# SEO Rockstars 2026: Day 2 - Simon Cornelius

**Video URL:** https://www.youtube.com/watch?v=lcK8NViWB_c

---

## Transcript

**[00:02]** Good stuff.

**[00:05]** All right.

**[00:06]** So, how many of you have ever heard me

**[00:09]** speak before or this is your first time?

**[00:13]** Okay.

**[00:16]** All right. Just so like half the room

**[00:18]** put their hand up and how many of you

**[00:19]** will never put up your hand no matter

**[00:21]** what I ask? Okay. All right. The rest of

**[00:25]** you. Okay. We probably looking at the

**[00:27]** previous presentation go, man, this is

**[00:28]** really good stuff. I So I don't have a

**[00:31]** big grandiose introduction about who I

**[00:33]** am. I'm Simon. I'm lazy and I'm cheap.

**[00:37]** So just remember I'm lazy and what?

**[00:40]** >> Okay. And I like interaction in the

**[00:42]** audience. Is that cool?

**[00:44]** >> Right. So I got some got some how what

**[00:49]** >> mic mic one two.

**[00:51]** >> There we go. Oh, operator error. Now I

**[00:53]** can hear me as well. Cool. All right.

**[00:56]** So, I am what? I'm lazy

**[01:00]** >> and cheap. Thank you. And I love

**[01:01]** interaction. So, the good news is that

**[01:04]** I've got a really good presentation

**[01:06]** today. I kept on adding more and more

**[01:09]** and more stuff. But the bad news is I

**[01:12]** got enough stuff to last for the rest of

**[01:13]** the day. So, I'm going to get through as

**[01:16]** much as I can, but you all have access

**[01:19]** to the presentation. It's in my speaker

**[01:22]** bio profile thing somewhere in there. If

**[01:25]** you want to find me, just find me. I

**[01:27]** know there's a few Simon Corneliuses

**[01:28]** around in the world. So, but anyway, I'm

**[01:31]** going to have fun doing this. And who

**[01:33]** wants to know how to make the first

**[01:35]** million with AI, right? Like just a few

**[01:39]** of you. So, there's a story behind this.

**[01:41]** I like stories because stories sell and

**[01:45]** facts tell. Where's my story going?

**[01:48]** There's a a blinding light here. Where's

**[01:50]** Where's our previous speaker, Jim?

**[01:52]** Where's Jim? Stories sell, right, Jim?

**[01:55]** >> Yes, sir.

**[01:56]** >> All right. And facts tell. So, I picked

**[01:59]** up this book recently called Making Your

**[02:01]** First Million with AI. I thought that's

**[02:05]** going to be a really good book. And I

**[02:07]** think I'm not to call anyone out, but

**[02:08]** where's Brian Kado here? Is he here?

**[02:11]** >> Brik,

**[02:13]** >> did you tell me about this book? how to

**[02:15]** make your first million with AI or some

**[02:18]** someone mentioned it. I so I bought it

**[02:20]** and I read the whole thing in like three

**[02:22]** hours. But the thing is it's written

**[02:25]** like in kid font and there's about 150

**[02:29]** words per page and it repeats itself

**[02:32]** over and over and over and over and over

**[02:36]** again. You can literally take and

**[02:38]** there's some good stuff in here, don't

**[02:39]** get me wrong, but there's some some good

**[02:40]** stuff in here. You could probably dilute

**[02:42]** to about two or three pages. I kid you

**[02:44]** not. Some good stuff in here, but very

**[02:46]** very high level. But it gets a brain

**[02:48]** thinking. But I wouldn't I wouldn't pay

**[02:51]** to buy this book again. But if you

**[02:53]** anyone wants it, it's theirs.

**[02:56]** Okay. The first person to put a hand up

**[02:58]** his story. All right. So, how to make

**[03:00]** your first million with AI.

**[03:05]** Oh, wait a minute. So, here's the guy.

**[03:08]** If you're at the back of the room, you

**[03:09]** don't have super biote vision, you need

**[03:11]** to be sitting at the front with the

**[03:13]** popular people. So AI made me my first

**[03:17]** million lines of code still waiting on

**[03:20]** the cash. So that's kind of my my line

**[03:23]** here is for all you people that think

**[03:26]** you can code, great. I try it because

**[03:29]** it's it's the only way you learn, right?

**[03:31]** when I learned to code back in last

**[03:34]** century, I'm dating myself now, is I

**[03:37]** made a bunch of mistakes. People told me

**[03:39]** my mistakes, but I got better and better

**[03:41]** and better, but you have to try. So, I'm

**[03:46]** going to just jump into some really cool

**[03:47]** stuff that, and this is not a

**[03:50]** one-sizefits-all, don't get me wrong.

**[03:52]** I'm here to inspire you. To inspire you.

**[03:55]** Is that okay? Because one thing is I am

**[03:59]** >> lazy.

**[03:59]** >> Lazy. And the other thing is I'm

**[04:01]** >> cheap. Cool. How many presentations do

**[04:05]** you all go to where you actually engage

**[04:07]** with the speaker in the audience? Not

**[04:09]** too. I'm a little bit different.

**[04:12]** So, let's take a simple SEO task. This

**[04:14]** is to get you thinking. This is not the

**[04:17]** solution for everyone. So, let's create

**[04:19]** a blog post, shall we? Anyone ever

**[04:21]** written a blog post before? Okay, four

**[04:23]** of you. Okay, cool. All right. Rest of

**[04:26]** you have a VA that does it, right? So

**[04:28]** what do we first do? Let me stand over

**[04:31]** here. I know I'm away from that now. So

**[04:34]** blog post process go come up with a

**[04:37]** topic

**[04:39]** research. Get the reader intent. Oh, by

**[04:42]** the way, Martin Marion taught me about

**[04:44]** the intent behind the intent. It took me

**[04:46]** four years to realize what he was

**[04:49]** talking about. Four years when I did

**[04:51]** like, oh my gosh, research niche. You

**[04:55]** get instructions and a purpose. You you

**[04:57]** give it to your writer. That might be

**[04:59]** Fiverr, an in-house team, AI, whatever.

**[05:03]** Here's the writing instructions. You

**[05:04]** review the writer's first draft, usually

**[05:06]** in a Google doc. More on that later.

