As a successful representative of the sharing economy in the short-term rental market, Airbnb has grown rapidly over the past years, with millions of tourists have used its service. The motivations of people choosing Airbnb has been studied and explained in different perspectives, such as the budget limit, cultural preferences. Our project explored people’s motivations in terms of physical attributes. We analyzed Airbnb listings’ locations, online reviews, and amenities, that has influenced Airbnb users’ past staying experiences and future decision-making. Moreover, when traveling around both China and the U.S, we found the Airbnb experiences in these two countries are very different. Therefore, our study analyzed the different Airbnb experiences between these two countries to help both improve their experience. We analyzed it by comparing one representative city, Beijing, in China, and one, Washington D.C., in the United States.
The questions we aimed to answer in our study were the following:
- Where do the most expensive houses locate in D.C and Beijing? How expensive these houses could be?
- How did the price for Airbnb rentals fluctuate across the year? How is the price tendency in D.C and Beijing in the next coming year?
- What localities in Beijing and in D.C. are rated highly? Where are the champions of Top-Rated houses?
- What aspects did people praise those champion houses?
- How the amenities and attributes vary in Airbnb houses in D.C. and Beijing? What causes these differences?
- What are the main property types in Beijing and in D.C.?
Where are the most expensive houses in D.C and Beijing? How expensive could these houses be?
We visualized the top 100 expensive houses in Beijing and D.C. on the map. Different colors refer to different neighborhoods. Larger circles imply higher prices, while smaller circles imply cheaper prices.
From the visualization, we could see followings:
- The most expensive houses in the Beijing group tightly in the center of the city.
- The most expensive houses in D.C. are located in the north and the center part of the city relatively evenly.
- Even comparing the top 100 expensive houses, the most expensive one, whose price is 68980 Chinese yuan, is seven times more expensive than the cheapest one, whose price is 9882 Chinese yuan, in Beijing.
- While the most expensive one in D.C., price at 5995 dollars, is only four times the cheapest one, the price at 1450 dollars.
These two images show the median price of each district to give people understandings of how expensive the area is and how large scale the area with that price will influence.
From the visualizations, we can see that Airbnb houses’ price level in Beijing is much higher than D.C.’s Airbnb houses.
The reason might be that the houses’ availability in Beijing is less than houses in D.C.
Another reason might be Beijing’s long-term house price is also more expensive than the price in D.C.
How did the price for Airbnb rentals fluctuate across the year? How is the price tendency in D.C and Beijing in the next coming year?
This image visualized the monthly average price of Airbnb houses from September 2019 to September 2020.
From the visualization, we found that:
1. Overall, D.C has an upward trend in price while Beijing has a descending trend.
2. In terms of price fluctuation, the highest point is 1.09 times the lowest point according to D.C, while the highest point is 1.03 times the lowest point in Beijing. Therefore, we conclude that the short-rental market in D.C. is more volatile than in Beijing across a year.
What localities in Beijing and in D.C. are rated highly? Where are the champions of Top-Rated houses?
We calculate the average rate of each neighborhood both in D.C and Beijing and filtered the top 15 neighborhoods in each city.
We found that:
1. The average rate in D.C. is relatively higher than the average rate in Beijing.
2. D.C.’s highly rated places are more scattered than the highly-rated places in Beijing.
How the amenities and attributes vary in Airbnb houses in D.C. and Beijing? What causes these differences?
- This visualization shows that most amenities are not varied in Airbnb houses in Beijing and D.C.
- However, the percentage of carbon monoxide detector, dishwasher, dryer, Iron, smoke detector, microwave, indoor fireplace in D.C are at least as twice as in Beijing. Those amenities’ varieties are affected by the culture, life customs, and climate differences between Beijing and D.C.
- The only thing Beijing wins significantly is the elevator. Because Beijing has more skyscrapers than D.C.
- This visualization shows that houses in Beijing are more likely to provide parking places no matter free or paid. Besides, since Airbnb houses in Beijing are relatively smoking friendly and suitable for events, we presume that they are also suitable for business use.
- Houses in D.C. are more for personal use. They tend to be more family and kids friendly, supporting self-check-in.
What are the main property types in Beijing and in D.C.?
For these two images, we put all these marks on the map of D.C and Beijing, to leave an overall impression of the density of different properties in these two cities.
We can easily find that:
- In Beijing, the majority of Airbnb houses located in the central districts, such as Chaoyang district, Dongcheng district, Haidian district.
- The apartment is the most popular property type in both Beijing and D.C.
When looking deeper into property types by locations, we found that while majority properties in Beijing are apartments and guest suite, D.C. ‘s main property types are apartment, house, guest suite, condominium, and townhouse.
This visualization shows the detail of each district. From the visualization, we found that:
1. In Beijing, Chaoyang, Dongcheng, Fengtai, and Haidian has the most types of houses;
2. In D.C, properties are scattered to different districts.
Besides those comparisons above, we also achieved a simple interactive interface that helps customers explore Airbnb houses in D.C. by themselves. Users could select attributes interested, and zoom in or zoom out the map by clicking the navigator on the top-left of the screen or using the mouse. For instance, the “Count” shows that all listings in D.C. are grouped into several clusters, and the number of available listings in each cluster displays on it. Clusters that have more listings are larger and in orange color, while clusters that have fewer listings are smaller and in light blue. The “Price” shows the average price of listings in each cluster. The “Review” shows the average reviews’ rates of listings in each cluster. More operation guidance is shown in the demo video below.
Conclusions
We summarize our findings in the table below:
How did we make it?
We used the dataset from InsideAirbnb, which contains all Airbnb reservation information from September 2018 to September 2019. We visualized comparative data in Tableau. We built the interactive visualization using CARTO VL, Mapbox API, and JavaScript. The whole project lasted three months, from September to December 2019.
What else can we explore in the future?
Due to the time issue, we have not realize some functions designed in our plan. Such us the popularity of the houses, the tendency of the houses’ evaluations in the past years, the connection between locations and the polarity of their opinions and emotions. These functions will be great supplements for our project.
Furthermore, this project can be extended to analyze the motivation of people writing reviews, their education level, years of using Airbnb, their home country, and the date they post the review. It might be interesting to find that some people may share some common attributes.
