We could do a few things. We could do an empty environment without WordPress. We could clone it. But I want to do WordPress because what we're going to do here and it's really, really cool. You'll see how cool this is. We're going to build a custom child theme on top of a well known theme. So Cadence is a really good theme. You have Generate Press. You have Block C. I think we're going to use Block C for this one. Let's pick the site name. We're going to do

3:00YouTube website build and that's just for my Kingston. We're going to actually go out and find a particular niche we're going to work in. We're going to find like what we're going to build. I don't have no idea what we're going to build so far. And let's just do San Jose and click continue. And here we just put in the credentials, right? Very, very easy. And from this point, we're going to have Kingston build us an environment that we can work in. And you're really going to like how I give Claude access to do all the things. Now, what we need to do first is come over to Claude code and we need to come down here to bypass permission.

3:30So you need to have this enabled if you want to go fast, right? We got to go fast. Otherwise, Claude code will ask us a billion questions and we really won't get anywhere with it. And this prompt here is pretty powerful. Let's come back over here to our slide deck. So this is the primer prompt. It's going to tell Claude the whole mission first. So the primer is you are Claude code, you're my building partner, et cetera. The mission is the mission. We're building a local microsite. Great. And here's the plan, right? And it says work in rules before each task. And now

4:00we're going to launch it in one plain sentence so I can follow along, keep your output short and readable, secrets, anything that lands in a .env file gets masked in every output forever, right? When something fails, so say so and plainly fix it. So what we're trying to do, at least for this build together, we're trying to tease out of Claude so we can learn together if something has a problem in this video. We'll be like, oh, that's the problem. And it explains to us. You don't necessarily have to do all this, right? But what we're going to do, let's come back here. We're going to just hit go.

4:30So this thing is going and we're using Fable 5 in this instance because, well, it's expensive, no doubt, but Fable 5 is super fast at doing this task. I really, really like it, right? So what it's going to do is boot the brain. Now you're going to create the skeleton of our LLM Wiki in this project folder. Now we've talked about this in the past, an LLM Wiki Wiki. It's really, really powerful. But before we get into that, let's just go out into the internet and let's find the actual niche that we're going to work in. So here's our opportunity. And when you're looking at it, it's a pretty darn big opportunity. The United States is humongous. And guess what? We can build a microsite for any service in any location. How do we figure out how to do that? Well, this is going to help us. So will that niche actually make money? Now, this is kind of a hard part. A lot of people struggle with this. What we need to do is find a niche that has a decent ticket price, not too much competition, right? Those two things can

5:30really help us. I'm going to come over to GBT and ask it, what are the most profitable sub niches within the plumbing industry for residential? And then we can see. So this is what I would ask for any sort of niche. I'm going to ask it, what are the most profitable sub niches within it? You could do it for roofing. You could do it for electricians. You could do it for any obscure type of industry. And the whole point is here, we're looking for a sub niche. I don't want to build a microsite, houstonplumbing.com.

6:00That doesn't make any sense. But when we start looking at something like trenchless sewer repair or just sewerrepairhouston.com, it could make a lot of sense. Let's check these out. What we're looking for is high ticket price. All these are pretty darn good. Water treatment and softener systems, tankless water heater installation. I mean, there's so much opportunity. I mean, so much. But you can't do for these strategies. You cannot go after the big dogs with the Google business profiles. You have to kind of sidestep Mike Tyson.

6:30To be able to win. You don't want to fight Mike Tyson head on. No way, no how. So what we're going to do, let's just pick one and test it. Let's come over here to sewer repair. Let's come here, sewer repair. Let's come over here to right here. Then where can we do it? Let's do it in Houston. I don't know. I was thinking Houston. Come over here, Houston, Texas. Score it. I'm very curious. Sewer repair. Let's see who shows up. It's going to pull in some live data. And what this does, it is truly hooked up to APIs, right? So that's using real data here. And it scores.

