Four Days.

Someone had been watching Andrew show off what he’d built.

They asked the question everybody asks now.

How long did that take?

Four days.

That’s the answer people want to hear. A few prompts, a long weekend, a product. It’s the story the whole internet is selling.

Andrew gave them the second half of the answer too.

The year.

That second half is this story.

Andrew had spent two years on live calls with founders. By his own count, about 1,900 of them.

People just starting out. People who’d been in business for years.

And the same thing kept coming up.

They’d ask ChatGPT or Claude to write something for their business, and what came back could have been written for anyone.

He’d ask a room about it straight out.

He was asking because he’d been there too.

Maybe you’ve been there yourself. You type out what you need, you get back something smooth and confident and empty, and you spend the next hour rewriting it until it sounds like you again.

Andrew’s diagnosis was simple, and it wasn’t the one most people were selling.

The problem wasn’t the prompt.

AI holds the whole internet. The good copy, the bad copy and everything in between. Ask it to write with nothing to go on, and in Andrew’s words, "it just takes the mean of that."

Average in. Average out.

So while everyone else was trading prompts, Andrew went after the thing the prompt was missing.

The question was what the right content actually was.

Most people, asked to describe their offer, describe the product.

What’s in the course. How many calls. What the software does.

Andrew had come to believe that was the smallest part of it.

The rest was the buyer. What they’re afraid of. What they want. What keeps them up at night. Then what makes you different from everyone else they could buy from.

So he built a way to pull all of that out of a business and put it in one place.

It started small. On one early client session, he described it as "the build a buyer GPT." A single custom GPT that asked two questions, and lived or died on the answers.

Then it grew.

One GPT became seven. Andrew called them his AI Conversion Keys. A core key for the foundation. An Intel key for the ideal customer profile, down to the filters you’d use to build a lead list. A Mind key that went so deep into buyer psychology that ChatGPT wouldn’t let him publish some of it.

Then the emotional drivers. The competitive edge. The unique mechanism. The hook.

You’d paste your website into the first one. Then take everything it gave you and paste it into the next. Then the next.

A traditional buyer persona, the kind most copywriters use, runs to about 25 components.

By the end of the seven keys, Andrew’s document had 126.

He called it the offer doc.

A document like that doesn’t come out right the first time.

Andrew learned its rules the slow way, and he was blunt about the price.

One offer doc per product. Never one for the whole business.

One offer doc per kind of buyer, too. If you sell the same supplement to a 65-year-old grandmother and to a kid fresh out of college, their pain points have nothing in common, and the document has to know that.

And the first key matters most, because everything after it stands on it. Andrew wanted that first answer plain enough that "you could hand to a teenage kid and they would understand what you sell, who you sell it to, and how you deliver it."

Then the document never really finished.

He’d feed his sales call transcripts back in next to it and ask what didn’t line up. What was inaccurate. What was new.

He did it to his own offer, too. When his offer wasn’t landing, he went back through those 1,900 calls looking for the common thread, found it, and used it to rebuild his own playbook.

Same method. Applied to himself first.

Andrew shared the offer doc in the communities he worked in.

Some of the feedback was short.

Some of it wasn’t.

Terry Chenowith posted in the community for everyone to see, addressed to Jay Feldman "and everyone else."

Look at that list.

ICPs. Emails. Texts. Social posts. Websites. Funnels. Buyer personas. Video sales letters.

One document, and Terry was building nearly everything else on top of it.

Six months of research by a team of four to eight marketers, as he put it, done in two hours.

And there’s one line in Terry’s post worth holding on to.

Terry wasn’t just using the method.

He was telling Andrew it was a product.

Andrew already knew.

On stage in Miami, in front of a room of founders, he laid out how long he’d been carrying it.

A year dreaming. Six months trying to build it. Four more finding the people who could.

That last part mattered. Andrew is not a developer, and he didn’t pretend to be. He told the same room why he brings in the people he does.

There was another worry, the one every expert with a method eventually has.

If you teach people exactly how you do it, why would they ever need you?

Andrew’s answer was to give the whole method away.

That day in Miami, he handed the room all seven Conversion Keys for free. The GPTs, the step-by-step, the guide to all 126 components.

Because he’d worked out what software would change.

And with the software?

Then he showed the room the difference.

He asked plain ChatGPT for a video sales letter for one founder’s offer, with nothing else to go on, and read out what came back.

Then he took the same VSL structure into his own tool. He didn’t have enough on that founder’s offer yet, so he attached the offer doc for his own offer and ran it there.

Same structure.

The difference was what the AI knew going in.

Access to the software, he told the room, normally cost $5,000.

He gave it to them anyway.

That software became Vibey.

Look at how the product describes itself, and you’ll find Andrew’s year sitting right on the first page.

The wizard walks you through six steps. Your product and your competitors. Your market. Your power offer. Your buyer persona. Your ideal customer. Your competitive edge.

It tells you to talk instead of type. It tells you the first draft doesn’t have to be perfect. Generate, review, edit, regenerate.

Every one of those is a lesson Andrew had learned by hand.

And then the line that sums up the whole thing:

Remember what Terry said.

This tool should be a high ticket offer we sell.

Now it was a platform anyone could sign up for. Funnels built from the offer doc. Emails built from the offer doc. Brand guidelines built from the offer doc.

The document didn’t stop at Vibey.

Later, when Andrew packaged his methods as AI skills, the seven keys came along, bundled into what he called "the offer MBA, which in and of itself contains the seven AI conversion keys."

Later still, on a client build, he was starting in the same place. Scrape the website, pull in the product, the audience, the pain points, the promise.

A GPT. Seven GPTs. A document with 126 parts. A platform. A skill.

Every time Andrew built something new, it started with the offer doc.

So, how long did it take?

Four days of code.

But those four days stood on a year of calls, a year of rewrites and wrong turns and rules learned the slow way, and a document that people like Terry and Toby had already put to work.

The four days were the part anyone could see.

The year is the part nobody can copy.