**[05:09]** There's an email. There's a feedback

**[05:10]** loop. You send it to the client for

**[05:12]** review. Google Doc. Email feedback loop.

**[05:15]** Are you tired yet?

**[05:18]** Then you create a blog post on the

**[05:20]** website. Oh, that's copy and paste from

**[05:22]** Google Doc, which means you have to

**[05:24]** remove the Google Doc formatting. Anyone

**[05:26]** been there before? Yeah. And then you do

**[05:29]** your SEO stuff, which is your title,

**[05:31]** description, link to the service page,

**[05:33]** and the first paragraph, and your

**[05:35]** category of the blog post. Then we got

**[05:37]** to do the featured image. Oh my gosh, I

**[05:40]** was tired after reading this. Go to

**[05:43]** deposit photos, shutter stock, and or

**[05:46]** valley. Remember the four fingers from

**[05:48]** Deli too? I know Chad does, right?

**[05:51]** >> So resize it. Got to compress it because

**[05:54]** it's four megabytes.

**[05:56]** Oh, then you have to give it a name

**[05:57]** because it's like shutter stop. One,

**[05:59]** two, three, four. Then you have to do

**[06:01]** your alt text.

**[06:06]** Click. Where's click? What's going on

**[06:09]** here?

**[06:11]** A little while later. I found out this

**[06:14]** is from like Spongebob. Yeah.

**[06:17]** A little while later.

**[06:20]** Publish. We Yeah, we published it.

**[06:23]** How long does that take you guys to do?

**[06:25]** I've heard like an hour, hour and a half

**[06:27]** to do that. Right. Right.

**[06:31]** So, what did I say I was in the

**[06:33]** beginning? I was

**[06:34]** >> lazy.

**[06:35]** >> Lazy and cheap.

**[06:37]** >> All right.

**[06:39]** So, let's start with step one. Come up

**[06:41]** with a topic. Cool.

**[06:45]** Push the AI easy button. Boom. Easy.

**[06:50]** >> So, who can do that? come up with a

**[06:52]** topic. Every one of you in the room can

**[06:55]** do. Can any every one of you push the AI

**[06:57]** easy button?

**[06:59]** >> Yes, you can. And it will research,

**[07:01]** write, review, email the client for

**[07:03]** review, create the blog post on the

**[07:05]** website, do all the SEO stuff, do the

**[07:07]** featured image, and publish it. Cool.

**[07:10]** So, I've actually implemented this. The

**[07:12]** only one piece I haven't done is email

**[07:15]** the client for review and feedback

**[07:17]** because most of my clients, they don't

**[07:19]** they trust me with content. So, I I

**[07:21]** leave them out of the loop, but you

**[07:23]** could easily do that.

**[07:25]** So, is that cool?

**[07:28]** >> I'm going to show you kind of how to do

**[07:30]** that. By the way, I'm not I don't have

**[07:32]** enough time. I don't have enough

**[07:34]** stamina, neither do you guys, to go

**[07:36]** through all the nerdy process, but I'm

**[07:38]** but if you want to connect with me

**[07:39]** afterwards, like Anna said, hey, let's

**[07:41]** sit down and nerd out. Cool. I'm I'm

**[07:44]** very accessible. I live around here.

**[07:47]** I I love sharing and networking with

**[07:49]** people. So, about images, they're one of

**[07:52]** the biggest pains in the bum I've had

**[07:54]** forever because I have to go to shutter

**[07:56]** stock or deposit photos, find an image,

**[07:59]** do all that stuff. It takes forever and

**[08:01]** I always miss that one step. Or you have

**[08:03]** the VA that misses a step or five and

**[08:06]** you don't know about it. Like, hey, what

**[08:08]** happened to this image here?

**[08:11]** So, here's how I create featured images

**[08:14]** now or any image. You ready for this?

**[08:17]** >> Yes.

**[08:18]** >> So, four people. Okay, are you ready?

**[08:21]** >> All right, cool. So, read the article.

**[08:23]** This is AI doing this. Read the article.

**[08:27]** Extract the visible entities. Underline,

**[08:30]** highlight, bold. Visible entities. Chad,

**[08:34]** you nodding your head or you just eating

**[08:36]** now? Because you're agreeing with me,

**[08:37]** right?

**[08:38]** >> Right. So, here's And now, no need to

**[08:41]** take pictures unless you're of me.

**[08:43]** Extract four to six ent visual entities

**[08:46]** from this article. You're asking AI to

**[08:48]** do this for you. I use Claude's haiku

**[08:50]** model because it doesn't need to be that

**[08:52]** that fancy. So, extract four to six

**[08:55]** visual entities. You can read the rest

**[08:57]** of it later.

**[08:59]** Write the prompt for image generation

**[09:02]** and send it to Google Nana Banana Pro,

**[09:06]** the latest and greatest version, not the

**[09:08]** 2.5, but the 30 Gemini preview. I

**[09:11]** believe it's only visible through the

**[09:13]** API or accessible through the API right

**[09:15]** now, which means it costs like 10 cents.

**[09:19]** So there's a link to Gemini Pro image

**[09:24]** preview. Next week or next month, it

**[09:26]** will be another one. So I just send

**[09:29]** everything there to Nano Banana Pro and

**[09:31]** it produces the image.

**[09:35]** I witness click.

**[09:38]** Oh yeah.

**[09:40]** >> Okay. So, this is what gets sent to

**[09:43]** Google Gemini. The prompts in the file.

**[09:45]** No need to copy it, but it sends all

**[09:47]** this over. It says you need to have this

**[09:50]** in the image. You like it, Bradley?

**[09:52]** >> I do.

**[09:52]** >> Stay tuned for the next slide. Okay. So,

**[09:55]** there's all this stuff. Like I said,

**[09:57]** it's in the slides. You're going to look

**[09:59]** at your your camera afterward. Look,

**[10:01]** want to take a picture of me just

**[10:05]** I mean

**[10:07]** All right. Okay. All right, that's

**[10:08]** enough. Don't run out of film. All

**[10:10]** right,

**[10:12]** Bradley, this one's for you.

