I Made A Digital Product With Claude In 3 Days

Everybody online right now is telling you to learn AI, pick up coding, or master some shiny new skill before you get left behind. But the first digital product that actually put money in Alston’s pocket had nothing to do with any of that. It was Microsoft Excel. Boring, ordinary, “nobody would pay for that” Microsoft Excel. He turned a skill he had been giving away for free for years into a $19 product that 60 people bought from around the world — people he had never met, in countries he had never visited. That product made him $1,671. Not because it was viral. Not because it was fancy. Because it solved one real problem for one specific person.

Then he retired it. And then, years later, he decided to rebuild the whole thing from scratch using AI in just three days — on camera, in real time — to prove something to you. The skill you have been writing off as too ordinary to matter might be the most valuable thing you already own. Here is exactly what he built, how he built it, what went wrong, and what it means for you.

What You’ll Walk Out With

Why a “boring” everyday skill is often more sellable than a cutting-edge one
The exact three-day process Alston used to rebuild his $1,671 Excel product with AI
The one-sentence rule that separates a real product from “someday” wishful thinking
How to use Claude Code without writing a single line of code
Why selling on your own page beats every marketplace, every time
What a $19 sale is actually buying you beyond the $19
The real reason most first products fail — and how to build one that doesn’t
A free tool to figure out what your $20 product should be at finder.platformproof.com

The Excel Product That Started It All

For years, Alston was the person people called when their spreadsheets broke. Coworkers, friends, strangers — they all showed up when a formula stopped working or a report looked wrong. He fixed their problems and moved on, never once thinking there was money in it. He was the free helpdesk that nobody hired.

Then he packaged that skill. Not as a giant course, not as a subscription, not as a consulting business. He identified one problem — messy data exports that office workers dread every single day — and built a simple $19 product around solving it. About 60 people bought it. Some paid $26, others $28, others $44 with the upsells and order bumps he added. When he added it all up it came out to exactly $1,671. And he was genuinely confused. Not because the number was huge. Because the number came from Microsoft Excel. The thing he had been doing for basically his entire adult life for free.

Meanwhile, every side hustle he had ever tried as an adult had made him either nothing or lost him money outright. The boring thing won. That was the lesson. A boring skill is not boring to the person who is drowning in it.

Why His First Product Failed and What Changed

Before the Excel product, Alston spent three months building a digital product that nobody bought. Not once. He kept adding more features, more lessons, more tabs. He built a sprawling course that covered everything he knew because he was afraid that simple would not be enough. He thought customers wanted more. He thought more complexity meant more value.

He was wrong. The complexity killed it. Nobody finished it. Nobody wanted it. Three months of lunch breaks and late nights after the kids went to bed, all for zero sales. He had to learn the hard lesson himself: if your product does everything, it solves nothing.

The Excel product was the complete opposite. One problem. One solution. One price. That was the entire product. And 60 people paid for it. The size of the product was not the difference. The clarity was the difference.

The 3-Day Rebuild Challenge

Over $176,000 in total from selling small digital products. That is the number Alston put on record, not to impress you, but to establish that what follows is not theory. He has made products that flopped, products that over-complicated everything, and products that sold steadily for months after he stopped thinking about them. He has lived the full cycle. And he wanted to take that experience and run a specific challenge: take the old Excel product, strip away everything unnecessary, and rebuild the whole thing from scratch over a single weekend using AI. Three days. No team. No coding experience required.

The challenge was partly personal. For years he was a little embarrassed that his most valuable skill was in a spreadsheet program used by millions of everyday office workers — not in some exciting new tech or AI or coding. Somewhere along the way he realized the problem was never the skill. The problem was that he kept trying to make it bigger and more impressive than it needed to be.

Day One (Friday): Pick One Problem

The first step is the one most people skip. On Friday, Alston forced himself to write a single sentence describing the entire product. Not a feature list. Not a course outline. One sentence.

His sentence: I want to help somebody who hates spreadsheets clean up a messy one in just 10 minutes.

That is the whole product. Not “Microsoft Excel.” Not “productivity.” Not “spreadsheets.” One specific problem — ugly data exports that get dumped on office workers every day — with one specific outcome: something clean, useful, and easy to read, without losing an entire afternoon doing it manually.

This is the rule he wishes someone had told him years earlier. If you cannot explain your product in one sentence, you do not have a product yet. You have an idea. You have “someday.” You have a pile of features without a container to hold them. When you have a product that solves exactly one problem, every other decision becomes clearer — what to include, what to cut, how to price it, how to describe it.

What is your one sentence? What is the problem that you could solve for someone in under an hour? That sentence is your product.

