Day 3｜When a Landing Page Quietly Becomes a System
Portal 32｜Prototype Before You Architect
A few days ago I was talking about Raspberry Pi Pico.
Eventually, I still want to head in that direction — but whether that’s in a year or several years… even I don’t know yet.
Lately, though, I’ve run into a completely different problem.
As my freelance projects have increased, many clients have turned out to belong to the same holdings group. What started as “one website per company” gradually became “multiple brands, multiple companies, shared billing, ongoing support, and related projects.”
At that point… my management system quietly descended into chaos.
Originally, when I only worked with individual companies, I managed everything with Notion and a tiny task app I built myself.
It worked.
Until it didn’t.
Eventually I reached the point where even I couldn’t immediately answer the question:
“Wait… which project is this change for again?”
So I did something unexpected.
I went back to Excel.
Yes… Excel.
From there I realized something even funnier.
Instead of forcing myself into a web application immediately, building a VBA system first would actually be faster.
Ironically, I had already decided I was “done with VBA.”
(Although… I still maintain VBA tools at my pharmacy job, so maybe I never really escaped it.)
Why not build the app immediately?
Because I already know what would happen.
If I build the application now, I’ll spend the next several months rewriting it as my workflow changes.
Right now I don’t actually know:
- Which information is truly necessary
- Which fields I’ll never use
- Which dashboards are genuinely useful
- Which inputs become annoying after entering them hundreds of times
Those answers only appear after using the workflow for a while.
VBA is perfect for this stage because changing a form or adding another field only takes a few minutes.
It lets the workflow evolve naturally before committing to a permanent architecture.
(Though… there’s always a chance the VBA version simply refuses to retire.)
How did things become this complicated?
One client originally asked for:
“We need a landing page that can pass advertising review.”
Simple enough.
So I rebuilt an existing landing page.
Then…
“Actually, let’s use this as the production site.”
The revision project was born.
Not long after…
“We’re changing the business direction a little.”
Another project appeared.
At this stage I had chosen Astro because it was only supposed to be a static landing page.
Then came…
“Can we add member registration?”
“Can we integrate payments?”
“Can you build our official LINE with L Message?”
Astro probably looked at me and quietly whispered:
“Wait… I thought I was just a landing page.”
Then another message arrived.
“We need another LP for another company in the group.”
Before I could even process that…
Then, while I was working as a pharmacist, my phone rang.
The caller’s LINE profile picture was Bando Eiji.
My first thought wasn’t,
“Hello.”
Get CodeyNode’s stories in your inbox
Join Medium for free to get updates from this writer.
It was,
“Wait… who is this?”
Turns out…
“Could you build a landing page for our bar?”
A bar?
Really?
Until then I’d mostly been working on athlete services, recruitment businesses, and support services for freelancers.
Now a bar had entered the story.
Then Slack management appeared.
Then AI collaboration.
Then even more ongoing work.
At some point I realized something.
I wasn’t spending most of my time building websites anymore.
I was spending my time remembering what I was supposed to be building.
Every morning started looking like this.
Me:
“…I’m alive today too.”
Merge:
“Mofu. Sleep is important, my friend. Build the management system first… and don’t forget to exercise.”
Probably good advice.
The workflow I’m experimenting with
Instead of managing projects as simple tasks, I’m starting to think of them as evolving records.
A single request often grows into multiple stages.
A completed landing page may later become a payment system.
A finished project may receive another request before the previous invoice cycle has fully closed.
That means a project can be:
- Financially completed
- Operationally still active
- Already invoiced
- Still receiving additional work
All at the same time.
So rather than managing individual tasks, I’m experimenting with a small management system built around a few simple ideas:
- Master tables for clients, companies, project types, and billing destinations
- Form-based input instead of manually editing spreadsheets
- Status management that reflects the project’s actual lifecycle
- Separate management for referral points and cash payments
- A dashboard showing what actually matters:
- current workload
- expected income
- invoices waiting to be sent
- payments waiting to arrive
- estimated monthly cash flow
Not because it’s elegant.
Because I finally reached the point where memory alone isn’t scalable anymore.
Building software starts with understanding the work
This project probably won’t end as a VBA application.
Eventually it’ll become a web application.
Maybe it’ll connect to Slack.
Maybe it’ll grow into something much larger.
But today isn’t the day to build the final product.
Today is about learning what the product actually needs to become.
Sometimes the fastest path to good software isn’t writing more code.
It’s understanding your own workflow first.
And strangely enough…
Excel is helping me do exactly that.