**[10:15]** All right, this was that previous slide

**[10:18]** went to Nano Banana whatever pro. Do you

**[10:22]** like it? What What's What sticks out to

**[10:23]** you, Bradley, in this image?

**[10:26]** Obviously, it's tree trimming.

**[10:29]** >> What I thought was real. Looks I mean

**[10:31]** looks believable.

**[10:33]** >> Yeah. You got the little We got this got

**[10:35]** the harness, the ladder, the safety. All

**[10:37]** the entities in

**[10:38]** >> all the all the visual entities in

**[10:40]** there. So for 10 cents and like push a

**[10:44]** button. Cool. Cool.

**[10:46]** >> All right.

**[10:47]** So, uh, let's talk. Can we get a little

**[10:50]** nerdy right now? Let's talk nerdy.

**[10:53]** I know you can't read this, but what

**[10:55]** I've set up is a whole string of AI

**[10:59]** agents doing this stuff for me. All

**[11:01]** right. So if you can imagine you got

**[11:04]** let's say you got uh each of these is a

**[11:06]** person in your organization.

**[11:08]** So we have one that creates a scope for

**[11:12]** the article, the one that actually

**[11:15]** writes it. You have someone that checks

**[11:17]** what they did. You have someone that

**[11:19]** revises the article if it needs to.

**[11:22]** Someone that formats it for the website

**[11:24]** and someone that creates an image and

**[11:26]** someone that publishes it. That could be

**[11:28]** individual people, right? You probably

**[11:30]** have those. So, what I've done is like

**[11:33]** I've trained each one of these with a

**[11:34]** prompt. And if you want to connect with

**[11:36]** me, we can talk about prompts and nerdy

**[11:37]** stuff, but these every one of these

**[11:40]** agents does their work. But do they you

**[11:44]** know how you have a VA and say, "Hey,

**[11:45]** write me an article like follow these

**[11:47]** instructions." Or you have say I know

**[11:50]** where's where's Mike? Mike will send out

**[11:52]** a job uh hunting for jobs and like let

**[11:55]** me know if you read these instructions

**[11:57]** like Philippine jobsonline.ph PH if you

**[12:01]** read the instructions write this in your

**[12:03]** response we've all done that right well

**[12:05]** how do you know that an AI has done that

**[12:07]** job you don't but if it if it hasn't

**[12:11]** right unless in the prompt file say when

**[12:14]** you are done tell output this so every

**[12:18]** step in the stage there's what I call a

**[12:21]** fingerprint it's like yes I've done the

**[12:24]** work and they go okay cool let's go on

**[12:25]** to the next stage so anyway this is not

**[12:27]** meant to confuse you it's meant to think

**[12:30]** of these as people. They're AI agents

**[12:32]** doing work as a person would normally.

**[12:34]** You have detailed instructions. So, it

**[12:36]** goes on bum bum bum bum. You could

**[12:38]** literally do and by the way

**[12:42]** the I have behind this there's a a

**[12:46]** business MD file which has all the

**[12:48]** business information that they're a

**[12:50]** plumber, they're a roofer, they're a

**[12:51]** tree arborist. Here's where they are.

**[12:53]** Here's their avatar. Here's their

**[12:55]** customer ideal customer profile. Here's

**[12:57]** where they work. So all this is in the

**[12:59]** background. So the writer and the scope

**[13:01]** agent know all this in advance. That

**[13:04]** cool, right? So this is literally one

**[13:07]** one click and you can do one at a time.

**[13:11]** You can do 10 at a time. You can do say

**[13:14]** like four clients at one time as well.

**[13:17]** So the only the only limit to this is

**[13:20]** your imagination, right? You would agree

**[13:21]** with me on that one, right? Because who

**[13:23]** wants to be like me? Lazy and Okay.

**[13:27]** Anyone want to be like me? By the way, I

**[13:30]** had a boss one time. He said, "Simon,

**[13:33]** you're the laziest person I know." And I

**[13:36]** knew what he was telling me right off

**[13:38]** the bat. He's like, "Thank you." I was

**[13:40]** one of like the top 10% in his

**[13:42]** organization or something. So, make me

**[13:44]** feel good.

**[13:47]** All right. So, I'm not going through

**[13:48]** each one of these. Like I said, if you

**[13:50]** want to connect with me afterwards, we

**[13:51]** can go through. But what does the scope

**[13:53]** agent do? The layout of the article.

**[13:55]** What sections come first? What comes

**[13:57]** second and third and fourth? The ICP,

**[13:59]** what sections do they have? Here's the

**[14:02]** detailed writing instructions. What is

**[14:04]** the user's intent? You have to answer

**[14:07]** that right out of the gate. Don't stray

**[14:11]** fan out. Anyone heard of fan out

**[14:13]** questions? Yeah. Fan out. Yeah. So, I I

**[14:17]** I saw a post in Signals Lab or something

**[14:19]** a while ago. It talked about fanout

**[14:20]** questions. So, I said, hm, what is that?

**[14:23]** So, I took the fan app questions and

**[14:25]** added it to my agent prompt.

**[14:30]** Don't repeat yourself. So, we've all

**[14:33]** seen AI slop. Yes.

**[14:35]** >> And 4,000 words. Where's Where's Sean? I

**[14:39]** know we we've seen some sloppy AI

**[14:41]** content, right? 4,000 words. Come on,

**[14:43]** man.

**[14:44]** >> Yeah.

**[14:46]** Like, how to fix a a leaking toilet. You

**[14:48]** had your plumbing example, right? Don't

**[14:50]** need seven and a half thousand words.

**[14:54]** What? Come on. Come on, man. Let's I'm

**[14:57]** trying to do my bite my best mic relo

**[15:04]** meet the reader search intent and

**[15:07]** nothing more.

**[15:09]** Right. Answer the fan out questions

**[15:11]** throughout the document. And apparently

**[15:13]** Google, we don't have any anyone here

**[15:15]** that knows Google really well, do we?

**[15:17]** Joy,

**[15:19]** right? Don't turn your content into

**[15:21]** bite-sized chunks. Like, just you don't

**[15:24]** need to force this stuff. Just be

**[15:25]** natural. It will come out once you

**[15:27]** finish answering the searches intent.