7:0040 to 42 out of 100 FFDB. That kind of sucks, right? Local demand competition too high. So we need to go back to the drawing board. Let's come back down here. We can do tankless water heater installation. Okay, that could work. Let's come up here. Let's type that one in. And that's a little bit better at 60 out of 100, but still not good enough for me to want to do it. And I can see a problem. And here's the problem, right? And you may have the same problem. So I'm going to say it. So when we're looking at this, we know this is a good

7:30niche, tankless water heater installation, right? But Houston, Texas, if we come over here to Houston, if I'm coming down here, is a huge population base. We don't often have to, well, we don't have to do anything, but we don't have to go after the megatropolis areas, right? We could come over here to this area, Beaumont, or we could come down here to Corpus Christi. You don't have to attack the major cities, right? So let's come back over to our tool. Let's come to, let's try Beaumont. Let's see if it has any effect. Beaumont, Texas. Let's hit search. And this back and forth thing with you and the data is very important because this is going to dictate a lot of our direction moving forward. So that gave us a 41 out of 100, a grade D. We're still not on the money. Now, Fable's done its task. We're going to come back to it. I always like to have Fable running on something, right? As I work in the background. But right now, Fable is just kind of chilling out over here. That's okay. We'll get into it.

8:30I want to show you what it looks like, what we need to avoid, right? Corpus Christi. So if I type in tankless water heater installation in Corpus Christi, we have ads, we have ads, we have Mat Pack, and then we can rank down here. Now, if we can rank down here, will we get leads? Yes. But we will get a lot more leads if we don't have all this sponsorship in competition. So we're trying to find a nice balance of all the things. You have to do your competitor research before you do anything else. Now, ask in GBT because we have to go back to the drawing boards. What are good niches specific to Houston, Texas, right?

9:00What are sub niches to primary categories? And it said that Houston has a lot of slabs. In other words, houses don't have basements. And mind you, typically with microsites, we are doing blue collar industries often. It's very easy compared to white collar because those are typically Y-M-Y-L. And then I found this one right here. The first one might as well try electronic slab leak detection. I already typed it in and it was a 41. But let's take pause. The Mat Pack opportunity was a B. And if it's ever a B, or an A, it could be a good thing. So let's type it into Google, electronic slab leak detection in Beaumont, coming down here. We have sponsorships. Okay, that's fine. We have the AI overview, but then we have Mat Pack below the organics. We can rank right here. Do you see this spot? Right here. We could rank this. So with my microsites, we rank right here, right below AI overviews, below a little bit of ads, and you can definitely get clicks. And you probably will get clicks on big.

10:00Okay, so we're going to go with this. We're going to go with electronic slab leak detection. But let's build it out a little bit further. Next up is the domain. Very easy. We need an exact match domain under 30 bucks. We're going to have the service plus the town as one word. And we're going to try variants where we can. So coming over here, we're going to come to Namecheap. And do you see this right here? Beast mode. You're going to click that. And then you're going to enter into this mode. And this is really cool because you can enter many, many domains right here. Instead of just checking one at a time.

10:30You can check like 20 at a time, or we're going to use GBT to do all this. So let's come over here. Let's come back to right here. Let's grab our keyword. And all of a sudden we're realizing electronic slab leak detection, beaumonttexas.com is ridiculous. That's too big, right? So what I'm thinking is, is there a potential for taking off electronic and just doing slab leak detection, right? Is it the same type of competition? Okay, the Mat Pack does show up, but there's still this slot right here.

11:00So I'd be tempted, and this is where your imagination has to come in. I would be tempted to have slableakdetection.com or variant of it, and then doing specific services under this. So let's do this. Let's just come here. I'm going to tell GBT if the darn thing ever loads. Right here, I'm going to say, I want to build an EMD for this concept. Give me a bunch of variants that I can paste into beast mode in Namecheap.

11:30So GBT is really good. And if you haven't used, I think it's called WhisperFlow, you'll see instead of me typing on this video, I'm actually talking into the computer and it just does it for me. Pretty darn cool. All right, Beaumont, I like it. So all these are good. So instead of thinking about this yourself, why not just do this? So if we were going to dominate this, we would do multiple websites, right? So let's come over here. Let's make sure we're in beast mode and paste them in.

