Project Overview

Business Problem

Cyclistic wanted to increase annual memberships by understanding how casual riders and annual members use its bike-sharing service differently. My objective was to analyze historical ride data and identify patterns that could support targeted marketing strategies.

Business Question

How do annual members and casual riders use Cyclistic bikes differently?

Tools Used

• Microsoft Excel
  • Pivot Tables
  • Pivot Charts
  • XLOOKUP
• R
  • ggplot2
• Python
  • matplotlib
  • pandas
  • folium

Data Sources

I analyzed Quarter 1 Divvy bike trip data from 2019 and 2020 provided for the Google Data Analytics case study.

The datasets contained approximately 786,000 rides and included information such as:

• Ride dates and times
• Station locations
• Membership status
• Bike identifiers
• Geographic coordinates
• Demographic information (2019 only)

Data Preparation

Before analysis, I merged both datasets into a common file.

Cleaning steps included

• Removed duplicate records
• Standardized column names so both datasets could be merged.
• Corrected inconsistent data types
• Converted ride durations into consistent units
• Filled missing values with appropriate placeholders
• Standardized membership labels
• Calculated ride durations
• Added weekday fields
• Validated station coordinates using an additional dataset from Divvy.
• Removed rides under 60 seconds as these signified false starts or unusable rides.
• Corrected negative ride durations based on times submitted by separate pieces of hardware.
• Added missing latitude and longitude values using XLOOKUP

The final merged dataset contained 785,896 rides.

Data Validation

To improve confidence in the analysis, I:

• Verified duplicate records
• Confirmed date ranges
• Validated ride durations against timestamps
• Compared original and modified datasets
• Verified station coordinates
• Checked for missing and inconsistent values

Analysis

I compared casual riders and annual members across several columns.

Ride Duration

• Casual riders took significantly longer rides than annual members.
• Median ride duration was also consistently longer for casual riders.
• Extreme ride-length outliers reinforced the importance of using median values in addition to averages.

Day of Week

• Annual members rode most frequently during weekdays.
• Casual riders rode most often on weekends.
• This suggests annual members primarily use Cyclistic for commuting, while casual riders use the service more for recreation.

Time of Day

Annual members showed clear commuting peaks:

• Morning commute
• Evening commute

Casual riders rode more consistently throughout the afternoon and early evening.

Stations

• Certain stations were heavily used by annual members while others attracted more casual riders.
• This suggests location-based marketing opportunities.

Demographics

Where demographic information was available (2019), annual members tended to be born between 1960–1990 with the highest concentration between 1980 – 1990, while casual riders were more concentrated between 1980–1990.

Because demographic information was removed from the 2020 dataset, these findings should be interpreted cautiously.

Key Insights

The strongest behavioral differences were:

• Annual members tend to commute.
• Casual riders tend to ride recreationally and take longer trips.
• Ride timing differs significantly.
• Weekend usage is much higher among casual riders.

Recommendations

Based on the 2019 and 2020 ride data, we found there are 67,982 opportunities to convert Casual Riders to Annual Riders.

To create the highest return on investment, I recommend building our marketing strategy towards:

• Riders who frequent high traffic areas, like the Ogilvie Transportation Center, Monroe Harbor, and Navy Pier.
• Riders who ride on weekends and between the hours of 7:30-9:00AM and 4:30-6:00PM
• Riders who were born between 1980 – 1990.