**[15:29]** You're done. Nothing else. By the way,

**[15:32]** content that used to rank for thousands

**[15:34]** of keywords in search console doesn't

**[15:37]** happen anymore. Doesn't happen anymore.

**[15:39]** Hat tip to Chris Castillo, one of my

**[15:41]** best friends. He brought this up

**[15:42]** recently. Like, dang. All right. So,

**[15:46]** some more bangers. By the way, does

**[15:47]** anyone know what bangers are besides the

**[15:49]** other Simon in the room?

**[15:51]** >> Bangers and mesh.

**[15:53]** >> Yeah. Yeah. So, bangers are in England

**[15:57]** like like a sausage, but of course also

**[16:00]** be these little firework things. They

**[16:03]** could also be that's a real banger

**[16:05]** that's really good. So, you know, if you

**[16:07]** go to England and you like you talk

**[16:09]** about bangers, they might be confused.

**[16:11]** You're talking about the food about I'm

**[16:13]** I'm a real banger at this or anything

**[16:15]** else. There's another connotation which

**[16:17]** I won't even get into

**[16:20]** that's not for this topic over here. So

**[16:22]** things I I like to do in the content. So

**[16:25]** I love to use tables. Tables and lists.

**[16:28]** Why?

**[16:31]** Google loves data. Why?

**[16:35]** >> Structured easy to read also by people

**[16:38]** as well. If you're looking for an easy

**[16:40]** to read content piece of content tables

**[16:42]** and lists. So bring it on short images.

**[16:44]** We talked about images. I love sections.

**[16:47]** You can read through this at your your

**[16:49]** leisure. How much time do I have left?

**[16:52]** 30 minutes. Cool. Only like 10% through

**[16:55]** my presentation.

**[16:57]** All right. Some So, more website

**[17:00]** bangers.

**[17:01]** We must have been on the same wavelength

**[17:03]** because I know we've been talking about

**[17:04]** reviews in here. So, let me ask let me

**[17:07]** tell you why I do this. Number one,

**[17:10]** because I'm

**[17:11]** >> lazy and also because I'm cheap. Look at

**[17:14]** you guys.

**[17:16]** So, I want to bring all my reviews into

**[17:19]** my website. Now, if I have, say, a

**[17:22]** hundred reviews for plumbing, how is it

**[17:24]** house plumbing? And some are about

**[17:26]** toilets, some about slab leaks, some

**[17:28]** about water heaters, I don't want to

**[17:31]** have every single one of those reviews

**[17:33]** on my toilet repair page, right? I just

**[17:36]** want the reviews which are about toilet

**[17:39]** repair. Yes. Right. So, how do I do

**[17:41]** that? I download download every single

**[17:44]** one of my reviews via code because I'm a

**[17:46]** nerd. I'm lazy. I'm cheap. Is I download

**[17:50]** all the reviews into the back end of the

**[17:51]** website as a custom post type CPT. Cool.

**[17:54]** Then I categorize them with AI. Said

**[17:57]** what is this review about? Generic about

**[18:01]** toilets about showers about water

**[18:03]** heaters. What is it about? And it tags

**[18:05]** it with a category. Now on the service

**[18:08]** pages, let's say it's water heater. You

**[18:10]** like that? Right. So, on your water

**[18:12]** heater page, you have a little code

**[18:13]** snippet that says, "Show me reviews just

**[18:16]** for water heaters." It pulls all those

**[18:18]** reviews in. So, the f the most recent

**[18:21]** ones show up first. So, every time you

**[18:23]** add a new review, that's from from the

**[18:25]** GMBB. Now, you got fresh fresh reviews

**[18:29]** coming in. Cool.

**[18:33]** All right.

**[18:36]** By the way, I used to hate public

**[18:37]** speaking. Now you can't stop me. I used

**[18:40]** to belong for now and all with all

**[18:42]** seriousness I joined a com not company a

**[18:47]** organization called toast masters

**[18:48]** international right if you have a

**[18:51]** passion and you just don't feel like you

**[18:53]** want to speak in front of people I

**[18:55]** literally when I gave my first speech

**[18:57]** was called an icebreaker I I kid you not

**[19:00]** I was in a cold broke out in a cold

**[19:01]** sweat I kid you not I had a jacket on

**[19:04]** luckily it was a jacket because I was

**[19:07]** literally just sweating So, it was it

**[19:10]** was scary, but it was the best thing

**[19:12]** that I've ever Didn't cost me hardly

**[19:14]** anything. So, if you ever want to be up

**[19:16]** in front of people, Toast Masters is the

**[19:19]** thing. It's it's a real banger.

**[19:22]** Okay. Uh service pages. Um Oh, so while

**[19:26]** I'm at generating the reviews, I'll put

**[19:29]** all the ADA compliance in there, do the

**[19:31]** schema for the reviews. So, the whole

**[19:34]** thing is coded out. So, I just do one

**[19:36]** short code and it's done. Okay. How cool

**[19:39]** is that?

**[19:40]** >> Cool.

**[19:43]** >> Simple to vibe code. Yeah,

**[19:45]** >> it is super super simple. If if you

**[19:48]** don't if this confuses the heck out of

**[19:50]** you and go like, "Oh my gosh, my brain

**[19:52]** hurts." Just reach out to any one of us

**[19:54]** in here who's done anything with Vive

**[19:56]** Code, you'll love it. All right,

**[20:00]** another one. I I've hardly seen hardly

**[20:03]** anyone. I I've done this presentation,

**[20:06]** this little snippet before. I hardly see

**[20:08]** anyone doing it, so why don't you? Where

**[20:12]** are you located?

**[20:15]** Oh, where? Here's our address. 123 Elm

**[20:17]** Street. Oh, big whoop. Well, where is

**[20:18]** it? This is GMBB.

**[20:22]** >> I don't know.

**[20:24]** So

**[20:26]** what I learned years ago is that you

**[20:29]** have to if someone's coming to your

**[20:32]** office, let's say it's a dentist or

**[20:33]** what? Let's use dentist is how do I get

**[20:36]** to your office? Well, you drive down

**[20:39]** this road and they say, "Oh yeah, you

**[20:41]** just go past the embassy suites by

**[20:43]** Hilton the Lovefield one where second on

**[20:45]** the left." Cool. Well, put that in your

**[20:47]** freaking website. This is where we are.

