Cyclistic Bike-Share Analysis

Google Data Analytics Professional Certificate — Capstone Case Study

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

---

Overview

This project analyzes 12 months of real bike-share trip data (April 2025 – March 2026) from Divvy Bikes in Chicago to identify behavioral differences between casual riders and annual members. The goal is to inform a marketing strategy aimed at converting casual riders into annual members.

This analysis was completed as the capstone project for the Google Data Analytics Professional Certificate. Unlike the course's suggested dataset (2019–2020), this project uses the most recent data available to ensure findings reflect current rider behavior.

---

Key Findings

Finding	Members	Casual Riders
Primary use case	Commuting	Leisure / recreation
Peak days	Tue – Thu	Saturday – Sunday
Peak hours	8 AM and 5 PM (bimodal)	3 – 5 PM (gradual curve)
Median ride duration	8.7 min	11.8 min (+36%)
Weekend share of rides	23.4%	37.7%
Summer share of annual rides	~33%	~42%

Top 3 recommendations:

1. Launch a weekend-focused membership tier to lower the conversion barrier for leisure riders
2. Concentrate conversion campaigns in summer, when casual ridership peaks
3. Deploy targeted activation at high-casual stations (lakefront, parks, tourist areas)

---

Repository Structure

cyclistic-bike-share/
│
├── data/
│   ├── raw/                  # Monthly CSV files (not included — see Data section)
│   └── processed/            # Cleaned dataset (cyclistic_clean.parquet)
│
├── charts/                   # All exported visualizations (PNG)
│   ├── 01_rides_by_type.png
│   ├── 02_rides_by_dow.png
│   ├── 03_avg_duration_dow.png
│   ├── 04_rides_by_hour.png
│   ├── 05_monthly_trend.png
│   ├── 06_ride_length_boxplot.png
│   ├── 07_rides_by_season.png
│   └── 08_bike_type.png
│
├── analysis.py               # Full analysis script (Prepare → Share)
├── writeup.md                # Case study write-up (Ask → Act)
└── README.md

---

Data

Source: Divvy Trip Data, provided by Motivate International Inc.
License: Divvy Data License Agreement
Period: April 2025 – March 2026 (12 monthly CSV files)
Raw size: ~1 GB, 5,242,349 trips
After cleaning: 5,103,110 trips (2.66% removed)

The raw CSV files are not included in this repository due to size. To reproduce the analysis, download the monthly files from the link above and place them in data/raw/.

---

How to Reproduce

Requirements

python >= 3.10
pandas
matplotlib
seaborn
pyarrow

Install dependencies:

pip install pandas matplotlib seaborn pyarrow

Run the analysis

git clone https://github.com/your-username/cyclistic-bike-share.git
cd cyclistic-bike-share

# Download and place CSV files in data/raw/ (see Data section above)

python analysis.py

The script will:

1. Load and validate all 12 CSV files
2. Clean the data and engineer features
3. Compute all aggregations
4. Generate and export 8 visualizations to charts/
5. Save the cleaned dataset to data/processed/cyclistic_clean.parquet

---

Methodology

This analysis follows the Google Data Analytics six-phase process:

Phase	Description
Ask	Defined the business task and stakeholders
Prepare	Downloaded, validated, and documented the data source
Process	Cleaned data, removed invalid trips, engineered features
Analyze	Identified behavioral patterns by user type across time, duration, and geography
Share	Created 8 visualizations communicating key findings
Act	Formulated 3 actionable recommendations for the marketing team

Full documentation is in writeup.md.

---

Tools

Tool	Use
Python 3.12	Primary analysis language
pandas	Data loading, cleaning, and aggregation
matplotlib	Chart generation
seaborn	Statistical visualizations
pyarrow	Efficient Parquet storage

---

Author

Giacomo
Google Data Analytics Professional Certificate — April 2026
LinkedIn · Portfolio