30 pages in three weeks, without cutting corners

The launch date was fixed before we started. Around 30 pages, motion in almost every section, and a client who would notice if anything felt rushed. We had three weeks, revisions included.

A deadline like that makes it tempting to build fast and fix later. I didn’t want a site that needed rework the month after launch, so I spent the first days on the parts nobody sees.

I set up the project structure and variables, and designed a new CMS architecture. The content was moving over from the client’s old Webflow site, so I restructured the collections instead of copying them as they were. That kind of fix is cheap in week one and expensive a year later.

I built the global components first — the pieces every page shares — and then went straight into pages. Each new section went into the component library as I built it, so the library grew with the site. A second developer worked from the same library, and anything already built could be reused rather than rebuilt. Two short review passes were enough to catch the small things.

The animations took the longest, and they were the easiest thing to cut. The client wanted to keep them, so we kept them. Because the components were already solid, I added the motion on top of a clean layout instead of patching the layout at the same time.

We launched on the date. The revision round was short, and most of it was about content, not the build. The site went live on a structure made to take new pages and content for years, not just to survive launch week.