**[20:50]** Take this like we are near the unique

**[20:52]** landmark. We are near the Embassy Suites

**[20:55]** by Hilton Dallas Lurfield. Just copy and

**[20:57]** paste from Google Maps. We're near the

**[21:00]** intersection of LemonAve because that's

**[21:04]** what Google Maps has and W Northwest

**[21:07]** Highway Loop 12. By the way, in Dallas

**[21:09]** in North Texas, most roads have like

**[21:12]** five names,

**[21:14]** >> right? Yeah.

**[21:15]** >> City has five names.

**[21:18]** >> Like highway five, right? you know that

**[21:20]** map in up in Plano McKini Allen's like

**[21:23]** highway five and Plano road and I don't

**[21:26]** know like which road are you on I don't

**[21:27]** know so take just copy and paste in

**[21:29]** Google maps and just put it in there

**[21:32]** take pictures and I don't know why you

**[21:35]** guys I'm not saying you guys but the

**[21:37]** majority don't do this take a picture of

**[21:40]** the street where you are here's my

**[21:42]** office from the from the street over

**[21:44]** here by by wherever right here's the

**[21:46]** street sign here's the parking here's

**[21:48]** how you get to our office. Sean

**[21:52]** had some troubles with

**[21:55]** you

**[21:57]** from streets, the office, everything

**[22:00]** around there and you just keep it into

**[22:02]** the it helps with the trust factor a ton

**[22:05]** because they're looking at that street

**[22:06]** viewing this and they're like, "Oh yeah,

**[22:09]** been out there." So what so for the

**[22:11]** people online and Sean are saying yeah

**[22:13]** just take pictures from around the area

**[22:15]** like around where the office is located

**[22:17]** add them to your GMBB or GBP and you're

**[22:20]** good to go.

**[22:22]** Cool.

**[22:23]** >> All right. Um

**[22:26]** >> here's another one that if you have

**[22:28]** plumbers or trades trees whatever you

**[22:31]** want to have you could do the same thing

**[22:33]** with this. Take a picture of the guys

**[22:36]** out there. Hopefully, it's good and it's

**[22:38]** not like a picture of their thumb or

**[22:40]** anything. You take the pictures of the

**[22:42]** job done. Oh, this is one that I I told

**[22:45]** one of my I have an HVAC client. He he

**[22:48]** started his business in July 4th of last

**[22:52]** year. He works part-time as a HVAC. He

**[22:55]** has a full-time job, but he works

**[22:57]** part-time as a HVAC. He started actually

**[23:00]** doing this. He went around neighborhoods

**[23:02]** and is where, let's say we're in um

**[23:04]** McKenna, let's say up there somewhere.

**[23:07]** and there's all these nice neighborhoods

**[23:08]** and you go take pictures of the

**[23:10]** neighborhood entrance sign and he go

**[23:12]** starts posting them to his GMBB profile

**[23:14]** and says, "Hey, I was out in Aubber

**[23:16]** Hills doing like doing an HVAC repair

**[23:19]** and puts it up there." Now you're

**[23:21]** sending signals to Google. Hey, I work

**[23:24]** at this place over here. Then he did

**[23:26]** like YouTube shorts of his jobs. He he

**[23:31]** made now he made let's see he doesn't

**[23:35]** mind me sharing this he was working

**[23:37]** part-time like 10 15 20 hours a week

**[23:41]** kind of right he made made like 60 grand

**[23:44]** in six months first six months now

**[23:47]** that's not that's not going to take uh

**[23:49]** bring him home from his job that was

**[23:51]** working 10 to 15 hours a week from zero

**[23:54]** to 60 grand in six months is that cool

**[23:58]** so there's another thing in here. I

**[24:00]** won't go through this one, but you could

**[24:02]** basically have a a one one second. You

**[24:04]** could basically have a way to upload

**[24:07]** images to the website from the text.

**[24:09]** Have AI take the images, write a little

**[24:11]** blurb, post it to social media, post it

**[24:14]** to your website, post it to GMBB. This

**[24:16]** is where we were doing XYZ in ABC

**[24:19]** Neighborhood. Some questions over there

**[24:23]** on YouTube.

**[24:24]** >> You're going into a neighbor. Wait,

**[24:26]** >> wait, wait. Oh, thank you.

**[24:29]** How to turn it on.

**[24:36]** >> Just clarif.

**[24:42]** >> Yeah.

**[24:42]** >> Come on. There we go. Thank you. So when

**[24:45]** you said add geo in YouTube, you did

**[24:47]** exactly what you said. You go out to

**[24:48]** take the picture of Arbor Hills, which

**[24:50]** is a neighborhood in Ky. Do you geo tag

**[24:53]** the YouTube shorts video to let's say

**[24:56]** that entrance area or to that

**[24:57]** neighborhood?

**[24:58]** >> No. Although the the images that you

**[25:02]** take that still images, they will have

**[25:04]** XF data and stuff in there right now.

**[25:07]** People say that Google strips the XF

**[25:09]** data. Well, it does, but what happens

**[25:11]** before it strips the XF data?

**[25:14]** >> It's it's evidence. No, what I'm saying

**[25:16]** on YouTube there is a underneath there's

**[25:19]** a I guess it's a category you can have

**[25:21]** and say where was it taken, where was

**[25:23]** the video taken and say Mckin or Dallas

**[25:25]** or wherever you are.

**[25:26]** >> Okay. So, but if you wanted to you could

**[25:28]** go as deep as say if you were in Arbor

**[25:30]** Hills to confir if you want to go that

**[25:32]** granular. Yeah, but you can put it in

**[25:35]** there but again it's like cheap and

**[25:37]** lazy. Well, how much do you do? Right.

**[25:39]** So, yeah, go for it. Right. Thank you.

**[25:42]** Yeah, there's another question ma right

**[25:45]** right behind you.

