Garden Church was planted in 2009 as Anthem CDA. In 2026 it changed its name to match what it had actually become, and a name change is the moment every inconsistency in a brand becomes visible at once.
A church ships more design per week than most startups. Sunday slides, series artwork, event graphics, social posts, bulletins, signage, the site. None of it made by designers. All of it made on a deadline that does not move.
The staff did not need a designer. They needed to stop needing one.
Thirty-nine pages across six sections: story, identity, typography, colour, photography, and voice. The document specifies proportion and sequence rather than preference.
It is written to be read by people who were never trained as designers. Photography direction names the high-priority subjects and, more usefully, the low-priority ones. Telling a volunteer what not to shoot is worth more than telling them what to shoot.
Built and shipped directly in code, serving both the current and the legacy domain so nobody arrives at a dead link after the rename. No mockups. No handoffs. What I designed is what shipped.
This is what I call Direct Design. Build directly in the medium. The browser is the canvas. The feedback loop collapses from weeks to minutes. Every decision is tested against real constraints, real content, real devices. Not a simulation of a website. The website.
View live site: gardencda.com
| Before | After |
|---|---|
| Two names in circulation | One lockup, supplied as a single unit, never redrawn |
| Colour chosen per asset | Colour specified as surface proportion, sixty percent green |
| A blank canvas every Saturday night | A template library covering the recurring weekly assets |
| Requests arriving as hallway asks | One intake route with named decision owners |
| Stage-lit crowd photography | Photography direction naming what to capture and what to avoid |






With more time I would have run the training session twice, once before the guidelines were final and once after, because the questions people ask when they are actually making something are different from the questions they ask reading a document.
The brand book was half the work. The other half was operational, and that half is what decides whether any of it survives.