Hindsight Is Simple

If I were starting Light School over from scratch, I would do things in a different order. Not because the current order was wrong. It worked. But I can see now where time was spent on things that could have waited.

The biggest change: I would start with the campus lessons, not the homepage.

Content Before Design

We spent time early on getting the homepage right. The layout, the hero section, the animations, the messaging. It looks good. But nobody comes to Light School because the homepage is pretty. They come because they need help with Claude Code.

The campus lessons are the thing people actually use. Every lesson solves a real problem. Every lesson gets someone unstuck. If I had built those first, people would have been getting value from day one, even if the homepage was just a list of links.

Ship the Useful Part First

This is not a new idea. People talk about shipping early and iterating. But it is easy to fall into the trap of wanting everything to feel "ready" before you share it. The homepage felt like the front door, so it had to be right. But the lessons were the actual rooms people needed to be in.

A lesson on how to set up Claude Code is useful on its own. It does not need a pretty homepage above it. It does not need a blog next to it. It just needs to exist and be findable.

What Else I Would Change

I would write the slash commands earlier. The tweet-to-lesson command and the lesson idea finder save so much time that every lesson written before they existed took longer than it needed to.

I would also set up the icon system on day one. We went through a few approaches before landing on the Noun Project API with the fetch script. If that had been in place from the start, there would have been less rework.

Small things: I would pick the partials structure earlier. Deciding that every page shares the same head, nav, footer, and scripts partial saved a lot of duplication, but we figured that out as we went instead of planning it upfront.

It Still Worked

None of this means we did it wrong. Building a project is messy. You learn what matters by building the wrong thing first. The homepage taught us what the site should feel like. The early lessons taught us what language to avoid. Every step informed the next one.

But if I did it again? Lessons first. Always lessons first.