The Uncomfortable Rebuild: Why I Blew Up My Website and Started Again
Summary
Portrush seafront, Thursday night. The kind of late March evening where the weather can't quite commit — cold enough to keep the windows up, bright enough to remind you that winter is losing the argument. The 400 watts of solar on the roof was pushing out nearly 150 watts, which for this time of year is a result worth noting. Spring is finally pulling its weight.
I'd had a rough day. The kind where by the time you close the work's laptop, the last thing you want to do is open another one again and start something difficult. But I'd been putting this off long enough, and the longer I left it, the more the site was going to keep quietly embarrassing me every time someone opened it on a phone.
So I made a brew, listened to the sea do its thing outside, and started pulling apart a website I'd spent two years building.
Not the most relaxing Thursday evening. But some things need doing before you're ready to do them.
The Squarespace Temptation
This is the story of my WordPress migration from Elementor to Kadence, and why it nearly didn’t happen.
Before we get to the rebuild, I need to confess something.
I nearly left WordPress entirely, I wrote about it in my previous post.
I had the tab open. I had the credit card out. I was fourteen minutes into a Squarespace trial, and I was genuinely starting to believe that a drag-and-drop lifestyle template was the answer to my problems. A YouTube influencer ~ bless his sponsored heart ~ had made it look like a weekend in Fermanagh. Peaceful. Simple. No plugin conflicts, no update notifications at 11pm, no staring at a mobile performance score that makes you want to lie down.
The honest truth is I wasn’t unhappy with WordPress. I was unhappy with my template. Elementor had served its purpose, but the site was slow, the mobile experience was embarrassing, and every time I looked at it I felt like I was looking at something I’d outgrown. Instead of fixing that, I was about to throw the whole thing out and start fresh on a platform that, the moment I actually poked around in it, revealed itself to be exactly what I should have expected.
A walled garden.
Beautiful from the outside. Bonnet welded shut.
Want to connect a custom workflow? That’s a premium tier. Want the granular SEO control I’d spent a year building with Rank Math? Here’s a simplified dashboard with fewer options and a cheerful interface that implies you probably shouldn’t be worrying about that sort of thing. Want to use the structured data I’d built into my Motorhome Tech posts — the queryable specs, the location-tagged park-up data, the ACF fields that make the site actually useful rather than just pretty? Good luck. That becomes a text block. Dead. Unsearchable. Trapped in a nice font.
I closed the tab. I kept the credit card in my pocket.
What I actually needed wasn’t a new platform. It was a better template and the willingness to do something uncomfortable.

Get Uncomfortable
There’s a thing I’ve said on LinkedIn more than once:
Comfort is the enemy of growth.
I say it about van life, about remote work, about learning new tools. It applies equally to websites, it turns out.
The decision to migrate from Elementor to Kadence Blocks wasn’t made lightly. It was going to mean rebuilding the homepage from scratch, reconfiguring the header, sorting out the shop layout, checking every page looked right on mobile, and doing all of this on a staging environment on Cloudways while also trying to, y’know, actually run the site. Three days, as it turned out. Not an afternoon. Three full days of incremental progress, the occasional backwards step, and at least one moment of staring at the screen wondering if I’d made a catastrophic mistake.
But here’s the thing about doing something difficult now to enjoy the benefits later, at some point, later arrives.
The performance scores went up. The mobile experience got clean. The site started looking like something I’d actually built with intention, rather than something I’d assembled in stages over two years and was now too afraid to touch.
A New Way of Working
The rebuild also coincided with a change I’d been circling for a while.
My previous AI partner and I had a good run. Useful tool. Got a lot done together. But when the news broke that they’d decided to cosy up with organisations whose definition of “responsible use” differs quite significantly from mine, that was the end of it. I’m not precious about much, but I am precious about that.
So I switched. Claude — Anthropic’s model — became the new AI in the passenger seat for this project. And if I’m honest, working through a three-day site migration with an AI that’s genuinely trying to help rather than just generating plausible-sounding answers made a noticeable difference. When I was debugging why the homepage header had vanished entirely, it stuck with the problem through CSS audits, database queries, plugin deactivations, and what felt like a full archaeological dig through two years of code, without once suggesting I just start a new chat.
We’ll come back to that.
The Cherry on Top
After three days of migration work, one issue remained. The homepage had no navigation menu. Not hidden. Not transparent. Just absent. What followed was a methodical, some might say obsessive, investigation through the CSS, the caching layers, the plugin stack, the database, and the theme files.
The answer, after approximately three hours, was a toggle at the bottom of the block editor sidebar.
“Disable Header” — switched on. Probably fat-fingered at some point during the rebuild and never noticed.
I turned it off. The header appeared. I sat quietly for a moment and listened to the seagulls flying overhead, the waves hitting the rocks.
The site works. The mobile layout is clean. The header is where it should be. And somewhere in a block editor sidebar, a toggle is sitting in the OFF position where it belongs.

Was It Worth It?
Yes. Obviously yes.
The uncomfortable version of this story is that I nearly talked myself out of it twice, once when I nearly moved to Squarespace, and once on day two of the migration when everything looked worse than when I started. The comfortable option both times was to stop, patch something, and leave it.
But the site I have now is faster, cleaner, and actually mine in a way the old one never quite felt. It looks like Digital Nomad NI. It works on a phone. It doesn’t embarrass me when I send someone the link.
Sometimes you have to blow the thing up to build it properly.
The Starlink is locked on. The header is back. Let’s get on with it.
Enjoyed this? Subscribe below. I write when I have something worth saying, which means your inbox stays quiet most of the time.
SUBSCRIBE