InspInspiration
We wanted to see how data could guide a real-estate investment decision. Given a $10 million budget, we asked whether Vancouver short-term rentals perform better as many lower-priced properties or as a few expensive ones.
What we learned High rental revenue doesn’t automatically make a good investment. Once purchase prices, transfer tax, operating costs, startup costs and cash reserves are included, the ranking of neighbourhoods changes. We also learned to question our own results: a segment with a high estimated yield may rest on only a handful of listings.
How we built it We built the analysis in R with the tidyverse. We cleaned Inside Airbnb’s Vancouver data and kept active, short-term, entire-home listings. We grouped them by neighbourhood, property type and bedroom count, and estimated annual revenue from reviews, minimum stays and nightly prices. To estimate purchase prices and property tax, we combined this with City of Vancouver assessment data, plus asking-price examples and sourced startup costs. Finally, we compared three $10M portfolio strategies and tested how lower occupancy and a higher Airbnb fee would affect returns.
Challenges Without real booking records, we had to estimate income. Reviews only approximate stays, and an unavailable calendar date can mean either a booking or a date the host blocked. Matching purchase prices to rental listings was also hard, because assessment data has no bedroom counts, so we used value percentiles as a stand-in. Updating the platform fee and adding startup costs noticeably changed our estimates, a reminder of how much the recommendation depends on its assumptions. We also treated the case as a simulation: in reality, Vancouver limits short-term rentals to the host’s principal residence.
Built With
- rstudio