**[25:48]** >> So, do you go through the trouble of

**[25:50]** changing any other um elements of the

**[25:53]** exit data? So, like ITC tags or titles,

**[25:57]** captions?

**[25:58]** >> No.

**[25:58]** >> No. Okay.

**[25:59]** >> No. I mean, why should if if my guy is

**[26:02]** out there in his truck? Oh, sorry. I

**[26:05]** missed out an important point. So, he

**[26:07]** parks his truck. So ABC HVAC in front of

**[26:11]** the Arbor Hills sign. So now you've got

**[26:14]** his truck with his with his logo, his

**[26:18]** his his um website, his brand name in

**[26:22]** front of Arbor Hills McKini. Right. So

**[26:25]** now you've got entities like backing up

**[26:27]** another entity. So it's evidence brand

**[26:30]** association. See? Thank you. Right. Is

**[26:32]** that cool? Right. So where's the mic?

**[26:42]** I assume these are service area

**[26:43]** businesses. Have you looked at all to

**[26:46]** see like the polygonal shape for your

**[26:49]** service area and if posting this kind of

**[26:52]** content will push the lines around on

**[26:54]** your service area at all?

**[26:56]** >> It seems like it should.

**[26:58]** >> Well, so all I can all I can tell you is

**[27:01]** this just a it's a test of one. So just

**[27:04]** take it with a grain of salt. But he's

**[27:07]** ranking like in let's call it M Mc Mckin

**[27:09]** is a really affluent area by the way. So

**[27:11]** I I'll use that as the example went from

**[27:13]** zero to having lots of ones across like

**[27:17]** call it 10 mile by 10 mile. So it

**[27:20]** definitely works. So I think the two

**[27:22]** things that he was doing and Mike Molina

**[27:24]** will also tell you do your YouTube video

**[27:27]** shorts be be active on social. So, he

**[27:29]** does his YouTube shorts. He does take a

**[27:31]** whole bunch of pictures and he gets

**[27:33]** reviews and he replies to the reviews.

**[27:37]** Okay. Does that answer your question?

**[27:41]** >> Okay. You think so? Okay.

**[27:43]** >> Yes.

**[27:44]** >> Yes. All right. Can you guys take this

**[27:45]** home and implement this tomorrow or

**[27:47]** tonight?

**[27:48]** >> For sure you can.

**[27:49]** >> Hard to get the clients to do it though.

**[27:51]** With tree service, I've been telling

**[27:52]** them this for five years and I could

**[27:54]** maybe 20% of them will actually do it

**[27:56]** consistently.

**[27:57]** >> Yeah. No kidding. Tell tell them that

**[27:59]** tell them that that house pros and

**[28:02]** they're forced to with house pros. It

**[28:04]** forces you when those guys

**[28:05]** >> No, but I sell services.

**[28:10]** >> Hey, real quick. Simon sir

**[28:12]** >> and Brad like you do lead gen, bro. You

**[28:15]** do AI. I mean, sorry, stock photo, flip,

**[28:18]** screenshot, then add in, right? I've

**[28:20]** been doing it for years, bro. When I

**[28:22]** take a van,

**[28:23]** >> I don't got no fleet, but Google thinks

**[28:25]** I have 74 vans in a certain geo. That's

**[28:28]** what they think cuz I put a number on

**[28:30]** the pan. Like a real business foot,

**[28:32]** right? You know what I'm talking about?

**[28:33]** Yeah.

**[28:34]** >> Yeah. Crew four was over here. Crew 19

**[28:36]** was over there.

**[28:39]** >> Yeah.

**[28:40]** >> So, I kind of What? How's my time?

**[28:44]** >> I'm good. I got another hour left. Cool.

**[28:47]** All right. I'm like

**[28:49]** >> another 400 slides to go. All right. By

**[28:51]** the way, I I I halfway joke about having

**[28:55]** so many slides, but I was sitting down.

**[28:58]** I was at home yesterday and doing some

**[28:59]** stuff like, "Let's add a few more

**[29:00]** slides." Like, you know how that goes,

**[29:02]** right? I I love to give because it's a

**[29:05]** community that when I first got started

**[29:08]** in SEO, there's a a Facebook group

**[29:10]** called Local Client Takeover, right?

**[29:13]** True story. So, I'm being authentic here

**[29:15]** and being vulnerable. Is I went in

**[29:17]** there, I knew I didn't even know what a

**[29:19]** GMBB was. didn't know what it stood for.

**[29:21]** I knew nothing about I didn't know the

**[29:23]** questions to ask. So, I was in there

**[29:26]** being one of these lurkers and going,

**[29:28]** "Dang, I need to know about this SEO. I

**[29:29]** probably can make some good money with

**[29:31]** SEO." So, I I went in there reading the

**[29:34]** stuff and someone had a question like

**[29:37]** that seems like a really good question

**[29:39]** to ask. Like, I have no idea what

**[29:40]** they're asking, but let me go off and

**[29:42]** like research it. So, I went to Google,

**[29:44]** went to YouTube, like, oh, that's what

**[29:47]** they're asking. And I came back and

**[29:50]** understand please this is not me trying

**[29:52]** to brag to you or to the people in local

**[29:55]** client takeover. I said oh by the way an

**[29:58]** this is the answer to your question I

**[30:00]** found is like x y and z and an goes well

**[30:02]** thank you so much that's really cool. Oh

**[30:04]** no problem. And then someone else had a

**[30:07]** question like and I did the same thing

**[30:09]** like why is it so hard? Why can't you

**[30:12]** just go to Google and YouTube and find

**[30:14]** this for yourself? What's what's wrong

**[30:15]** with your hand? Why can't you find this?

**[30:18]** After about a month,

**[30:20]** it was like so crazy. People saying,

**[30:22]** "Simon, you know this stuff really well.

