Rebuilding My Own Site from WordPress to Astro Felt Like Leaving an Old Friend

I've been telling clients to move away from WordPress for a while now. Preaching the benefits of static sites, faster load times, less maintenance, better security. All true. All sensible. And yet when it came to rebuilding my own site — soliddigital.io — I put it off for months.

Not because it was technically difficult. Because it felt wrong.

WordPress and I go way back. Over a decade. It's how I learned web development. It's how I built my first client sites. It's how I paid my bills in the early days when "web designer in Bradford" was my entire business model and every job was a WordPress theme and a prayer.

Ripping that out and replacing it with something shiny and new felt like trading in a car that still runs. Sure, the new one's faster. But you've got memories in the old one.

The WordPress Years

My WordPress setup had been through everything. Started as a custom theme, got rebuilt twice, survived a migration between hosting providers that nearly gave me a heart attack, and somehow still worked. Kind of.

The thing is, it was never just a portfolio site. It was my testing ground. Every new plugin I wanted to evaluate for a client project got trialled on my own site first. ACF was the backbone — I used Advanced Custom Fields for practically everything, building out flexible layouts and custom post types the proper way. No Elementor, no page builders. I avoided those like the plague. If I couldn't build it with ACF and clean templates, I didn't want it on my site.

But that meant it accumulated years of experiments. Plugins I'd installed in 2018 to test for a client and never removed. A staging environment that hadn't been touched since 2021. Three different contact form plugins because I was evaluating them all and apparently couldn't commit to one. Gutenberg with ACF custom blocks because if I was going to use the block editor, I was doing it my way.

It wasn't neglected. It was lived in. There's a difference — though the PageSpeed score didn't seem to care.

The Tipping Point

The moment I actually committed to the rebuild was embarrassingly on-brand. A client asked me why their WordPress site was slow. I ran their PageSpeed score. Then — out of morbid curiosity — I ran mine.

64 on mobile.

Sixty. Four.

I'm out here telling Bradford businesses their sites need to be faster, and my own site was loading like it was being served from a filing cabinet in someone's shed. The cobbler's shoes, as they say. Except worse, because I literally do this for a living.

That was the push. Not some grand strategic vision. Just pure, undiluted embarrassment.

The Actual Rebuild

I won't bore you with every technical decision — I've covered the WordPress to Astro comparison in detail already. But rebuilding your own site is a different experience to rebuilding someone else's.

For a start, you're the worst client you'll ever have. Indecisive, scope-creeping, constantly second-guessing yourself. "Should the font be slightly larger? Should I add a blog section? Should the accent colour be warmer? Actually, let me redesign the whole services page from scratch for the fourth time."

I redesigned that services page five times. Five. A client would have been invoiced for three of those.

The content migration was the bit I dreaded most. Years of blog posts, case studies, page content — all living in a MySQL database, formatted with WordPress shortcodes and ACF flexible content fields that made perfect sense inside WordPress and absolutely none outside of it. In the end, I decided to start fresh. Clean slate. The old posts had served their purpose and most of them weren't bringing in traffic anyway. Nobody was reading my 2019 post about "Top 5 Reasons Your Business Needs a Website" — certainly not now that AI can generate that same post in eleven seconds. Better to start the blog again with content that's actually worth reading than drag across years of filler for the sake of it.

The Moment It Clicked

About three days into the rebuild, something shifted. I'd set up the basic structure in Astro, dropped in a few pages, and hit preview.

It loaded instantly.

Not "fast." Not "pretty quick for a dev environment." Instantly. Like the page was already there before I'd asked for it. Because, in a sense, it was — Astro had already built the HTML. No database query, no PHP, no theme framework initialising sixteen things I didn't ask for.

I sat there for a moment, clicked between pages a few times like an idiot, and thought: "Oh. Right. This is what websites are supposed to feel like."

That was the moment WordPress stopped being an old friend and started being an old habit.

What I Don't Miss

The update notifications. That orange circle in the dashboard telling me seventeen plugins need updating. The anxiety of clicking "update all" and wondering if the site would still work afterwards. The answer was usually yes. Usually.

The speed optimisation dance. Caching plugin, image optimisation plugin, minification plugin, lazy loading plugin — four plugins just to make the site do what it should have done by default. And even then, 64 on mobile. Brilliant.

The security paranoia. Wordfence alerts at 2am. Brute force login attempts from IP addresses in countries I've never been to. That one time I found a PHP file in my uploads folder that definitely wasn't mine. Fun times.

The hosting bill. Fifteen quid a month for managed WordPress hosting. Not a fortune, but for a static site that gets maybe a few hundred visits a month? Cloudflare Pages does it for free. Actually free, not "free tier that'll start charging you if someone shares your blog post" free.

What I Do Miss

Honestly? The familiarity. Logging into that dashboard was like putting on an old jumper. Everything in its place. Posts on the left, pages below, settings at the bottom. I could navigate it with my eyes closed.

Astro doesn't have a dashboard. It has code. Which is fine — I'm a developer. But there was something comforting about WordPress's admin panel. Like a cockpit where everything had a button, even if half the buttons were from plugins you'd forgotten you'd installed.

I also miss the five-minute fix. Client needs a typo changed? Log in, edit, publish. With Astro, I edit a file, commit, push, wait for the build. It's ninety seconds. Hardly a hardship. But WordPress had that instant gratification thing nailed.

That said, I've set up Keystatic now, which gives me a proper editing interface. So even that's fading.

The Numbers

Because this wouldn't be a proper post without them:

Four hundreds and a ~98. On mobile. I didn't even know PageSpeed went that high — I'd certainly never seen it on a WordPress site I'd built. The desktop score is a clean sweep. 100 across the board. Performance, accessibility, best practices, SEO. All green. All perfect.

The old WordPress version? I never ran the full audit because I didn't want to know. The performance score was 64 on mobile and I stopped looking.

The last row is the one that matters most to me personally. Sleep is underrated.

Can You Make WordPress Fast? Yes. But...

Look, I'm not saying WordPress can't be fast. It absolutely can. I've done it for clients. Strip the theme back to basics, avoid page builders, use ACF for structured content, pick your plugins carefully, set up proper caching, optimise your images, put it on decent hosting. You can get a WordPress site into the 80s and 90s on PageSpeed. I've done it.

But here's the thing: that takes effort. Ongoing effort. You're fighting against WordPress's natural tendency to accumulate bloat. Every plugin update, every new feature, every "I just need this one thing" request from a client adds weight. You're constantly optimising to counteract complexity that doesn't need to be there in the first place.

With Astro, the performance is the default. You don't optimise towards fast — you start there and have to actively try to make it slow. That's a fundamentally different starting point, and for a simple business site, it's the better one.

Could I have made my old WordPress site score 90+ on mobile? Probably. If I'd stripped out the test plugins, rebuilt the theme from scratch, and spent a weekend on performance tuning. But at that point, I'm spending more time making WordPress behave like a static site than it would take to just... build a static site.

Would I Go Back?

No. Not for my own site, and not for most client projects.

WordPress served me well for years. It taught me web development, it paid my bills, it built my business here in Bradford. I genuinely owe it a lot. But the web's moved on, and the tools have moved on, and clinging to something because it's comfortable isn't a business strategy.

The site's faster. It's cheaper to run. It's easier to maintain. It looks better. And I no longer have to explain to clients why my own site doesn't follow the advice I give them.

Sometimes the best thing you can do for an old friend is admit you've both changed — and move on.