Big Idea

The argument in one line.

Posting volume tells you nothing about audience quality; the only signal that matters is whether the people commenting on your content match the buyer profile you actually want to reach.

Who This Is For

Read if. Skip if.

READ IF YOU ARE…

You post on LinkedIn regularly but cannot tell whether your audience is buyers or competitors watching you.
You have an offer and want data on which content angles attract the right commenters.
You are learning Claude Code and want a practical end-to-end project that produces something immediately usable.
You are willing to spend roughly $5-6 on Apify scraping credits to get 30 days of real audience data.

SKIP IF…

You do not have an existing LinkedIn presence with enough comment history to analyze.
You want strategy theory rather than a hands-on screen-share build.

TL;DR

The full version, fast.

Most LinkedIn creators optimize hooks when the real problem is audience mismatch. This video walks through a five-step Claude Code pipeline: define your ICP from your offer description, scrape 30 days of posts and comments via Apify, classify every commenter as buyer/peer/creator/competitor/unknown, render a branded HTML dashboard, and generate a content brief with three angles to lean into and three to cut. The live result: 44% buyer commenters, 40% competitors, and the discovery that format was not the problem.

Free for members

Chat with this breakdown — free.

Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.

Create a free account →

Chapters

Where the time goes.

00:00 – 00:36

01 · Claude Code Reads Your LinkedIn

Hook and promise: Claude analyzed 1,500 posts and identified which ones to stop writing.

00:36 – 00:55

02 · Setting Up Claude Code

Open Claude Code desktop app, create a new project folder called LinkedIn Analyzer.

00:55 – 01:57

03 · The Full Build Game Plan

Six-step plan: define ICP, scrape posts and comments, analyze job titles, classify against ICP, render dashboard, extract content patterns.

01:57 – 03:00

04 · How to Define Your ICP

ICP Discovery prompt: feed Claude your offer description, it returns buyer psychology, trigger event, exact language, negative ICP.

03:00 – 04:03

05 · How to Set Up Apify

Apify is a scraper marketplace with a Claude Code MCP connector. Free $5 per week. Connect via the connectors menu.

04:03 – 05:18

06 · Scraping Posts and Comments

30 days of LinkedIn posts and comments scraped to CSV. Result: 45 posts, 1,586 comments, 1,135 unique commenters. Cost: $5.67.

05:18 – 06:18

07 · How to Classify Commenters

Commenter Classifier prompt reads CSV, checks job title and bio against ICP, assigns BUYER/PEER/CREATOR/COMPETITOR/UNKNOWN with confidence.

06:18 – 06:31

08 · Analyzing the Audience Results

44% buyers, 40% competitors and peers. Hooks landing in the industry echo chamber. Fix: shift CTAs away from founder-flex content.

06:31 – 07:02

09 · How to Build the Dashboard

Audience Dashboard prompt: buyer % headline, commenter breakdown chart, best/worst posts by buyer engagement, 30-day trend line. Single HTML file.

07:02 – 07:44

10 · Adding Your Design System

A design system file (bold retro typography, orange/cream/black) passed to Claude Code so the dashboard renders on-brand.

07:44 – 08:22

11 · Reviewing the Live Dashboard

42% ICP engagement displayed. Full post-performance table, buyer trend line, best/worst posts with audience split percentages.

08:22 – 09:36

12 · How to Generate Content Brief

Content Brief Generator: top 5 vs bottom 5 posts by buyer engagement, pattern extraction, 3 angles to lean into, 3 to cut, reusable template appended to the dashboard.

09:36 – 09:57

13 · Content Angles and Template

Key finding: format was right, messaging was wrong. Beginner language attracts buyers; flexing repels them. Draft Monday post generated.

09:57 – 10:13

14 · How to Get All Prompts

CTA: all prompts and pre-built dashboard inside Claude Code Club ($9/mo on Skool).

Atomic Insights

Lines worth screenshotting.