11:58And then we're going to hit generate. And what Namecheap is going to do, it's going to see which ones are available. And the cool part about using LLMs like GBT for this, it kind of understands, like user intent, and it knows what an EMD is. And it knows through previous conversations, what I'm going after. And you'll see, because this is a very specialized niche, most of them are going to be available, right? But when you get into more competition, if you see green, it means it's available.

12:24But when you get into more competition, harder things to rank, then you are going to have a lot of red. And using this tactic right here is going to help you a lot. Now, something we should consider using is the Google free keyword planner, right? So right here is one variant that I like, Slab, Leek, Prose, Beaumont. But what I want to do is let's try Beaumont, Slab, Leek. Let's just type this in. Let's do Slab, Leek, Beaumont. Let's type that in as well. So I want to verify via some data if there is any volume. Okay. There's not much volume there. Let's think about what we're going to do here. Let's do Slab, Leek, Houston. Let's do Houston, Slab, Leek. Let's do Slab, Leek, Texas, right? And we're going to kind of build up from what we have. All right. So now we're working with things. Slab, Leek, Plano, Texas. Okay. 100 potential monthly searches. Houston, Slab, Leek. All right. We have under Slab, Plumb, and Repair Houston. Very interesting. All right. So now what we're

13:24doing is we're verifying with what we already supposed to be a good plan, right? We already were like, let's go for it. Back it up with data. Like be willing to be wrong. It's okay to be wrong, right? Let's be wrong all the time so we can fix our problems, right? And then we'll iterate. So what I want to do, I want to come here to Houston, Slab, Leek. I want to use this term and come back over here and type it in. Let's see what it comes up with. We have, this is awesome. We have sponsored, right? Sponsored, sponsored. We have

13:54AI overview. And then we have the organic. So all of a sudden we came full circle, right? Ridiculous. But we did. We came full circle. Mat Pack is down here and it's way down there. So all of a sudden we found a better solution. Houston, which is a bigger city than what we were going after is actually better. So what we're going to go after is HoustonSlabLeek.com. Now this is what FTP access looks like. SFTP to be exact. And what you're going to do is you're going to give Claude Code your host.

14:24Your port, your username and password. And what this does, it allows it to change and do everything basically as a super user for WordPress, right? That's the amazing ability of Claude nowadays. Very, very cool. So let's go back. Let's see what it said. All right, cool. Let's come up here on the FTP access. Don't paste credentials into chat, right? Chat history isn't the place for secrets. So it's going to create a folder on our local machine where we're going to upload the .emv. It's called an environment file that's supposed to be set.

14:54It's separate from anything live that it can just reference one time and just roll with it, right? Or multiple times. So what it's going to do here, File Tree. Nice. It's just giving us an overview because remember in the beginning when we went up here, we told it to show us what it's doing. And that was for YouTube purposes. Like that whole first prompt was for YouTube so you can see what's going on. But it's also good for you if you're new at this or even if you're not to have Claude systematically give you the breakdown of what it's up to.

15:24It's just going to roll with it, right? So it's good to have like stop places, purposeful stop places where it's going to be like, all right, this is what I just did. Here's what I'm doing and spell it out in plain English. Now this is what the .emv file looks like. It's just a notebook file and it's going to be in your primary folder. So you pick a folder like we did in the beginning where everything lives, where your LLM Wiki lives, all the things, right? And you're going to input right here, right here, right here, right here, all of your secrets. And from that, it'll give Claude

15:54this is the superpower. It'll give Claude access to do anything on your WordPress website. Now next up is saving points like a video game. So we got to save before each boss fight because the boss wins sometimes and we can reload your save. So we don't want to play video games on hardcore mode. I was never into that. Hardcore mode, if you don't know, is like you play the video game. If you get defeated, you can't restart. You'd have to like literally have to go all the way back to the beginning of the game. We don't want to do that. And that's where GitHub version control comes in. So we have a private GitHub repo. We're going to

