In this video, I'm going to teach you

how to use Claude Code so that you can

do your marketing, your sales, your

customer experience, your product work,

everything that you can imagine that is

tangentially related to growing your

business. This is often called GTM

engineering or go-to-market engineering.

And in this video, I'm going to show you

how to get started in the next 10

minutes. And once you set this up,

you're going to have the ability to be

able to actually get Claude Code to do

work for you. I'm talking about doing

keyword research for you, then writing

the blog post, then actually publishing

the blog post. I'm talking about doing

research on what you should be talking

about within your Facebook ads, actually

making those Facebook ads, publishing

those to Facebook, and then analyzing

which of those ads are the low

performers and the high performers. This

isn't a skill. This is actual work

getting done. And that's the biggest

contrast. And with that said, I'm Cody

Schneider. I'm the co-founder of a

company called graph.com. I've been

doing marketing for over 15 years. And

in this video, I'm going to give you

everything you need so that you can

start doing go-to-market engineering

using Claude Code. I'm going to show you

how to set it up. I'm going to show you

why you should go immediately to this

instead of Claude Co-work or in 8 n. And

I'm going to teach you how you can even

take this further and actually build

personal software and personal agents

for yourself to help get your job done.

And real quick, before we get started,

this episode is brought to you by

graph.com, the AI data analyst for

go-to-market teams. Unify all your data

into a single place and then use it to

analyze that data to turn it into

information and then build reports and

live dashboards based off of it. You can

also connect it via MCP so that you can

use it within Claude Code. All right,

let's jump into it right now. All right,

to begin, let's just talk about what GTM

engineering actually is. Just to give

context. So, it stands for go-to-market

engineering. And traditionally, it used

to mean just what would be involved in

the process of like cold email or reach

out. Basically, everything in

relationship to that. Clay really kind

of solidified or owned this word

initially, clay.com, for this GTM

engineering role. But now it's quickly

evolving past that, further past that,

into okay, everything related to

marketing. So, this could be customer

experience, this could be product, this

could be paid ads, this could be SEO,

anything that's involved in that

go-to-market process that you're using

AI to do the engineering, the software

development, the build-outs for. This is

how this is fitting into it. So, just to

talk through what this actually means in

concept, like when I say GTM

engineering, what am I actually

describing? It's all of the middle work

that previously you would be like

hands-on keyboard touching or touching a

mouse to get done. And what your role

now is turning into is you have ideas,

all that middle work you hand off the

Claude code, and then you're the polish,

you're the endpoint. And we're going to

show you we're going to I'm going to

walk you through this whole process on

like how to actually do this, but I want

you to just think in your mind, okay,

I'm just trying to get Claude code to do

the work for me. That is what I'm trying

to get to happen. It could be other

agents as well. This is just what we're

going to be doing today. All right, and

that's all the context that you need to

get started. Now, let me show you how to

actually set this up so you can start

doing GTM engineering and then give you

an example of what's possible with this.

So, to get started, the first thing that

you need to do is build a folder that

you're going to be working out of. So, I

work out of this folder called Graft

Growth Agents, and just for this today,

I'm going to create a new folder called

demo. And then once I've created that

folder, I'm then going to go and I'm

going to work either out of the Claude

desktop app or I'm going to work out of

the Claude terminal. So, I just select

that folder, and at this point I need to

do two things to be able to start

actually getting Claude to start doing

work for me. I need to set up what's

called an environment file where all of

my API keys for my entire stack is going

to be stored. And then I need to set up

what's called a Claude.md file that

tells Claude, whenever I provide a new

API key, add it to that environment file

so that all of these get stacked

together so they can be reused in the

future. So, you can work out of the

Claude desktop. I personally don't do

that. I work out of the terminal so that

I can have multiple windows open

simultaneously and be jockeying between

them, going back and forth. So, let me

show you how I actually work. I'm going

to go and I'm going to open up a

terminal. And once I'm in that terminal,

I'm going to then get into that folder

that we just made, that demo folder. I'm

going to do CD documents {forward slash}

demo. And now that I'm in that folder,

and this CD command is just a bash

command that gets you there. This is

again, super easy to get started and to

do. I'm going to just type Claude.