Optimizing hooks when you have an audience-mismatch problem is like fixing your delivery when no one wants the product.
Classifying commenters by job title and bio rather than engagement count is the only way to separate buyers from spectators.
A single $5 Apify scrape of 30 days of LinkedIn data produces 1,135 unique leads worth far more than the cost if even one converts.
44% buyer commenters sounds good until you notice 40% are competitors and peers watching the space.
Format rarely explains LinkedIn performance gaps; the variable that actually moves buyer engagement is who you are speaking to.
Beginner-facing copy outperforms founder-flex copy for attracting paying customers from educational communities.
A design system file passed to Claude Code lets it generate branded HTML dashboards and landing pages without manual CSS work.
Reverse-engineering your ICP from your offer description forces Claude to surface the exact language buyers use to describe their own problem.
The negative ICP is as important as the positive one because those people are already in your comments distorting your signals.
A content brief generated from your own data is more actionable than a generic strategy session because it is grounded in what already worked.
The pipeline here is reusable monthly, turning audience quality into a trackable metric rather than a gut feeling.
Writing for the person who has not started yet, rather than for your peers, is a concrete content directive that removes guesswork about what to post next.

Takeaway

Your comments are a lead list you have never mined.

WHAT TO LEARN

Most creators optimize the wrong variable -- post format stays consistent while the audience receiving it drifts further from people who will ever pay.

01Claude Code Reads Your LinkedIn

A hook built around a specific surprising data point creates immediate credibility and sets a clear expectation for what the video will deliver.

03The Full Build Game Plan

Showing the complete plan before executing any step reduces anxiety because the scope feels known and the destination is visible.

04How to Define Your ICP

Reverse-engineering your ICP from your offer forces you to surface the trigger event: what happened in the last 30-90 days that made someone ready to buy now.
The negative ICP field prevents the mistake of writing content that attracts people who look like buyers but will waste your time.

06Scraping Posts and Comments

1,135 unique commenters from 45 posts is a real lead pool -- most creators treat these people as engagement metrics rather than as identifiable potential customers.

07How to Classify Commenters

Classification at scale reveals the ratio of buyers to industry watchers, which is invisible when you look at engagement totals alone.

08Analyzing the Audience Results

44% buyers with 40% competitors is a specific actionable finding -- it tells you the hooks are working but the CTA framing is pulling the wrong crowd.

11Reviewing the Live Dashboard

A self-contained HTML file with embedded data is portable and shareable without any server or database -- Claude Code can generate it in one prompt given a CSV and a design file.

12How to Generate Content Brief

Comparing your top-5 and bottom-5 buyer-engaging posts surfaces the variable that actually matters -- the contrast is what you cannot see by reading individual posts.
A reusable post template modeled on your own winners is more reliable than a generic template because it is tuned to your voice and your audience.

13Content Angles and Template

Beginner-facing language removes the primary psychological barrier to buying from an education offer -- the fear of being too new is the objection, not price.
Writing for the person who has not started yet is a concrete directive that removes guesswork about what to post next.

Glossary

Terms worth knowing.

ICP (Ideal Customer Profile): A precise description of the person most likely to buy your offer, defined by job title, life situation, trigger event, and exact language they use to describe their problem.
Apify: A marketplace of pre-built web scrapers (called Actors) that can pull structured data from LinkedIn, Instagram, YouTube, and thousands of other sites via API.
MCP connector: A Model Context Protocol integration that lets Claude Code call external tools like Apify scrapers directly from the chat without leaving the interface.
Commenter classifier: A Claude prompt that reads a CSV of commenters and assigns each one a category (BUYER, PEER, CREATOR, COMPETITOR, UNKNOWN) based on job title and bio against a stored ICP.
Negative ICP: People who superficially resemble your ideal buyer but will not convert, typically peers or consultants in the same industry. Identifying them prevents wasted sales energy.
Design system: A file containing fonts, colors, spacing rules, and typographic templates that Claude Code can read and apply when generating HTML assets, ensuring on-brand output without manual styling.

Resources

Things they pointed at.

03:00toolApify ↗

09:57productClaude Code Club ↗

Quotables

Lines you could clip.

00:11

“You might think you need better hooks to reach buyers, but the real problem is you do not know which posts are already attracting them.”

