Welcome to BnB Facts! I started this project in 2020 as a way to combine my interest in the short-term rental industry with my background in data science and analytics. With over a decade of experience in the field, I wanted to explore what the data could tell us about Airbnb and uncover insights that might be useful to hosts.
Over the course of the project, I published several in-depth analyses using Python and R, including:
- Uncovering factors which drive the occupancy rate using a gradient boost classifier model and interpreting the results using Shapley values
- Determining how hosts price their listings using a lasso regression
- How guests define value using logistic regression
- How to write a listing description using natural language processing (NLP) and network visualizations
- Most common reasons for bad reviews using NLP, sentiment analysis, and topic modelling.
- Creating a segmentation for Airbnb hosts using k-means clustering
- Understanding how guests score the overall rating as a function of other categories ratings using relative importance.
While BnB Facts is no longer an active project, I’ve decided to keep the site online as an archive of the research. Some of the data and findings may be dated, but I hope the analyses remain useful to anyone interested in the short-term rental industry.
If you’d like to get in touch, feel free to send me a message on Instagram at @bnbfacts.