**[30:24]** Can you help me out?" Like, dude, I've

**[30:26]** been here for like I'm I'm still still

**[30:28]** crawling. I'm still wet in my diapers

**[30:30]** type of thing. But it kind of gave me a

**[30:33]** reputation, which was really weird for

**[30:36]** me because I didn't see myself as a

**[30:38]** guru, but I I was trying to give back to

**[30:41]** the community. And that's what I'm here

**[30:43]** today to give to you is give back. So,

**[30:45]** there's a bunch more slides in here. You

**[30:47]** won't see them here live, but you'll see

**[30:49]** them in the in the file later on. And

**[30:50]** like I say, if you want to connect with

**[30:51]** me, find me. I'm I'm not that active on

**[30:54]** Facebook, but you can still find me. So,

**[30:57]** I won't reply like in five minutes, but

**[30:59]** maybe a day or two, maybe. Cool. Uh, all

**[31:03]** right. Again, lazy and cheap. I don't

**[31:06]** use WordPress themes anymore in new

**[31:07]** sites because

**[31:10]** a I'm lazy and b I'm cheap more like

**[31:15]** lazy like divvy like Sean loves divvy. I

**[31:18]** used to like evader and I used to like

**[31:20]** divvy but man it's like by the time I

**[31:22]** want to change the color in the

**[31:24]** background of something like oh my gosh

**[31:27]** how hard can it be? Here's the thing. AI

**[31:32]** can write really, really good HTML.

**[31:37]** So if you give it precise instructions,

**[31:39]** guess what? AI can update HTML much

**[31:43]** faster than you can without using page

**[31:46]** builders, right? So why don't we do it?

**[31:50]** So I've now replaced all the Yoast and

**[31:53]** SEO Press plugins. I do a bunch of nerdy

**[31:56]** short codes for really cool stuff and I

**[31:59]** use AI a lot. You know the drill.

**[32:04]** Okay, so that's kind of like section

**[32:07]** one. How am I doing on time?

**[32:08]** >> You're doing good.

**[32:09]** >> It's good. What does that mean?

**[32:13]** >> 30 minutes left.

**[32:14]** >> 20 at least 20. Then we're going to take

**[32:16]** a break. So

**[32:19]** >> cool. All right. Back in the old days,

**[32:23]** anyone kind of familiar? Who's who's old

**[32:25]** like me?

**[32:26]** >> Right. So, back in the back in like this

**[32:29]** is where you show your grandkids this

**[32:31]** like hey back in the day we used to have

**[32:33]** like a phone that was plugged into the

**[32:35]** wall like for real like how did they go

**[32:39]** outside without a cell phone like like

**[32:41]** they just can't fathom it. So, believe

**[32:45]** it or not they used to use like IBM XTS

**[32:49]** is that 286 chips from Intel. You have

**[32:52]** someone over here writing their specs

**[32:54]** for the future of someone

**[32:57]** writing it, someone reviewing it,

**[32:59]** someone testing it. Reams and reams and

**[33:02]** reams of paper. This is probably not too

**[33:06]** far out from what it used to look like.

**[33:08]** A bunch of people in a room, you know,

**[33:10]** got the 80s stuff going on there. It it

**[33:13]** was like this. I used to work at a

**[33:15]** company. We would have a group called uh

**[33:19]** they were like systems engineering or

**[33:22]** something and they would create this

**[33:23]** product requirement document the big big

**[33:26]** thing. Then we'd give it to the

**[33:27]** engineers and the engineers would write

**[33:29]** feature requirements. Then we give it to

**[33:31]** the coders and the coders would code it.

**[33:33]** Then it would go to testers and they

**[33:35]** would test it. And it was a good process

**[33:37]** to have when you when you're coding at

**[33:39]** scale. It worked and it worked really

**[33:41]** well. So, but we don't have we don't

**[33:45]** need all these people today because we

**[33:47]** have AI.

**[33:49]** Cool. So,

**[33:52]** I I dislike vibe coding. Not to upset

**[33:56]** anyone in this room because you can get

**[33:58]** results with it, but there's also a

**[34:00]** level of frustration, right? You if

**[34:02]** you've been there before. Chad, do you

**[34:03]** vibe code?

**[34:06]** >> Depends on how you define it, but I

**[34:07]** definitely use code these days.

**[34:10]** >> Yeah. So here's here's where most people

**[34:13]** start because they've heard about the

**[34:15]** latest fad going on with AI and it can

**[34:17]** do all this wonderful coding like hey

**[34:19]** let me write this this this like feature

**[34:21]** to do X Y and Z. Bam off it goes and

**[34:24]** then go like

**[34:27]** it's not like how I wanted it to turn

**[34:29]** out. Then you go back and make some

**[34:31]** tweaks like h I guess it's good enough.

**[34:35]** Well here's what you need to do, right?

**[34:38]** Let's get nerdy, but I want to make you

**[34:40]** feel uncomfortable, then give you a

**[34:42]** solution. Is that okay? Right. Is it

**[34:44]** okay being uncomfortable to get better?

**[34:46]** Okay. How many of you work out?

**[34:50]** Brian, why don't you work out? Uh, by

**[34:52]** the way, shout out to to Brian Hung at

**[34:54]** the back. He's like one of the most

**[34:55]** nerdiest fitness guys on the planet I've

**[34:58]** known. He's like, "Oh my gosh." So, so I

**[35:02]** love stories because stories tell or

**[35:05]** sorry, stories sell, facts tell.

**[35:08]** the So, we're in this group every

**[35:10]** Friday. We're talking about health. I'm

**[35:11]** sitting there in my office chair going

**[35:13]** like, I haven't moved for like three

**[35:14]** hours. I need to get up and move around.

**[35:17]** So, like Brian goes, "Yeah, I've been

**[35:19]** doing this and that." Okay, that's cool.

**[35:22]** I'm feeling a little bit bad, but also I

**[35:24]** know in my head I need to move around a

**[35:26]** bit more. Go like, "You know what? I

**[35:28]** went to I went to see a nutritionist

**[35:30]** because it was covered in my insurance."

