Why We Built Standalone Microsites Instead of Another Bloated CMS
A campaign page has one job: load fast, say one thing, and take one action. A general-purpose CMS is built for a different job, which is managing hundreds of pages with plugins, editors and workflows. This teardown explains why we built standalone microsites, and how to judge page performance on your own pages instead of taking our word for it.
The design target
We treat roughly 300 milliseconds to first meaningful content on a warm cache as a design target for a microsite, not as a promised measurement. Your result depends on your visitors' networks, your images and your own scripts. The architecture below is what makes the target realistic; you should still measure.
What you remove by going standalone
- A theme layer. Fewer templates means fewer places for layout shifts to start.
- Plugin scripts. Each plugin adds JavaScript the visitor must download and run before the page feels ready.
- A shared database call on every view. A published release is a document, not a query.
What you give up
A microsite is not a replacement for a content-heavy site with hundreds of articles and complex navigation. If your team needs editorial workflows across a large archive, a CMS is the right tool. Use a microsite when the page exists for one campaign and one conversion.
How to measure your own pages
- Test on a mid-range phone over a throttled connection, not on your office network.
- Record time to first meaningful paint and total transferred bytes.
- Remove one script at a time and re-measure. The cost of each script is usually larger than expected.
- Repeat after every design change, because performance regresses quietly.
The editorial rule that keeps pages fast
Every element on the page must earn its place by serving the single conversion. If you cannot name the action an element supports, delete it. Fast pages are mostly a content decision, and the architecture only keeps that decision from being undone.