In depth analysis of a company's bikesharing data from summer in New York City was done in attempt to understand trends in rides and demographics.
##Links to my Tableau Public
##Resources Data Sources: 201908-citibike-tripdata.csv Tableau Desktop 2022.1.2 Visual Studio Code version 1.68.1
##Overview In depth analysis of a company's bikesharing data from summer in New York City was done in attempt to understand trends in rides and demographics. Trends would be used to predict customer use and need in a similar business model in a different city. The summer was chosen as it was thought to be the best example of a busy time, showing the most availability the new company would need. The type of user based on age and gender was analyzed for potential input in observed trends. The use of bikes was analyzed to show the number of bikes used, where the most popular bikesharing locations were, and which bikes would likely need repair more often due to higher use.
##Results
- Based on analysis of ride share data, it was found that a signicant portion of users were male and/or subscribers as seen in the pie charts below. Throughout further analysis of the data this trend remained apparent. Both male and subscriber groups have a large influence in the over-arching trends as the make up the vast majority of users studied. The graphs of user demographics are as follows:
- The average trip duration for rides was visualized to depict the length of checkout time for most users. From visualization 2, it is clear the majority of users are men as there is a significantly larger spike in the early trip duration time. This large spike for mens checkout time largely resembles the spike in general checkout times, showing males have a significant impact on trends for usage. Average users spend less than 20 minutes with a bike checked out. Filtering by hour of trip duration is available for both checkout time graphs if desired. The graphs of checkout times are as below:
- When studying the most common stop times of rides, it is shown that the highest concentration of bike rental returns is from 8-9am & 5-7pm Monday-Friday. This could potentially be due to people renting bikes for their transit to/from work. More dispersed stop times on Saturday & Sunday could be due the the weekend giving users more flexibility in their schedule. The stop time by hour of the day is shown as below:
- When analyzing the stop times as well as gender of user, there is a greater concentration in stop times in male users, once again highly influencing stop time trends seen previously in general user stop times and check out duration. There are similar patterns for males and females, the difference in gradience could be attributed to there being less females than males as users. Due to similar trends accross genders, gender is likely not a major factor in influencing stop time. The heatmap for stoptime sorted by gender is as shown below:
- The same trends are supported when visualizing the user trips by gender and subscription but sorted by weekday rather than stop time. Male subscribers show the most trips. The highest concentration of rides occurs on Thursdays and Fridays as seen below:
- Bike usage was analyzed by comparing the number of rides per bike Id to total rides. Certain bikes maintain a significantly higher number of rides than others. This could potentially be due to their location in higher ride concentration areas. These bikes will likely need repair before other, and should be monitored or serviced to ensure quality control. The bike usage by Id number is as follows:
##Summary
Overall, the following trends were found when studying the bike share data. Most rides will be under 20 minutes, this could mean serveral trips could be completed by a bike in a day which is promising for higher revenue with low supplies in the beginning stages of the new company. Males tend to encompase a large portion of users, so advertisements could either focus on women to expand the groups usage, or on marketing that connects to men to establish the same customer base in the new location. Majority of rides occur around the times that people would be going to or getting off work. The total number of bikes needed would have to be large enough to support user need during these times are they are the busiest. Repairs should additionally be avoided at these times to avoid unneccessary loss of high profit times. There are popular locations for start/stop of rides. These locations may be connected to high concentration of office or business areas as people go to work or neighborhoods and apartments as people come home. Similar areas should be attempted to be identified in the new business area so the high demand areas can have adequate supplies of bikes. Popular ride starting/stopping location tend to be similar so the company should not have to worry about transporting bikes back to start areas to ensure supply.
Additional visualizations could be created to show the relationship between bike usage by ID in the most popular start/stop areas. Rotation of these bikes to less popular areas could potentially reduce the amount of repairs needed in high demand areas. Rides per station id could be visualized to show the number of rides at the most and least popular stations.
##Contact Me
Email: [email protected]
LinkedIn