I'm building Renviq, software for small car, motorbike and boat rental businesses. We wanted a number we could stand behind: how many of these businesses can actually be booked online?
So we pulled 1,759 rental businesses from Google Maps across five Spanish tourist areas (Mallorca, Ibiza, Málaga, Alicante, Tenerife), downloaded each homepage, and looked for a booking engine. Result: 8% bookable. 92% can't take an online booking. A beautiful headline for someone selling booking software.
It was wrong. A plain download runs no JavaScript, so it misses every booking widget injected by a script, loaded in an iframe or living on a sub-page. We re-checked a stratified sample of 119 in a real browser, three passes each. The corrected figure: 68.9% cannot be booked online end to end (95% CI 56.9–80.8%, and it's a floor, because our "bookable" bar is generous). Of the ones that can, only one in seven runs an actual booking engine; the rest bolted a shop cart onto a car.
What I took from it:
- A homepage crawl overstates "no online booking" by a lot. If you see a study built on one, discount it.
- The error ran in the direction that flattered us. That's exactly when to check twice.
- Publishing the correction cost nothing, and the corrected number is one I can defend to anyone who checks the method.
Full method and data: https://renviq.io/en/guides/spain-rental-online-booking-study
The data is also on Zenodo under CC BY 4.0: https://doi.org/10.5281/zenodo.23789643