Day Two (Saturday): Building With Claude Code

Step two is where AI did the heavy lifting. Alston opened Claude Code and talked to it in plain English — the same way you would talk to a patient co-worker sitting across from you. No technical expertise. No coding. He told it to build a simple tool and a short guide that would help somebody clean up a messy spreadsheet in just a few clicks.

The first version was rough. A button did not work. The layout was messy. One section crashed completely. Here is the part nobody else puts in their YouTube videos: he never opened a single line of code. Not once. He just described what was wrong. “This button does not work.” “Move this from here to there.” “Change this color.” Back and forth with Claude like two people having a conversation across a cubicle, making incremental adjustments until it actually worked.

A couple of hours. A few rounds of trial and error. The result was something real and usable that he could turn around and sell.

Alston used to be a software developer. By his own admission, a bad one. Work that took other developers a few hours took him a week. So when he says that using Claude Code was faster than anything he could have built before, that is not marketing language. It is a direct comparison. The tool finally got good enough to package what he knows faster than he could do it himself — and without the weeks of frustrating progress he used to have to grind through.

Here is the critical thing he said, though. AI can package what you already know. It cannot hand you the knowing. Claude does not know which spreadsheet problems waste people’s time. Alston knows that — because he spent years watching people struggle with those exact problems and fixing them thousands of times. AI turned that accumulated knowledge into a product in a weekend. Without the knowledge, there is no product. The tool works because the person using it already has the raw material.

Not sure what skill you should be selling?

Answer a few questions and get a digital product idea plus a launch roadmap at finder.platformproof.com.

Day Three: The Sales Page and Why You Should Own It

The product is built. Now you need somewhere to sell it. This step matters more than most people realize, and the decision you make here determines whether you are building something you own or something you rent.

Alston built the sales page on a simple webpage he has complete control over. He priced it the same as the original: $19. Not because $19 is a magic number, but because he wanted the customer relationship to belong to him — not to a marketplace, not to an algorithm, not to a platform that can change its rules at any time. He has made that mistake before. Building on someone else’s platform means someone else can cut off your access to your own customers at any moment.

Here is what a $19 sale actually buys you. It generates revenue, yes. But it also captures the customer’s email address. It gives you permission to sell the next thing. Those 60 buyers from the original Excel product were not 60 transactions — they were 60 people who raised their hand and said they have this problem and they need help. That list is an asset you can keep building on. The product kept selling months after he built it, not because of luck or virality, but because the system he built kept running in the background while he was doing other things.

Old Version vs. New Version: What AI Actually Changed

On the left: the original Excel product from 2024. The clunky one. Took months to build — long Saturdays, lunch breaks, late nights after the kids went to bed, starting over multiple times, second-guessing every decision. Made $1,671.

On the right: the new version built over a single weekend with Claude Code. Same skill. Same knowledge. Same boring Microsoft Excel experience. Different result.

AI did not make Alston’s skills valuable. They were already valuable. What AI changed is the timeline. The thing that used to take months can now be created and sold in three days. Before, the barrier was not the skill — it was time and complexity. It was the feeling that building something real would take so long that it was not worth starting. That barrier no longer exists in the same way.

The comparison is not old tools versus new tools. It is someday versus done. The only difference between the version that made $1,671 and the version you have not built yet is that one of them got finished.

Step-by-Step: How to Build Your First Digital Product in a Weekend

Here is the exact process, condensed from the three-day challenge:

Friday — Pick one problem: Write one sentence. “I want to help [someone] do [specific thing] in [specific timeframe].” If it takes more than one sentence, you have more than one product. Narrow it down.
Friday — Validate the problem is real: This is not a step to skip. Is this something people actually get stuck on? Have you watched real people struggle with it? Have you solved it dozens or hundreds of times? If yes, proceed.
Saturday — Build with Claude Code in plain English: Describe what you want built. When something breaks, describe what is broken. No technical knowledge required. Expect the first version to be rough — that is normal, not a signal to quit.
Saturday — Give it a second pass: Done beats perfect, but done still needs to work. Go through the product yourself and catch the obvious bugs before a customer does.
Sunday — Build the sales page on a platform you own: Your own website, your own domain, your own checkout. Not a marketplace. Capture the customer’s email at the point of purchase.
Sunday — Price it between $17-$29 for a first product: Alston used $19. Low enough that price is not the decision, high enough that you are not undervaluing it. Add an order bump or upsell if you have a natural next step to offer.
Monday — Publish and tell people about it: You cannot sell something nobody knows about. The system does not run in the background until you turn it on.

Honest Drawbacks

The three-day framing is real, but it assumes something important: you already have the knowledge. If you are starting from zero knowledge on a topic, Claude Code cannot invent expertise you do not have. The AI is a production accelerator, not a knowledge generator. If you do not know which problems are worth solving in your field, no AI tool can fill that gap for you.

