Skateland West opened on San Antonio's Loop 410 in 1985. Their WordPress site came along years later and had been quietly getting in the way ever since. Here's how I moved them off it, and the handful of decisions that mattered most.
The short version is in the case study. This post covers the decisions: what I threw away, what I kept, and the calls where the interesting work actually happened.
Why not WordPress.
WordPress isn't the problem. This particular WordPress install had calcified. It had a stack of plugins patched around each other over the years, a page builder nobody on staff knew how to use, and a theme modified by hand so many times that it was brittle to touch. Updating the Friday-night schedule reliably meant a phone call to someone who knew the workarounds.
I looked at cleaning it up in place. For roughly the same effort, I could get them onto something they'd still want to be using five years from now, so the rebuild won.
The stack.
- Next.js on the front end. It's fast, well supported, and not going anywhere.
- Payload CMS with Postgres behind it. The editor is built around what the staff actually change, and the content structure is written in code, so it's reviewable.
- My own server, running everything in Docker behind Caddy, with monitoring and nightly off-site backups.
Nothing on that list is fashionable, and that's on purpose. Each piece is dull enough to maintain for years without surprises.
The decisions that mattered.
01. Owning it vs. a hosted website builder.
A hosted builder would have made the migration faster. It would also have put the rink's content behind someone else's login, with a monthly bill that grows as the site does. The family was clear that they wanted to own it. A site on a server I manage, with a database backup they can download any time, answered that better.
02. Keeping the owner's own words.
The old site's copy was written by the owner herself. It wasn't polished, but it was hers and it sounded like her. I moved the wording over almost word for word and edited only where a sentence was no longer true. The new site sounds like the rink because I didn't try to improve the rink's voice.
03. Bookings: stay with what works.
My first plan included a custom booking tool on the site that fed the event calendar directly, and I offered to build it. But the rink already ran bookings through a provider the staff knew how to use. A new tool would have meant another login for the team and another system for me to maintain, so I left the provider in place. The offer still stands for whenever they want to bring bookings in-house.
04. The newsletter: simpler than planned.
The newsletter went out from a personal Gmail account, addressed from a spreadsheet. My first version replaced that with a self-hosted newsletter tool. Before launch I swapped it for something simpler: sign-ups from the site now land straight in a Google Sheet, with spam protection on the form. If they outgrow it, moving to a full email platform is a one-file change. Until then, there's one less tool for them to learn and one less thing for me to maintain.
05. A quiet launch.
The new site ran at a staging address for months while the family reviewed it. On May 12, ahead of the 41st season and before the old hosting contract ran out, I moved the rink's domain over. There was no announcement. The goal was for regulars to show up Friday night and notice only that the event page finally worked on their phone.
The hosting math.
The old setup cost $398.25 a year for Turbify's managed WordPress hosting, with domain and basic email bundled in. That's cancelled now. The rink pays me $250 a year, one flat fee covering hosting, backups, updates, monitoring, and small changes. That's $148 less than Turbify charged for hosting alone. Their site shares a server with a handful of other client sites, and the whole server costs about $16 a month. That's how a small business can get custom software looked after without enterprise pricing.
The more interesting number never showed up on the old bill: how much of the staff's time the old site was costing them. Since launch, the family has updated their own pricing, schedule, birthday-party page, and seasonal promos without calling anyone.
The point.
Skateland didn't need a fancier website. They needed their software to stop getting in the way of what they're already great at: two generations of a family running a roller rink on Loop 410, with neon signage and seven-dollar Thursdays. The best thing I could do was make the website quieter, and then keep it running.