That's all I have to do to get started

to start using Claude. And once I'm

here, I can literally tell Claude, "Hey,

build that environment file and add to

the Claude.md that anytime I provide an

API key to add it to that environment

file." And this is again, just the

foundational framework of the what's

necessary to get Claude to be able to

start doing this work for you, this

middle work for you. So, how I actually

do this is I use transcription software,

it's called Super Whisper, shoutout

Neil. And right now I'm about to press

option space bar, and that's going to

start this transcription software. I'm

going to say, "All right, make an

environment file for me, please, within

this directory." And whenever I say

directory, that just means folder, just

for you guys that are listening. And

then also add a Claude.md file and

include in for include instructions

within that Claude.md file that anytime

that you get an API key or I provide an

API key, you add that to the environment

file so that it can be used in the

future as well. And now I'm just going

to hit send and I'm going to let it do

that work for me. All right, at that

point, that is all you have to do to get

started working on this.

Infrastructurally, you are good to go.

And just to cast the vision, I would now

then go open like five different

terminals like this, and I would be

bouncing back and forth between all of

them. And I'm just jockeying the agents

to get that work to be done for me. So,

what's an example of that work? Let me

show you an actual actual example. What

we're going to do is research keywords.

We're going to then write a blog post.

We're then going to publish a blog post,

and that whole thing is going to be done

by Claude in the next 5 minutes. So, to

do that, let's make a new chat, and I'm

going to CD into another folder, and

this is the one that I actually work out

of that has all of my API keys stored in

it. And then I'm going to launch Claude,

and then I'm going to tell it what I

want it to do. So, I'm going to tell it

use the Keywords Everywhere API and find

all of the versus keywords that you can

for Looker Studio verse Y. So, it'll be

like Looker Studio ver- or X verse Y

style keywords. All right, so I just

transcribed that and hit send message,

and I'm now going to start letting that

work. So, what's about to happen is

Claude knows that the Keywords

Everywhere API is available. It's going

to go and do research for what it needs

to do to find the keywords that I told

it to look for, those X verse Y

keywords. And then once I have those

keywords, I'm then going to go and I'm

going to have it write a blog post based

off of what is ranking currently on page

one of Google for that target keyword

phrase. And again, at this point, I

would actually start working on

something in the background. I would

have another window open, and I would be

bouncing back and forth between those.

Or just to break the fourth wall, what

I'm actually doing right now is editing

the intro of this video in the script

while I'm waiting for this to work. But

this is the core concept, this is the

core idea here is you're basically

trying to get these agents to go and do

that work for you, the actual touching

of the keyboard work, and then you are

just that conductor and that curator of

the actions that it's doing. And if

you're wanting to learn how to do GTM

engineering like this, please go to

gtmengineeringcourse.com.

It's totally free. It's brought to you

by me and my company graph.com. It's

literally just tutorials like this

showing you how to actually do this

work. And if you also want more content

like this, go and subscribe to my

YouTube channel. And at this point, it's

brought those keywords back in. And now

let's go and actually make a blog post

out of this and publish this to the

website using this tooling. So, I'm just

going to pull out this Looker Studio

verse Metabase, and then I'm going to

prompt it to write this blog post based

off of the specifications that I want it

to hit. So, I'm going to say, "Write a

blog post about this target keyword. The

article should be 1,500 words long. I'm

going to

I'm going to provide source material for

you. The article should be 1,500 words

long. I'm going to provide source

material for you to write based off of."

And then the source material that I'm

going to provide is going to be what's

ranking on page one of Google for that

target keyword phrase already. And the

reason for this is because what's

currently ranking, Google is basically

signaling this is what we think a good

search result looks like. Kind of

hilarious, Meta Base has the both of

these variations, which I find funny.

And the reason that we're doing this is

because Google is already signaling that

these are the types of pages that it

thinks are a good search result for us

to write based on. Now, you can get way

more technical into this. I'm not going

to go there in this video. The best way

that you can do this actually is you

basically have an AI interview you for

like 30 minutes. You take that

transcript, and it's your perspective,

your opinions, your point of view on the

specific content category topic, and

then you incorporate that into the

source material as well. So, the

scraping that you did, you incorporate

that along with your own tone of voice,

and then you probably have a style guide

that you would incorporate as well. And

then have it write based on all of that,

and that's going to create the best

output. But again, today, this is going

to be pretty good, and these will rank.

