Skift Take
Past events gave us enough incentive to look into how vendors gather and glean data on short-term rentals. Where there is transparency, there can be a degree of trust.
Much ink has already been spilled (including our own) over questionable data that showed plunging Airbnb host revenue in several big cities. A tweet about it took on a life of its own, leading to a slew of news articles, social media posts, and investor concerns.
Other data providers said the tweet was wrong: Yes, host revenue was down a little, but nowhere as much.
With so much riding on this data, we wanted to understand how these firms gather, sort and process their numbers.
While most data companies scrape the internet for new listings, most of them also source data from integrations with Airbnb, Vrbo, Expedia, property managers, and individual hosts. They then apply proprietary algorithms.
For instance, AirDNA’s machine learning model is good at identifying blocked versus booked nights. PriceLabs prides itself on nailing hyperlocal data that can tell hosts how other listings are priced in their neighborhood. Beyond is testing pricing based on search data.
“All these companies are attempting the same kind of problems and they are thorny problems to solve,” said Drew Patterson, co-founder and CEO at travel startup Thermal. “Running data sets around short-term rentals is a complicated exercise, and it’s hard to get right. A lot of things can compromise analysis.” Patterson co-founded Transparent Intelligence, which was a