16:24first commit before the first edit, right? We're going to get this done before we do anything. And this allows us, if anything goes wrong here, we have version control, right? Everything is version controlled. Very, very good. So this is what it looks like a locked delivery truck between my computer and the land. So here's my computer where Claude edits the files. And then we have the locked truck, the SFTP. It carries the file safely and nobody can peek inside. And then the land is like king stuff, right? So Claude code gets the SFTP keys to king stuff. So Claude edits, it commits to a private git, and we deploy the SFTP push to king stuff and refresh. It's there. Now, if anything goes wrong, and here's the code. If anything goes wrong in that time, because AI is not perfect, we can just go back to the last instance that we had it working and say, let's just redo all this and start here. Now we're going to set up the LLM wiki a notebook so the AI never forgets this project. So an AI without the notebook is like, man, it forgets. It costs way more money to operate an

17:24LLM without this. So we might as well give it a notebook paired with it. I call it a second brain. Pretty darn easy to do. So set up the LLM wiki before we build anything. And listen, as you learn all this, I'm assuming you want to do this yourself, you're going to realize you can do custom websites for clients now for any business, do it at this method, right? So the LLM wiki is going to be the strategy. So it's going to be our SEO rules or AEO rules. It's going to be one repo, brain and body, and it's going to be a full deep dive on everything. And now this is what the

17:54prompt looks like when we paste it in. It's right here. Copy this prompt. This is the wiki brain, right? And you can customize this however you want, but it's going to ask you a bunch of questions such as the business itself. So who actually answers the phone, the niche itself? What do you want to talk about in your niche? How far out do you want to go? Your non-negotiables, your territory, your brand's face, et cetera. And you can give it other things too. Like this is your opportunity to build out the brain of your entire system. And then from this point on, we're pretty much ready to rock. Now let's come over here.

18:24Let's come here. F11, come again right here. Google already published the answer key. And this is going to be a fun one, right? So how most people build is guess what Google wants, build it, wait months, and then wonder what happened. Now what we're going to do is we're going to do the open book test. So the site's ranking the top three right now are the answer key. We're going to read them, learn the pattern, and do it better. And Claude reads the top three winners before we build. So we're literally going to have Claude go out to the internet and look at the competition.

18:54And we're going to be able to emulate and do it better. We're going to look at page inventory. We're going to look at structure and navigation, language patterns, the gaps, all of it. So all we have to do is take this, copy this prompt down into Claude and let it do its thing. Now, if you click down here, three running tasks, you're going to have it pop up here. And you'll see that there are three separate agents going out and targeting different search results. This one's targeting Houston. This one's targeting Katie. This one's targeting Sugarland. It is going out to the internet.

19:24And it's figuring out who's ranking for what. And that's super powerful because that task right there used to take SEOs like myself a lot of time, right? A lot of time. All right, so it says these two are still running. This one just came back with rich intel, heading, finding. Only one of the three Katie winners publishes any price at all, which plays right into our strategy because I said I want to talk about pricing. Filling its raw notes now with the Houston Sugar Scouts finish. Another one just came through. Very, very cool. Now, Fable does cost a lot.

19:54You can probably use Opus called code Opus for this too, but I just like Fable. I feel like if it comes across a door, right? That is a closed door. It'll try to open it. Whereas Opus, if it sees a closed door, it's just like, nah, I'm going to give up here. All right. So it just gave us a lot of information here. And I think it's worth like pause in here and reading it. So the recon is complete nine competitor sites profiled against three towns. Everything filed and pushed. So what we did is we looked at nine separate websites. Very cool. And it gave us a recon report slash MD.

20:24So basically what it's doing is remember the brain that we were building. It's building the brain out. It's crazy powerful, crazy powerful. Common patterns, the price of entry, right? Every winner has a dedicated slab leak page with the town in the title in H1. Okay, great. So some SEO stuff. Now here's the gaps, right? Our unfair advantage. What none of them do, right? Have published pricing. So no one's talking about pricing. So maybe that's our edge and actual specialization. No one's doing that. Review numbers on the page.

