He Quietly Used Claude to Build a $3K/Month Side Project

$3,700 a month from one product. 120 sales at $25 a copy. Four sales a day. Built on a Saturday morning six months ago by a guy you have never heard of who has no AI agency, no audience, and no course. If you have watched every “make money with Claude” video this week and left feeling like the AI lane already belongs to 23-year-olds in San Francisco, this story will change that picture.

His name is Edmund Young. He goes by @edmund on Indie Hackers. His product is called Easy Folders, a Chrome extension that adds folder search and a prompts manager to Claude and ChatGPT. His revenue chart is public. His GitHub is public. His numbers survived every check Alston ran across two weeks of research, which is more than could be said for the 13 other Claude case studies that fell apart on closer look. This post walks through the full story and the three-step weekend plan anyone can copy.

What You’ll Walk Out With

The verified numbers behind Edmund’s $3,700/month product and what they actually mean
Why Claude didn’t create Edmund’s skill, just changed how long it took to ship it
Four questions that tell you whether your own job skill is already a product hiding in plain sight
How to build your own Claude setup in one weekend with named commands and custom helpers
Where to sell a file product (Gumroad) vs. a software product (your own site) and why Etsy is the wrong answer
The honest drawbacks of this approach and which situations it will not work for
A free quiz at finder.platformproof.com that identifies which of your existing skills Claude can turn into a product in five minutes flat

The Math That Made Alston Stop Scrolling

Run this once before anything else. $25 times 120 sales equals $3,000 a month. That is four sales a day. Nothing unusual about that number until you find out Edmund made all 120 of those sales from a product he built on a single Saturday morning six months prior. He does not have a team. He does not have a customer service person. He has one product, one page, and the math runs itself every month while he is at his day job.

Month one of his Indie Hackers chart showed about $400 in revenue. Month three showed $1,200. Month six showed $3,700. At the six-month mark his total was $42,000. Every number on that chart is public, attached to his @edmund profile on Indie Hackers, and verifiable by anyone willing to go look it up right now. This is not a screenshot shared in a private Telegram group. This is a public page with disclosed monthly revenue and a visible chart of how it grew month after month.

What Claude Actually Did (and What It Didn’t)

Here is the part most make-money-with-Claude videos skip. Claude did not create Edmund’s skill. Before Claude existed, Edmund was already building Chrome extensions at his day job. That skill was sitting on his hard drive years before AI became a conversation topic. What Claude changed was the time it took him to turn that skill into a shipped, live, selling product.

The hours of drafting, formatting, writing product descriptions, writing FAQ copy, and doing all the surrounding setup work for a product launch: Claude turned all of that into a Saturday morning. The product is the same product it would have been without Claude. It just would have taken Edmund weeks instead of one weekend to get it out the door.

If you took Claude away from Edmund today, he would still have the skill. He would just be a slower version of himself. That is the right relationship to have with this tool. Claude is not the product. Your skill is the product. Claude is the thing that shrinks the time between having the skill and having something up for sale that the math can run on every month.

Why 14 of 15 Claude Case Studies Fell Apart

Before arriving at Edmund’s story, Alston looked at 14 other Claude case studies that seemed promising. All 14 fell apart on closer inspection. Some had revenue numbers that didn’t match the dates on the page. Some had product pages that returned a 404 error when checked directly. Some were obvious AI-generated case studies dressed up as real people with real income screenshots.

Edmund’s story survived three specific checks. First, his Indie Hackers product page with publicly disclosed monthly revenue. Second, the revenue chart on that same page showing month-by-month growth from $400 to $3,700. Third, a public GitHub profile where he keeps the Claude setup he built for himself, including all 14 commands and 11 helpers, each one named and readable by anyone who opens the page.

That level of public verification is not common in the AI income space, where most of the case studies are either fabricated or exaggerated to the point of being useless. Edmund’s numbers are boring in the best way. They are consistent, public, and growing at a pace that a working adult with a full-time job could actually replicate without quitting anything or hiring anyone.

The Four Questions That Identify Your Sellable Skill

The gap between “I have a skill” and “I have a product” usually comes down to whether the skill is specific enough to be worth $25 to someone else. These four questions help you check that before you spend a weekend building anything.

Question 1: Has a coworker walked over to your desk in the last six months and asked you to explain something more than once? Not just once. More than once. That moment, when someone asks for the same explanation a second time, is the moment your way became different from the default way. That gap between how you do it and how everyone else does it is where the product lives.

Question 2: When something breaks at work, do people text you first instead of the manager? The org chart says one thing. The text messages say something else. If you are the unofficial expert, the person people contact before they contact the person they are supposed to contact, that is a real skill with a real audience that would pay to borrow your shortcut.