The first version will also be rough. Alston said this directly. A button did not work. A section crashed. You will need to iterate. The weekend timeline is achievable, but “done in a weekend” should not be confused with “perfect in a weekend.” Budget time for the second pass.

$19 is also not retirement money. The original product made $1,671 total. That is a meaningful proof point, not a business by itself. The value is in building the asset — the product, the page, the email list — that compounds over time. One sale gives you permission to sell the next thing. The income grows as the list grows.

Finally: selling on your own platform requires more setup than a marketplace. You are responsible for the checkout, the delivery, the customer email. That is the trade-off for owning the customer relationship. Platforms are simpler on day one and more fragile over time. Your own page is more work upfront and far more durable.

Find Your X

The question Alston asked in his last video — “What is actually stopping you from making your first $1,000 online?” — got the same answer from more than half the people who responded. They do not know what to sell.

If that is you, the answer is almost certainly sitting in the thing you have been writing off as too ordinary. What do your coworkers keep asking you to help with? What have you done thousands of times without thinking twice? What problem do you solve in ten minutes that takes someone else an entire afternoon? Start there. If you want a structured way to work through it, head to finder.platformproof.com — answer a few questions and you will come out with a product idea and a launch roadmap.

Frequently Asked Questions

What was Alston’s original Excel product, exactly?

It was a simple Excel course focused on one specific problem: helping people who hate spreadsheets clean up a messy data export in just a few clicks. It was not a comprehensive Excel training. It did have some upsells — an order bump, templates, and a longer course for people who wanted more depth. The base price was $19, with buyers paying anywhere from $19 to $44 depending on what they added.

How much did the original Excel product make?

About 60 buyers from around the world purchased it, generating a total of $1,671. The product continued to sell for months after Alston built it without ongoing promotion. He has since retired the original version.

Do you need to know how to code to use Claude Code?

No. Alston used it entirely in plain English. He never opened a line of code during the entire build. He described what he wanted, described what was broken when something went wrong, and iterated from there. His description was that it felt less like programming and more like working with a very patient developer.

What is the one-sentence rule?

If you cannot describe your product in a single sentence, you do not have a product yet. You have an idea or a collection of features without a clear purpose. The sentence should name who you are helping, what specific problem you are solving, and what the outcome looks like. Everything else in the product either serves that sentence or gets cut.

Why did Alston not sell the product on a marketplace like Etsy or Gumroad?

He specifically wanted the customer relationship to belong to him, not to a platform. When you sell on a marketplace, the platform controls who sees your product, what rules you follow, and what happens to your account. Selling on your own page means you capture the buyer’s email address and can sell to them again. The $19 transaction is the beginning of a relationship, not the end of one.

Can AI replace the knowledge you need to build a useful product?

No — and Alston is direct about this. AI can package what you already know and turn it into a product much faster than you could do it manually. It cannot invent expertise you do not have. The reason the Excel product worked is that Alston had spent years watching real people struggle with real spreadsheet problems and had fixed those problems thousands of times. That accumulated knowledge was the raw material. Claude was the production tool.

How long does it actually take to build a digital product this way?

Alston’s challenge was three days: Friday for the problem definition, Saturday for building with Claude Code, Sunday for the sales page. He framed this as a contrast to his original product, which took three months. The three-day timeline assumes you already have the knowledge and are focused on one specific problem. Scope creep — adding more features, more lessons, more complexity — is the main thing that blows up the timeline.

What should you do if you do not know what skill to sell?

Alston’s suggestion is to look at what your coworkers keep asking you to help with — the thing you do without thinking that takes other people all day. Then write it down in comments on his videos and he will tell you what product you could build from it. He pointed to a commenter named kur t 5 s who mentioned helping busy professionals manage their time — his immediate read was that they could build a simple time-hack worksheet and sell it for $20. If you want a more structured answer, finder.platformproof.com walks you through the same process with a tool built specifically for this.

Sources

Alston Godbolt, “I Made A Digital Product With Claude In 3 Days,” YouTube, https://youtu.be/-AjK2Ejrq0o
Transcript: original Microsoft Excel product made $1,671 from approximately 60 buyers at $19 base price with upsells to $26, $28, and $44
Transcript: Alston’s total from small digital products over time: over $176,000
Transcript: Product built using Claude Code in plain English with no code written directly
Transcript: Three-day build timeline — problem definition (Friday), product build (Saturday), sales page (Sunday)

Helping 1 million working adults make their first $3,000 online with the skills they already have. Alston Godbolt, Platform Proof.