20:54So it shows. So basically when I gave it the prompt, I could have chose what towns for it to seek out. So you can choose which towns to go. And you don't have to do just three towns. You could have done 20 different towns if you want to do crazy SEO research. But it said, I chose Katie and Sugarland based upon its, and I'm saying, Claude shows this itself. Basically, that's what it's telling us. Because I didn't choose.

21:24For you, which I'm all for, right? Now we're going to take it a step further because Claude is going to propose different structures with this, right? We're going to pick one live. So we have three basic options. We can do the mirror, the competitor informed, or the hybrid. So this prompt right here, right, is going to basically say, look, if we have these nine websites, you just checked out what is the structure of those websites? And what do you think is the best fit, right? So instead of building blindly, like most SEOs, most developers, most local business owners do, use your competition as inspiration. We have these crazy tools nowadays. We might as well use them. Here we go. It's going to give us its take, but it's going to also visualize things for us. So option one, the structure, the mirror structure could be like this. Slab leak detection, reroute and re-pipe in, tunnel in, et cetera. It says, here's the risks. Here's the wins, honest build time, three to four hours, right? But for Claude Fable, it's not going to be that long. Option B, competitor informed. Okay. And

22:24then we have the hybrid approach. And here's my argument from the data. It says the recon shows the entry bar is quality per page, not page count, blah, blah, blah, blah, blah, blah. I'm going to say, do what you think is best. And what's cool is we can give Claude the ability to do that. Like think Fable probably knows better than me in this instance. If I just gave it tons of data, which would probably take me an hour to read, I'm just going to give it the ability to do its thing. Now we have our WordPress build. Look at this pretty thing.

22:54We have our WordPress build and something happened that I was interested in with Claude. So it gave me all these recommendations and it said, I still need these things, these WordPress credentials. It basically told me that it needs a user login and password. And I said, no, you don't because you have the SFTP credentials. So sometimes you're going to have to give AI a little bit of a pushback. And it said, blah, blah, blah. And then it said, no, I do need them. And I said, wrong. You should have access to do all the things, right?

23:24And it said, you may well be right. Let me check something confirmed. You were right. So cool part is number one, if you give it SFTP access, like we did earlier in the video, it can do everything it needs to do. It can change the themes. It can change the content and do everything. And number two is you have to be able to talk back to the machine, right? But here's where it's fun. Let's try something. So I asked it to pick whichever option it wants. Like I said, you pick. I don't care. I remember we had three options up

23:54here. You pick the site structure. It's going to do well. Option C is locked in, recorded in structure.md. So now it grew the brain again. And it says the decision in one breath, 10 pages each shipped deep in conversion complete. Now it says, per the primer, I build nothing until build prompt arrives ready when you are. Now here's the fun part, right? Claude is going to write every single page to the house rules. We have house rules around here. We're not going to let it just do anything at once, right? So the liftable answers to the top.

24:24We're going to have name, service, town, identical everywhere where it needs to be local knowledge and location pages and claim every claim survivable. So no fake reviews, no invented team, all that type of stuff. And here's the prompt. So we're going to take the start thing. This is the big one, right? And we're going to come down here and just paste it. And like before, it's always cool to see it cook. So it's research and three things with three separate agents. And that's the beauty of Fable and these frontier models, right? It can send multiple agents to do specific jobs if you tell it to, right?

24:54You have to convey that you want this to happen. So it's research in slab leak costs and Texas insurance. It's research in Houston soil and housing facts. It's research in these specific areas. It's doing intense research for us in the tax time that it takes me to actually build the website. Like you've been watching this video for 20 plus minutes, right? Thank you for watching. Please subscribe all those things. But when you figure out this process, like the second website goes so much faster. Now it's starting to build it. And this is what we're working with.

25:23It's absolutely disgustingly, terribly awful. And it's awful not because the text is awful. It's awful because the design is awful. So I think we made a little bit of a mistake. What I should have done is this next step first. So I should have flip flopped these two steps. I think we have awesome copy and an awesome structure of a website, right? And we're going to build off of it. But what we should have done is come down here. So we should. And this is what how I love to build websites. And I just randomly pick these.

