Growth Log: Building My Own Website As A Marketer

For most of my career, "the website" lived on the other side of a ticket.

Design here, growth there, development somewhere far away in another sprint. My job was to bring users in, optimize funnels, analyze cohorts, and ask for changes I couldn't make myself.

In 2026, that wall is gone.

For this version of my personal site, I decided to build the stack myself: Replit for the environment, GitHub for version control, Cloudflare Workers and Pages for deployment - with AI as my quiet pair-programmer in the background. I'm still a marketer, not a software engineer. But the tools have shifted enough that I can finally touch the thing I've been optimizing for years.

This article is the first entry in my Growth Logs: documenting how I'm using new tools and a beginner's mindset to grow faster across business, music, and everything I build.

The New Reality: Marketers Who Build

I'm part of the generation of marketers who grew up on dashboards, not terminals. My "stack" used to be analytics, ad platforms, ESPs, and spreadsheets.

But the ground has moved:

AI can generate working HTML, CSS, and JavaScript from prompts.

Replit gives you a live, browser-based environment where you can see changes instantly.

Cloudflare lets you deploy to a global edge in minutes, not weeks.

I'm not trying to become a full-time developer. I'm trying to become a growth person who can:

• Spin up and test ideas without waiting in a backlog.
• Understand what's technically possible before I ask a team to build it.
• Respect the craft of development and design by feeling the complexity first-hand.

AI didn't remove developers from the equation; it removed my excuses.

My Stack: Replit, GitHub, Cloudflare Workers and Pages

Replit as the Playground

Replit is my playground and editor. No local setup, no "it works on my machine". I open the browser, load my repl, and start building.

I use AI prompts like:

"Create a minimal responsive layout for a personal site with a hero, navigation, and three sections."

"Refactor this CSS so it's easier to maintain."

Replit gives me instant feedback, and AI fills in the gaps when I don't remember exact syntax.

GitHub as the Source of Truth

GitHub is where the code actually lives.

Every change is versioned and reversible.

I can see my own history of experiments.

If I ever hand this project to a proper dev, they're not stuck untangling a mystery zip file.

Version control is the difference between "playing with code" and "maintaining a real project".

Cloudflare Workers and Pages for Deployment

Cloudflare Workers and Pages handle the rest:

Static assets and front-end code go to Cloudflare Pages.

Any dynamic logic or APIs live in Cloudflare Workers at the edge.

On paper, it sounds technical. In practice, AI bridges most of the scary parts:

"Help me create a basic Cloudflare Worker that responds with JSON."

"Show me how to configure this project for Cloudflare Pages."

"Explain this error in simple terms and suggest a fix."

I'm still reading, debugging, and learning - but I'm no longer blocked by the setup phase.

Respecting Developers and Designers

It's important to say this clearly: what I'm doing is not the same as the work of experienced developers and designers.

They think in:

• Architectures and system design
• Performance, accessibility, and security
• Long-term maintainability and technical debt

I'm mostly thinking: "Can I get this to work without breaking everything?"

AI makes it easier for non-technical people to build things. It does not replace:

• Years of experience making hard trade-offs in complex systems
• Taste and judgment around UI, UX, and interaction
• Deep understanding of how different technologies behave at scale

What AI does is lower the activation energy.

It lets someone like me:

• Prototype faster
• Speak more clearly with technical teams
• See the invisible work developers and designers have always been doing

When I finally wired my own deployment pipeline and broke it three times in a row, my respect for every engineer I've worked with quietly doubled.

AI as Helper, Not Hero

The best way I've found to describe this new workflow is simple:

Humans set direction. AI handles scaffolding.

I decide:

• What story the site needs to tell
• How growth, music, and ventures should be connected
• What a visitor should feel and do in the first 10 seconds

AI helps with:

• Boilerplate code, routing, and layouts
• Repetitive components and refactors
• Explaining why something broke and how to fix it

Sometimes the AI is wrong. Sometimes the code is inefficient. Sometimes the design feels off. That's fine. My job is to bring judgment, not to worship the output.

Why This Matters for Growth Marketing

So why is this a Growth Log and not just a tech diary?

Because this shift changes the shape of the growth role.

Faster Experiments, Shorter Feedback Loops

I can now:

• Test UX ideas and flows myself before handing them to a team
• Prototype landing pages and microsites instead of writing long specs
• Connect marketing ideas to technical reality much faster

That means more experiments, tighter feedback loops, and less "lost in translation" between slides and shipping.

Better Collaboration With Product and Engineering

By touching the stack, even at a basic level, I:

• Ask better questions
• Make more realistic requests
• Understand the cost of "just a small change"

That makes collaboration with developers and designers more respectful and more productive.

Owning More of the Growth Surface Area

The same mindset I use for growth - hypothesis, build, measure, learn - now applies all the way down to the infrastructure of my own site.

It's not just "optimize the funnel someone else built".

It's "design the funnel, build it, deploy it, and then optimize it".

Growth Metrics for This Experiment

Every Growth Log should have at least a few concrete metrics. For this one, I'm tracking:

• Time to first deploy: from "new project" to live site on Cloudflare
• Number of iterations shipped in the first month
• Basic performance metrics: page speed, uptime, and error rates
• Early traffic and engagement: organic visits, time on site, click-through to key pages

The goal isn't to brag about numbers. The goal is to build a habit: every experiment deserves a baseline and a follow-up.

The Bigger Shift: Tech Is Becoming More Human

The world is changing.

Growth marketers are becoming builders.

Developers are getting AI copilots that remove the boring parts so they can focus on the hard problems.

Tools like Replit, GitHub, and Cloudflare are quietly making serious tech accessible to people who used to live only in slide decks.

I don't think everyone in growth needs to become a developer.

But I do think more of us need to touch the stack, even just a little. Not to replace anyone - to understand, to collaborate better, and to ship ideas that would have stayed stuck in a notebook a few years ago.

This site is my proof of concept.

Growth Log #1 is done. On to the next experiment.