Tight reframe, no setup needed→ TikTok hook↗ Tweet quote

06:03

“Your content is working as a buyer magnet -- 44% of everyone who comments fits your ICP, but a striking 40% are competitors and peers.”

Surprising stat with immediate implication→ IG reel cold open↗ Tweet quote

09:24

“When I flex about running my business, buyers actually collapse.”

Counterintuitive, short, proven by data→ newsletter pull-quote↗ Tweet quote

09:55

“You have saved 40 AI tutorials this month, you have opened zero. There is only one thing that works.”

AI-written self-referential pattern interrupt→ TikTok hook↗ Tweet quote

The Script

Word for word.

Read-along

Don't just watch it. Burn it in.

See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.

00:00Claude Code read over 1,500 LinkedIn posts and told me which ones I need to stop writing. In this video, we're going to build a full audience dashboard so we can figure out how to turn commenters into buyers.

00:11If we haven't met yet, I'm Duncan. I grew a following of over 200,000 in the last twelve months, and I run one of the top communities for learning quad code and building income. You might think that you need better hooks to reach buyers, but the real problem is you don't know which posts are already attracting them.

00:26At the end of this video, Claude's going to give us three proven content angles and a template that we can lean on every single time. So focus in, close all your open tabs, and let's build. So the first thing I'm gonna do is I'm going to come into Claude code in the desktop app and make sure we're just inside of code up here.

00:42The next thing you need to do is to create a new folder for quad code to work inside. So I'm just gonna go ahead and click open folder, come into my desktop and go ahead and just call this LinkedIn analyzer and hit open.

00:53So here's how I think we should approach this today. It's always helpful to go into any build with a game plan. So if we're gonna figure out if our posts are reaching the right people, the first thing we need to do is define our ideal client.

01:04So who do we actually want to be talking to? The next thing we're gonna do is we're going to scrape my LinkedIn posts and all of the comments to figure out who's commenting on them. Then from there, we can take the data on who's commenting and analyze their job titles and their about sections just to kind of understand who's actually leaving comments.

01:21And then we can use Claude to classify each commenter against our ICP. So are the people who are actually commenting on our posts the people who we want to be reaching? From there, we can actually have Claude code create our dashboard with all of the metrics and information that we might need.

01:36And then the last thing we're gonna do is we're going to take our buyer engaging posts. So any of the posts that Claude recognizes like, yeah, this actually did hit our target audience, we're going to extract any of the patterns that worked to target those people and generate a brand new content brief for us that can inform our next round of LinkedIn posts to make sure that we're always targeting our actual buyers.

01:56So here's a prompt that we can use to figure out who our ideal client actually is even if you have no idea. If wanna you get access to all the prompts that we use today as well as the dashboard already built for you, I'll leave a link in the description. I'm going to describe my offer.

02:10Your job is to reverse engineer my ICP. So here, all you have to do is paste in the thing that you're actually selling. And then from here, give me who is most likely to buy this, what's their job title, their life situation, where are they in their journey, what's the trigger event.

02:23So what happened in their life in the last thirty to ninety days that made them ready to buy now? Pull their exact language, so how they describe their problem to themselves, what have they already tried, why did it fail, and then the negative ICP is actually super important. So who looks like my buyer but actually isn't, and these are the people who are going to waste your time.

02:41And then the single sentence they'd use to describe what they want if they could have anything. So I'm going to copy this, and if we come over into Claude, I'm just going to paste this into code and we can just come up here and edit my offer. And I can just say, I run the Claude Code Club, it's a $9 a month community.

02:56It's where people can learn how to use Claude Code and grow their income using AI. So while that's cooking, we can get started on the next piece. And this is a site called Appify, and Appify is awesome.

03:07It's basically a marketplace for something called scrapers. Scrapers basically just take any data that exists on the web and lets you access it. So we can look through Instagram, we can look through LinkedIn, we can look through YouTube.

03:18So what's really great is that Applefy already has a connector to Cloud Code, so Cloud Code can do all of the work for you and it's super easy to set up. So just click the plus, come into connectors, go to add connectors, and up here you can just search for Appify and click this and click enable.

