Case study · № 001
Kwarto's Guest House
My family's guesthouse in Purok 3, Tubod, San Juan — and the first site I built. 8.2 on Booking.com from 47 guest reviews at the time of writing.
A guesthouse that only existed on other people's pages
Before the site, Kwarto's lived on Booking.com and word of mouth. A listing works — it brings guests — but it isn't ours. Every booking pays a commission, the photos sit inside the same template as every other property, and a guest who searches for the guesthouse by name lands on a booking widget, not on the family running the place.
I built the site to give the guesthouse one page of its own: our photos, our words, and a direct way in — with the Booking.com listing still there for people who prefer it.
kwartosguesthouse.com
The site opens with the island itself, then gets to the point: a Booking.com badge shown proudly — 8.2 from 47 reviews — and two clear paths for a guest: “Book Your Stay” or “Explore Siquijor”.
It started as the same brochure recipe I sell on this site. Then the guesthouse outgrew it. Today it's a full booking platform: seven rooms with live rates and availability, booking paid right on the page by QR Ph — any bank or e-wallet app, commission-free — real-time guest chat with saved replies and after-hours auto-answers, and an admin dashboard where the family manages rooms, walk-ins, and every booking. Guests can even install it like an app on their phone.
What shipped since
The list further down is the second one I've published here. In August 2026 I spent four rounds of work on it. Here is what changed, in the order it shipped:
01
A guest who pays now gets it in writing. The deposit goes through PayMongo and the confirmation lands by email and in the chat thread, instead of a redirect page the guest has to trust. A card that gets declined now says why.
02
The app carries less. The Firebase pieces the first screen doesn't need load later, the admin dashboard moved into its own file that only the family downloads, and a dead animation module came out. The main script went from 958KB to 746KB.
03
Every room has its own page. Seven rooms, seven addresses, a rooms index and real 404 pages — Google can reach a room directly now instead of only the homepage.
04
The first screen paints before the app loads. Every external script is deferred and the opening screen sits in the HTML itself. On a throttled phone test the biggest element appears in 9.9 seconds, down from 13.5 in May. Better. Still not the 4 seconds I want.
The honest part
What's still missing
I promised these case studies would tell the truth. The first version of this list — no rooms or rates, no contact details, no map, no amenities — has since shipped, every item, and so has most of the second one above. Here is the honest list as of August 2026:
01
Booking.com is still reconciled by hand. I wrote the tool that reads their reservations export and blocks those dates on our own calendar, but I'm the one who runs it. Nothing syncs on its own, so a booking made there can collide with one made here in between.
02
The site is still slow on a weak signal. 1.2MB and 52 requests to open the homepage, and the analytics and Meta pixel scripts together weigh more than my own code does.
03
A guest can't cancel or change a booking on the site yet. That's still a message to us, and refunds are handled by hand.
That list is the next round of work, and it's also the point: a website isn't a trophy you hang once, it grows with the business. I'd rather show you a real site mid-journey than a perfect mockup of one that doesn't exist.
Want one like it — with your gaps closed from day one?
A brochure site — your rooms, rates, map, and contact details, photographed and built by me — is ₱17,500 as one of three founding slots, half the regular ₱35,000, and I write a case study this honest about yours. And if your place is ready for what Kwarto's has now — bookings and payments on your own site, commission-free — that's the booking web app, scoped together. It is from ₱120,000, or ₱60,000 as one of two founding slots.