A small real estate outfit in the Philippines needed to look like a firm you would trust with eight figures. I built them two things: an agency site that establishes credibility, and a listing system that sells individual properties. Design, front end, data layer and project management, all of it me.
The agency site convinces you the company is real. The listing pages convince you about one house. Two jobs, one visual system.
Client name and real listings withheld. Placeholder content in the mockups; the structure and type treatment are as built.
Nobody buys property from a website. The site's job is to get them to the viewing.
Listing sites love to bury the number behind an enquiry form. That filters out serious buyers, not tyre kickers. The price sits next to the location in the largest accent type on the page.
The dark enquiry box sits next to the description rather than at the bottom of the page. The moment someone finishes reading about the house is the moment they want to ask about it.
Floor area, lot area, bedrooms: three figures in a row with icons, not a spec table. They are the first things a buyer checks and the last thing they want to hunt for.
The agency site puts the team on the homepage with names and roles. For a small operation competing against national brokerages, a real named person is the credibility a logo cannot buy.
Suggested properties close every listing page. Most visitors will not want this house. Giving them three more keeps them on the site instead of back on a search results page.
Same green, same type pairing, same card treatment across the agency site and every listing page. A small operation cannot afford to look like two different companies, so the listing pages read as part of the same firm rather than a separate microsite.
There was no team. That shaped the work more than any design decision.
A framework I could move fast in and the client could hand to anyone later. A bespoke build would have looked better on my end and been worse for them. Deployed on Vercel so updates ship without a server conversation.
Parts of this stack were new to me when I took it on. The deadline did not care. Shipping something real while still learning it is how most of what I know arrived, and it is a reasonable way to work as long as you are honest about it with the client.
Scoping, chasing property photos and copy, deciding what shipped in round one and what waited. Being the only person on it meant the schedule and the design were the same conversation, which is faster and lonelier.
Both sites shipped and went live on Vercel. No conversion data on this one, which is honest: it was a small client with no analytics practice and I was not going to invent numbers for a portfolio. What I take from it is the range, one person handling design, build, data and schedule, and the discipline of making a small firm look established without overstating what it is.