Data Visualization with Tableau
CitiBike is the nation's largest bike share program, with 20,000 bikes and over 1, 300 stations across Manhattan, Brooklyn, Queens, the Bronx and Jersey City. It was designed for quick trips with convenience in mind, and it’s a fun and affordable way to get around town. The fortunate news is that there is an ideal tool for this need, called Tableau Tableau is one of the most popular and effective tools for data visualization in todays' professional world. It allows data visualization professionals to create data stories that are visually appealing and easy to undersand for a non-technical audience. In addition, Tableau provides the tool to create powerful analytic dashboards and tell a clear story that can be easily shared with others. Tableau can be simple, requiring, little-to-no-coding, or quite complex, requiring some experiences to write custom code in. Tableau is very flexible.
For this project we are creating a data visualization for the New York City. The idea is to analyze the data, see the operations of the business and become informed about how well the bike sharing program really works. The initial information is that the program is so successful, we could expand it to other cities. With a strong, clear story, backed-up by data we can create a proposal on how the business could work in other cities as well.
The Tableau story of the NYC CitiBike can be found in the following link:
This page contains the following information:
- Data Source:
- CitiBike Trip History Data from August 2019 - .csv file in folder
- Software:
- Tableau Public 2020.3
- Languages & Environment:
- Pandas, Jupyter Notebook, Python 3.7
On first page of the story is the dashboard in the Tableau Public platform that contains basic overview
information about the data provided in the csv data set. The purpose of this page is to become familiar with the data
and to introduce the audience what kind of data we will be dealing with in the further analysis.
This page contains the following information:
- Type of business, time frame and location of the data; CitiCike, New York City, for the Month of August, 2019.
- Number of total rides: 2,344,224
- Customer type: Customers and Subscribers
This graph has the number of checkout bikes (number of trips) on the y-axis and trip duration on the x-axis.
X-axis is divided into hours and minutes for more analysis.
This graph has the number of checkout bikes (number of trips) on the y-axis and trip duration on the x-axis.
The color indicates the number of trips with orange representing male, blue for female, and red representing
the unknown genders. The highlite of the chart is that males checkout bikes at a much higher fequency than
female or unknown gender, especially between 1 hour to 1-17 hour durations on long trips.
This graph shows number of checked out bikes on the y-axis and trip duration on the x-axis.
This graph shows the number of trips per hour and per weekday that bikes were checked out. The hours are on the y-axis and weekdays on the x-axis. The colors speak to the volume. We can see that the busiest times are in the moring hours on weekdays from 6am to 9am and in evening hours on weekdays between 5pm and 7pm. Also Saturdays and Sundays are business in the middle of the day between 10am and 6pm. Again, males are the most frequent users.
This graph shows the number of trips per hour and per weekday that bikes were checked out, by user type (customers and subsribers) and by gender. The weekdays and user type are on the y-axis and the gender on x-axis. For subscribers, males has the highest number of trips especially on Thursdays and Fridays, followed by trips on Monday and Tuesdays.
This graph indicates, with dots, the accumulated trip utilization (by hours and minutes using the calculated field). Small bubbles are approx. 200,000 time units, Medium bubbles are approx. 300,000 time units, Large bubbles are approx. 2,000,000 time units. From this visualization and using the bike ids, maintenance and replacements can be planned.
These next two graphs show the most popular starting and ending locations. The larger bubbles represent the larger numbers.
The story of the NYC CitiBike starts off with the basic information about the users and the trips they take. From the first page we can learn a lot about users for bike share company and helps us understand data in further analysis and operations. Business Decisions made from the data might include these categories:
Bike Maintenance plays a big role in bike share business. Heat maps are a great visualization for large amounts of data and provide us clear story about the data. The charts suggest that the least busy times are between 11pm and 5am. These would be the best times to stagger the maintenance work throughout the fleet of bikes.
Customers and Subscribers Customers are the most important part in the business. CitiBike NYC is doing well because they understand the different users types (customers and subscribers) and their very different habits and needs. With this information, the best decisions can be made to support all needs and keep the busines growing.
Gender The charts reveal that there are very different behaviors for each gender. A deeper dive into the details could help formulate a marketing strategy and grow the business.
Trip Duration The most common checkout time is between 3-8 hours. The data lets us study the one way trips and two way trips too. We can strategize the pricing structure and possibly provide day passes or other pricing options for 2 way trips.
Further Analysis Recommendations
I would use this dashboard with a test group. I would see what questions came up with the audience
and further develop the details, based on that feedback.
Some ideas might include:
- Weather impact on Utilization (especially on low volume days, possibly remove those outliers).
- Covid impact on Utilization from 2019 before, during 2020 conditions and also 2021 to understand any possible new needs that have surfaced.
- How many bikes are and the "end of life" and would need to be replaced. Additional counts and charts could help with this capital planning work.