The requirement that mattered
Not "build a CMS". The requirement I was given was that a specific person could add a gallery on a specific afternoon without messaging anyone. Requirements that concrete are rare, and I protect them from scope creep.
Nexel visual CMS
01
The problem
The team's complaint was specific and correct: adding gallery images and videos was too complicated. Not impossible — complicated enough that content updates queued behind a developer, which meant the site went stale between releases for no good reason.
02
Approach
Aim for the ease of use people already expect from tools like Framer and Wix, rather than building a technically elegant content API that still requires someone technical to operate. The measure is whether a non-developer can publish a gallery without asking anyone.
03
What it does
- Realtime preview
- Seeing the change as it will appear, rather than saving and hoping — which is what makes non-technical editors confident enough to publish.
- Simpler gallery management
- The specific pain point that started the project, treated as the primary requirement rather than a feature among many.
- Easier video uploads
- Video is where most CMS workflows break down for non-technical users, so it gets first-class handling.
- Full-site editing
- Editing scoped to the whole site rather than a fenced-off content area, because the content people want to change is rarely only inside the blog.
01
02
03
04
04
The hard parts
- Flexibility fights simplicity
- Every configuration option added for an edge case makes the common path harder. A CMS for non-developers has to be opinionated to be usable.
- Preview accuracy is the whole trust model
- If preview and published output ever differ, editors stop trusting the tool and go back to asking a developer.
01
02
Detail
Stack
- Editing
- Media
Case study
The decisions behind the build