03:33It's going to ask you to sign in with your Appify account. What's cool is Appify gives you a free $5 per week to use so you can get started today. So I'm not gonna read through all this specifically, but it gives me some information like who's most likely to buy, like marketers or ops people.

03:46Here's the exact language that they use. How do people actually make money with AI? Or I keep hearing about Cloud Code, but I don't get how to use it.

03:53So if I start including this language in my post, I am way more likely to actually reach these people. Here's what they've already tried and why it failed, and here are people who look like they might be my ICP but aren't. But the next thing we do is we can actually tell Claude to use Appify to scrape the last thirty days of posts from my LinkedIn along with all the comments on each post and just save everything as a CSV so that we have it.

04:15Now all you need to give it is your LinkedIn profile URL. So I'm gonna copy that and I'm just gonna come back to here and paste this in. And if I come over to LinkedIn, go to view profile, can just grab my profile URL and paste this back in here.

04:29So while this is going, I just wanna let you know that if you want to learn Claude code, but you're scared that you're going to break something or make a mistake, like you don't need to worry. You can literally just talk to Claude in normal language, and if something doesn't look right, just tell it. Like you can't really break it and you can't really make a mistake.

04:45And so if you're new to Apify and you actually want to check what's happening, just come over to runs here and you can see what's going on. And a lot of people often ask about cost. So this run finished, it grabbed over 1,500 comments and it did cost $5.67 to run.

04:59But if I'm able to convert even one of these people into a client, it is well worth it. It found 1,586 comments, which is awesome.

05:07It went ahead and built a CSV file for me, which exists inside of the folder. It is all 45 of my unique posts. 1,135 of them were actually different people, and so those are different leads for my business.

05:18And so the next thing we need to do is run this commenter classifier. So using the data in the LinkedIn audit, classify every commenter against my ICP. It's going to read the CSV file with all of the information on my commenters.

05:31It's going to check this against the ICP. You don't need to paste this in because it exists in the chat. It's going to look at their job title and bio and about section and it's actually going to classify them as a buyer, a peer, a creator, a competitor, or unknown.

05:43So I'm going to copy this, come back into Cloud Code, paste this in, and send this off. And Cloud Code is going to do all this analysis for me. This will likely take a couple minutes, so I'll come back when it's done and we can start building our dashboard.

05:55And so this finished, it only took about two minutes and here's what's really cool. Your content is working as a buyer magnet. 44% of everyone who comments fits your ICP, but a striking 40% are competitors and peers, so other people selling AI, coaches or consultants, which means your hooks are landing as much in the industry watching industry echo chamber as with actual prospects.

06:18So tightening our CTAs like away from the how I built my business angle would actually shift the mix towards buyers more. So it's actually saying that I really just need adjusting of my CTAs to get more buyers. So now we can go ahead and build the dashboard.

06:31And why we use Cloud Code for all of this is because Cloud Code can actually take action and build things for us, which is so much better than the simple chatbot. So basically using all of the data that we have, build me an audience quality dashboard. Dashboard.

06:43Needs to have the headline stats, a post performance table with all of my posts listed out, a commenter breakdown, so showing who all of our comments are, our best and worst posts, so this is one of the most impactful things. So top three posts by buyer engagement and the bottom three, and a trend line. So I'm going to copy this, but before I build it, I'm going to do one more thing.

07:03So I created this entire design system based off of a single Pinterest image. If you wanna see how I did this to create branded Instagram carousels, check out this video up here. But what's so cool about these design systems is I can use it for a dashboard, I can use it for a website or a landing page and so much more.

07:19So not only am I gonna use all the data, I'm going to give QuadCode access to this design file and this design system to build out our dashboard so that it looks on brand. So if I come into the desktop, come into our LinkedIn analyzer, and I'm just gonna go ahead and name this design system, come in here and paste in our two files.

07:36And just at the bottom here, I can say inside of the folder, I also gave you access to a design system. I want you to design the dashboard in this style. And I'm just gonna hit enter.

07:45So check this out. This is awesome. It built this beautiful dashboard for me.

