You know that friend who gives everyone great advice and then ignores all of it in his own life? For a long time, that was me and this website.
I spend my days building websites for other people. Meanwhile, the site you’re reading right now was held together with duct tape, old plugins, and prayers. The cobbler’s kids have no shoes. The web designer’s gaming blog was still loading like it was 2019.
So I finally did something about it. This post is the full breakdown: what was broken, what I changed, what it cost, and what I learned rebuilding my own site after years of building everyone else’s.
Why I Finally Snapped
Real talk: the old site wasn’t terrible. It was worse than terrible. It was fine.
Fine is the most dangerous thing a website can be. Terrible gets fixed. Fine gets ignored for three years.
Here’s what “fine” actually looked like under the hood, straight from a GTmetrix report I ran before touching anything: a D grade, a 54% performance score, an 8.1 second fully loaded time, and 856ms of the page just sitting there blocking the main thread before you could interact with anything.
- A theme I picked in an afternoon and never questioned again
- Mobile layouts that technically worked, the way a shopping cart with one bad wheel technically works
- Search traffic growing in spite of the site, not thanks to it
That last one is what got me. I could see in Search Console that posts like our Letterboxd alternatives for gamers piece and the Nine Sols article were ranking on page one and two. Google was sending people to a site I was quietly embarrassed by.
Sound familiar? If you run any kind of site next to a full-time life, I’m guessing it does.
What I Kept (And Why That Matters)
First instinct when redesigning anything: burn it all down. Resist that instinct.
I kept three things on purpose:
WordPress. Yes, I build client sites on Next.js and Webflow all the time. TAG stays on WordPress. The content workflow works, Rank Math handles the SEO plumbing, and migrating a few hundred posts to a new platform is a great way to lose rankings you spent years earning. The boring choice was the right choice.
Every URL. Not one permalink changed. When you have posts ranking for terms people actually search, changing URLs is like deranking yourself in placement matches on purpose. If a redesign ever forces a URL change, you set up redirects. But the best redirect is the one you never need.
The content. No mass deleting, no “fresh start.” The archive is the asset. The wrapper was the problem.
What Actually Changed
The short version: design, speed, and structure.
Design. New theme, built lean. Fewer fonts, real spacing, dark-mode-friendly, and layouts that assume you’re reading on your phone during a lunch break. Our whole audience is people squeezing gaming into real life. The site should respect that.
Speed. Cut the plugin count from around 52 down to 36. Compressed and lazy-loaded images. Real caching instead of none. The difference shows up in the numbers: that D grade and 8.1 second load turned into a PageSpeed Insights report that now passes Core Web Vitals on real user data, with a performance score of 62 on mobile and 86 on desktop.
Structure. Clearer categories, related posts that are actually related, and internal links between guides so you can get from a Palworld base guide to the tier lists without using the search bar like a grappling hook.
The Tools and the Bill
Everyone hides the numbers in these posts. Here are mine.
- Hosting: Cloudways, $7.64 a month. Same bill I had before the redesign, nothing new.
- Theme and build: custom, my own hours, with Claude Code doing a lot of the heavy lifting on the PHP and theme work.
- Plugins that survived the purge: Rank Math for SEO, Breeze and Redis Object Cache for speed, Imagify and Advanced Media Offloader for images.
- Design tools: Figma, free tier.
- Total out of pocket: $200, one month of Claude Code Max. That’s the whole bill on top of hosting I was already paying.
The real cost was time. About four weeks of actual, heads-down work, stretched across a couple of months because plenty of weeks I didn’t touch it at all. Which is exactly what I’d tell you to expect if you’re doing this yourself around a job: it’s a marathon boss fight, not a speedrun.
What I Learned Doing It For Myself
Building for clients taught me the skills. Building for myself taught me the lessons.
Your own site is the hardest client. No deadline, no invoice, no one waiting. You have to manufacture the accountability yourself. I finally treated TAG like a paying project, put it on the calendar, and shipped it the way I ship client work.
Speed is a feature readers feel but never mention. Nobody emails you to say the site loads fast. They just stop leaving.
Done beats perfect. There are still ten things I want to change. There always will be. Shipping the 90% version now beats shipping the 100% version never.
The Part Where I Mention What I Do
Some of you know this, some don’t: rebuilding sites like this is my actual job. I run Plinth, a web design studio where I build custom sites for founders and small brands. This rebuild used the exact same process I use for client work, from the audit to the launch checklist.
If your site is stuck in “fine” and it’s costing you, that’s literally what I fix all day. But honestly, even if you never hire anyone, steal the process from this post. Keep your URLs, cut your plugins, respect the phone reader. You’ll be ahead of most of the internet.
The site you’re reading is the proof of work.
Rebuilt your own site or thinking about it? Tell me what’s holding you back in the comments, or come yell at me about my font choices. Both are welcome.
About the Author: Fred is one half of Two Average Gamers, a community-focused gaming site dedicated to helping regular folks enjoy gaming without the toxicity. He builds websites for a living at Plinth and still took three years to fix his own. Do as he says, not as he did.
