One developer spent six months scraping luxury hotel prices every single day — pulling a full year of rate data across 2,000+ hotels daily — and turned the dataset into a working site.
The project targets hotels with Travel Advisor rates, and the patterns that fell out of the data are specific enough to be useful:
- Last-minute drops are rare. For the hotels that do discount, 2–5 days before check-in is the sweet spot.
- Weekend vs. weekday pricing splits cleanly. Some hotels pack out on weekends; others charge more on weekdays. The data makes the pattern obvious per property.
- “Low season” isn’t always low. Random price spikes hit during periods travelers assume are cheap — the developer got burned personally, arriving in Thessaloniki during Ochi Day (a Greek national holiday) with no booking and near-zero availability.
- City-level trend charts don’t exist elsewhere. Aggregating even 30 hotels in a city like Paris produces a readable year-ahead price curve showing cheap days, expensive days, and the spikes worth dodging.
The interesting part isn’t the scraping — pulling hotel rates daily is grunt work. The interesting part is that nobody seems to publish this view. Booking sites show you prices for your dates. They don’t show you a year-long price chart for an entire city so you can pick dates around the data. That’s the gap this project fills.
NOTE
The dataset covers hotels with Travel Advisor rates specifically — not every property on every OTA. The sample skews toward luxury and upper-midrange.
This is a clean example of vibe coding producing something with real utility: a solo builder, a scraper, a dataset nobody else bothered to assemble, and a front end that makes it browsable. No team. No funding round. Just a script running daily for six months.
2,000+ hotels, six months of daily pulls, one person. The project is live and the developer shared it on r/vibecoding.