Question 3: If a new hire had to do this thing tomorrow, would you need to spend a full afternoon writing it down before they could do it without you? The writing-down part is the product. If the process lives only in your head, the gap between you and everyone else is wide enough to charge for. That afternoon of documentation becomes the thing someone pays $25 to skip.

Question 4: Have you been doing this thing for so long that you’ve forgotten when you learned it? The things that are automatic for you are the things someone else would pay $25 to learn. The harder a skill is to date, the more embedded it is. The more embedded it is, the more someone else is willing to pay to borrow your shortcut rather than spend years getting there the long way.

If you answered yes to three of the four, the skill is already in your head. The product is the documentation of how you do it, formatted and packaged for the person who wants to do it too and doesn’t want to spend years figuring it out on their own.

Step 1: Name the Old Skill First

The first step is not about Claude. It is about identifying the thing you had before Claude existed. For Edmund, it was Chrome extensions. He had been building them at his day job for years before AI became a dinner-table conversation topic. Claude didn’t teach him that. The skill was already there, already proven, already something his coworkers relied on.

Your version might be a spreadsheet that only makes sense to you, the one where if you got hit by a bus on Tuesday, the new hire would spend six weeks trying to reverse-engineer it and still couldn’t fully figure it out. It might be the checklist nobody else on your team can run as cleanly. It might be the way you write client emails, or the onboarding flow you have refined over seven years in the same role. Whatever it is, name it before you open Claude. The skill comes first. Claude is the thing you apply to the skill, not the other way around.

Step 2: Build Your Own Claude Setup in One Weekend

Edmund’s GitHub profile is titled “Edmund’s Claude code.” On it, he has 14 name shortcuts saved in Claude. The technical term is commands. They are buttons you build once for the steps you take every week at your job. If your skill is bookkeeping, you might name one “draft monthly salary.” If your skill is project management, you might name it “weekly status report.” If your skill is sales ops, you might name one “pipeline cleanup.” The name is whatever maps to the repetitive piece of your specific job.

He also built 11 small Claude helpers. Each one was set up once for a specific piece of his product-building process. One for the way he ships a new version. One for the way he writes code. One for the way he diagnoses and fixes something when it breaks. These are not general Claude prompts copied from a Reddit thread. They are built for his specific way of doing his specific work, which is what makes them actually useful instead of mostly useless.

The first time he built the system took a weekend. Edmund’s recommended window is Friday night to Sunday afternoon if you need a hard deadline. But every product he ships after that first weekend uses the same setup, so the time investment drops to near zero for everything that comes after. The first weekend is the cost. Every product you ship for the rest of your life is the return on it.

Your setup will look different because your skill is different. The structure is the same: one weekend to build it, then reuse it for everything you ship after that. If your skill produces a file or document, your commands will match how you write and format it. If your skill is software, your commands will match your build and ship process. The specificity is what makes it compound over time rather than sit idle.

Not sure which skill you have that Claude can turn into a product?

Take the free five-minute quiz at finder.platformproof.com. No credit card required. The same email unlocks every worksheet from every video on the channel.

Step 3: Ship It to the Right Platform

Once the skill is named and the Claude setup is running, the file needs somewhere people can pay for it. Edmund’s product is a Chrome extension, so it needs to be installed, not downloaded. He runs his own product page because that is the correct home for software that needs to live in a browser.

If your product is a file, a guide, a spreadsheet, a template, or an SOP, the easiest starting point is Gumroad. Setup takes about 30 minutes. You upload the file, write the description, set the price, and you are live. The critical thing about Gumroad is that you own the buyer’s email address. When you ship version two, you can contact every person who bought version one. That list is more valuable than the storefront.

Don’t use Etsy. Etsy owns the customer relationship. If you leave the platform, change the product, or get removed for any reason, you walk away with nothing. The whole point of building a product you own is to actually own something. Use Gumroad first, your own website second. If the product is software like Edmund’s, skip Gumroad entirely and go straight to your own site where the install experience can be controlled properly.

Honest Drawbacks

This plan works. But it doesn’t work instantly and it doesn’t work for every situation. Here are the cases where it won’t.

If your skill is genuinely common, the price ceiling is low and the market is crowded. Edmund’s Chrome extension solved a specific problem with how Claude and ChatGPT handle folder organization. That is a narrow problem with a defined audience. A generic productivity guide competes with thousands of others and usually sells for less. The more specific your skill, the more you can charge for the product built from it.

