A 7-step framework any entrepreneur can follow to build custom software without writing a single line of code.
Three years ago, my company Full Focus made what felt like a smart bet. Our customers had been asking for a digital version of the Full Focus Planner — the physical planner we’ve sold over 1.4 million copies of. So we hired a development team to create it. We invested over a million dollars. We spent eighteen months waiting.
At the end of that period, we had nothing to ship. We wrote off the entire project. It was one of the most painful lessons of my career. And frankly, my career is built on mistakes like that one. That’s the seabed for wisdom.
But here’s the thing. That million-dollar failure turned out to be the setup for a breakthrough I never saw coming.
The Distance Is Gone
If you’ve ever had an idea for an app that would solve a real problem in your business — and then hit a wall because you don’t code, can’t afford a development team, or don’t have eighteen months to wait — I want you to know something.
That wall is gone. AI has collapsed the distance between vision and reality. A year ago, building custom software required a team, a budget, and a timeline measured in months. Today, I build fully functional apps by myself. I’m seventy years old. No technical background. Zero experience with code.
I recently shared this process during a sparkSESSION with Jonathan Mast and his AI community. Jonathan has had more impact on my AI journey than any other single person — he gives it everything he has, every single day. During that session, I walked the group through the exact framework I use to go from a problem to a published app.
I want to share that framework with you today. But first, a warning. Vibe coding is the most addictive thing I’ve ever experienced. You envision something, describe it, and — boom — there it is. My wife and I have been married for forty-seven years. We have a great relationship. But I can tell there are moments when she gets a little annoyed as I sit there adding just one more feature. The magic phrase — the one you’ll find yourself saying over and over — is this: Wouldn’t it be cool if …
And the next thing you know, you’re building that thing.
The 7-Step Framework for Building Your Own Apps
Here’s the system I’ve developed. It’s the same process I used to build Promptlee, my prompt management app, and Respondlee, my iMessage auto-reply app. Neither required a traditional developer. Both are live and functional. And Promptlee is already generating revenue.
Step 1: Identify the Problem You Want to Solve
This is where most people go wrong. As powerful as these tools are, if you don’t define the problem first, the project will sprawl, go out of scope, and become a mess.
My advice? Scratch your own itch. I built Promptlee because my AI prompts were scattered everywhere — Word docs, Notion, random files on my computer. Every time I needed one, I couldn’t find it. That’s an enormous inefficiency for someone who is supposed to be a productivity expert.
So I asked myself: What if I had one beautiful app where all my prompts, links, and collections lived in one place?
That’s the problem I set out to solve. Start there.
Step 2: Set Up Your Project Folder
Before you touch a vibe coding tool, get organized. Create a dedicated folder for your project. Store your planning documents, screenshots, and reference images there — everything you’ll need during the build.
It sounds simple. It is. But skipping this step creates chaos later when you’re iterating fast and can’t find what you need.
Step 3: Create Your Product Requirements Document
This is the step that separates people who build apps that work from people who burn through credits chasing a vague idea.
I use Claude to help me write what’s called a PRD — a Product Requirements Document. You don’t need to know the technical term. Here’s all you need to know: it describes exactly what your app should do, how it should look, and what features it needs.
Think of it as the blueprint for a house. You wouldn’t hand a construction crew a napkin sketch and say “build this.” You’d give them architectural plans. A PRD is your architectural plan.
Here’s what I include in mine:
- The problem the app solves
- The core features it needs
- The user experience — how should it feel to use?
- Design preferences — colors, fonts, layout references
- Integrations — does it need to connect to Stripe, email, etc.?
The better your PRD, the fewer credits you’ll burn and the closer your first build will be to what you actually want.
Step 4: Sign Up for Your Vibe Coding Platform
I use Replit. I’ll be honest — I didn’t research every alternative. I tried it, liked it, and stuck with it. What I appreciate is that I didn’t have to figure out database connections, Stripe integrations, or domain management separately. Replit handles all of that.