07:49It shows 42 of all engagement comes from my ICP. It gives me a full breakdown.

07:54It gives me the entire trend of my buyer. It shows my best post and my worst post, then it actually goes ahead and breaks down every single post from the last thirty days including the audience split and what percent of the commenters were actual potential buyers. So that's pretty awesome, but this data alone isn't enough.

08:11The last piece of this is I need to figure out how to create more content that speaks to my buyer directly, resonates with them, and actually ends up converting them into actual customers. So here's a prompt to do this.

08:23It's going to basically look at all of the data we have. It's going to pull the top five posts and the bottom five posts. It's going to analyze the two groups to figure out what's attracting buyers, what topics and what format patterns, what hook style and length, and then what the low performing posts have in common so that I don't do that again.

08:40And then at the end, it's actually going to give me a content brief. So three content angles I should lean into, three content types I should cut, and a reusable post template modeled on my best buyer attracting posts. So that is pretty cool.

08:53So I'm going to copy this, come back over here into Quad, paste this in, and let's send this off. So this is awesome. It drafted up this entire brief for me and added it to the bottom of my dashboard.

09:04So what we found out is that the format of my posts are right, and then basically who I am speaking to isn't always accurate. My best and worst post basically use the same format and the lead magnet structure, but when I talk to beginners like no coding required or you can't break it or ship this today or free tools that buyers actually show up significantly more.

09:24When I flex about running my business like I replace my team, buyers actually collapse. And then the peers and competitors fill the comments. So basically what it's telling me is that I need to write for the person who hasn't started yet.

09:35Here are three angles to lean into, here's everything that I should avoid, and it even gave me a reusable template I could use that was modeled on my best performing post. That includes a hook, a one to three word reframe, bullet point list which you've seen before and so on, and then it even drafted a post that I can post on Monday.

09:52You've saved 40 AI tutorials this month, you've opened zero. There's only one thing that works. If you wanna get access to all the prompts I used today and the HTML dashboard along with all of the lessons and done for you skills so you can learn Claude code and grow your income starting today, just check the link in the description.

10:07If you wanna see how I use Claude code to build an agentic operating system, check out this video right here. I'll see

The Hook

The bait, then the rug-pull.

What if the reason your LinkedIn posts are not converting is not the hook -- it is that 40% of your commenters are competitors watching you? This video builds the pipeline to find out: a Claude Code-powered audience quality dashboard that classifies every commenter against your ICP and hands you a content brief before Monday.

Frameworks

Named ideas worth stealing.

01:57model

ICP Discovery Prompt

Who is most likely to buy
Job title and life situation
Trigger event (last 30-90 days)
Exact language they use
What they have already tried
Negative ICP
Single sentence they would use to describe what they want

Feed Claude your offer description; it reverse-engineers a full buyer psychology profile including the exact words your prospect uses to describe their own problem.

Steal forAny offer page, ad copy, or email sequence needing voice-of-customer language

05:18model

Commenter Classifier

BUYER
PEER
CREATOR
COMPETITOR
UNKNOWN

Reads a CSV of LinkedIn commenters, checks each person against the ICP, assigns a classification with confidence level and one-line reason.

Steal forAny platform where comments or signups need audience-quality scoring

08:22model

Content Brief Generator

Top 5 posts by buyer engagement
Bottom 5 posts by buyer engagement
What buyer-attracting posts share
What low-performing posts share
3 angles to lean into
3 content types to cut
Reusable post template

Compares your best and worst buyer-attracting posts to extract patterns, then generates a content brief and reusable template modeled on your actual winners.

Steal forMonthly content planning for any creator with existing post history

CTA Breakdown

How they asked for the click.

VERBAL ASK

09:57product

“If you want access to all the prompts I used today and the HTML dashboard along with all the lessons and done-for-you skills, just check the link in the description.”

Soft sell at end only. Claude Code Club at $9/mo on Skool. No mid-roll pitch. Clean execution.

MENTIONED ON CAMERA

03:00toolApify ↗

09:57productClaude Code Club ↗

FROM THE DESCRIPTION

PRIMARY CTAWhere the creator wants you to go next.