If you answer no to three of the four questions above, the skill may not be differentiated enough yet. That is not a permanent condition. It means the product is not ready right now. Keep the questions somewhere visible for the next six months and notice what coworkers ask you about more than once.

Month one will not be $3,700. Edmund’s month one was $400. That is a real number and a solid result for a first month, but it took six months of consistent sales to reach the $3,700 figure. If you are expecting four sales a day starting in week one, you will quit before the chart starts to climb.

The deciding step is harder than the building step. Alston’s point in the video is that you don’t need to learn Claude before you start. You need to pick the one part of your job Claude is going to make ten times faster. That selection is where most people get stuck, not the actual building of commands or helpers. The quiz at finder.platformproof.com is built specifically for that stuck moment.

Find Your X

The five-minute quiz at finder.platformproof.com walks through the skill identification process and tells you which of your existing skills is most likely to become a Claude-powered product. It is free, requires no credit card, and the same email gives you access to every worksheet from every video on the channel. If the math in this post is what got your attention, the quiz is the next step.

Frequently Asked Questions

Do I need to be a developer to do what Edmund did?

No. Edmund’s specific product is a Chrome extension, which does require coding knowledge. But the underlying pattern works for any skill. Spreadsheets, SOPs, project templates, HR onboarding checklists, client email writing, financial models: all of these are products in the same shape. Edmund’s version is the developer version. It is not the only version. The deciding factor is whether you can answer yes to three of the four skill questions, not whether you can write code.

What is Easy Folders and where can I find it?

Easy Folders is a Chrome extension that adds folder search and a prompts manager to Claude and ChatGPT. Edmund sells it through his own product page. His public Indie Hackers profile is under @edmund and shows the full revenue history. His GitHub profile includes his complete Claude setup under the name “Edmund’s Claude code,” with all 14 commands and 11 helpers available to read.

How much did Edmund make in his first month?

About $400 in month one. Month three was $1,200. Month six was $3,700. The six-month total was $42,000. All of these numbers are publicly disclosed on his Indie Hackers page, which anyone can open and verify independently right now. There is no screenshot, no claim, and no unverifiable figure in this case study.

Why is Gumroad better than Etsy for digital products?

Gumroad gives you the buyer’s email address. Etsy does not. When you sell on Etsy, the customer relationship belongs to Etsy’s platform. If you leave, change the product, or get removed, you walk away with nothing. On Gumroad, every buyer’s email is yours to keep permanently. When you ship version two of the product, you can email everyone who bought version one. That list is the most durable asset the whole operation produces.

What are Claude commands and how do I build them?

Claude commands are name shortcuts you save inside your Claude account. They function like buttons you build once and reuse. When you type the command name, Claude runs the full instruction attached to it. Edmund built 14 of them, each named after a specific step in his product-building workflow. You would build yours for the repetitive parts of your own job: the monthly report, the client status update, the pipeline review, or whatever you do on a loop every week. The goal is to make each step in your work something Claude can execute in one command instead of something you type from scratch every time.

How long does it take to build a Claude setup like Edmund’s?

One weekend, based on Edmund’s timeline. The recommended window in the video is Friday night to Sunday afternoon if you need a hard deadline. The first time you build the setup is the full cost. After that, every product you ship uses the same commands and helpers, so the marginal time investment for each new product drops to near zero. The first weekend buys you the next year of shipping faster than you would without it.

What is Indie Hackers and why does Edmund post his revenue there?

Indie Hackers is a public community where independent product builders share their revenue numbers and product journeys. Posting revenue is optional but common on the platform, and the community norms around transparency are what make the numbers credible. Edmund’s page shows his monthly revenue, his product description, and the full growth trajectory from launch. It is a community built around real numbers and honest reporting, which is what makes Edmund’s case study verifiable in a way that most online income claims simply are not.

What is the Offer Engine mentioned at the end of the video?

Offer Engine is a $17 one-time paid product mentioned briefly at the end of the video. It is described as a deeper plan for turning your existing skill into a Claude-powered product. It is separate from and different to the free quiz at finder.platformproof.com. The video describes it as worth the price for the math alone if the revenue numbers are what got your attention watching. There is no subscription and no recurring charge: one payment, full access.

Sources

Edmund Young’s Indie Hackers profile (@edmund) with public monthly revenue chart and product page for Easy Folders
Edmund’s public GitHub profile (“Edmund’s Claude code”) showing 14 named commands and 11 Claude helpers
Easy Folders Chrome extension: adds folder search and prompts manager to Claude and ChatGPT interfaces
Platform Proof YouTube: “He Quietly Used Claude to Build a $3K/Month Side Project” (video ID dzu5owMKOfk)