25:53I'm going to give it these two websites, right? Rotor rooter and Mr. Rotor rooter plumbing, whatever. Big websites. And I'm going to say now we need to redesign and address the child theme and design in a manner that takes the best aspects of each of these URLs. Basically, if you didn't know, Claude has like a vision ability. It can take screenshots. It can use your browser. It can see things, not just the code. It can visually inspect. And if we want to have a really nice website that actually looks nice, not like this, this will never rank.

26:23Right. And if a customer or potential customer comes on here, they're not going to be too happy with this. They're going to be like, what the heck is this thing? So we need it to look professional. Now, this is interesting. This is the current site map, right? So we have the root domain, which is the homepage. And it's slab leak repair in Houston, Texas. And you can kind of see how this lays flat. So we would call that a flat URL structure because everything is as close as possible to the root domain. There is no slab detection or a slab leak detection slash

26:53the area. This is flat. Everything is right there next to the root domain. Now we're going to build this out further, right? This is a good start. And you don't have to get a one, you know, one prompt website that's, it exists. It's cool, but it's okay to do it this way because the system, the brain we're building is learning. And then if I wanted to build another website on top of this afterwards for a different city in Texas, how easy would that be? It would be super easy because it knows our preferences. So right now it's waiting on the design stuff.

27:23It's going to go out into the world, look at those two websites, we gave it and then build us a proper website. It's still working, right? It's still working. Whenever this is flashing down here, it's working. But I can click this file, created two files, edited a file, used 20 tools. And what tools did it use? It used Claude browser, right? So I was saying, you can give it access to your browser. So it can go use its vision. It's one of the most powerful things it has, right? And right now it started to create. So it used its vision. It went out there, it figured out the style CSS,

27:53so the CSS is like the structure and vibe of a website. It's figuring out all these things itself. It's hyper powerful. You just have to give it the right tools and the frameworks. Now, after a little bit of back and forth, this is what we're working with. Not bad at all for first iterations. Pretty cool. So we have Houston slab repair. I quickly made that logo in chat GPT, turned it into a PNG. There you go. The cool part is we can come over here. This is a WordPress build. So the typical media stuff, you can drop files into here, right? You can drop files.

28:23And have it referenced in Claude and say, this is my logo. Use it, boom it, pick it up. No problem. We have services. So slab leak detection, slab leak repair costs, signs of a slab leak. Probably going to reincorporate these into different parts of the website because the services would probably be slab leak detection and repair, right? So we could change that up. But what I care about right now is just the feel and the vibe, right? I probably would put a contact form right here, but no problem.

28:53I would change this out. You can use GPT to change images. Looks like there's some sort of European or Asian text there. But I like this feel. This is nice. It's simple. And listen, I mean, I've built with this method, beautiful websites, like pretty websites. Local business websites don't have to be pretty. In fact, that deters leads. You need a very straightforward website. And I think this is it right here, right? The consumer feels comfortable with this type because this is what local businesses do.

29:23What do you need right now? Find the leak. Okay. So we get price in here. Know the cost. And we have a cost guide. The cool part is, remember when the agents went out and did their research? If I click here, this is federal construction data on X, Y, and Z. So which foundation type is most common in your region? So what it did was, instead of just building, you know, blindly more or less, it went out, it grabbed data and then it came back with that data. I like this overhead image. It's just stock imagery. So I asked it to go grab stock imagery and it did.

