Bodrum Property Group

The brief

A Manchester property investment company with no web presence and a founder whose whole business model runs on being trusted by strangers with money. What it needed was not clever: it needed to look established, explain one specific investment strategy clearly, and take an enquiry without losing it.

I built the CMS first

The received wisdom on client sites is to give them something they can edit themselves, so I did. Jekyll with Decap CMS 3.0 behind Netlify Identity, and three properly typed collections:

• Properties — price, address, bedrooms, bathrooms, area, an image upload, and a Status field constrained to Available / Lease Option / Completed so it could not be typed wrong.
• Testimonials — name, role, avatar letter, text.
• Videos — title, description, file upload, tag.

A real admin, deployed, that he could log into and use. It took the work it takes.

He didn’t want it. And he was right not to — he was never going to log in to edit a property card, and a CMS nobody uses is not a feature, it is a thing that breaks quietly while you aren’t looking. A plugin goes stale, an Identity service changes its terms, and a year later the site is a liability instead of an asset.

So thirteen days after building it, I took it out.

Why a single file

One index.html containing the markup, the styles and the small amount of JavaScript the page needs. Images in a folder beside it. The contact form posts to Formspree, which means a static file can still receive enquiries without anything running on a server.

No build step, no dependencies, no framework. Open index.html in a browser. That’s it.

That line is from the repo’s README, and it is the whole maintenance story. It only reads as a confident decision because the other version exists — the backend is archived, not deleted. If he ever changes his mind, it works. Throwing it away would have made the simpler site a smaller claim. Any text change is a find-and-replace by someone who has never seen the project. In a year, it will still work, because there is nothing in it that can rot.

What changed along the way

The interesting edits were not design ones. They were all the same category of mistake — placeholder content that looked finished enough to forget about. Stock property photos with invented prices. A stock hero image instead of the founder. A gallery section with nothing in it.

Each one got caught and replaced, and each time the page got more convincing in a way that no amount of layout work would have achieved. The gold stat numbers were the one genuine design error, and it was the obvious kind: using the accent colour on the content that most needed to be legible.

What I would do differently

Retire the GitHub Pages copy. It was the original host, the site moved to GoDaddy, and nobody turned the old one off — so there are now two live copies of the same site, and only one of them gets updated. That is a slow-motion problem, and it is the kind of thing that is trivial to fix today and confusing in two years.