A few things to know going in:
- The basic plan is about $20 a month. But that’s not your real cost — you’ll burn through credits. I’ve probably invested $400–$500 in Promptlee at this point. Still a fraction of what a development team would charge for the same build.
- Use “Plan” mode before “Build” mode. Planning takes a fraction of the credits that building does. Get clear before you tell the tool to start writing code.
- Start with Economy mode if you’re budget-conscious. It’s slower, but the quality is comparable.
Step 5: Build
This is where the magic happens. Upload your PRD and your prompt can be as simple as: Build the app described in the attached PRD. The tool generates a working version in the preview panel. From there, it’s iteration.
Here’s what surprised me — the process is a lot like sculpture. You start to see the general shape, and then you get more detailed as the app takes form. You’re not coding. You’re directing.
A few tips I’ve picked up:
- Use screenshots liberally. If you see a design you like on another site, screenshot it and say “make it look like this.”
- Think in components — account registration, Stripe integration, a feedback system. I keep a system design document that describes how I want each of these to work. Every new app starts from that baseline.
- In Replit, you can select any element in the preview and describe what you want changed. You’re directing, not coding.
One more tip: once you’ve made all the design decisions for your first app, create a system design markdown file. Upload it alongside your PRD for future projects. It gives the tool a baseline so you don’t repeat the same corrections twice.
Step 6: Publish Your App
Replit makes publishing straightforward. There’s a development environment where you iterate, and a production environment where the world sees your app. I publish early — even before sharing the URL with anyone — because things sometimes behave differently in production, and I want to catch issues fast.
You can purchase a custom domain directly through Replit. No GoDaddy. No DNS configuration. Pick a name, and it’s live within minutes.
Step 7: Monetize
This is the part that excites me most. If you’re scratching your own itch and building something elegant — something that solves a real problem — other people will pay for it.
I’ve been thinking about this a lot lately. I have friends who invest in rental real estate. That never appealed to me. I don’t know anything about real estate, and I don’t want the headaches. But a portfolio of small, useful software apps? That’s digital real estate.
Imagine twenty apps, each generating a thousand dollars a month in recurring revenue. No property taxes. No backed-up sinks. No difficult tenants. Just useful tools solving real problems for real people.
For Promptlee, I used Donald Miller’s StoryBrand framework to create a brand script, had Claude write the sales copy, then told Replit to build the sales page — pricing table, Stripe integration, testimonials, FAQs, footer. All of it built through description, not code.
Jonathan made a point during our session that stuck with me. He said he’s built simple apps for ten or twelve dollars that solved problems that had been nagging him for years. And he asked the question we should all be asking: Why wouldn’t you do this?
I think that’s exactly right.
The “Wouldn’t It Be Cool If” Mindset
Here’s what I’ve learned. Vibe coding isn’t really about technology. It’s about vision. It’s about being willing to say, “Wouldn’t it be cool if this existed?”—and then actually building it.
Let me be honest with you. The advertising would have you believe you’ll dictate an app, press a button, and it springs to life fully formed. That’s not how it works. There’s iteration. There’s learning. There’s frustration. But the time investment is manageable. The financial investment is modest. And the upside is real.
You don’t need a technical background. I’m living proof. You don’t need a development team. You don’t even need to know what JSON means. I built an export feature for one of my apps and I still don’t really know what it is. But my users wanted it, so I built it.
What you need is a clear problem, a solid plan, and the willingness to start.
Start Here
If this resonates with you, take two steps today.
- First: Identify one problem in your life or business that an app could solve. Write it down. Be specific.
- Second: Open Claude and type: ”Help me create a product requirements document for an app that does [your description].” You’ll be amazed at what comes back.
And if you want to go deeper, subscribe to the AI Business Lab newsletter at aibusinesslab.ai. The free version gives you the strategy. The Insider+ version gives you Implementation Kits — step-by-step guides with custom GPTs and templates you can use right away.
You can also try Promptlee at getpromptlee.com. It started as a tool I built for myself. Now it’s serving a growing community of people who want to get organized with AI.
The distance between your vision and reality has never been shorter.
What will you build?