But if you want to take this further,

that's how you can make this like an

unbelievable piece of content is by

doing what I just said. And at this

point, I've extracted all of that

content, that source material. I'm going

to put that into the context and write

based on this. And it's going to go and

write that 1,500-word article. Once we

have that, we'll then go and publish

that to our CMS. We use Strapi

internally, but this could work into

Webflow, this could work into WordPress,

this could work anywhere that you

publish content. It's the same process,

the same strategy, the same idea. And

right now, I'm just talking about a

single article. But you can do this for

every one of these keywords. I can tell

Claude, go and write this, do the same

process for every keyword, and publish

this for every keyword. And so, I want

you to expand your mind on like what's

possible here, and like how much work

you can get done by going through this

process. And I know there's going to be

people listening to this right now and

be like, "Oh, this is AI slop. There's

no way that you can create good content

out of this." I totally disagree. I

think that that is a skill issue. You

can 100% get good content to come out of

this. You are just not providing it

guardrails of what good content looks

like. And if you think that your job is

not going to be augmented in some way by

this, you're getting high on your own

supply. I promise you this is going to

affect every person that's doing any

type of growth role. All right, so it's

finished writing that article. I'm now

going to prompt it. Now publish this

article to our blog using the Strapi

API. And then the other thing I'll say

on people that say this doesn't work

with this content doesn't work,

especially on the SEO side, is they're

not even refreshing the content. They're

not improving it. And I'm going to show

you how we can literally have it improve

the content based off of the data that

we're getting back from Google Search

Console for those specific articles. All

right, I just published that blog post.

I just completed it. So let's go look at

that now. And it's live on the site. So

at this point, I would now want to start

tracking this campaign that I'm doing

this versus campaign. So this is how I

would accomplish that. I would go to

graph.com and I would select the data

source Google Search Console. And then I

would just tell it to make a dashboard

for this. So it's going to just make a

line chart for me. And while it's

working on that, I'm going to show you

how I can use the Graph MCP to improve

these articles as well using Claude

code. So within that same chat, I can

now say use the Graph MCP and find the

other versus articles in the blog. I

want you to use the Google Search

Console data connection and pull the top

five versus pages. I want to find the

keywords that are related to those top

five pages and give me insights on how I

can optimize those pages so that they'll

perform better. Look at both the content

and also the data from Google Search

Console. All right, and coming back to

that dashboard that's now built, I can

drag that over and filter this by daily,

weekly, monthly. I can also share this

with my team by creating a link. But

instead what I'm going to do is add a

scorecard so I can see the count of of

the impressions. So I'm going to say add

a scorecard showing the impressions for

blog content URLs that contain versus.

And then I hit send message and I'm

going to go back over here and see what

it's working on. So it's going to do

this analysis for me. And let's pull

this over to this side. And then I'm

going to click into this chart. I want

to add clicks as well, so I'm going to

tell it to do that. Please add clicks to

this chart as well. And actually, just

for the sake of time, let's just find

the top five versus articles, and then

show me the keywords related to each of

those URLs from Google Search Console.

Google Search Console is the data source

that you I want you to pull via the

Graph MCP. And then give me some insight

on how I can optimize those blog posts

based off of the keyword data that we're

getting from Google Search Console. And

now I'm back over at that dashboard.

Clicks are way smaller, it's way lower.

So, I'm going to say, "Move clicks to

the right side of the Y axis." So, left

side will be impressions, right side

will be clicks. And we can see now,

through the Graph MCP, it's pulling in

that data that I asked for, and it's

going to do that analysis for me within

Cloud Code. And this is how you can

continuously improve your marketing

using all this tooling in parallel, and

also handle the reporting for all of the

work that you're doing on the

go-to-market side. You can see it made

that change, shifting that over to the

right side. And you can get started with

Graph for free at graph.com. It has a

14-day free trial that comes with 10

seats for your team. Everything that I

just showed you, you can do in Graph.

And if you want more content like this,

please, again, go to

gtmengineeringcourse.com

and join the newsletter. It's free. I

will just send out videos like this. And

also, go subscribe to the YouTube

channel if you want more content like

this. If you have questions, comment on

this And if you have questions, comment

wherever you're watching this. And if

you want even more content around

go-to-market engineering, go follow me

on Twitter, or follow me on LinkedIn.

And finally, to conclude, it just did

this analysis [laughter]

for me

based off of my live data.

I still can't get over this whenever it

happens. It's basically telling me how I

can can improve this. So,

absolutely crazy. All right, I'll talk

to you next time.