**[35:32]** And a nutritionist said, "Hey, would you

**[35:34]** like to have to go see a personal

**[35:36]** trainer twice a week? doesn't cost

**[35:37]** anything. Like, okay, cool. So, I see a

**[35:40]** personal trainer now for six weeks. I've

**[35:42]** been seeing him since uh April of last

**[35:45]** year. Doesn't cost me anything. But man,

**[35:48]** do I feel so much better now. So, I

**[35:51]** mean, it beats sitting around all day

**[35:53]** long. So, non SEO, non nerdy tip, don't

**[35:56]** forget about your health, guys. It's the

**[35:58]** only health that you have. You'll have

**[35:59]** you you you'll never get I know there's

**[36:02]** probably a better way of saying this,

**[36:03]** but this you could be in the worst shape

**[36:05]** today and then six months from now be in

**[36:07]** the best shape ever. So I'm probably in

**[36:09]** the best shape I've been in about 10

**[36:10]** years, right? And I feel great. So thank

**[36:13]** you, Brian. Back to this.

**[36:17]** Run a PR product requirement doc. Is

**[36:20]** that right, Chad?

**[36:21]** >> Yeah.

**[36:22]** >> Turn it into tasks. We we talk in

**[36:24]** acronyms. Turn to task. Write the code.

**[36:26]** We have another LLM verify it and then

**[36:30]** there's this thing called Ralph Wigum

**[36:32]** which I'll explain on the later slide.

**[36:34]** Does anyone before like last week or

**[36:36]** last month does anyone know who Ralph

**[36:38]** Wigum is?

**[36:40]** >> Simpsons.

**[36:43]** All right. So here's the shift that

**[36:45]** we're seeing in in effort allocation. By

**[36:48]** the way, this slide was created just a

**[36:51]** cool tip. If you go to Gemini

**[36:54]** Pro think three, there's a drop down and

**[36:57]** then you can do what's it called? Um

**[37:01]** canvas and you say I want to build a

**[37:03]** presentation and you drop in everything

**[37:05]** it will present a PowerPoint slide or

**[37:08]** create a PowerPoint slide for you. It's

**[37:09]** really good.

**[37:11]** So most people when they buy code little

**[37:13]** bit of planning some coding and they're

**[37:16]** like oh man 70% debugging and testing

**[37:18]** and getting frustrated.

**[37:21]** The right way to do a big pro a bigger

**[37:23]** project is let's plan and spec it out.

**[37:27]** Let's implement it and that's maybe like

**[37:29]** 10 20% of recode. So now you're asking

**[37:32]** me

**[37:39]** how do I build a PRD? I know that's only

**[37:41]** a tip of mind, right? How do I build one

**[37:43]** of these PRDs? Well, everything's in the

**[37:45]** slice, by the way, unless you're just

**[37:46]** taking pictures of me secretly on my

**[37:48]** own. Okay, so you build a PRD. It

**[37:51]** launches in an agent. You can use claw,

**[37:54]** you can use cursor, you can any one of

**[37:57]** these is pretty good. It builds a PRD.

**[38:00]** It builds a set of discrete tasks for

**[38:03]** for the coder to do. And each task has

**[38:06]** define inputs and output success

**[38:08]** metrics. You got to inspect what you

**[38:11]** expect all the time. So then the AI

**[38:14]** codes it then the output is inspected or

**[38:17]** audited by a new AI session. That's

**[38:19]** really important. So you have here's the

**[38:22]** thing, right? If any of you have ever

**[38:23]** like tried to like do something on the

**[38:26]** website and you're like, "Yeah, this

**[38:27]** should work." And you sit there and

**[38:28]** you're going around and around and

**[38:30]** around and like I can't see what's going

**[38:32]** on. Have we've been there before? Like

**[38:35]** you just can't see the forest for the

**[38:36]** tree. Like why on earth is this not

**[38:39]** working? It should work. You spent an

**[38:42]** hour then you say hey an can you take a

**[38:45]** look at this like what's going on and an

**[38:46]** goes you missed a semicolon like like

**[38:51]** duh when you're doing doing schema if

**[38:53]** you've missed a semicolon in schema you

**[38:55]** know what I'm talking about comma or

**[38:57]** something so feedback so you have a

**[39:01]** little bit of human verification now if

**[39:03]** you want this PRD generator you go here

**[39:07]** it's a it's a GitHub thing so I I love

**[39:09]** GitHub it's in the slide BK so you can

**[39:12]** see all the slides. So, oh, he was just

**[39:15]** taking a picture of me to post up like,

**[39:17]** "Hey, I saw this really cool guy like

**[39:20]** cool accent." It's like, "Man, this guy

**[39:22]** is is is the is a real banger over

**[39:24]** here." Not don't say that. Don't say

**[39:26]** that.

**[39:28]** So, go to GitHub. You can see it. You

**[39:30]** can install it. It's it's a skill. You

**[39:31]** can put it in Claude Code. I love Claude

**[39:34]** Claude Code. So, you can put it in

**[39:36]** there. It helps you make coding so much

**[39:38]** more fun. So, who is Ralph?

**[39:42]** I I don't know why the guy had this. I I

**[39:45]** don't know. Maybe he's a he's some nerd

**[39:48]** who likes the Simpsons. So, you can read

**[39:52]** all this. But basically, we have this

**[39:54]** feedback loop, right? When you code

**[39:56]** something, how do you know it's correct?

**[39:59]** If you have the same coder inspecting

**[40:02]** their own work and going, "Oh, it's has

**[40:04]** some mistakes. Let me go fix it again

**[40:06]** and again and again," they get tired.

**[40:09]** They make mistakes. They don't see it.

**[40:11]** So what the Ralph Wiggum

**[40:14]** plug, not plugin, because there's one

**[40:16]** for Claude doesn't do this. But what it

**[40:18]** does, it takes a new agent, as I say, it

**[40:21]** takes an as a new tester and an looks

**[40:24]** through it again. It's like, oh, I found

**[40:25]** these issues. So then you feed it back

**[40:28]** and it goes to Bradley like Bradley, is

**[40:30]** this okay? And Bradley goes, yeah,

**[40:31]** that's good. Okay, fantastic. So that's

**[40:33]** Ralph Wiggum. kind of nerdy, but it's

**[40:36]** it's important to know about

**[40:39]** uh AI context. Rock, I explained this

**[40:43]** before is that when you're chatting with

**[40:45]** a with an AI, when it's got through

**[40:48]** maybe half of its context, it's going

**[40:50]** like it gets lazy. It gets confused.

**[40:52]** It's like, man, I'm tired. It makes

**[40:54]** mistakes. So,