From Construction Site to Website: How One Article Changed My Direction
How a 36-year-old civil engineer went from late-night anxiety scrolling to building an English finance website. The origin story.
From Construction Site to Website: How One Article Changed My Direction
This is Part 2 of “A Civil Engineer’s Website Building Diary.” Part 1 showed the results. This one tells the origin story.
The Late-Night Scroll
Late 2025. Both kids asleep. Me on the couch, scrolling through my phone.
The AI wave was everywhere. Every social media feed. Every dinner conversation. ChatGPT writing essays, Midjourney creating art, someone claiming they made $100K in a month with AI tools.
At first, I brushed it off. I’m a civil engineer. I draw blueprints and pour concrete. What does AI have to do with me?
But anxiety has a way of seeping in.
Construction wasn’t booming. Projects were shrinking. Colleagues were whispering about career changes. At 36, with two young children and a mortgage, “just pivot” isn’t really an option.
That night, I stumbled on an article that would change everything.
”Building Websites as a Retirement Plan”
The article was by a Chinese blogger. The core thesis was disarmingly simple:
Ordinary people can build English content websites, attract Google search traffic, and earn passive income through ads. The key is finding a niche with real demand and manageable competition.
He proposed a methodology: take a domain (like personal finance, health, or education), combine it with “money-making suffixes” (calculator, tracker, planner, template, generator), and use keyword tools to find high-volume, low-competition opportunities.
As an engineer, I immediately recognized this framework.
This was a feasibility study.
Find demand (search volume). Assess competition (SERP analysis). Calculate ROI (CPC and costs). Make a decision.
The thinking was the same. Only the domain had changed — from construction sites to websites.
The Learning Curve
Excitement aside, I didn’t know basic internet terminology:
- SEO: Search Engine Optimization — making your pages rank in Google
- Keywords: What users type into Google’s search bar
- Search Volume: How many people search a term per month
- CPC (Cost Per Click): How much advertisers pay per click — indicates commercial value
- SERP: Search Engine Results Page
Every single concept was new to me.
The article recommended Semrush for keyword research — a professional tool starting at $129/month. That was out of my budget for an experiment.
I needed a free alternative.
Building a Free Toolkit
After some research, I assembled three free tools:
1. Keyword Surfer (Free Chrome Extension)
Install it, search anything on Google, and it shows monthly search volume and CPC right in the search results. No account needed.
2. Google Trends
Free tool to check whether a keyword’s interest is rising, falling, or seasonal.
3. auto_collect.py (Built by Cursor AI)
This is where the magic happened.
The methodology required searching keywords one by one in Google and manually recording data. I had 100+ candidate keywords. Manual work would take days.
So I told Cursor: “I have a list of keywords. Write a Python script that automatically opens Chrome, searches each keyword, reads the Keyword Surfer data (search volume and CPC), analyzes whether the top SERP results are inner pages or home pages, and saves everything to a CSV file.”
Cursor wrote it. It worked.
The Moment It Clicked
The first time I ran auto_collect.py, I was stunned.
Chrome opened by itself. No one touching the keyboard or mouse.
It started typing keywords into Google — automatically. One by one.
For each search, it read the Keyword Surfer data displayed on the page.
Then it analyzed the top results — were they inner pages (low competition) or home pages of major sites (high competition)?
Finally, it saved everything to CSV.
100+ keywords, analyzed in about an hour.
I remember thinking: “AI can actually do this for me?”
This was my first real encounter with AI’s power. Not reading about it. Not watching a demo. Watching code I didn’t write run on my own computer, doing useful work.
It was like watching AutoCAD auto-generate drawings for the first time after years of hand-drafting.
What the Data Revealed
One keyword stood out:
| Keyword | Monthly Searches | CPC | SERP Type |
|---|---|---|---|
| saving money tips | 301,000 | $2.75 | Mostly inner pages |
301K monthly searches — massive, real demand. $2.75 CPC — advertisers value this space. Inner pages in SERP — competition is manageable for a new site.
I also found high-value uncovered long-tail keywords:
- “what is passive income” (14,800 searches/month)
- “money management tips” ($10.02 CPC)
- “how to budget money” (27,100 searches/month)
But data was only part of it. Personal finance matched my own real needs. At 36, with two kids and an uncertain job market, I genuinely needed to learn about budgeting, saving, and financial planning. The 50/30/20 rule, emergency funds, practical money tips — these weren’t topics I researched for SEO. They were things I wanted to understand for my own life.
I decided to learn, practice, and share simultaneously — turning my own financial education journey into content that could help others too.
The direction was clear: build an English personal finance site anchored on “saving money tips.”
Why Cursor
With the direction set, I needed a way to actually build a website.
I can’t code. I barely knew what HTML was.
I evaluated several options:
- WordPress: Too heavy, too much to learn
- Wix/Squarespace: Template-locked, limited SEO flexibility
- Hiring a developer: Expensive, and I couldn’t maintain it afterward
Then I found Cursor — an AI-powered code editor.
What sold me: you describe what you want in plain language, and it writes the code. Not just snippets to copy-paste, but a full collaborative workflow:
- I describe the requirement
- Cursor proposes a technical approach (with trade-off analysis)
- I approve or ask questions
- Cursor writes the code
- I check the result
- We iterate
It felt like pair programming with a senior developer who never gets impatient.
Cursor recommended Astro + MDX + Tailwind CSS. I had no idea what any of those were, but it explained the pros and cons in a comparison table. I picked the “best for SEO” option.
That choice turned out to be excellent. Astro generates pure static HTML that Google’s crawlers love.
The Three-Part Formula
Looking back, the entire origin story comes down to three elements:
- Direction — One article showed me the path
- Data — Free tools validated the opportunity
- Capability — Cursor gave me the ability to execute
Remove any one, and nothing happens.
Knowing the path but not validating it? You stay in “interesting idea” territory. Having data but no way to build? You hire someone or give up. Having tools but no direction? You build something nobody needs.
Direction + Data + Tools = Action.
Next: The complete keyword research deep dive — how auto_collect.py works, what the data revealed, and how you can replicate this process for free.
Previous: A Civil Engineer Built an English Finance Website Using AI Site: https://www.moneytipshub.com/