29:53Where do you work? We have this, that, and Houston. Okay. Three areas. It'd be very, very easy. And I think that should be our next step to increase our service location areas. So each microsite has like a central location. In this case, it's Houston. And then you actually bubble it out with the suburbs and cities around it. Let's look at how they rendered. So slab leak repair in Katy, Texas. Looks nice. We have research papers here. USDA's official soil survey. So if we click here, I guess this was a link that goes out to that. If we come back over here to University of Texas research, these are nice research papers that will show in PDFs, right? That will show Google. Oh, wow. They actually researched this stuff and they're providing a good service for their clients. Now, what we're going to do, we're going to ask it one thing. I want to show you how this works because once you have the system set up, it's super easy to iterate, super easy to edit, right? There's no going back and forth with a designer nowadays who may or may not be readily

30:53available on Fiverr or Upwork and all these things, right? We can just talk to Claude. So I asked it. Now I want you to build out 10 more service location pages in closest proximity with decent population from Houston, basically creating a circle around our primary service area. And then on the contact page, I want you to have a call to action with a form fill that has basic questions, five questions altogether that utilizes the WP forms plugin, make it stylistic and work well for what we're going here. It's very cool that you can just talk to it like a regular human, right? And it will understand

31:23and just because, because remember we're using bypass permissions down here, it will just do what it thinks is best. Said it did the contact form first and here it is. How hard was that? Prefer to type instead of call and that looks nice. It looks really nice. Now I want to make sure it installed the WordPress plugin correctly. Come down here. Let's see. Okay. It has WordPress forms API. Very cool. All right. So one thing that people don't often realize is Claude can install plugins to a WordPress build. If you give it the F S T P access, right? Very, very nice. So now what we're waiting for is the service location pages. Let's refresh this thing. Let's see if it's done. Sometimes call it takes a second. All right. Not done yet. So let's come over here. Let's see. So it says created. Let's click on here. All right. It's research and you can see it to the top research, five Northern Houston suburbs. So basically what it's doing is figuring out first, which are the best suburbs to go after. And the cool part is it's going to use some logic here, right?

32:23Here it says Pasadena turns out to be the strongest slab leak market of the five homes built around 77, 64% of housing pre 1990. It's crazy, right? It's like using logic. I would just be like, find the closest ones. But it's like, no, we're going to go find the house year of the thing and figure out which are truly the best fit. There you go. You can see when we do the service areas, you'll have a bunch of service locations that Claude decided were the best fit. Very, very cool. And that's how you can expand the image of one of these microsites very easily. A lot of people ask me how big is a microsite? Well, microsite can be humongous. It's just a tiny amount of work to do it, right? Now that we have our Claude instance with the brain attached to it, it'd be very easy to hook on another website, another website, another website. And that's where it gets really fun. The first part, it's kind of like a pain in the butt, right? I can probably do it in 10 minutes to set up a new brain and all the things you saw me do in the first section of this video. And then you're just rolling, right?

33:23You're designing the website and going back and forth because you get to talk to the probably the best designer in the world, right? I know some designers might get mad at me for saying that, but Claude is an incredible designer, especially if you give it examples to work from. So what we should do now is optimize for the AI readers. And it's pretty simple when you look at it. It's just another copy this into Claude. Very simple, but what it does, it's going to do proper schema for us. And schema is how the robots more or less read data online. So you can show up where you need to show up.

33:53I'm going to have all those prompts here in the classroom of the Rank Expand Academy. What I wanted to do, it'll be here and here, right? Two big modules, Microsite Empire and the Microsite Playbook. But what I wanted to do in this video was just basically show you how to get the foundation set up. It's pretty darn fun when you learn that set up because there is no limitation on what you can build. If you're accustomed to WordPress, most people still like to work in WordPress, that's okay. You can do that. And you can have Claude be your super designer, your super

34:23everything, your super SEO. And realize that is not the only way to do Microsites. So in the Academy, we have our primary method. I think this is the funnest method I showed you, but it's not the most effective way. It's not the cheapest way. It's not the most effective way, but it is probably the funnest. And for bespoke websites, like one-off custom websites, you could sell a website like that and put some little design effort into it. Like I did no effort on design. I just got a bare-boned vanilla, right? If I spent an hour extra, that thing would be

34:53common, it would be beautiful, it would be polished, right? That's the bespoke way to build a Microsite or any website. But we have other methods. If you want to scale to 10, 20, 30, there's cheaper, there's better ways to do it. But I hope you like this video. Please subscribe, let me know what you think. And I'll check you on the next one.