Difference of usage between member and casual users of Cyclistic.

Context

Cyclistic is a bike rental company. The purpose of this analysis is to discover how annual and casual member use the service differently. The knowledge will be used in order to increase the conversion from casual to annual member.

Data

The data for this research was provided directly by the company. There is a csv report for every month, but only the last 12 months will be used. Every observations is anonymous and is describing a trip bike from station A at given start time to station B at given end time with the type of bike and type of user.

Yet the data still needed some cleaning, here are the different cleaning operation made :

  • Deleted when :
    • wrong gps informations
    • trip duration is less than 30 second and has a distance of 0
    • trip with an average speed of more than 50km/h (14m/s)

ROCCC status

  • Reliable, the data comes from a reputable source.
  • Original, comes directly from the source with only an anonymization step on top.
  • Comprehensive, all the important fields to solve our question are in the dataset
  • Current, the data is up to date as the range i chose was from 2023-12 to 2024-11.
  • Cited, the dataset is nicely prepared and shared by Divvy.

The pristine dataset contains 6.268.787 observations, after cleaning, it remains with 4.039.500 observations and it has 18 variables

Bike usage between user type

We can see a clear difference of usage between the member and the casual users during typical weekdays. The usage tend to be similar during the weekends. The number of trips increase while the season is getting better between April and October.

This chart is showing us a big difference in use between the annual and casual member of Cyclistic. Annual members have an average trip length around 10 and 15 minutes when the casual members have a more large average trips length around 12 and 28 minutes depending on the season of the year.

Thanks to the two previous charts, we can see that the season of the year correlate to the use of the bikes. We see longer rides and more trips between April and October.

This hour by day chart for casual and annual member also show a correlation between high usage from annual member and business hour. Although the weekend is more evenly distributed between the two category of user.

We can see major differences betweenthe top stations between the two groups.

Annual members top 3 stations :

  1. Kingsbury St. & Kinzie st. [ 26.558 ]

  2. Clinton St. & Washington Bldv. [ 25.501 ]

  3. Clinton St. & Madison St. [ 23.232 ]

Casual members top 3 stations :

  1. Streeter Dr. & Grand Ave. [ 44.391 ]

  2. DuSable Lake shore & Monroe St. [ 23.653 ]

  3. DuSable Lake shore & North Blvd. [ 23.079 ]

The top stations of each group are clearly in different area. The annual members top stations are located in the city where the casual member’s top stations are located near the lake michigan.

Summary of data exploration

Annual members :

  • Main creator of trips

  • Short duration of use

    • ~10 and 13 minutes on average during weekdays
    • A small increase of the duration during the weekend

Casual members :

  • Use the bikes between April and October

  • Longer use than annual members on average

    • depending on the month the average is very different
    • the weekend and monday usage is usually longer

Common :

  • Hours of usage are similar between groups.
  • April to October is the main season of usage

Differences :

  • Top stations very different locations

Stats and number :

  • Trips for members :

    • casual [ 1.402.329 ] (34.7%)

    • annual [ 2.637.171 ] (65.3%)

  • Average length trip for :

    • casual [ 22.43 min. ]

    • annual [ 12.26 min. ]

Conclusion

From the data, we can read that there is two big category of trip :

  1. Commute to work

    1. annual member mainly

    2. pic of usage weekdays 7am to 9am and 4pm to 7pm

  2. Weekend at the lakeshore trip

    1. casual members mainly

    2. pic of usage during weekends

    3. very low usage during winter

We did this analysis to

To convert casual users into an annual membership, we could therefore :

  1. Think about a seasonnal or weekend only membership that could better suit the casual rider usage

    1. strong usage from April to October

    2. mainly during weekends

  2. Use location hints for marketing :

    1. Top three stations for geographic marketing

      1. Streeter Dr. & Grand Ave. [ 44.391 ]

      2. DuSable Lake shore & Monroe St. [ 23.653 ]

      3. DuSable Lake shore & North Blvd. [ 23.079 ]

    2. Start campaigns around the starting season of casual riders (April)

  3. Since casual riders tend to frequent cultural and leisure destinations, we could explore creating a membership program in partnership with venues near the most popular stations for casual riders.

Sources & References

Code & Packages :

Programming : R - Rstudio 
Package tidyverse
Package dplyr
Package scales
Package lubridate
Package maps
Package ggmap

Data :

The Divvy bike rental company in Chicago is providing all data for this case study at : https://divvy-tripdata.s3.amazonaws.com/index.html